Amendments to the Accelerated Filer and Large Accelerated Filer Definitions — SEC — US

1 15 U.S.C. 78a et seq.
2See Acceleration of Periodic Report Filing Dates and Disclosure Concerning website Access to Reports, Release No. 33–8128 (Sept. 5, 2002) [67 FR 58480 (Sept. 16, 2002)]. The definitions in Rule 12b–2 are not enumerated, including the definition of ‘‘accelerated filer and large accelerated filer.’’ The paragraphs under the ‘‘accelerated filer and large accelerated filer’’ definition, however, are enumerated. Paragraph (1) defines ‘‘accelerated filer,’’ paragraph (2) defines ‘‘large accelerated filer,’’ and paragraph (3) discusses entering and exiting accelerated filer and large accelerated filer status. Also, although Rule 12b–2 defines the terms ‘‘accelerated filer’’ and ‘‘large accelerated filer,’’ it does not define the term ‘‘non-accelerated filer.’’ SECURITIES AND EXCHANGE COMMISSION 17 CFR Part 240 [Release No. 34–85814; File No. S7–06–19] RIN 3235–AM41 Amendments to the Accelerated Filer and Large Accelerated Filer Definitions AGENCY: Securities and Exchange Commission. ACTION: Proposed rules. SUMMARY: We are proposing amendments to the accelerated filer and large accelerated filer definitions to promote capital formation for smaller reporting issuers, by more appropriately tailoring the types of issuers that are included in the categories of accelerated and large accelerated filers and revising the transition thresholds for accelerated and large accelerated filers. The proposed amendments would exclude from the accelerated and large accelerated filer definitions an issuer that is eligible to be a smaller reporting company and had annual revenues of less than $100 million in the most recent fiscal year for which audited financial statements are available. In addition, the proposed amendments would increase the transition thresholds for accelerated and large accelerated filers becoming non-accelerated filers from $50 million to $60 million and for exiting large accelerated filer status from $500 million to $560 million. Finally, the proposed amendments would add a revenue test to the transition thresholds for exiting both accelerated and large accelerated filer status. As a result of the amendments, certain low-revenue issuers would not be required to have their assessment of the effectiveness of internal control over financial reporting attested to, and reported on, by an independent auditor, although they would continue to be required to make such assessments and to establish and maintain the effectiveness of their internal control over financial reporting. DATES: Comments should be received on or before July 29, 2019. ADDRESSES: Comments may be submitted by any of the following methods:
Electronic Comments

Use our internet comment form
(https://www.sec.gov/rules/ proposed.shtml); or
Send an email to rule-comments@
sec.gov. Please include File No. S7–06– 19 on the subject line. Paper Comments
Send paper comments to Secretary,
Securities and Exchange Commission, 100 F Street NE, Washington, DC 20549–1090. All submissions should refer to File Number S7–06–19. This file number should be included on the subject line if email is used. To help us process and review your comments more efficiently, please use only one method. We will post all comments on our internet website (https://www.sec.gov/rules/ proposed.shtml). Comments are also available for website viewing and printing in our Public Reference Room, 100 F Street NE, Washington, DC 20549 on official business days between the hours of 10:00 a.m. and 3:00 p.m. All comments received will be posted without change. Persons submitting comments are cautioned that we do not redact or edit personal identifying information from comment submissions. You should submit only information that you wish to make available publicly. We or the staff may add studies, memoranda, or other substantive items to the comment file during this rulemaking. A notification of the inclusion in the comment file of any such materials will be made available on our website. To ensure direct electronic receipt of such notifications, sign up through the ‘‘Stay Connected’’ option at www.sec.gov to receive notifications by email. FOR FURTHER INFORMATION CONTACT: John Fieldsend, Special Counsel, or Jennifer Riegel, Special Counsel, in the Division of Corporation Finance, at (202) 551– 3430, U.S. Securities and Exchange Commission, 100 F Street NE, Washington, DC 20549–3628. SUPPLEMENTARY INFORMATION: We are proposing amendments to 17 CFR 12b– 2 (‘‘Rule 12b–2’’) under the Securities Exchange Act of 1934 (‘‘Exchange Act’’).1
Table of Contents
I. Introduction
A. Background
B. Summary of the Proposed Amendments
II. Discussion of the Proposed Amendments
A. Historical and Current Relationship Between the SRC and Accelerated and Large Accelerated Filer Definitions B. ICFR Requirements
C. Proposed Amendments To Exclude
Low-Revenue SRCs From the
Accelerated and Large Accelerated Filer Definitions D. Proposed Amendments to the Transition Provisions in the Accelerated and Large Accelerated Filer Definitions E. Request for Comment
III. Economic Analysis
A. Introduction
B. Baseline
1. Regulatory Baseline
2. Characteristics of Accelerated Filer
Population
3. Timing of Filings
4. Internal Controls and Restatements
C. Discussion of Economic Effects
1. Affected Issuers
2. Comparison Populations
3. Potential Benefits of Eliminating the
ICFR Auditor Attestation Requirement for Affected Issuers
4. Potential Costs of Eliminating the ICFR
Auditor Attestation Requirement for Affected Issuers
5. Potential Benefits and Costs Related to
Other Aspects of the Proposed Amendments
6. Alternatives to the Proposed
Amendments
D. Request for Comment
IV. Paperwork Reduction Act
A. Summary of the Collections of Information B. Burden and Cost Estimates Related to the Proposed Amendments
1. ICFR Auditor Attestation Requirement
2. Filing Deadlines; Disclosure Regarding
Filing Availability and Unresolved Staff Comments
3. Total Burden Reduction
C. Request for Comment
V. Small Business Regulatory Enforcement
Fairness Act
VI. Initial Regulatory Flexibility Act Analysis
A. Reasons for, and Objectives of, the Proposing Action B. Legal Basis
C. Small Entities Subject to the Proposed
Rules
D. Projected Reporting, Recordkeeping, and Other Compliance Requirements E. Duplicative, Overlapping, or Conflicting Federal Rules F. Significant Alternatives G. Request for Comment
VII. Statutory Authority and Text of Proposed
Rule Amendments
I. Introduction
A. Background
In 2002, the Commission introduced a reporting regime that categorized issuers subject to the Exchange Act reporting requirements as non-accelerated, accelerated, and large accelerated filers.2 Under this regime, accelerated

See paragraphs (1) and (2) under the ‘‘accelerated filer and large accelerated filer’’ definition in Rule 12b–2. If an issuer does not meet the definition of accelerated filer or large accelerated filer, it is considered a non-accelerated filer. See Table 1 in
Section II.C. below for the definitions of
‘‘accelerated filer’’ and ‘‘large accelerated filer.’’ 3Accelerated and large accelerated filers are required to provide the disclosure required by Item 1B of 17 CFR 249.310 (‘‘Form 10–K’’) and Item 4A of 17 CFR 249.220f (‘‘Form 20–F’’) about unresolved staff comments on their periodic and/or current reports. Also, accelerated and large accelerated filers are required to provide certain disclosures about whether they make filings available on or through their internet website. See 17 CFR 229.101(e)(4). 4See Table 6 in Section III.B.1 below. 5See 17 CFR 240.13a–15(f) and 17 CFR 240.15d– 15(f) (defining ICFR). 6 15 U.S.C. 7262(a). 7See 17 CFR 240.13a–15 and 17 CFR 240.15d–15. 8 15 U.S.C. 7262(b). 9See 15 U.S.C. 7262(b). 10See 15 U.S.C. 7262(c). 11See 15 U.S.C. 7262 (SOX’s subheading is, ‘‘AN ACT To protect investors by improving the accuracy and reliability of corporate disclosures made pursuant to the securities laws, and for other purposes.’’). 12See Section III.C below. 13See, e.g., Smaller Reporting Company Regulatory Relief and Simplification, Release No. 33–8876 (Dec. 19, 2007) [73 FR 934 (Jan. 4, 2008)] (‘‘SRC Regulatory Relief Release’’) (discussing small business issuers and Regulation S–B). 14See id. 15 In addition, an SRC also was not required to provide the disclosure required by Item 1B of Form 10–K, and a non-accelerated filer was not required to provide the disclosure required by Item 4A of Form 20–F about unresolved staff comments on its periodic and/or current reports. Further, nonaccelerated filers were not required to provide certain disclosures about whether they make filings available on or through their internet website. See 17 CFR 229.101(e)(4). 16See Smaller Reporting Company Definition, Release No. 33–10513 (June 28, 2018) [83 FR 31992 (July 10, 2018)] (‘‘SRC Adopting Release’’). 17See 17 CFR 229.10(f)(1)(i), 17 CFR 230.405 (‘‘Rule 405’’), and Rule 12b–2. 18Public float is defined in paragraph (3)(i)(A) of the SRC definition in Rule 12b–2, which states that public float is measured as of the last business day of the issuer’s most recently completed second fiscal quarter and computed by multiplying the aggregate worldwide number of shares of its voting and non-voting common equity held by nonaffiliates by the price at which the common equity was last sold, or the average of the bid and asked prices of common equity, in the principal market for the common equity. See also 17 CFR 229.10(f) (2)(i)(A) and Rule 405. An entity with no public float because, for example, it has equity securities outstanding but is not trading in any public trading market would not be able to qualify on the basis of a public float test. 19To avoid situations where an issuer frequently enters and exits SRC status, each test includes two thresholds—one for initially determining whether an issuer qualifies as an SRC and a subsequent, lower threshold for issuers that did not initially qualify as an SRC. 20Annual revenues are measured as of the most
recently completed fiscal year for which audited financials are available. See 17 CFR 229.10(f)(2)(i)(B), Rule 405, and Rule 12b–2. 21See 17 CFR 229.10(f)(1), Rule 405, and Rule 12b–2. The prior revenue test included issuers with no public float and annual revenues of less than $50 million. See SRC Adopting Release, note 16 above, at 31995. The lower transition thresholds under the revenue test for an issuer that did not initially qualify as an SRC were revised from less than $40 million of annual revenues and no public float to less than $80 million of annual revenues and either no public float or a public float of less than $560 million. See Item 17 CFR 229.10(f)(2)(iii)(B), Rule 405, and Rule 12b–2. 22SRC Adopting Release, note 16 above, at 32005. 23 Id. at 31992. 24This amendment, among other things, preserved the existing thresholds in those definitions and did not change the number of issuers subject to the ICFR auditor attestation requirement. 25Although rare, under our existing rules, some issuers that meet the large accelerated filer definition may be eligible to be an SRC because of the expanded revenue test in the SRC definition. An issuer is eligible to be an SRC and a large accelerated filer if it: (1) Previously qualified as a large accelerated filer because its public float was $700 million or more; (2) its revenues for the most recent fiscal year were less than $100 million; and Continued and large accelerated filers are subject to shorter filing deadlines for quarterly and annual reports and are subject to some disclosure 3 and other requirements that do not apply to non-accelerated filers. The only difference between the requirements for accelerated and large accelerated filers is that large accelerated filers are subject to a filing deadline for their annual reports on Form 10–K that is 15 days shorter than the deadline for accelerated filers.4 A significant requirement that applies to accelerated and large accelerated filers, but not to non-accelerated filers, is the requirement that an issuer’s independent auditor must attest to, and report on, management’s assessment of the effectiveness of the issuer’s internal control over financial reporting (‘‘ICFR’’).5 Section 404(a) of the Sarbanes-Oxley Act (‘‘SOX’’) 6 requires almost all issuers, including smaller reporting companies (‘‘SRCs’’), that file reports pursuant to Exchange Act
Section 13(a) or 15(d) to establish and
maintain ICFR and have their management assess the effectiveness of their ICFR.7 In addition, SOX Section 404(b) 8 requires those issuers to have the independent accounting firm that prepares or issues their financial statement audit report attest to, and report on, management’s assessment of the effectiveness of their ICFR (‘‘ICFR auditor attestation’’).9 SOX Section 404(c),10 however, exempts from the ICFR auditor attestation requirement issuers that are neither large accelerated nor accelerated filers. Congress introduced the ICFR auditor attestation requirement as part of a package of regulations intended to improve the accuracy and reliability of corporate disclosures.11 Although there are benefits to the ICFR auditor attestation requirement, there are also costs and burdens, which we discuss in more detail below.12 Initially, the categories of issuers under the accelerated and large accelerated filer reporting regime existed separately from categories that the Commission created to provide regulatory relief to smaller entities.13 However, in 2007, when the Commission combined its separate disclosure regime for small business issuers with the regime for larger issuers, it attempted to align the SRC and non-accelerated filer categories, to the extent feasible, to avoid unnecessary complexity.14 As a result, an SRC generally was not an accelerated or large accelerated filer and did not have to comply with the accelerated or large accelerated filing deadlines or the ICFR auditor attestation requirement.15 This alignment changed in June 2018 when the Commission adopted amendments 16 to the SRC definition 17 to expand the number of issuers that qualify for scaled disclosure accommodations. The revised SRC definition allows an issuer to use either a public float 18 test or a revenue test (‘‘SRC revenue test’’) to determine whether it is an SRC. The amendments increased the threshold in the public float test for an issuer to initially qualify as an SRC from less than $75 million to less than $250 million.19 The Commission also expanded the revenue test to include issuers with annual revenues 20 of less than $100 million if they have no public float or a public float of less than $700 million.21 Before the amendments, the revenue test in the SRC definition applied only to issuers with no public float. In the SRC Adopting Release, the Commission estimated that raising the threshold used in the public float test and expanding the revenue test in the SRC definition would result in an additional 966 issuers being eligible for SRC status in the first year under the new definition.22 The Commission intended the amendments to promote capital formation for smaller reporting issuers by reducing compliance costs for the newly-eligible SRCs while maintaining appropriate investor protections.23 In conjunction with these amendments, the Commission also revised the accelerated filer and large accelerated filer definitions in Rule 12b2 to remove the condition
that, for an issuer to be an accelerated filer or a large accelerated filer, it must not be eligible to use the SRC accommodations.24 One result of these amendments is that some issuers now are categorized as both SRCs and accelerated or large accelerated filers.25

(3) its public float as of the end of the most recent second quarter is less than $560 million (or, for the first year after the new SRC rules are effective, is less than $700 million), such that it is eligible to be an SRC, but does not fall below the $500 million transition threshold necessary to exit large accelerated filer status. See SRC Adopting Release, note 16 above, at 31994 n.31 and 32001 n.128. We are proposing to revise the ‘‘large accelerated filer’’ definition so that an issuer that would be eligible to be an SRC under the SRC revenue test would not also qualify as a large accelerated filer. 26See SRC Adopting Release, note 16 above, at 32001. 27The issuer also would not have to provide the disclosure required by Item 1B of Form 10–K and Item 4A of Form 20–F about unresolved staff comments on its periodic and/or current reports or the disclosure required by Item 101(e)(4) of Regulation S–K about whether it makes filings available on or through its internet website. See 17 CFR 229.101(e)(4). 28See SRC Adopting Release, note 16 above, at 32001. 29As discussed in Section II.C below, the existing conditions for qualifying as an accelerated filer are that an issuer: (1) Had an aggregate worldwide public float of $75 million or more, but less than $700 million, as of the last business day of the issuer’s most recently completed second fiscal quarter; (2) has been subject to the requirements of 15 U.S.C. 78m (Exchange Act Section 13(a)) or 15 U.S.C. 78o(d) (Exchange Act Section 15(d)) for a period of at least twelve calendar months; and (3) has filed at least one annual report pursuant to those sections. For a large accelerated filer, conditions (2) and (3) are the same, but condition (1) is that an issuer had an aggregate worldwide public float of $700 million or more, as of the last business day of the issuer’s most recently completed second fiscal quarter. Also, as discussed in note 25 above, some issuers that meet the ‘‘large accelerated filer’’ definition may be eligible to be an SRC. 30The thresholds provided below are based on the initial thresholds of each definition; however, due to the transition provisions of the accelerated and large accelerated filer definitions, additional issuers may also be both an SRC and an accelerated or large accelerated filer. 31Amendments to Smaller Reporting Company Definition, Release No. 33–10107 (June 27, 2016) [81 FR 43130 (July 1, 2016)] (‘‘SRC Proposing Release’’). These issuers have some, but not all, of the benefits of scaled regulation and, in particular, are required to comply with earlier filing deadlines for annual and quarterly reports and the ICFR auditor attestation requirement. At the time of the SRC Adopting Release, as noted in that release, the Chairman directed the staff to formulate recommendations to the Commission for possible rule amendments that, if adopted, would have the effect of reducing the number of issuers that qualify as accelerated filers to promote capital
formation by reducing compliance costs for certain registrants, while maintaining appropriate investor protections.26 As part of the staff’s consideration of possible amendments to recommend, the Chairman directed the staff to consider, among other things, the historical and current relationship between the SRC and accelerated filer definitions. B. Summary of the Proposed Amendments We are proposing to amend the accelerated and large accelerated filer definitions in Rule 12b-2 to exclude any issuer that is eligible to be an SRC under the SRC revenue test. The effect of this proposal would be that such an issuer would not be subject to accelerated or large accelerated filing deadlines for its annual and quarterly reports or to the ICFR auditor attestation requirement.27 Other issuers that are eligible to be SRCs but are not excluded from the accelerated or large accelerated filer definition would need to satisfy all of the requirements applicable to an accelerated or large accelerated filer, including the ICFR auditor attestation requirement. Additionally, we are proposing to revise the transition provisions set forth in the ‘‘Entering and exiting accelerated filer and large accelerated filer status’’
section applicable to the Rule 12b-2
accelerated and large accelerated filer definitions. The proposed amendments would revise the public float transition threshold for accelerated and large accelerated filers to become a nonaccelerated filer from $50 million to $60 million. Also, the proposed amendments would increase the exit threshold in the large accelerated filer transition provision from $500 million to $560 million in public float to align the SRC and large accelerated filer transition thresholds. Finally, the proposed amendments would allow an accelerated or a large accelerated filer to become a non-accelerated filer if it becomes eligible to be an SRC under the SRC revenue test.
II. Discussion of the Proposed
Amendments
A. Historical and Current Relationship Between the SRC and Accelerated and Large Accelerated Filer Definitions Prior to the SRC amendments, the SRC category of filers generally did not overlap with either the accelerated or large accelerated filer categories.28 In addition, the accelerated and large accelerated filer definitions explicitly excluded any issuer eligible to use the SRC accommodations. Now, however, as illustrated in Figure 1 of this section, because the public float tests in the SRC and accelerated filer definitions partially overlap, and the accelerated and large accelerated filer definitions no longer specifically exclude an issuer that is eligible to be an SRC, an issuer meeting the accelerated filer definition 29 will be both an SRC and an accelerated filer 30 if it has:

A public float of $75 million or
more, but less than $250 million, regardless of annual revenues; or
Less than $100 million in annual
revenues, and a public float of $250 million or more, but less than $700 million. When the Commission proposed the amendments to the SRC definition,31 it did not propose to exclude the newlyeligible SRCs from the accelerated or large accelerated filer definitions but solicited comment on this point. Some EP29MY19.000</GPH>

32See, e.g., letters from Acorda Therapeutics, Inc. et al. (Aug. 23, 2016) (‘‘Acorda, et al.’’); Advanced Medical Technology Association (Aug. 20, 2016) (‘‘AMTA’’); Biotechnology Innovation Organization, (Aug. 30, 2016) (‘‘BIO’’); Calithera Biosciences (Aug. 8, 2016) (‘‘Calithera’’); CONNECT (Aug. 4, 2016) (‘‘CONNECT’’); Corporate Governance Coalition for Investor Value (Aug. 30, 2016) (‘‘Coalition’’); Council of State Bioscience Associations (Aug. 26,
2016) (‘‘CSBA’’); Independent Community Bankers
of America (Aug. 29, 2016) (‘‘ICBA’’); The Dixie Group, Inc. (July 11, 2016) (‘‘Dixie’’); MidSouth Bancorp, Inc. (Aug. 24, 2016) (‘‘MidSouth’’); Nasdaq (Aug. 30, 2016) (‘‘Nasdaq’’); National Venture Capital Association (Aug. 25, 2016) (‘‘NVCA’’); NYSE Group (July 25, 2016) (‘‘NYSE’’); and Seneca Foods Corporation (Aug. 2, 2016) (‘‘Seneca’’). However, some commenters expressed concern about amending the public float thresholds. See letters from BDO USA, LLP (Aug. 29, 2016); Center for Audit Quality and Counsel of Institutional Investors. (Aug. 30, 2016) (‘‘CAQ/CII’’); CFA Institute (Aug. 30, 2016) (‘‘CFA’’); and Ernst & Young LLP (Sept. 8, 2016) (‘‘EY’’). References to comment letters in this release refer to comments on the SRC Proposing Release, available at https:// www.sec.gov/comments/s7-12-16/s71216.htm, unless otherwise specified. 33See, e.g., letters from Acorda, et al.; AMTA; BIO; Calithera; CONNECT; Coalition; CSBA; ICBA; MidSouth; Nasdaq; NVCA; NYSE; and Seneca. 34See BIO; Coalition; Nasdaq; NVCA; and NYSE. 35 15 U.S.C. 78m(b)(2)(B) (referring to ‘‘internal accounting controls’’ rather than ICFR). 36 15 U.S.C. 7241. 37See 17 CFR 229.308, 17 CFR 240.13a–15, 17 CFR 240.15d–15, Form 20–F, Form 40–F, 17 CFR 270.30a–2, and 17 CFR 270.30a–3. 38 15 U.S.C. 78m(b)(2)(B)(i)–(ii). 39 15 U.S.C. 78m(b)(2)(B)(iii)–(iv). 40See 17 CFR 240.13a–14 or 17 CFR 240.15d–14 (requiring certification) and 17 CFR 229.601(b)(31) (prescribing certification content). 41See 17 CFR 229.308, 17 CFR 240.13a–15, 17 CFR 240.15d–15, Item 15 of Form 20–F, and Certifications 4 and 5 of Form 40–F. Effective ICFR is designed to provide reasonable assurance that an issuer’s financial disclosures are reliable and prepared in accordance with U.S. Generally Accepted Accounting Principles (‘‘U.S. GAAP’’) or International Financial Reporting Standards (‘‘IFRS’’). See 17 CFR 240.13a–15(f) and 17 CFR 240.15d–15(f). Effective ICFR includes policies and procedures designed to maintain records that, in reasonable detail, accurately and fairly reflect the transactions and dispositions of the assets of the issuer; provide reasonable assurance that transactions are recorded as necessary to permit preparation of financial statements in accordance with generally accepted accounting principles, and that receipts and expenditures of the issuer are being made only in accordance with authorizations of management and directors of the issuer; and provide reasonable assurance regarding prevention or timely detection of unauthorized acquisition, use or disposition of the issuer’s assets that could have a material effect on the financial statements. See 17 CFR 240.13a–15(f) and 17 CFR 240.15d–15(f). These controls can help prevent or detect financial misstatements, whether intentional or unintentional. Id. 42See 17 CFR 240.13a–15(d) and 17 CFR 240.15d–15(d). See also 17 CFR 229.308(c). A registered investment company (‘‘RIC’’) must disclose in each report on Form N–CSR any change in its ICFR that has
materially affected, or is reasonably likely to materially affect, its ICFR. See Item 11(b) of Form N–CSR [17 CFR 249.331; 17 CFR 274.128]. 43 15 U.S.C 80a–8. 44 15 U.S.C. 7263. Notwithstanding the exemption pursuant to SOX Section 405, RICs are required to provide the certifications pursuant to SOX Section 302 and to maintain ICFR. See 17 CFR 270.30a–2 and 270.30a–3; see also Management’s Report on Internal Control Over Financial Reporting and Certification of Disclosure in Exchange Act Periodic Reports, Release No. 34–47986 (June 5, 2003) [68 FR 36635 (June 18, 2003)]. RICs that are management companies, other than small business investment companies, are also required to file a copy of their independent public accountant’s report on internal controls. See Form N–CEN (17 CFR 274.101); see also Investment Company Reporting Modernization, Release No. IC–32314, notes 879–881 and accompanying text (Oct. 13, 2016) [81 FR 81870 (Nov. 18, 2016)]. Additionally, business development companies (‘‘BDCs’’) are subject to the rules adopted by the Commission to implement SOX Section 404. BDCs are a type of closed-end investment company that is not registered under the Investment Company Act and, therefore, not within the exemption provided by SOX Section 405. 45See Asset-Backed Securities, Release No. 33– 8518 (Dec. 22, 2004) [70 FR 1506, 1510 n.41 (Jan. 7, 2005)] (‘‘ABS Release’’). See also 17 CFR 240.13a–15(a) and 17 CFR 240.15d–15(a) and Instruction J to Form 10–K. 46See Items 15(b) and (d) of Form 20–F and Certifications 4 and 5 of Form 40–F. 47An EGC is defined as an issuer that had total annual gross revenues of less than $1.07 million during its most recently completed fiscal year. See Rule 405; Rule 12b–2; 15 U.S.C. 77b(a)(19); 15 Continued commenters recommended that the Commission increase the threshold in the accelerated filer definition to be consistent with changes to the SRC definition,32 reduce compliance costs associated with the ICFR auditor attestation requirement,33 and maintain uniformity across the rules.34 B. ICFR Requirements Issuer obligations with respect to internal accounting controls and ICFR derive primarily from the Foreign Corrupt Practices Act (‘‘FCPA’’), which added Section 13(b)(2)(B) to the Exchange Act; 35 SOX Sections 302 36 and 404(a); and related rules.37 Exchange Act Section 13(b)(2)(B) requires every issuer that is required to file reports pursuant to Exchange Act
Section 13(a) or 15(d) to devise and
maintain a system of internal accounting controls sufficient to provide reasonable assurances that transactions are executed in accordance with management’s general or specific authorization and recorded as necessary to permit preparation of financial statements in conformity with generally accepted accounting principles or any other criteria applicable to such statements and to maintain accountability for assets.38 Additionally, Exchange Act Section 13(b)(2)(B) requires that the issuer’s system of internal accounting controls provide reasonable assurances that access to assets is permitted only in accordance with management’s general or specific authorization and that the recorded accountability for assets is compared with the existing assets at reasonable intervals and appropriate action is taken with respect to any differences.39 Similarly, pursuant to SOX Section 302, the Commission adopted rules requiring the principal executive and financial officers of certain issuers filing reports pursuant to Exchange Act
Section 13(a) or 15(d) to certify that,
among other things, they are responsible for establishing and maintaining ICFR, have designed ICFR to ensure material information relating to the issuer and its consolidated subsidiaries is made known to such officers by others within those entities, and evaluated and reported on the effectiveness of the issuer’s ICFR.40 Also, pursuant to SOX
Section 404(a), the Commission adopted
rules requiring each annual report required by Exchange Act Section 13(a) or 15(d) to include a statement that it is management’s responsibility to establish and maintain adequate ICFR and to provide management’s assessment of the effectiveness of the issuer’s ICFR.41 Issuers must evaluate and disclose any change to their ICFR that occurred during each fiscal quarter.42 Although SOX Section 404 generally requires and directs the Commission to adopt rules regarding internal accounting controls and ICFR that apply to every issuer that is required to file reports pursuant to Exchange Act
Section 13(a) or 15(d), RICs under
Section 8 of the Investment Company
Act of 1940 (‘‘Investment Company Act’’) 43 are specifically exempted from SOX Section 404 by SOX Section 405.44 In addition, the Commission’s rules implementing the FCPA and SOX
Section 404 exempted other types of
issuers, such as asset-backed securities (‘‘ABS’’) issuers, from the ICFR obligations.45 The Commission also determined that foreign private issuers (‘‘FPIs’’) and Canadian multijurisdictional disclosure system (‘‘MJDS’’) issuers must have their management assess and report annually on the effectiveness of their ICFR as of the end of their fiscal year and evaluate and disclose any change in their ICFR that occurred during the period covered by the annual report.46 In addition to the responsibility of the issuer’s management to establish and maintain an effective internal control structure and procedures for financial reporting, the independent accounting firm that prepares or issues a financial statement audit report also helps support effective ICFR. SOX Section 404(b) requires any issuer subject to the rules the Commission adopted related to SOX Section 404(a), other than an emerging growth company (‘‘EGC’’),47 to

U.S.C. 78c(a)(80); and Inflation Adjustments and Other Technical Amendments under Titles I and II of the JOBS Act, Release No. 33–10332 (Mar. 31,
2017) [82 FR 17545 (Apr. 12, 2017)].
Similar to other issuers, BDCs that qualify as an EGC or as a non-accelerated filer are not subject to the auditor attestation requirement in SOX Section 404(b). Unlike the Commission’s SRC definition, the statutory definition of EGC does not exclude BDCs. See 15 U.S.C. 78c(80). Given the existing regulatory regime for BDCs and the context of the Jumpstart Our Business Startups (‘‘JOBS’’) Act of 2012, Public Law 112–106, Sec. 103, 126 Stat. 306 (2012), we believe that BDCs can qualify as EGCs. BDCs invest in startup companies and EGCs for which they make available significant managerial experience, and are subject to many of the disclosure and other requirements from which Title I of the JOBS Act provides exemptions, including executive compensation disclosure, say-on-pay votes, management discussion and analysis, and SOX Section 404(b). 48See PCAOB Accounting Standard (‘‘AS’’) 2110, Identifying and Assessing Risks of Material Misstatement, paragraphs .18–.40. 49See Study and Recommendations on Section 404(b) of the Sarbanes-Oxley Act of 2002 For Issuers With Public Float Between $75 and $250 Million 97–99 and 102–104 (Apr. 2011) (‘‘2011 SEC Staff Study’’), available at https://www.sec.gov/ news/studies/2011/404bfloat-study.pdf. 50See, e.g., letters from CAQ, CFA, and Deloitte (Aug. 23, 2016). 51See, e.g., letters from Acorda et al., AMTA, BIO, Calithera, Coalition, CONNECT, CSBA, Dixie, and Seneca. One commenter estimated that it will spend more than $400,000 annually on compliance with SOX Section 404(b) upon expiration of its EGC status. See letter from Calithera. Another commenter estimated that relief from SOX Section 404(b) would result in a 35% reduction in compliance costs. See letter from Seneca. 52See, e.g., letters from Acorda et al., BIO, CSBA, ICBA, and NVCA. One commenter stated that expanding relief from the ICFR auditor attestation requirement to issuers with a public float of less than $250 million would encourage capital formation because the reduced audit and disclosure requirements may encourage companies that have been hesitant to go public to do so. See letter from ICBA (citing a 2005 ICBA study that estimated that audit fees for publicly held bank holding companies would drop dramatically—some by as much as 50%—if these companies were exempted from the ICFR auditor attestation requirement). 53See, e.g., letters from Acorda et al., BIO, CONNECT, CSBA, and Seneca. 54See, e.g., letters from Acorda et al. and CONNECT. 55See Section III.C.1 below. 56This figure is based on staff analysis of data from the Center for Research in Security Prices database for December 1998 versus December 2018. The estimates exclude RICs and issuers of American depositary receipts (‘‘ADRs’’). 57This figure is based on staff analysis of data from Standard & Poor’s Compustat and Center for Research in Security Prices databases for fiscal year 1998 versus fiscal year 2017. The estimates exclude RICs and issuers of ADRs. 58See the large accelerated
filer definition in Rule 12b–2. 59See proposed subparagraph (1)(iv) of the definition of accelerated filer and proposed subparagraph (2)(iv) of the definition of large accelerated filer in Rule 12b–2. have the accounting firm that prepares or issues its financial statement audit report attest to, and report on, management’s assessment of the effectiveness of the issuer’s ICFR. Under the current Public Company Accounting Oversight Board (‘‘PCAOB’’) risk assessment standards,48 the independent auditor for the ICFR attestation considers certain information that is similar to information it considers for purposes of the issuer’s financial statement audit. SOX Section 404(c) exempts non-accelerated filers from SOX Section 404(b)’s ICFR auditor attestation requirement. The ICFR auditor attestation requirement is intended to enhance the reliability of management’s disclosure related to ICFR. It also may help an issuer identify and disclose a significant deficiency or material weakness in ICFR that had not been identified or properly characterized by management.49 In response to the SRC Proposing Release, some commenters indicated that the ICFR auditor attestation requirement strengthens the quality and reliability of issuers’ ICFR, which enhances investor protection.50 At the same time, the ICFR auditor attestation requirement is associated with certain costs that may be significant, particularly for lowrevenue issuers. In response to the SRC Proposing Release, several commenters indicated that this requirement is the most costly aspect of being an accelerated filer 51 and that audit fees and other costs associated with the ICFR auditor attestation requirement can divert capital from core business needs.52 Some commenters asserted that these costs are especially burdensome for emerging and growing biotechnology issuers,53 with a few of these commenters specifying that the costs of the requirement represent over $1 million of capital diversion from such issuers.54
C. Proposed Amendments To Exclude
Low-Revenue SRCs From the
Accelerated and Large Accelerated Filer Definitions We are proposing amendments to revise the accelerated and large accelerated filer definitions to exclude from those definitions issuers that are eligible to be an SRC under the SRC revenue test. Permitting these issuers to avoid the burdens of being an accelerated or large accelerated filer may enhance their ability to preserve capital without significantly affecting the ability of investors to make informed investment decisions based on the financial reporting of those issuers. Additionally, the benefits of having those issuers comply with the accelerated and large accelerated filer requirements may be more limited than for other issuers. Further, the proposed amendments are targeted at issuers whose representation in public markets has decreased over the years, and may be a positive factor in the decision of additional companies to register their offering or a class of their securities, which would provide an increased level of transparency and investor protection with respect to those companies. As discussed below,55 the number of issuers listed on major exchanges with market capitalizations below $700 million decreased by about 65%,56 and the number of listed issuers with less than $100 million in revenue decreased by about 60% 57 from 1998 to 2017. The issuers targeted by the proposed amendments would not incur the cost of the ICFR attestation until they exceed the SRC revenue test. Under the existing accelerated filer definition in Rule 12b–2, an issuer must satisfy three conditions to be an accelerated filer. First, the issuer must have a public float of $75 million or more, but less than $700 million, as of the last business day of the issuer’s most recently completed second fiscal quarter. Second, the issuer must have been subject to the requirements of Exchange Act Section 13(a) or 15(d) for a period of at least twelve calendar months. Third, the issuer must have filed at least one annual report pursuant to those same Exchange Act sections. Similarly, to be a large accelerated filer, an issuer must meet the second and third conditions just described and have a public float of $700 million or more as of the same measurement date.58 We are proposing to add a new condition to the definitions of accelerated filer and large accelerated filer that would exclude from those definitions an issuer eligible to be an SRC under the SRC revenue test.59 The table below summarizes the current and proposed conditions to be considered an accelerated and large accelerated filer under Rule 12b–2.

60Under the proposed amendments, an FPI that qualifies as an SRC under the SRC revenue test and is eligible to use the scaled disclosure requirements available to SRCs would qualify for the exclusion under the accelerated filer definition. This position is consistent with past guidance we have provided to FPIs. See Smaller Reporting Company Regulatory Relief and Simplification, Release No. 33–8876 (Dec. 19, 2007) [73 FR 934 (Jan. 4, 2008)] (‘‘2007 SRC Adopting Release’’) (noting that an FPI may also qualify as an SRC and has the option to make filings on forms available to U.S. domestic issuers if it presents financial statements pursuant to U.S. GAAP). 61See paragraphs (1) and (3)(iii)(A) of the SRC definition in Rule 12b–2. 62See ABS Release, note 45 above, at 1501 n.41. See also Instruction J to Form 10–K. 63See note 44 above. 64See Section III.C.6.b below. 65See Fair Value Measurement (Topic 820), Financial Accounting Standards Board (‘‘FASB’’) Accounting Standards Update No. 2010–06 (Jan. 2010). 66 17 CFR 210.6–01 et seq. 67See 17 CFR 210.6–07. 68Form N–CEN requires that the report be based on the review, study, and evaluation of the accounting system, internal accounting controls, and procedures for safeguarding securities made during the audit of the financial statements for the reporting period. The report should disclose any material weaknesses in: (a) The accounting system; (b) system of internal accounting control; or (c) procedures for safeguarding securities which exist as of the end of the registrant’s fiscal year. See Instruction 3 to Item G.1 of Form N–CEN. 69Although the proposed amendments do not specifically address BDCs, we are soliciting comment on whether alternative approaches would be appropriate and the relative costs and benefits of such alternatives.
TABLE 1—CURRENT AND PROPOSED ACCELERATED FILER AND LARGE ACCELERATED FILER CONDITIONS
Current accelerated filer conditions Proposed accelerated filer conditions The issuer has a public float of $75 million or more, but less than $700 million, as of the last business day of the issuer’s most recently completed second fiscal quarter. Same. The issuer has been subject to the requirements of Exchange Act Section 13(a) or 15(d) for a period of at least twelve calendar months. Same. The issuer has filed at least one annual report pursuant Exchange Act
Section 13(a) or 15(d).
Same.
The issuer is not eligible to use the requirements for SRCs under the revenue test in paragraphs (2) or (3)(iii)(B), as applicable, of the ‘‘smaller reporting company’’ definition in Rule 12b–2. Current large accelerated filer conditions Proposed large accelerated filer conditions The issuer has a public float of $700 million or more, as of the last business day of the issuer’s most recently completed second fiscal quarter. Same. The issuer has been subject to the requirements of Exchange Act Section 13(a) or 15(d) for a period of at least twelve calendar months. Same. The issuer has filed at least one annual report pursuant Exchange Act
Section 13(a) or 15(d).
Same.
The issuer is not eligible to use the requirements for SRCs under the revenue test in paragraphs (2) or (3)(iii)(B), as applicable, of the ‘‘smaller reporting company’’ definition in Rule 12b–2. The proposed new conditions would only be available to issuers that are eligible to be an SRC under the SRC revenue test.60 Issuers that are eligible to be an SRC that have a public float between $75 million and $250 million 61 would be accelerated filers if their annual revenues are $100 million or more, and thus they would remain subject to all of the requirements applicable to accelerated filers. We are proposing to refer to ‘‘paragraphs (2) or (3)(iii)(B), as applicable’’ of the SRC definition in the proposed rule text instead of referring to the actual numerical thresholds specified in those paragraphs. We preliminarily believe that referring to the SRC definition would be the clearest and most efficient way to codify the requirement given that the thresholds could change in the future. The SRC definition excludes ABS issuers, RICs, BDCs, and majorityowned subsidiaries of issuers that do not qualify as an SRC. ABS issuers are exempt from ICFR reporting obligations.62 While RICs are also exempt from SOX Section 404,63 BDCs are not exempt. BDCs and majorityowned subsidiaries of a non-SRC parent are subject to the ICFR auditor attestation requirement to the same extent as other accelerated and large accelerated filers. As a result, even if these issuers were to fall within the public float and revenue thresholds in the SRC revenue test, they cannot rely on the SRC revenue test because they are excluded from the SRC definition. We estimate that 28 BDCs would meet the same public float and revenue thresholds as the issuers affected by the proposed rules, which constitutes about 60% of the total number of BDCs.64 We further estimate that one majorityowned subsidiary of a non-SRC parent may meet the same thresholds. We considered potential amendments to the definition of accelerated filer and large accelerated filer that would specifically address BDCs. Unlike investors in low-revenue noninvestment company issuers, investors in BDCs may place greater significance on the financial reporting of BDCs, many of which hold illiquid portfolio securities valued using level three inputs of the U.S. GAAP fair value hierarchy.65 The SRC revenue test would not be meaningful for BDCs because BDCs prepare financial statements under Article 6 of Regulation S–X 66 and generally do not report revenue. Instead, BDCs report investment income (dividends, interest on securities, fee income, and other income) and realized and unrealized gains and losses on investments on their statements of operations.67 RICs also prepare financial statements under
Article 6 of Regulation S–X. Even
though RICs are not subject to SOX
Section 404, RICs are subject to an
independent public accountant’s report on internal controls requirement through Form N–CEN.68 Expanding BDCs’ ability to be considered nonaccelerated filers, in contrast, would reduce auditor review of internal controls for a significant majority of BDCs. Accordingly, the proposed amendments to the definitions of accelerated and large accelerated filer do not specifically address BDCs.69 The tables below summarize the current and proposed relationships

70Tables 2 and 3 include only the initial SRC and accelerated filer thresholds and exclude the transition thresholds. A large accelerated filer may be eligible to be an SRC only through the transition threshold, so the table does not reflect the relationship between SRCs and large accelerated filers. 71See AS 2301, The Auditor’s Response to the Risks of Material Misstatement, paragraph .16. 72See id., paragraph .17. 73See 2110, note 48 above, paragraphs .18–.40. 74See id., paragraph .18. 75See id., paragraph .20. 76See generally AS 2201, An Audit of Internal Control Over Financial Reporting That Is Integrated with An Audit of Financial Statements. This standard relates to testing of design and whether the controls are implemented are part of the ICFR auditor attestation requirement. 77See AS 1305, Communications About Control Deficiencies in an Audit of Financial Statements, and id., at paragraphs .78–.80. 78See generally AS 2710, Other Information in Documents Containing Audited Financial Statements. between SRCs and non-accelerated and accelerated filers.70
TABLE 2—EXISTING RELATIONSHIPS BETWEEN SRCS AND NON-ACCELERATED AND ACCELERATED FILERS
Existing relationships between SRCs and non-accelerated and accelerated filers Status Public float Annual revenues SRC and Non-Accelerated Filer ....................................... Less than $75 million ....................................................... N/A. SRC and Accelerated Filer .............................................. $75 million to less than $250 million ............................... N/A. $250 million to less than $700 million ............................. Less than $100 million. Accelerated Filer (not SRC) ............................................. $250 million to less than $700 million ............................. $100 million or more.
TABLE 3—PROPOSED RELATIONSHIPS BETWEEN SRCS AND NON-ACCELERATED AND ACCELERATED FILERS
Proposed relationships between SRCs and non-accelerated and accelerated filers Status Public float Annual revenues SRC and Non-Accelerated Filer ....................................... Less than $75 million ....................................................... N/A. $75 million to less than $700 million ............................... Less than $100 million. SRC and Accelerated Filer .............................................. $75 million to less than $250 million ............................... $100 million or more. Accelerated Filer (not SRC) ............................................. $250 million to less than $700 million ............................. $100 million or more. The proposed amendments would increase the number of issuers that are exempt from the ICFR auditor attestation requirement by increasing the number of non-accelerated filers. Although the proposed amendments could, in some cases, result in investors receiving less or different disclosure about material weaknesses in ICFR at low-revenue SRCs than under our current rules, based on our experience with these matters, including in the cases of EGCs, SRCs, and other smaller reporting issuers, we believe it is unlikely there would be a significant effect on the ability of investors to make informed investment decisions based on the financial reporting of those issuers. A non-accelerated filer that meets the SRC revenue test would remain subject to many of the same obligations as accelerated and large accelerated filers with respect to ICFR, including the requirements for establishing, maintaining, and assessing the effectiveness of ICFR and for management to assess internal controls. Additionally, pursuant to the PCAOB’s recently adopted risk assessment standards in financial statement audits, in many cases auditors are testing operating effectiveness of certain internal controls even if they are not performing an integrated audit. For instance, an auditor may rely on internal controls to reduce substantive testing in the financial statement audit. To rely on internal controls, the auditor must obtain evidence that the controls selected for testing are effectively designed and operating effectively during the entire period of reliance.71 Also, an auditor must test the controls related to each relevant financial statement assertion for which substantive procedures alone cannot provide sufficient appropriate audit evidence.72 The proposed amendments would not relieve an independent auditor of its obligation to consider ICFR in the performance of its financial statement audit of an issuer, if applicable, regardless of whether the issuer is subject to the ICFR auditor attestation requirement, as is the case today with respect to issuers that are nonaccelerated filers.73 For example, the risk assessment requirement in a financial statement audit is similar to that in an ICFR attestation audit. In a financial statement audit, the auditor is required to identify and assess the
risks of material misstatements. The auditor is, therefore, required to ‘‘obtain a sufficient understanding of each component of [ICFR] to (a) identify the types of potential misstatements, (b) assess the factors that affect the risks of material misstatement, and (c) design further audit procedures.’’ 74 This understanding includes evaluating the design of the controls relevant to the audit and determining whether the controls have been implemented.75 A similar evaluation is required in an ICFR attestation.76 Also, evaluation and communication to management and the audit committee of significant deficiencies and material weaknesses in ICFR are required in both a financial statement audit and an ICFR attestation audit.77 When the auditor becomes aware of a material weakness, it has the responsibility to review management’s disclosure for any misstatement of facts, such as a statement that ICFR is effective when there is a known material weakness, including in a financial statement audit.78 Further, as discussed above, auditors may also test operating effectiveness of internal controls in a financial statement audit, such as when the auditor determines to rely on those controls to reduce the substantive testing. We note that, because certain of the information considered by the independent auditor for the ICFR attestation is also considered by the auditor for purposes of the issuer’s financial statement audit, some of the audit fees and the other audit-related costs associated with the ICFR auditor

79See Section III.C.3 below.
80See Section III.C.3.b below.
81See Section III.C.3.c below.
82Although the proposed amendments would not eliminate the attestation requirement for all accelerated filers that are SRCs, we are soliciting comment on whether such an approach would be appropriate and the relative costs and benefits of such an approach for both issuers and investors. 83See Section III.A below. 84See Section III.C.1 below. 85See, e.g., letters from Acorda et al., BIO, CONNECT, CSBA, and Seneca. 86See Section III.C.4.a below. 87See, e.g., SRC Regulatory Relief Release, note 13 above, 2007 SRC Adopting Release, note 60 above, and SRC Adopting Release, note 16 above. 88For example, Title I of the JOBS Act amended SOX Section 404(b) to exempt EGCs from the ICFR auditor attestation requirement. In addition, Section 72002 of the Fixing America’s Surface Transportation Act of 2015 requires the Commission to revise Regulation S–K to further scale or eliminate requirements to reduce the burden on EGCs, accelerated filers, SRCs, and other smaller issuers, while still providing all material information to investors. See Public Law 114–94, 129 Stat. 1312 (2015). 89See proposed paragraphs (3)(ii) and (iii) of the ‘‘accelerated and large accelerated filer’’ definition in Rule 12b–2. 90See proposed paragraph (3)(iii) of the ‘‘accelerated and large accelerated filer’’ definition in Rule 12b–2. 91See proposed paragraph (3)(ii) of the ‘‘accelerated and large accelerated filer’’ definition in Rule 12b–2. 92See proposed paragraph (3)(iii) of the ‘‘accelerated and large accelerated filer’’ definition in Rule 12b–2. attestation requirement are included in the issuer’s financial statement audit costs. However, for issuers with less complex financial systems and controls, such as issuers with lower revenues, this may be less likely to be the case under the proposed amendments. For these issuers, the auditor could determine that, in the absence of an ICFR auditor attestation requirement, it may be a more effective and efficient financial statement audit approach to not rely on and have to test the operating effectiveness of certain controls, such as those related to revenue recognition. Therefore, eliminating the ICFR auditor attestation requirement could have a greater impact in the reduction of costs for such issuers. As discussed in more detail in the Economic Analysis section below,79 there are a number of component costs of the ICFR auditor attestation requirement. In general, the largest individual cost component relates to audit fees that would typically not be incurred in audits in which an ICFR attestation is not required.80 We estimate that such audit fees would average approximately $110,000 per year for accelerated filers with revenues of less than $100 million. The ICFR auditor attestation requirement is also associated with additional costs,81 and we estimate that these non-audit costs would average approximately $100,000 per year for accelerated filers. We believe that the proposed amendments would eliminate these two types of costs for
issuers that are eligible to be an SRC under the SRC revenue test. Although certain requirements and costs of the ICFR attestation overlap with those associated with a financial statement audit, we continue to believe that the ICFR auditor attestation requirement incrementally can contribute to the reliability of financial disclosures, particularly for issuers that typically have more complex financial reporting requirements and processes. Accordingly, the proposed amendments would not eliminate the requirement for all accelerated filers that are SRCs. Instead, the proposed amendments reflect a more tailored approach that recognizes that the impact of the ICFR auditor attestation requirement on the reliability of an issuer’s financial disclosures is not necessarily the same across all issuers, including all SRCs.82 As noted in this section above, and discussed in greater detail below,83 the compliance costs associated with the ICFR auditor attestation requirement may be disproportionately burdensome for the issuers that are eligible to be an SRC under the SRC revenue test and, as with all compliance requirements, these costs may divert funds otherwise available for reinvestment by these issuers because they have less access than other issuers to internallygenerated capital. In this regard, the issuers we expect to be affected by the proposed amendments are concentrated in a few specific industries. For example, 36.1% of the issuers that are eligible to be an SRC under the SRC revenue test are in the ‘‘Pharmaceutical Products’’ or ‘‘Medical Equipment’’ industries,84 and a number of commenters noted that the attestation requirement is especially burdensome for biotechnology issuers.85 We believe these and other low-revenue issuers would particularly benefit from the cost savings associated with non-accelerated filer status and could re-direct those savings into growing their business without significantly affecting the ability of investors to make informed investment decisions based on the financial reporting of those issuers. Further, the benefits of the ICFR auditor attestation requirement may be smaller for issuers with low revenues because they may be less susceptible to the risk of certain kinds of misstatements, such as those related to revenue recognition. Also, it is possible that low-revenue issuers may have less complex financial systems and controls and, therefore, be less likely than other issuers to fail to detect and disclose material weaknesses in the absence of an ICFR auditor attestation. Additionally, we note the financial statements of low-revenue issuers may, in many cases, be less critical to assessing their valuation given, for example, the relative importance of their future prospects.86 Providing this benefit to low-revenue SRCs is consistent with our historical practice of providing scaled disclosure and other accommodations for smaller issuers 87 and with recent actions by Congress to reduce burdens on new and smaller
issuers.88 Issuers that are eligible to be an SRC under the SRC revenue test no longer would be required to comply with accelerated or large accelerated filer requirements, reducing these issuers’ compliance costs and thereby enhancing their ability to preserve capital without significantly affecting the ability of investors to make informed investment decisions based on the financial reporting of those issuers. D. Proposed Amendments to the Transition Provisions in the Accelerated and Large Accelerated Filer Definitions We are also proposing to amend the transition thresholds for issuers exiting accelerated and large accelerated filer status. First, the proposed amendments would revise the public float transition threshold for accelerated and large accelerated filers to become a nonaccelerated filer from $50 million to $60 million.89 Second, the large accelerated filer public float transition provision would be revised from $500 million to $560 million.90 Finally, the proposed amendments would add the SRC revenue test to the transition threshold for accelerated 91 and large accelerated filers.92 Under the current rules, once an issuer is an accelerated or a large accelerated filer, it will not become a non-accelerated or accelerated filer until its public float falls below a specified lower threshold than the public float threshold that it needed to become an accelerated or large accelerated filer initially. The purpose of this lower threshold is to avoid situations in which an issuer frequently enters and exits accelerated and large accelerated filer status due to small fluctuations in its public float. Currently, an issuer initially becomes an accelerated filer after it first meets certain conditions as of the end of its fiscal year, including that it had a public

93See paragraph (3)(ii) of the ‘‘accelerated and large accelerated filer’’ definition in Rule 12b–2. 94See paragraph (3)(iii) of the ‘‘accelerated and large accelerated filer’’ definition in Rule 12b–2. 95For example, under the current rules, if an issuer that is a non-accelerated filer determines at the end of its fiscal year that it had a public float of $75 million or more, but less than $700 million, on the last business day of its most recentlycompleted second fiscal quarter, it will become an accelerated filer. On the last business day of its next fiscal year, the issuer must re-determine its public float to re-evaluate its filer status. If the accelerated filer’s public float fell to $70 million on the last business day of its most recently-completed second fiscal quarter, it would remain an accelerated filer because its public float did not fall below the $50 million transition threshold. Alternatively, if the issuer’s public float fell to $49 million, it would then become a non-accelerated filer because its newly-determined public float is below $50 million. As another example, an issuer that has not been a large accelerated filer but had a public float of $700 million or more on the last business day of its most recently completed second fiscal quarter would then become a large accelerated filer at the end of its fiscal year. If, on the last business day of its subsequently completed second fiscal quarter, the issuer’s public float fell to $600 million, it would remain a large accelerated filer because its public float did not fall below $500 million. If, however, the issuer’s public float fell to $490 million at the end of its most recently-completed second fiscal quarter, it would become an accelerated filer at the end of the fiscal year because its public float fell below $500 million. Similarly, if the issuer’s public float fell to $49 million, the issuer would become a non-accelerated filer. 96See note 16 above. float of $75 million or more but less than $700 million as of the last business day of its most recently completed second fiscal quarter. An issuer initially becomes a large accelerated filer in a similar manner, including that it had a public float of $700 million or more as of the last business day of its most recently completed second fiscal quarter. Once the issuer becomes an accelerated filer, it will not become a non-accelerated filer unless it determines at the end of a fiscal year that its public float had fallen below $50 million on the last business day of its most recently completed second fiscal quarter.93 Similarly, a large accelerated filer will remain one unless its public float had fallen below $500 million on the last business day of its most recently completed second fiscal quarter.94 If the large accelerated filer’s public float falls below $500 million but is $50 million or more, it becomes an accelerated filer. Alternatively, if the issuer’s public float falls below $50 million, it becomes a
non-accelerated filer.95 The table below summarizes the existing transition thresholds and how an issuer’s filer status changes based on its subsequent public float determination.
TABLE 4—SUBSEQUENT DETERMINATION OF FILER STATUS BASED ON PUBLIC FLOAT UNDER EXISTING REQUIREMENTS
Existing requirements
Initial public float determination Resulting filer status Subsequent public float determination Resulting filer status $700 million or more ......................... Large Accelerated Filer .................... $500 million or more ........................ Large Accelerated Filer. Less than $500 million but $50 million or more. Accelerated Filer. Less than $50 million ....................... Non-Accelerated Filer. Less than $700 million but $75 million or more. Accelerated Filer .............................. Less than $700 million but $50 million or more. Accelerated Filer. Less than $50 million ....................... Non-Accelerated Filer. The proposed amendments would revise the transition threshold for becoming a non-accelerated filer from $50 million to $60 million and the transition threshold for leaving the large accelerated filer status from $500 million to $560 million. We preliminarily believe it would be appropriate to increase these transition thresholds because doing so would make the public float transition thresholds 80% of the initial thresholds, which would be consistent with the percentage used in the transition thresholds for SRC eligibility. In the SRC Adopting Release,96 we amended the SRC rules so that the SRC transition thresholds were set at 80% of the corresponding initial qualification thresholds. Revising these transition thresholds to be 80% of the corresponding initial qualification thresholds would align the transition thresholds across the SRC, accelerated filer, and large accelerated filer definitions. Additionally, revising these thresholds would limit the cases in which an issuer could be both an accelerated filer and an SRC or a large accelerated filer and an SRC, thereby reducing regulatory complexity.
TABLE 5—SUBSEQUENT DETERMINATION OF FILER STATUS BASED ON PUBLIC FLOAT UNDER PROPOSED AMENDMENTS
Proposed amendments to the public float thresholds Initial public float determination Resulting filer status Subsequent public float determination Resulting filer status $700 million or more ......................... Large Accelerated Filer .................... $560 million or more ........................ Large Accelerated Filer. Less than $560 million but $60 million or more. Accelerated Filer. Less than $60 million ....................... Non-Accelerated Filer. Less than $700 million but $75 million or more. Accelerated Filer .............................. Less than $700 million but $60 million or more. Accelerated Filer. Less than $60 million ....................... Non-Accelerated Filer. In addition, the proposed amendments would add the SRC revenue test to the public float transition thresholds for accelerated and large accelerated filers. We are proposing that an issuer that is already an accelerated filer will remain one unless either its public float falls below $60 million or it becomes eligible to use

97Under the proposed amendments, an accelerated filer with revenues of $100 million or more that is eligible to be an SRC based on the public float test contained in paragraphs (1) and (3)(iii)(A) of the SRC definition could transition to non-accelerated filer status in a subsequent year if it had revenues of less than $100 million. For example, assuming the proposed amendments were in effect, an issuer with a December 31 fiscal year end that has a public float as of June 29, 2018 of $230 million and annual revenues for the fiscal year ended December 31, 2017 of $101 million would be eligible to be an SRC under the public float test, but because the issuer would not be eligible to be an SRC under the SRC revenue test it would be an accelerated filer (assuming the other conditions described in Table 1 were also met). At the next determination date (June 28, 2019), if its public float as of June 28, 2019 remained at $230 million and its annual revenues for the fiscal year ended December 31, 2018 were less than $100 million, that issuer would be eligible to be an SRC under the SRC revenue test (in addition to the public float test) and thus it would also become a non-accelerated filer. On the other hand, assuming the proposed amendments were in effect, an issuer with a December 31 fiscal year end that has a public float as of June 29, 2018 of $400 million and annual revenues for the fiscal year ended December 31, 2017 of $101 million would not be eligible to be an SRC under either the public float test or the SRC revenue test and would be an accelerated filer (assuming the other conditions described in Table 1 were also met). At the next determination date (June 28, 2019), if its public float as of June 28, 2019 remained at $400 million, that issuer would not be eligible to be an SRC under the SRC revenue test unless its annual revenues for the fiscal year ended December 31, 2018 were less than $80 million, at which point it would be eligible to be an SRC under the SRC revenue test and also become a nonaccelerated filer. 98See SRC Adopting Release, note 16 above, at note 31 (‘‘For purposes of the first fiscal year ending after effectiveness of the amendments, a registrant will qualify as a SRC if it meets one of the initial qualification thresholds in the revised definition as of the date it is required to measure its public float or revenues (the ‘measurement date’), even if such registrant previously did not qualify as a SRC.’’) the SRC accommodations under the revenue test in paragraphs (2) or (3)(iii)(B), as applicable, of the SRC definition. An issuer that is initially applying the SRC definition or previously qualified as an SRC would apply paragraph (2) of the SRC definition. Once an issuer determines that it does not qualify for SRC status, it would apply paragraph (3)(iii)(B) of the SRC definition at its next annual determination. As discussed above, paragraph (2) of the SRC definition states that an issuer qualifies as an SRC if its
annual revenues are less than $100 million and it has no public float or a public float of less than $700 million. Paragraph (3)(iii)(B) of the SRC definition states, among other things, that an issuer that initially determines it does not qualify as an SRC because its annual revenues are $100 million or more cannot become an SRC until its annual revenues fall below $80 million.97 Therefore, under the proposed amendments, an accelerated filer would remain an accelerated filer until its public float falls below $60 million or its annual revenues fall below the applicable revenue threshold ($80 million or $100 million), at which point it would become a non-accelerated filer. Similarly, we are proposing conforming amendments to the large accelerated filer transition provisions that describe when an issuer that is already a large accelerated filer transitions to either accelerated or nonaccelerated filer status. As discussed above, to transition out of large accelerated filer status at the end of the issuer’s fiscal year, an issuer would need to have a public float below $560 million as of the last business day of its most recently completed second fiscal quarter or meet the revenue test in paragraph (2) or (3)(iii)(B), as applicable, of the SRC definition. A large accelerated filer would become an accelerated filer at the end of its fiscal year if its public float fell to $60 million or more but less than $560 million as of the last business day of its most recently completed second fiscal quarter and its annual revenues are not below the applicable revenue threshold ($80 million or $100 million). The large accelerated filer would become a nonaccelerated filer if its public float fell below $60 million or it meets the revenue test in paragraph (2) or (3)(iii)(B), as applicable, of the SRC definition. For a large accelerated filer to meet the SRC revenue test, generally, its public float would need to fall below $560 million as of the last business day of its most recently completed second fiscal quarter and its annual revenues would need to fall below the applicable revenue threshold ($80 million or $100 million). One exception to this requirement is that an issuer that was a large accelerated filer whose public float had fallen below $700 million (but remained $560 million or more) but became eligible to be an SRC under the SRC revenue test in the first year the SRC amendments became effective would become a non-accelerated filer even though its public float remained at or above $560 million.98 If the SRC revenue test were not added to the accelerated filer and large accelerated filer transition provisions, an issuer’s annual revenues would never factor into determining whether an accelerated filer could become a non-accelerated filer, or whether a large accelerated filer could become an accelerated or nonaccelerated filer. For example, if the SRC revenue test is not added to the transition provisions, an accelerated filer with a public float
that remains more than $60 million but less than $700 million and with annual revenues of $100 million or more would not be able to become a non-accelerated filer even if its annual revenues subsequently fall below $80 million. E. Request for Comment We request and encourage any interested person to submit comments regarding the proposed amendments, specific issues discussed in this release and other matters that may have an effect on the proposals. We note that comments are of the greatest assistance if accompanied by supporting data and analysis of the issues addressed in those comments.

Should we exclude an issuer that
is eligible to be an SRC under the SRC revenue test from the accelerated and large accelerated filer definitions, as proposed? Why or why not? Are there investor protection benefits in distinguishing an issuer that is eligible to be an SRC under the SRC revenue test from an SRC that does not meet the revenue test and therefore would be an accelerated or large accelerated filer? Should we use different criteria to identify issuers to exclude from the accelerated and large accelerated filer definitions? If so, what criteria should we use and why?
With respect to the ICFR auditor
attestation requirement, is the issuer’s level of revenues relevant to the complexity of its financial systems and controls and the nature of its ICFR? If so, how does that complexity affect the benefits and costs of ICFR auditor attestation? How do the benefits and costs of the ICFR auditor attestation requirement vary with the complexity of an issuer’s financial reporting? Are the financial statements of low-revenue issuers less susceptible to the risk of material misstatements or control deficiencies such that the effect of an ICFR auditor attestation may be less significant than for other types of issuers? Would the proposed approach allow low-revenue issuers to benefit from cost savings without significantly affecting the ability of investors to make informed investment decisions based on the financial reporting of those issuers?
As an alternative, should we
instead exclude all SRCs from the accelerated and large accelerated filer definitions? Why or why not? What would be the effects, including the benefits and costs, of such an approach for issuers and investors? What would be the effects on the reliability of such issuers’ financial reporting or their

99A $250 million or less public float threshold would be consistent with the SRC definition, and we estimate that the average of the investment income of BDCs with market capitalization ranging from $75 to $700 million is $50 million. See Section III.C.6 below. susceptibility to the risk of material misstatements or control deficiencies? What would be the effects on these issuers’ willingness to be public companies? How would such an alternative affect investor protection? Are there additional considerations relevant to such issuers that we should consider? If we were to adopt such an approach, should we adjust the public float and annual revenue thresholds in the accelerated filer definition to be the same as those in the SRC definition? That is, should the accelerated filer definition include only issuers with a public float of $250 million or more but less than $700 million that had revenues of $100 million or more in the previous year? Would this approach have an effect on the transition between accelerated filer and non-accelerated status? If so, what would be the effect? If we were to adopt this approach, should we revise the transition thresholds for large accelerated, accelerated, and/or non-accelerated filers? Alternatively, should we exclude SRCs from the definition of accelerated filer without changing the thresholds in the definition itself? Why or why not? Would these approaches have different effects that we should consider?
4. In the SRC Adopting Release, the
Commission established the SRC revenue test to include issuers with annual revenues of less than $100 million if they have no public float or a public float of less than $700 million. The proposed amendments would use the SRC revenue test’s $100 million annual revenue threshold to determine whether an issuer would qualify as an accelerated or large accelerated filer. Should the proposed amendments use the SRC revenue test’s $100 million annual revenue threshold? Why or why not? Should there be a different annual revenue threshold for determining whether an issuer is an accelerated or large accelerated filer? Why or why not?
5. Would it be more appropriate to
determine filer status for any given year by using the average of an issuer’s public float, or applying some other metric, such as the issuer’s volumeweighted average price (‘‘VWAP’’)? What would be the appropriate way to calculate an issuer’s VWAP? If filer status were determined through the use of a VWAP calculation, should shares held by affiliates be included in the calculation of the issuer’s market value or public float? Why or why not? Should a VWAP calculation reflect the average VWAP over a longer period of time? If so, what longer period of time (e.g., three consecutive trading days, one week, one month, or one quarter), or different metric, would be more appropriate? What costs and benefits would be associated with use of a longer period of time or a different valuation standard? For example, if an average of an issuer’s public float over a longer period of time is used, are there additional costs to issuers to compute their aggregate worldwide number of shares of common equity held by nonaffiliates on each of the respective days? If we used a longer period of time or different valuation standard in the accelerated filer definitions, should we similarly revise other provisions that require an issuer to calculate its public float on a single day, such as in the Rule 12b–2 definition of an SRC?
6. Should all SRCs that meet the
accelerated filer definition be excluded from only the accelerated reporting deadlines? Would investors be adversely affected by expanding the population of issuers that would report later than they do today?
7. Should we increase the nonaccelerated filer transition threshold
from $50 million to $60 million and/or the large accelerated filer transition threshold from $500 million to $560 million, as proposed? Why or why not? Should we revise the non-accelerated filer transition threshold to one other than $60 million and/or the large accelerated filer transition threshold to one other than $560 million? If so, what threshold would be appropriate?
8. Should we align the transition
thresholds in the accelerated filer and large accelerated filer definitions with the SRC revenue test transition threshold, as proposed? Why or why not? Instead of aligning the transition thresholds, should we consider other approaches to the transition thresholds in the accelerated filer and large accelerated filer definitions? For example, should we adjust the transition provisions of the large accelerated filer definition to permit all issuers with a public float below $700 million and annual revenues below $100 million to become non-accelerated filers even if such issuers would not meet the transition thresholds to qualify as SRCs? Why or why not? For example, what would be the effects of any such alternatives on the frequency with which an issuer enters and exits large accelerated, accelerated, or nonaccelerated filer status due to small fluctuations in public float or revenues?
9. Should we adjust the transition
provisions of the accelerated filer and large accelerated filer definitions to include the respective public float and annual revenue thresholds in the definitions, rather than referencing the SRC revenue test? Why or why not?
10. We request comment on
alternative approaches that would include or exclude additional issuer types from the accelerated and large accelerated filer definitions. For example, should we exclude FPIs from the proposed amendments? Why or why not? Should we permit BDCs and majority-owned subsidiaries of nonSRCs, which are excluded from the definition of SRC, to be non-accelerated filers if they meet the SRC revenue test thresholds? Why or why not? The SRC revenue test thresholds are based, in
part, on an issuer’s annual revenues.
Are there alternative metrics that should be applied for BDCs instead of revenue? For example, should we use investment income received by the BDC rather than revenue? Should we include realized gains and losses from the sale of portfolio securities? Should unrealized gains and losses affect a BDC’s revenue for this purpose, and if so, how? Should we use the net increase or decrease in net assets resulting from operations? Alternatively, should we also exclude BDCs if they meet the public float test in the SRC definition alone? Should we have a specific BDC test of $250 million or less in public float and $50 million or less in investment income? 99 Why or why not? Are there other alternatives we should consider, such as providing an independent accountant’s report on internal controls similar to the one required by Form N–CEN? If we were to require a Form N–CEN report, should we apply the requirement only to those BDCs that were previously required to provide a report under SOX Section 404(b)?
11. Should we provide a definition for
the term ‘‘non-accelerated filer?’’ If so, should we define it as a filer that is not an accelerated or large accelerated filer? Why or why not? Should we use some other definition?
12. The proposed rule would refer to
‘‘paragraphs (2) or (3)(iii)(B)’’ of the SRC definition instead of referring to the actual numerical thresholds specified in those paragraphs. Should we include the actual numerical thresholds? Why or why not?
13. For the low-revenue issuers that
would be newly exempted from the ICFR auditor attestation requirement under the proposed amendments, would an auditor engaged for the purpose of a financial statement only audit be as likely to test the operating effectiveness of certain of the issuer’s internal

100 15 U.S.C. 78c(f).
101 15 U.S.C. 78w(a)(2).
102Non-accelerated filers also are not required to provide disclosure required by Item 1B of Form 10– K and Item 4A of Form 20–F about unresolved staff comments on their periodic and/or current reports or disclosure required by Item 101(e)(4) of Regulation S–K about whether they make filings available on or through their internet websites. 103As discussed below, issuers may not always choose to voluntarily obtain an ICFR auditor attestation even when the total benefits of doing so would exceed the total costs because they may not internalize some of the market-level benefits of compliance and because the incentives of managers may not be aligned perfectly with those of shareholders. 104See Section III.C.1 for detail on the data sources and methodologies underlying these estimates. controls to reduce the amount of substantive testing it performs as it may do under our existing rules? Given the potential for such testing as well as the risk assessment standards that apply to a financial statement only audit, to what extent would the consideration of internal controls by the auditors of these issuers change as a result of the proposed amendments?
14. Should we consider any changes
in how and where issuers report their accelerated filer status, public float, or revenue? Should we consider any new disclosure requirements associated with the proposed amendments? For example, should we permit or require issuers that voluntarily comply with SOX Section 404(b) to disclose that information, such as on the cover page of their periodic filings? If so, should we require issuers that voluntarily comply with SOX Section 404(b) to include the ICFR auditor attestation with the filing?
15. In lieu of, or in addition to, the
proposed amendments, should we consider amendments that would result in ICFR attestation audits being required at a reduced frequency? For example, should we require the proposed affected issuers to provide an ICFR auditor attestation only once every three years? If required once every three years, what financial reporting periods should we require the ICFR attestation audit to cover? Currently, the ICFR attestation audit is required to cover only the current period. Should we require the ICFR attestation audit to cover only the current period or should it include all three years?
III. Economic Analysis
We are mindful of the costs and benefits of the proposed amendments. Exchange Act Section 3(f) requires us, when engaging in rulemaking that requires us to consider or determine whether an action is necessary or appropriate in the public interest, to consider, in addition to the protection of shareholders, whether the action will promote efficiency, competition, and capital formation.100 Exchange Act
Section 23(a)(2) requires us, when
adopting rules, to consider the impact that any new rule would have on competition and prohibits any rule that would impose a burden on competition that is not necessary or appropriate in furtherance of the purposes of the Exchange Act.101 The discussion below addresses the economic effects of the proposed amendments, including their anticipated costs and benefits, as well as the likely effects of the proposed amendments on efficiency, competition, and capital formation. We also analyze the potential costs and benefits of reasonable alternatives to what is proposed. Where practicable, we have attempted to quantify the economic effects of the proposal; however, in certain cases, we are unable to do so because either the necessary data are unavailable or certain effects are not quantifiable. In these cases, we provide a qualitative assessment of the likely economic effects. A. Introduction As discussed above, we are proposing amendments to the definition of ‘‘accelerated filer’’ that will expand the number of issuers that qualify as nonaccelerated filers. Currently, issuers with no public float or public float of less than $75 million are generally nonaccelerated filers. The proposed amendments would generally extend non-accelerated filer status to issuers with a greater public float if they are eligible to be SRCs and their revenues are less than $100 million. As nonaccelerated filers, these issuers would not be required to obtain an ICFR auditor attestation pursuant to SOX
Section 404(b). They also would be
permitted an additional 15 days and five days, respectively, after the end of each period to file their annual and quarterly reports, relative to the deadlines that apply to accelerated filers.102 The proposed amendments also would revise the transition provisions for accelerated and large accelerated filer status, including increasing the public float thresholds to exit accelerated and large accelerated filer status from $50 million and $500 million in public float to $60 million and $560 million in public float. As discussed above, the ICFR auditor attestation requirement was introduced together with other changes to the financial reporting control environment with the intention of improving the accuracy and reliability of corporate disclosures. Section III.C.4.a discusses the evidence that the imposition of the ICFR auditor attestation requirement has been associated with benefits to issuers and investors. However, this requirement has also been associated with significant compliance costs. Relative to other issuers that are subject to this requirement, the affected issuers may find the costs to be particularly burdensome, while the ICFR auditor attestation requirement may, on average, provide fewer benefits related to the accuracy and reliability of these issuers’ financial statements. Further, issuers exempted from this requirement may choose to voluntarily obtain an ICFR auditor attestation if investors demand it or the issuers otherwise deem it, from their perspective, to be the best use of their resources.103 The proposed amendments are therefore intended to reduce compliance costs for these issuers without significantly affecting the ability of investors to make informed investment decisions based on the financial reporting of those issuers. In particular, we estimate that the affected issuers have median annual revenues of about $40 million and a median number of employees of about 125, while their median public float is about $145 million.104 The costs of providing an ICFR auditor attestation include some fixed costs that do not scale proportionately with size, and may therefore be disproportionately burdensome for smaller issuers. For the affected issuers, these costs may represent a meaningful percentage of their cash flows. Importantly, because these issuers have limited access to internally-generated capital, compliance costs may be more likely to displace spending on other things such as investment, research, or hiring than for other issuers subject to the ICFR auditor attestation requirement. Exempting these issuers from this requirement would allow them the discretion to invest their funds in the way they believe is most value-enhancing. At the same time, the ICFR auditor attestation requirement may, on average, provide fewer benefits related to these issuers versus other issuers subject to this requirement. We find preliminary evidence consistent with the argument that, compared to other issuers subject to the ICFR auditor
attestation requirement, the affected issuers may be less susceptible to the risk of certain kinds of misstatements (such as those related to revenue recognition). Although we expect that exempting these issuers may result in some adverse effects on the effectiveness of their ICFR and their

105Specifically, the requirements apply to all issuers that file reports pursuant to Section 13(a) or 15(d) of the Exchange Act. 106See Section 13(b)(2)(B) of the Exchange Act. 107Although there is substantial overlap between an issuer’s disclosure controls and procedures and ICFR, there are elements of each that are not subsumed by the other. 108See 17 CFR 240.13a–14 and 17 CFR 240.15d– 14. 109See 17 CFR 240.13a–14(b) and 17 CFR 240.15d–14(b). 110See 17 CFR 240.13a–15 and 17 CFR 240.15d–
15. A newly public issuer is also not required to
provide a SOX Section 404(a) management report on ICFR until its second annual report filed with the Commission. See Instructions to Item 308 of Regulation S–K. 111See Management’s Report on Internal Control Over Financial Reporting and Certification of Disclosure in Exchange Act Periodic Reports, Release No. 33–8238 (June 5, 2003) [68 FR 36635 (June 18, 2003)]. These evaluations of ICFR, as well as any associated auditor assessments of ICFR, should be based on a suitable, recognized control framework. The most widely used framework for this purpose is the one set forth in a report of the Committee of Sponsoring Organizations of the Treadway Commission (‘‘COSO’’). 112See AS 2110, note 48 above. See also the discussion below in this section about this auditing standard. 113See 17 CFR 240.10A–3. 114Part 363 of the FDIC regulations requires that the auditor of an insured depository institution with consolidated total assets of $1 billion or more (as of the beginning of the fiscal year) examine, attest to, and report separately on the assertion of management concerning the effectiveness of the institution’s internal control structure and procedures for financial reporting. 115Up to about seven percent of exempt issuers voluntarily provided an ICFR auditor attestation from 2005 through 2011. See U.S. Gov’t Accountability Office, GAO–13–582, Internal Controls: SEC Should Consider Requiring Companies to Disclose Whether They Obtained an Auditor Attestation (July 2013) (‘‘2013 GAO Study’’). restatement rates, we preliminarily believe that these effects are unlikely to result in a rate of restatements for the affected issuers that exceeds that for the issuers that would remain subject to this requirement. Moreover, in many cases, the market value of the affected issuers may be driven to a greater degree by their future prospects than by the current period’s financial statements. We find evidence consistent with this argument, which could further mitigate the extent of the adverse effects of eliminating the ICFR auditor attestation requirement for these issuers. The discussion that follows examines the potential benefits and costs of the proposed amendments in detail, with consideration for the likelihood that the effects of the ICFR auditor attestation have changed over time with changes in auditing standards and other market conditions. B. Baseline To assess the economic impact of the proposed amendments, we are using as our baseline the current state of the market under the existing definition of ‘‘accelerated filer.’’ This section discusses the current regulatory requirements and market practices. It also provides statistics characterizing accelerated filers, the timing of filings, disclosures about ineffective ICFR, and restatement rates under the baseline.

Regulatory Baseline
Our baseline includes existing statutes and Commission rules that govern the responsibilities of issuers with respect to financial reporting, as well as PCAOB auditing standards and market standards related to the implementation of these responsibilities. In particular, accelerated and large accelerated filers are subject to accelerated filing deadlines for their periodic reports relative to nonaccelerated filers. These deadlines are summarized in Table 6 below. All registrants can file Form 12b–25 (‘‘Form NT’’) to avail themselves of an additional 15 calendar days to file an annual report, or an additional five calendar days to file a quarterly report, and still have their report deemed to have been timely filed.
TABLE 6—FILING DEADLINES FOR
PERIODIC REPORTS
Category of filer
Calendar days after period end Annual Quarterly Non-Accelerated Filer ................... 90 45 Accelerated Filer ... 75 40 Large Accelerated Filer ................... 60 40
Section II.B. above discusses in detail
the issuer and auditor responsibilities with respect to disclosure controls and procedures and ICFR for issuers of different filer types. These responsibilities reflect the FCPA requirements with respect to internal accounting controls as well as a number of different changes to the financial reporting control environment that were introduced by SOX. In particular, all issuers 105 are required to devise and maintain an adequate system of internal accounting controls 106 and to have their corporate officers assess the effectiveness of the issuer’s disclosure controls and procedures 107 and disclose the conclusions of their assessments, typically on a quarterly basis.108 In addition, all issuers are required to have their corporate officers certify in each of their periodic reports that the information in the report fairly presents, in all material respects, the issuer’s financial condition and results of operations.109 All issuers other than RICs and ABS issuers 110 are also required to include management’s assessment of the effectiveness of their ICFR in their annual reports.111 Further, all issuers are required to have the financial statements in their annual reports examined and reported on by an independent auditor, who, even if not engaged to provide an ICFR auditor attestation, is responsible for considering ICFR in the performance of the financial statement audit.112 Also, an auditor engaged in a financial statement only audit may choose, though it is not required, to test the operating effectiveness of some internal controls in order to reduce the extent of substantive testing required to issue an opinion on the financial statements. Finally, all issuers listed on national exchanges are required to have an audit committee that is composed solely of independent directors and is directly responsible for the appointment, compensation, retention and oversight of the issuer’s independent auditors.113 Importantly, all of these responsibilities with respect to financial reporting and ICFR apply equally to non-accelerated as well as accelerated and large accelerated filers. Beyond these requirements, accelerated filers and large accelerated filers other than EGCs, RICs, and ABS issuers are required under SOX Section 404(b) and related rules to include an ICFR auditor attestation in their annual reports. In addition, certain banks, even if they are non-accelerated filers, are required under Federal Deposit Insurance Corporation (‘‘FDIC’’) rules to have their auditor attest to, and report on, management’s assessment of the effectiveness of the bank’s ICFR and reporting procedures (the ‘‘FDIC auditor attestation requirement’’).114 Some issuers that are not required to comply with SOX Section 404(b) voluntarily obtain an ICFR auditor attestation.115 Estimates of the number of issuers of

116The estimates in this table are based on staff analysis of self-identified filer status for issuers filing annual reports on Forms 10–K, 20–F, or 40– F in calendar year 2017, excluding any such filings that pertain to fiscal years prior to 2016. Staff extracted filer status from filings using a computer program supplemented with hand collection and compared the results for robustness with data from XBRL filings, Ives Group Audit Analytics, and Calcbench. Foreign issuers in this table represent those filing on Forms 20–F or 40–F and do not include FPIs that choose to file on Form 10–K. EGC issuers are identified by using data from Ives Group Audit Analytics and/or by using a computer program to search issuer filings, including filings other than annual reports, for a statement regarding EGC status. The estimates generally exclude RICs because these issuers rarely file on the annual report types considered. This table also excludes 135 issuers, mostly Canadian MJDS issuers filing on Form 40–F (which does not require disclosure of filer status or public float), for which filer type is unavailable. 117The estimated number of non-accelerated filers includes approximately 586 ABS issuers, which are not required to comply with SOX Section
404. Staff estimates that very few, if any, ABS
issuers are accelerated or large accelerated filers. ABS issuers are identified as issuers that made distributions reported via Form 10–D. 118See note 76 above. 119AS No. 5 was renumbered as AS 2201, effective Dec. 31, 2016. See Reorganization of PCAOB Auditing Standards and Related Amendments to PCAOB Standards and Rules, PCAOB Release No. 2015–002 (Mar. 31, 2015). 120See Auditing Standard No. 5, An Audit of Internal Control Over Financial Reporting That Is Integrated with An Audit of Financial Statements, and Related Independence Rule and Conforming Amendments, PCAOB Release No. 2007–005A (June 12, 2007). See also Public Company Accounting Oversight Board; Order Approving Auditing Standard No. 5, An Audit of Internal Control Over Financial Reporting that is Integrated with an Audit of Financial Statements, a Related Independence Rule, and Conforming Amendments, Release No. 34–56152, File No. PCAOB 2007–02 (July 27, 2007) [72 FR 42141 (Aug. 1, 2007)]. 121 Id. 122See Commission Guidance Regarding Management’s Report on Internal Control Over Financial Reporting Under Section 13(a) or 15(d) of the Securities Exchange Act of 1934, Release No. 33–8810 (June 20, 2007) [72 FR 35323 (June 27, 2007)]. See also Amendments to Rules Regarding Management’s Report on Internal Control Over Financial Reporting, Release No. 33–8810 (June 20,
2007) [72 FR 35309 (June 27, 2007)].
123See, e.g., Study of the Sarbanes-Oxley Act of 2002 Section 404 Internal Control over Financial Reporting Requirements (Sept. 2009) (‘‘2009 SEC Staff Study’’), available at https://www.sec.gov/ news/studies/2009/sox-404_study.pdf; Rajib Doogar, Padmakumar Sivadasan, & Ira Solomon, 48(4) J. of Acct. Res. 795 (2010). 124See, e.g., Joseph Schroeder & Marcy Shepardson, Do SOX 404 Control Audits and Management Assessments Improve Overall Internal Control System Quality?, 91(5) Acct. Rev. 1513 (‘‘Schroeder and Shepardson 2016 Study’’); Lori Bhaskar, Joseph Schroeder, & Marcy Shepardson, Integration of Internal Control and Financial Statement Audits: Are Two Audits Better than One? Acct. Rev. (forthcoming 2018) (‘‘Bhaskar et al. 2018 Study’’), available at http://aaajournals.org/doi/abs/ 10.2308/accr-52197. 125See Jeanette Franzel, Board Member, PCAOB, Speech by PCAOB board member at the American Accounting Association Annual Meeting, Current Issues, Trends, and Open Questions in Audits of Internal Control over Financial Reporting (2015), available at https://pcaobus.org/News/Speech/ Pages/08102015_Franzel.aspx. 126See Mark Defond & Clive Lennox, Do PCAOB Inspections Improve the Quality of Internal Control Audits?, 55(3) J. of Acct. Res. 591 (2017) (‘‘Defond and Lennox 2017 Study’’). 127See, e.g., Tammy Whitehouse, Audit Inspections: Improvement? Maybe. Costs? Yes, Compliance Week (April 14, 2015), available at https://www.complianceweek.com/news/newsarticle/audit-inspections-improvement-maybecosts-yes#.W5LW7mlpCEd. 128See Auditing Standards Related to the Auditor’s Assessment of and Response to Risk and Related Amendments to PCAOB Standards, PCAOB Release No. 2010–004 (Aug. 5, 2010) (‘‘PCAOB Release No. 2010–004’’). See also Public Company Accounting Oversight Board; Order Approving Proposed Rules on Auditing Standards Related to the Auditor’s Assessment of and Response to Risk and Related Amendments to PCAOB Standards, Release No. 34–63606, File No. PCAOB 2010–01 (Dec. 23, 2010) [75 FR 82417 (Dec. 30, 2010)]. 129See AS 2110, note 48 above, paragraphs .18– .40. each filer type are provided in Table 7 below.
TABLE 7—FILER STATUS FOR ISSUERS FILING ANNUAL REPORTS IN 2017 116
Nonaccelerated 117 Accelerated Large accelerated Total ........................................................................................................................... 3,899 1,497 2,138 Foreign ................................................................................................................ 240 146 255 EGC .................................................................................................................... 1,201 375 0 Audits of ICFR and the associated ICFR auditor attestation reports are made in accordance with AS 2201,118 previously known as Auditing Standard Number 5 (‘‘AS No. 5’’).119 This standard, which replaced Auditing Standard Number 2 (‘‘AS No. 2’’) in 2007, was intended to focus auditors on the most important matters in the audit of ICFR and eliminate procedures that the PCAOB believed were unnecessary to an effective audit of ICFR.120 Among other things, the 2007 standard facilitates the scaling of the evaluation of ICFR for smaller, less complex issuers.121 It was accompanied by Commission guidance similarly facilitating the scaling of SOX Section 404(a) management evaluations of ICFR.122 Relative to AS No. 2, AS 2201 facilitates the scaling of audits of ICFR by, for example, encouraging auditors to use top-down risk-based approaches and to rely on the work of others in the attestation process. The adoption of AS 2201 in 2007 has been found to have lowered audit fees.123 However, several studies have provided evidence that, at least initially, audits of ICFR under the revised standard may not have been as effective in improving the quality of ICFR as those under AS No. 2.124 PCAOB inspections of auditors began, around 2010, to include a heightened focus on whether auditing firms had obtained sufficient evidence to support their opinions on the effectiveness of ICFR.125 There is some evidence that these inspections have led to an improvement in the reliability of ICFR auditor attestations,126 but also concerns about whether they have resulted in increased audit fees.127 In 2010, the PCAOB adopted enhanced auditing standards related to the auditor’s assessment of and response to risk.128 The enhanced risk assessment standards have likely reduced the degree of difference between a financial statement only audit and an integrated audit (which includes an audit of ICFR) because the standards clarify and augment the extent to which internal controls are to be considered even in a financial statement only audit. In particular, the risk assessment standards applying to both types of audits require auditors, in either case, to evaluate the design of certain controls, including whether the controls are implemented.129

130See Inspection Observations Related to PCAOB ‘‘Risk Assessment’’ Auditing Standards (No. 8 through No.15), PCAOB Release No. 2015– 007 i–iii (Oct. 15, 2015). 131See PCAOB Release No. 2010–004, note 128 above, at 7 and A10–41. As discussed above, even in a financial statement only audit, if the auditor becomes aware of a material weakness in ICFR, it is required to inform management and the audit committee of this finding and has the responsibility to review management’s disclosure for any misstatement of facts, such as a statement that ICFR is effective when there is a known material weakness. See notes 77 and 78 above. 132See Proposed Auditing Standards Related to the Auditor’s Assessment of and Response to Risk and Conforming Amendments to PCAOB Standards, PCAOB Release No. 2008–006 A9–8 (Oct. 21, 2008). 133 Information on these and other FASB Accounting Standards updates is available at https://www.fasb.org/jsp/FASB/Page/SectionPage& cid=1176156316498. 134See, e.g., Kevin Moffitt, Andrea Rozario, & Miklos Vasarhelyi (2018), Robotic Process Automation for Auditing, Journal of Emerging Technologies, 15(1) Acct. 1 (describing how, for example, a robotic process automation program can be ‘‘set up to automatically match purchase orders, invoices, and shipping documents [and] can check that the price and quantity on each of the documents match [to] help auditors validate the effectiveness of preventive internal controls . . . .’’). 135See, e.g., Protiviti survey results, Benchmarking SOX Costs, Hours and Controls (2018) (‘‘Protiviti 2018 Report’’). 136See, e.g., 2011 SEC Staff Study, note 49 above, at 106 (stating that ‘‘. . . once effective controls are in place at the issuer, the auditor is more likely to continue to test them even if [it is] not issuing an auditor attestation during a particular year in order to rely on them for purposes of reducing substantive testing in the audit of the financial statements, particularly for issuers that are larger and more complex’’). 137Because of the accelerated filer transition provisions, some accelerated filers have float below $75 million. The public float of these issuers would previously have exceeded $75 million, causing them to enter accelerated filer status, but has not dropped below the $50 million public float level required to exit accelerated filer status. Based on the results of inspections in the several years after the adoption of the new risk assessment auditing standards, the PCAOB expressed concern about the number and significance of deficiencies in auditing firm compliance with these standards, but also noted promising improvements in the application of these standards.130 While the risk assessment standards may reduce the degree of difference between a financial statement only audit and an integrated audit, there remain important differences in the requirements of these audits as they relate to controls. For example, in an integrated audit, but not a financial statement only audit,
the auditor is required to identify likely sources of misstatements.131 Also, the extent of the procedures necessary to obtain the required understanding of controls generally will be greater in an integrated audit due to the different objectives of such an audit as compared to a financial statement only audit.132 We also note that there have been some recent changes in accounting and auditing that are part of our baseline and could increase the uncertainty of our analysis due to their effects on factors such as audit fees, restatements, and ICFR. For example, three new reporting standards have been issued recently by FASB, on the topics of revenue recognition, leases, and credit losses, which could temporarily increase audit fees as issuers and auditors adjust to the new standards.133 Recent changes in audit technology, such as the potential for automated controls testing and process automation,134 may result in improvements in ICFR regardless of the ICFR auditor attestation requirement. Such automation could also reduce audit fees, including the costs of an audit of ICFR, but the uptake of these technologies has been slow.135 Finally, auditors have had many years of experience with integrated audits, as well as risk assessment standards that require the consideration of ICFR even in the absence of an ICFR auditor attestation. This experience may affect their execution of financial statement only audits of issuers for whom the ICFR auditor attestation requirement is eliminated. For example, given their experience, auditors may be more likely to detect control deficiencies or to increase their auditing efficiency by reducing substantive testing in favor of testing some related controls even when an ICFR auditor attestation is not required.136
2. Characteristics of Accelerated Filer
Population
Per Table 7, there were approximately 1,500 accelerated filers in total in 2017.
Figure 2 presents the distribution of
public float across these issuers.137 BILLING CODE 8011–01–P

138The estimates in the figure are based on staff analysis of data from XBRL filings. See note 116 above for details on the identification of the population of accelerated filers. The distribution of public float among accelerated filers is skewed towards lower levels of float, but higher levels of float are also significantly represented.
Figure 3 presents the distribution of
revenues across those accelerated filers that have less than $1 billion in revenues. While the full population of accelerated filers has revenues of up to over $8 billion, about 90% of accelerated filers have less than $1 billion in revenues. We restrict the
figure to this subset in order to more
clearly display the distribution in this range. EP29MY19.001</GPH>

139The estimates of revenues are based on staff analysis of data from XBRL filings, Compustat, and Calcbench. The revenue data used is from the last fiscal year prior to the annual report in calendar year 2017, because the SRC revenue test is based on the prior year’s revenues. See note 116 above for details on the identification of the population of accelerated filers. 140These estimates are based on staff analysis of data including SIC codes from XBRL filings and Ives Group Audit Analytics, using the Fama-French 49-industry classification system. See http:// mba.tuck.dartmouth.edu/pages/faculty/ken.french/ Data_Library/det_49_ind_port.html. See note 116 above for details on identification of population of accelerated filers. BILLING CODE 8011–01–C The distribution of revenues for accelerated filers is heavily skewed towards lower levels of revenue, with roughly three-quarters of accelerated filers having revenues of less than $500 million and more than a third having revenues of less than $100 million. Other than a clustering of issuers with zero or near zero revenues, there are no obvious breaks in the distribution. While a large range of industries are represented among accelerated filers, a small number of industries account for the majority of these issuers. The ‘‘Banking’’ industry accounts for about 14.2% of accelerated filers, followed by ‘‘Pharmaceutical Products’’ (12.8%), ‘‘Financial Trading’’ (7.7%), ‘‘Business Services’’ (6.7%), ‘‘Computer Software’’ (4.5%), ‘‘Electronic Equipment’’ (4.3%) and ‘‘Petroleum and Natural Gas’’ (4.0%).140
3. Timing of Filings
As discussed above, non-accelerated, accelerated, and large accelerated filers face different filing deadlines for their periodic reports. In Table 8, we present the timing in recent years of annual report filings by these different groups of issuers relative to their corresponding deadlines.
TABLE 8—FILING TIMING FOR ANNUAL REPORTS IN YEARS 2014 THROUGH 2017, BY FILER STATUS 141
Non-accelerated Accelerated Large accelerated Annual report filing deadline .................................................................... 90 days ...................... 75 days ...................... 60 days. Average days to file ................................................................................. 101 days .................... 70 days ...................... 56 days. Percentage filed:
By deadline ....................................................................................... 73% ........................... 91% ........................... 95%. Over 5 days early ............................................................................. 45% ........................... 64% ........................... 63%. After deadline ................................................................................... 27% ........................... 9% ............................. 5%. Over 15 days after deadline ............................................................. 11% ........................... 4% ............................. 3%. EP29MY19.002</GPH>

141The estimates in this table are based on staff analysis of EDGAR filings. These statistics include all annual reports on Forms 10–K, 20–F, and 40– F filed in calendar years 2014 through 2017 other than amendments. Given the effect of weekends and holidays, filings are considered to be on time if within two calendar days after the original deadline. The ‘‘5 days early’’ and ‘‘over 15 days after’’ categories are similarly adjusted to account for the possible effect of weekends and holidays. See note 116 above for details on the identification of filer type. 142Unless otherwise specified, statistics and analysis regarding restatements are not restricted to those restatements requiring Form 8–K Item 4.02 disclosure. 143The estimates in this table are based on staff analysis of Ives Group Audit Analytics data. ICFR effectiveness is based on the last amended management or auditor attestation report for the fiscal year. Percentages are computed out of all issuers of a given filer type with the specified type of report available in the Ives Group Audit Analytics database. See note 116 above for details on the identification of filer type. 144Per the second column of the first panel of
Table 9, the rate of ineffective ICFR among
accelerated filers has ranged from 7.8 to 9.4% for the years 2014 through 2017, for an average per year of 8.8%. 145Per the third column of the first panel of Table 9, the rate of ineffective ICFR among large accelerated filers has ranged from 3.1 to 4.9% for the years 2014 through 2017, for an average per year of 4.1%. 146Per the first column of the first panel of Table 9, the rate of ineffective ICFR among nonaccelerated filers has ranged from 38.4 to 41.2% for the years 2014 through 2017, for an average per year of 40.1%.
Table 8 documents that accelerated
and large accelerated filers file their annual reports, on average, four or five days before the applicable deadline. Nine percent and five percent, respectively, of accelerated and large accelerated filers submit their annual reports after the initial deadline, with roughly half of these filers surpassing the 15-day grace period that is obtained by filing Form NT. Non-accelerated filers are less likely to meet their initial deadline or extended deadline, with the average non-accelerated filer submitting its annual report 11 days after the initial deadline and 11% of non-accelerated filers filing after the 15-day grace period obtained by filing Form NT.
4. Internal Controls and Restatements
We next consider the current rates of ineffective ICFR and restatements 142 among issuers that are accelerated filers under the baseline relative to other filer types. Throughout our analysis, we use the term restatement to refer to a restatement that is associated with some type of misstatement. As discussed above, non-accelerated filers and EGCs are statutorily exempted from the ICFR auditor attestation requirement. Table 9 presents the percentage of issuers reporting ineffective ICFR in recent years by filer type.
TABLE 9—PERCENTAGE OF ISSUERS REPORTING INEFFECTIVE ICFR 143
Ineffective ICFR year reported in: Non-accelerated (percent) Accelerated (percent) Large accelerated (percent) Management Report 2014 .................................................................................................................... 40.3 7.8 3.1 2015 .................................................................................................................... 41.2 8.8 3.7 2016 .................................................................................................................... 38.4 9.3 4.5 2017 .................................................................................................................... 40.3 9.4 4.9 Average/year ...................................................................................................... 40.1 8.8 4.1 Auditor Attestation 2014 .................................................................................................................... n/a 8.0 3.3 2015 .................................................................................................................... n/a 8.8 3.7 2016 .................................................................................................................... n/a 8.9 4.5 2017 .................................................................................................................... n/a 9.6 4.8 Average/year ...................................................................................................... n/a 8.8 4.1 Based on management’s SOX Section 404(a) reports on ICFR from recent years, on average, about eight or nine percent of accelerated filers reported at least one material weakness in ICFR in a given year.144 This represents a moderately higher rate than that among large accelerated filers, approximately four percent, on average, of which reported ineffective ICFR,145 and a substantially lower rate than that among non-accelerated filers, more than a third of which reported ineffective ICFR each year.146 For issuers subject to the ICFR auditor attestation requirement, the rates of ineffective ICFR reported by management and by auditors are similar. This may not be surprising, as management will be made aware of any material weaknesses discovered by the auditor and vice versa. We next consider the persistence of material weaknesses across these issuer categories. Table 10 presents the percentage of issuers that reported two, three, or four consecutive years of ineffective ICFR culminating in 2017, by filer type.
TABLE 10—PERCENTAGE OF ISSUERS REPORTING CONSECUTIVE YEARS OF INEFFECTIVE ICFR IN MANAGEMENT REPORT,
BY 2017 FILER STATUS 147
Ineffective ICFR years: Non-accelerated Accelerated Large accelerated As % of issuers 2016–2017 (at least 2 years) ............................................................................. 27.5 4.3 1.6 2015–2017 (at least 3 years) ............................................................................. 20.6 2.2 0.4 2014–2017 (4 years) .......................................................................................... 15.4 1.3 0.2 As % of issuers with 2017 ineffective ICFR 2016–2017 (at least 2 years) ............................................................................. 68.6 48.9 39.0 2015–2017 (at least 3 years) ............................................................................. 51.4 25.0 9.8

147The estimates in this table are based on staff analysis of Ives Group Audit Analytics data. ICFR effectiveness is based on the last amended management report for the fiscal year. Percentages in the first panel are computed out of all issuers of a given filer type in 2017 with SOX Section 404(a) management reports available in Ives Group Audit Analytics database, while percentages in the second panel are computed out of issuers of a given filer type reporting ineffective ICFR in their SOX Section 404(a) management report for 2017 (see the fourth row of Table 9). See note 116 above for details on the identification of filer type. 148The estimates in this table are based on staff analysis of Ives Group Audit Analytics data. Percentages are computed out of all issuers of a given filer type with a SOX Section 404(a) management report available in the Ives Group Audit Analytics database. Accelerated and nonaccelerated categories exclude EGCs that are in these filer categories. See note 116 above for details on the identification of filer type. 149See Table 14 below. 150 Id.
TABLE 10—PERCENTAGE OF ISSUERS REPORTING CONSECUTIVE YEARS OF INEFFECTIVE ICFR IN MANAGEMENT REPORT,
BY 2017 FILER STATUS 147—Continued Ineffective ICFR years: Non-accelerated Accelerated Large accelerated 2014–2017 (4 years) .......................................................................................... 38.4 14.8 4.9 Compared to non-accelerated filers, we find that a smaller percentage of accelerated and large accelerated filers report material weaknesses that persist for multiple years, with about one percent of accelerated filers and about 0.2% of large accelerated filers reporting ineffective ICFR for four consecutive years, representing about 15% of the accelerated filers and about five percent of the large accelerated filers that reported ineffective ICFR in 2017. A larger percentage of non-accelerated filers persistently report material weaknesses, with about 15% of these issuers, or more than one-third of those reporting ineffective ICFR in 2017, having reported material weaknesses for four consecutive years.
Table 11 presents the rate of
restatements among each of these filer types, excluding EGCs, and for EGCs separately. For each year, we consider the percentage of issuers that eventually restated the financial statements for that year. The reporting lag before restatements are filed results in a lower observed rate in the later years of our sample, particularly for 2016 (and even more so for 2017, which we do not report for this reason), as issuers may yet restate their results from recent years.
TABLE 11—PERCENTAGE OF ISSUERS ISSUING RESTATEMENTS BY YEAR OF RESTATED DATA 148
Restated
Non-accelerated
(ex. EGCs)
(percent)
Accelerated
(ex. EGCs)
(percent)
Large accelerated
(percent)
EGC
(percent)
Total Restatements:
2014 .................................................................................. 10.9 11.4 13.8 17.0 2015 .................................................................................. 8.9 11.1 11.8 15.5 2016 .................................................................................. 5.9 7.2 6.6 8.0 Average/year .................................................................... 8.5 9.9 10.8 13.5 8–K Item 4.02 Restatements:
2014 .................................................................................. 3.3 2.9 2.1 4.9 2015 .................................................................................. 2.6 3.1 1.4 4.7 2016 .................................................................................. 1.7 2.1 1.0 2.5 Average/year .................................................................... 2.5 2.7 1.5 4.0 The first panel of Table 11 presents the percentage of issuers that make at least one restatement, of any type, while the second panel presents those that make at least one restatement requiring Form 8–K Item 4.02 disclosure. The latter type of restatement (‘‘Item 4.02 restatements’’) reflects material misstatements, while other restatements deal with misstatements or adjustments that are considered immaterial. We find that EGCs, which are not subject to the ICFR auditor attestation requirement and generally are also younger issuers than those in the other groups, restate their financial statements at higher rates than other issuers, whether we consider all restatements or only Item 4.02 restatements. For non-accelerated filers, which also are not subject to the ICFR auditor attestation requirement, we find that the percentage of issuers reporting restatements or Item 4.02 restatements is similar to that for accelerated filers who are subject to the ICFR auditor attestation requirement. We note that there is a greater proportion of lowrevenue issuers, which we find below to have lower rates of restatement than other issuers,149 in the non-accelerated filer category than in other categories. Below, when we separately consider issuers with revenues below $100 million, we find that the nonaccelerated filers in this category are more likely to restate their financial statements than accelerated filers in the same revenue category.150
C. Discussion of Economic Effects
The costs and benefits of the proposed amendments, including impacts on efficiency, competition, and capital formation, are discussed below. We first address the population and characteristics of issuers that would newly qualify as non-accelerated filers under the proposed amendments, and then introduce certain categories of issuers that are used for comparison purposes. We next discuss the anticipated costs and benefits associated with the proposed change in applicability of the ICFR auditor attestation requirement. Following this discussion, we consider the costs and benefits associated with the proposed changes with respect to filing deadlines, exit thresholds, and other required disclosures. Finally, we consider the relative benefits and costs of the principal reasonable alternatives to the proposed amendments.

Affected Issuers
We estimate that the proposed amendments would result in 539 additional issuers being classified as non-accelerated filers, and therefore no longer subject to the filing deadlines and ICFR auditor attestation requirement applicable to accelerated

151The number of affected issuers is based on staff estimates of: (i) The number of accelerated filers in 2017 that have prior fiscal year revenues of less than $100 million and are eligible to be SRCs (i.e., excluding ABS issuers, RICs, BDCs, and subsidiaries of non-SRCs); (ii) the number of large accelerated filers in 2017 that have a public float of less than $560 million and prior fiscal year revenues of less than $100 million and are eligible to be SRCs; and (iii) the number of accelerated filers in 2017 that have a public float of at least $50 million but less than $60 million. The estimate of the number of affected issuers does not include large accelerated filers that have a public float of at least $560 million but less than $700 million even though such issuers could become non-accelerated filers under the proposed amendments if they became eligible to be SRCs under the SRC revenue test in the first year the SRC amendments became effective due to the limited horizon of this accommodation. See note 98 above (describing the accommodation provided in the SRC Adopting Release). Revenue data is sourced from XBRL filings, Compustat, and Calcbench. See note 116 above for details on the identification of the population of accelerated and large accelerated filers. 152 Id. 153 Id. 154 Id. 155Banks are identified as issuers with SIC codes of 6020 (commercial banks), 6021 (national commercial banks), 6022 (state commercial banks), 6029 (NEC commercial banks), 6035 (savings institutions, fed-chartered) or 6036 (savings institutions, not fed-chartered). 156 If these banks are no longer subject to the SOX
Section 404(b) auditor attestation requirement, their
auditors may follow the AICPA’s auditing standards in lieu of the PCAOB’s auditing standards for the FDIC auditor attestation. See Section 18A of
Appendix A to FDIC Rule 363 and the AICPA’s
AU–C Section 940.
157This estimate is based on staff analysis of XBRL filings using a computer program supplemented by hand collection and data from Ives Group Audit Analytics. The majority of these potential additional issuers are Canadian MJDS filers that are not required to disclose filer type or public float, though there are also domestic issuers and other foreign issuers for which some of the required data is not available. See note 116 above. 158Staff extracted information regarding whether issuers reported having securities registered under
Section 12(b) of the Exchange Act from the cover
page of annual report filings using a computer
program supplemented with hand collection. See note 151 above for details on the identification of the population of affected issuers. 159This estimate is based on staff analysis of data from the Center for Research in Security Prices database for December 1998 versus December 2018. The estimate excludes RICs and issuers of ADRs. 160 Id. 161This estimate is based on staff analysis of data from Standard & Poor’s Compustat and Center for Research in Security Prices databases for fiscal year 1998 versus fiscal year 2017. The estimate excludes RICs and issuers of ADRs. 162Because of the accelerated filer transition provisions, some of the affected issuers have public float of at least $50 million but below $75 million. See note 137 above. filers.151 Of these issuers, an estimated 525 issuers are accelerated filers (or large accelerated filers that have public float of less than $560 million) that would be newly classified as nonaccelerated filers because they have annual revenues of less than $100 million and are eligible to be SRCs.152 An additional 14 issuers are accelerated filers that would be newly classified as non-accelerated filers despite having revenues of at least $100 million because they have a public float of at least $50 million but less than $60 million.153 The total number of affected issuers includes an estimated 36 foreign private issuers and 181 EGCs.154 It also includes an estimated 76 banks with $1 billion or more in total assets that are not EGCs.155 Because the estimated 181 EGCs are not required to comply with the ICFR auditor attestation requirement under SOX Section 404(b), we estimate that the remaining 358 affected issuers would be newly exempt from this requirement. Of these 358 issuers, we expect that the 76 banks identified above would be subject to the FDIC auditor attestation requirement,156 while the remaining 282 issuers would not be subject to any such auditor attestation requirement. Our estimate of the number of affected issuers excludes issuers for which we were unable to determine filer classification or revenues, which could represent up to approximately an additional 100 affected issuers.157 We estimate that approximately 90% of the affected issuers (whether including or excluding EGCs) have securities that are listed on national exchanges.158 The affected issuers represent a type of issuer whose representation in public markets has decreased relative to the years before SOX. Over the past two decades, the number of issuers listed on major exchanges has decreased by about 40%,159 but the decline has been concentrated among smaller size issuers. Specifically, the number of listed issuers with market capitalization below $700 million has decreased by about 65%,160 and the number of listed issuers with less than $100 million in revenue has decreased by about 60%.161
Figure 4 presents the distribution of
public float across the full sample of affected issuers.162 BILLING CODE 8011–01–P

163The estimates in this figure are based on staff analysis of data from XBRL filings. See note 151 above for details on the identification of the population of affected issuers. Relativeto the distribution for all accelerated filers presented in Figure 2, the sample of affected issuers is more strongly skewed toward lower levels of public float, with higher levels of public float only thinly represented. However, some of the affected issuers do have public float approaching the top of the range for accelerated filers.
Figure 5 presents the distribution of
revenues across the 525 accelerated filers (or large accelerated filers with public float of less than $560 million) that would be newly classified as nonaccelerated filers because they have revenues of less than $100 million. EP29MY19.003</GPH>

164The estimates in this figure are based on staff analysis of data from XBRL filings, Compustat, and Calcbench. The revenue data used is from the last fiscal year prior to the annual report in calendar year 2017, because the SRC revenue test is based on the prior year’s revenues. See note 151 above for details on the identification of the population of affected issuers. 165Approximately 13% of the estimated 525 affected issuers with revenues of less than $100 million and approximately 11% of the estimated 347 non-EGC affected issuers (which would be newly exempt from the SOX Section 404(b) ICFR auditor attestation requirement) with revenues of less than $100 million have zero revenues. 166These estimates are based on staff analysis of data from Compustat. See note 151 above for details on the identification of the population of affected issuers. 167 Id. For the 282 affected issuers that would be newly exempt from all ICFR auditor attestation requirements (i.e., those that are not EGCs and are not banks subject to the FDIC auditor attestation requirement), the median total assets and median number of employees are somewhat lower at about $110 million and 110 employees, and the median issuer age is similar at about 19 years. 168For the 282 affected issuers that would be newly exempt from all ICFR auditor attestation requirements (i.e., those that are not EGCs and are not banks subject to the FDIC auditor attestation requirement), the proportion of ‘‘Banking’’ issuers drops to 5.7%. By contrast, the proportion in other industries does not change by more than a few percentage points. 169These estimates are based on staff analysis of data including SIC codes from XBRL filings and Ives Group Audit Analytics, using the Fama-French 49-industry classification system. See http:// mba.tuck.dartmouth.edu/pages/faculty/ken.french/ Data_Library/det_49_ind_port.html. See note 151 above for details on the identification of the population of affected issuers. BILLING CODE 8011–01–C Other than a concentration of issuers with zero or near zero revenues,165 these affected issuers are fairly evenly distributed over different levels of revenue up to $100 million in revenues. The additional 14 affected issuers with revenues of at least $100 million but a public float of less than $60 million have revenues ranging from $120 million to $1.2 billion, with a mean of about $500 million in revenues. The affected issuers are estimated to have median total assets of about $175 million, a median number of employees of about 125, and a median age of about 11 years.166 For those issuers that would be newly exempt from the SOX Section 404(b) ICFR auditor attestation requirement, the median total assets, median number of employees and median issuer age are estimated to be slightly higher at about $190 million, 160 employees and about 18 years.167 The affected issuers are heavily concentrated in the ‘‘Pharmaceutical Products’’ (30.2%), ‘‘Banking’’ (20.2%),168 and ‘‘Financial
Trading’’ (10.2%) industries, followed by ‘‘Medical Equipment’’ (5.2%), ‘‘Business Services’’ (4.3%), ‘‘Electronic Equipment’’ (3.9%) and ‘‘Petroleum and Natural Gas’’ (3.0%).169 If the distribution of eligible issuers does not change over time, the proposed amendments could lead to a noticeable decrease in the presence of ‘‘Pharmaceutical Products’’ and ‘‘Banking’’ issuers in the pool of accelerated filers.
2. Comparison Populations
The proposed amendments would extend the exemption from the ICFR auditor attestation requirement to certain issuers that would be classified as accelerated filers under current rules and that have revenues of less than $100 million. To analyze the effects of this EP29MY19.004</GPH>

170That is, the accelerated filers in this analysis exclude EGCs as well as ABS issuers and RICs. 171The issuers in these analyses exclude those that do not provide a SOX Section 404(a) management report on ICFR (i.e., ABS issuers, RICs, and certain newly public issuers prior to filing their second annual report). 172The analyses also include, per year, 725 to 851 higher-revenue accelerated filers (other than EGCs), 384 to 424 higher-revenue non-accelerated filers (other than EGCs), and 37 to 223 higher-revenue EGCs. The sample size varies across years and is based on issuers of a given filer type with revenue data and a SOX Section 404(a) management report available in the Ives Group Audit Analytics database. See note 116 above for details on the identification of filer type. 173For those issuers that would be newly exempt from the ICFR auditor attestation requirement, the median total assets and median number of employees are estimated to be about $190 million and about 160 employees. See Section III.C.1 above. 174Age is measured based on the number of years of data available in the Compustat database, as is common in the academic literature, and likely exceeds the number of years after the issuer’s initial public offering. 175We considered limiting our analysis to more narrow subsamples of these groups of issuers. For example, EGCs that have less than $100 million in revenues and are also accelerated filers would likely be more comparable to the affected issuers. However, we have identified only 19 such issuers in 2014, growing to 166 in 2017, which is not a sufficient number to allow us to statistically differentiate between, for example, the rates of restatements across different types of issuers. Therefore, in order to preserve a sample size sufficient for robust inference, we do not apply further filters to the issuers in these analyses beyond requiring that the necessary data be available. 176See, e.g., David Hay, W. Robert Knechel, & Norman Wong, Audit Fees: A Meta-analysis of the Effect of Supply and Demand Attributes, 23(1) Contemporary Acct. Res. 141 (2006) (reviewing a large body of research on audit fees and determining that studies consistently find a positive relation between various measures of client size and audit fees, where the most common measure used was total assets, and that this relation accounts for a large proportion of the variation in audit fees); Charles Cullinan, Hui Du, and Xiaochuan Zheng, Size Variables in Audit Fee Models: An Examination of the Effects of Alternative Mathematical Transformations, 35(3) Auditing: A J. of Prac. and Theory 169 (2016). 177See, e.g., Jeffrey Doyle, Weili Ge & Sarah McVay, Determinants of Weaknesses in Internal Control Over Financial Reporting, 44(1⁄2) J. of Acct. and Econ. 193 (2007) (finding a negative association of material weaknesses in ICFR with size, based on market capitalization, and with age, based on the number of years in the CRSP database) and Hollis Ashbaugh-Skaife,
Daniel Collins, & William Kinney, The Discovery and Reporting of Internal Control Deficiencies Prior to SOX-Mandated Audits, 44(1⁄2) J. of Acct. and Econ. 166 (2007) (finding a negative association of material weaknesses in ICFR with size, based on market capitalization, but not finding a similar association with age, based on the number of years in the CRSP database, after controlling for other factors). For more recent evidence, see Weili Ge, Allison Koester, & Sarah McVay, Benefits and Costs of Sarbanes-Oxley Section 404(b) Exemption:
Evidence from Small Firms’ Internal Control Disclosures, 63 J. of Acct. and Econ. 358 (2017) (‘‘Ge et al. 2017 Study’’) (applying a model of the determinants of material weaknesses in ICFR based on these previous studies to data from 2007 through 2014, and finding a negative association of material weaknesses in ICFR with size, based on market capitalization, and with age, based on the number of years in the Compustat database). 178See, e.g., Susan Scholz, Financial Restatement Trends in the United States: 2003–2012, Ctr. for Audit Quality White Paper (2014), available at https://www.thecaq.org/financial-restatementtrends-united-states-2003-2012 (where size is measured based on total assets). 179See, e.g., Gopal Krishnan, Emma-Riikka Myllyma¨k, & Neerav Nagar, Does Financial Reporting Quality Vary Across Firm Life Cycle?, Working Paper (finding a higher rate of restatements for issuers in the ‘‘introduction’’ stage of their life cycle relative to the ‘‘mature’’ stage, where life cycle stages are identified based on cash flow patterns), available at https://papers.ssrn.com/sol3/ papers.cfm?abstract_id=3233512. 180For the EGCs in our sample, based on data from Ives Group Audit Analytics, we estimate that those in their first two years after their initial disclosure of EGC status in 2014 through 2016 have approximately a 15% rate of restatements of their financial statements from these years, while those in their third and fourth years after initial disclosure of EGC status have approximately an 11% rate of restatements in these years. change, we would ideally compare, for the issuers that would be newly exempted, the effectiveness of their ICFR, their audit fees, and other key outcomes when they are subject to the ICFR auditor attestation requirement with the outcomes when they are not subject to this requirement. However, because the category of issuers that would be newly exempted is currently subject to the ICFR auditor attestation requirement, we are unable to assess their likely experience in the absence of this requirement by analyzing these issuers in isolation. Therefore, in addition to examining low-revenue accelerated filers that are subject to the ICFR auditor attestation requirement,170 we also consider the experience of other low-revenue issuers that are not subject to this requirement: Non-accelerated filers (other than EGCs) and EGCs.171 Our analyses of data from 2014 through 2017 include, per year, 367 to 423 low-revenue accelerated filers (other than EGCs), 995 to 1,170 lowrevenue non-accelerated filers (other than EGCs), and 136 to 647 low-revenue EGCs.172 Non-accelerated filers (other than EGCs) and EGCs with revenues below $100 million have similar revenues and similar responsibilities regarding their internal controls (including being subject to the SOX
Section 404(a) management ICFR
reporting requirements) as the affected issuers, but are not subject to the ICFR auditor attestation requirement. Importantly, however, the issuers in these two comparison groups are not fully comparable to the affected issuers. While the affected issuers all have a public float of at least $50 million, and an estimated median of about $145 million in public float, non-accelerated filers and the majority of the EGCs in our sample have public float of less than $75 million. The median total assets are estimated to be about $20 million for low-revenue non-accelerated filers (other than EGCs) and $50 million for low-revenue EGCs, and the median number of employees is estimated to be about 60 for low-revenue nonaccelerated filers (other than EGCs) and about 50 for low-revenue EGCs. These estimates represent roughly one-fourth of the median total assets and one-third of the median number of employees reported above for the affected issuers that would be newly exempt from the ICFR auditor attestation requirement.173 In addition, while the affected issuers have generally been reporting companies for more than five years, and those that would be newly exempt from the ICFR auditor attestation requirement have a median age of 18 years,174 EGC status generally is limited to issuers in the first five years after their initial public offering. The issuers in both comparison groups will thus tend to be smaller, and the EGCs younger, than the affected issuers, which may reduce the reliability of estimates of the potential effects on audit fees, the effectiveness of ICFR, and restatement rates that are derived in part based on comparisons to these issuers.175 We note that smaller issuers generally incur lower audit fees.176 Also, research has associated having a lower market capitalization with having a greater likelihood of material weaknesses in ICFR, with some studies finding a similar association for issuers with less experience as a publicly-traded company.177 Studies have similarly found that smaller issuers are often associated with a higher rate of restatements.178 One study,179 as well as our own analysis,180 suggests that issuers that are very early in their lifecycle, as are EGCs, may also have a higher rate of restatements. These associations may result in a greater disparity between the audit fees, rates of ineffective ICFR, and rates of restatement between the category of affected issuers and the two comparison samples than would be expected if these samples were more comparable in terms of their size and age. We believe that the experience of the issuers in these comparison groups is still informative to our analysis but note that they may be more likely to provide an upper bound rather than a direct reflection of the likely outcomes for the affected issuers as a result of the proposed amendments.
3. Potential Benefits of Eliminating the
ICFR Auditor Attestation Requirement for Affected Issuers The ICFR auditor attestation requirement has been associated with

181See John Coates & Suraj Srinivasan, SOX after Ten Years: A Multidisciplinary Review, 28(3) Acct. Horizons 627 at 643–645 (2014) (‘‘Coates and Srinivasan 2014 Study’’) (discussing these possible effects and summarizing related studies). 182 Id. 183 Id. 184See Ana Albuquerque & Julie Zhu (2018), Has
Section 404 of the Sarbanes-Oxley Act Discouraged
Corporate Risk-Taking? New Evidence from a Natural Experiment, Mgmt. Sci. (forthcoming) (using the staggered implementation of SOX Section 404 to better identify its effects on smaller reporting issuers, with public float of less than $150 million, and finding no evidence of a decrease in the investment and risk-taking activities for issuers that were subject to SOX Section 404 versus those that were not), available at https://papers.ssrn.com/sol3/ papers.cfm?abstract_id=3049232. 185See Huasheng Gao & Jin Zhang, SOX Section 404 and Corporate Innovation,’’ J. of Fin. and Quantitative Analysis (2018) (forthcoming), available at https://papers.ssrn.com/sol3/ papers.cfm?abstract_id=3130588. 186See, e.g., Peter Iliev, The Effect of SOX Section 404: Costs, Earnings Quality, and Stock Prices, 45 J. of Fin. 1163 (2010) (‘‘Iliev 2010 Study’’) (finding that a disproportionate number of issuers had a public float of just under $75 million in 2004, when auditor attestations of ICFR and management ICFR reports were first required for accelerated filers, but not in earlier years). 187See F. Gao, J.S. Wu,, & J. Zimmerman, Unintended Consequences of Granting Small Firms Exemptions from Securities Regulation: Evidence from the Sarbanes-Oxley Act, 47(2) J. of Acct. Res. 459 (2009) and M. E. Nondorf, Z. Singer, & H. You, A Study of Firms Surrounding the Threshold of Sarbanes–Oxley Section 404 Compliance, 28(1) Advances in Acct. 96 (2012). See also F. Gao, To Comply or Not to Comply: Understanding the Discretion in Reporting Public Float and SEC Regulations, 33(3) Contemporary Acct. Res. 1075 (2016) (presenting evidence that companies that expected higher compliance costs may have used discretion in defining affiliates in order to report lower float). 188See Dhammika Dharmapala, Estimating the Compliance Costs of Securities Regulation: A Bunching Analysis of Sarbanes-Oxley Section 404(b), Working Paper (2016), available at http:// papers.ssrn.com/sol3/papers.cfm?abstract_ id=2885849. 189 Id. This paper estimates a net cost of compliance for companies near the threshold of $4 million to $6 million for a few years of compliance (i.e., $1 million to $2 million per year). The analysis leading to this estimate relies on the relation between public float and market capitalization for other companies to approximate the stock market value foregone by those that are estimated to be manipulating their public float downwards. However, we note that the ratio of market capitalization to public float for other companies may simply reflect their propensity towards having affiliated ownership rather than being a reliable basis with which to measure the cost incurred by manipulating public float. 190We note that the estimates in this study rely on a number of critical assumptions and estimations. See Ge et al. 2017 Study, note 177 above (estimating the effect on audit fees by comparing the audit fees of non-accelerated filers to those of accelerated filers with market capitalization
of $300 million or less; and estimating the effect on earnings by estimating the percentage of non-accelerated filers that may newly disclose ineffective ICFR upon entering an ICFR auditor attestation requirement, based on changes in the rate of disclosure of ineffective ICFR by issuers that transition into accelerated filer status, and applying to this estimate a further estimate of the difference in return on assets that could be associated with such disclosure and any related remediation, based on the results of a multivariate regression relating issuers’ change in return on assets to a number of factors, including whether or not they disclosed and remediated ineffective ICFR). significant costs. Exempting the affected issuers from this requirement therefore is likely to have the benefit of reducing compliance costs for these issuers. Given the disproportionate burden that the fixed component of compliance costs impose on smaller reporting issuers, as well as the likelihood that many of the affected issuers face financing constraints, these costs savings may enhance capital formation and competition. The discussion below explores the anticipated cost savings and their potential implications in detail. We begin by summarizing evidence on the non-compliance costs and net costs of the ICFR auditor attestation requirement. We then estimate the anticipated effects on audit fees and on other compliance costs of eliminating this requirement for the affected issuers, using reported audit fees, survey data, and existing studies. Finally, we discuss the implications of the cost savings and other potential benefits. a. Evidence on Possible Indirect Costs and Net Costs of ICFR Auditor Attestation Requirement The ICFR auditor attestation requirement may impose costs on issuers and investors beyond the direct costs of compliance. For example, an increased focus on ICFR as a result of the ICFR auditor attestation requirement could have negative effects on issuer performance, if it creates a distraction from operational matters or reduces investment or risk-taking.181 Along these lines, studies have documented a decrease in investment and risk-taking by U.S. companies compared to companies in other countries around the passage of SOX.182 However, others have demonstrated that these findings are merely the continuation of a trend that began many years before the passage of SOX183 and that they do not appear to be driven by the applicability of the ICFR auditor attestation or SOX
Section 404(a) management ICFR
reporting requirements.184 Another study associates the SOX Section 404 requirements with a decrease in patents and patent citations, but the findings are limited to the early years of implementation of these requirements and the study is not able to distinguish to what extent the effects are attributable to the SOX Section 404(a) management ICFR reporting requirements versus the SOX Section 404(b) ICFR auditor attestation requirement.185 Our analysis separately considers the costs and benefits of extending the exemption from the ICFR auditor attestation requirement. While we are unable to quantify the extent to which the expected cost savings exceed any loss of benefits associated with the ICFR auditor attestation requirement, we note that researchers have attempted to estimate such ‘‘net costs’’ of the requirement in specific contexts. For example, studies have demonstrated that smaller reporting issuers find the total compliance costs associated with the ICFR auditor attestation requirement to be significant by providing evidence that non-accelerated filers may seek to avoid crossing the $75 million public float threshold and becoming accelerated filers.186 Issuers near or below this threshold have been found to be more likely than comparable issuers to take actions that may reduce or avoid an increase in their public float, such as disclosing more negative news in the second fiscal quarter (when public float is measured), increasing payouts to shareholders, reducing investment in property, plant, equipment, intangibles and acquisitions, and increasing the number of shares held by insiders.187 One study uses this avoidance behavior to estimate the net costs of compliance with the ICFR auditor attestation requirement for issuers close to the $75 million public float threshold.188 The study concludes that the overall costs, net of any benefits, of the ICFR auditor attestation requirement for these issuers is roughly $1 million to $2 million per year, but we note that the methodology used to translate the avoidance behavior into a dollar cost may be unreliable.189 One study attempts to quantify and compare certain costs and benefits of exempting non-accelerated filers from the ICFR auditor attestation requirement, focusing on those costs and benefits that the study deems to be measurable, and finds that the cost savings associated with exempting these issuers (an estimated $388 million in aggregate audit fee savings) have been less than the lost benefits (e.g., an aggregate $719 million in lower earnings) in aggregate present value terms.190 Studies have also used stock market reactions to changes in the applicability of the ICFR auditor attestation requirement to estimate its net costs or benefits, because the stock market valuation should incorporate both expected costs and expected benefits from a shareholder’s perspective. We

191 15 U.S.C. 78f.
192See Iliev 2010 Study, note 186 above. This study also finds a net reduction in value for small domestic issuers from the SOX Section 404 requirements, but is not able, for these issuers, to isolate the effect attributable to the ICFR auditor attestation requirement versus the SOX Section 404(a) management ICFR reporting requirement. 193See Kareen Brown, Fayez Elayan, Jingyu Li, Emad Mohammad, Parunchana Pacharn, & Zhefeng Frank Liu, To Exempt or not to Exempt NonAccelerated Filers from Compliance with the Auditor Attestation Requirement of Section 404(b) of the Sarbanes-Oxley Act, 28(2) Res. in Acct. Reg. 86 (2016) (‘‘Brown et al. 2016 Study’’). See also Christina Leuz & Peter Wysocki, The Economics of Disclosure and Financial Reporting Regulation:
Evidence and Suggestions for Future Research, 54(2) J. of Acct. Res. 525 at 566–569 (2016) (‘‘Leuz and Wysocki 2016 Study’’) (summarizing mixed evidence from earlier event studies related to SOX that were unable to differentiate the effects of the ICFR auditor attestation requirement from other requirements imposed by SOX). 194See, e.g., Robert Carnes, Dane Christensen, & Phillip Lamoreaux, Investor Demand for Internal Control Audits of Large U.S. Companies: Evidence from a Regulatory Exemption for M&A Transactions, 94(1) The Acct. Rev. 71 (2019) (‘‘Carnes et al. 2019 Study’’). 195See Hongmei Jia, Hong Xie, & David Ziebart, An Analysis of the Costs and Benefits of Auditor Attestation of Internal Control over Financial Reporting, Working Paper (2014) (‘‘Jia et al. 2014 study’’), available at https://www.lsu.edu/business/ accounting/files/researchseries/20141027JXZ.PDF. 196See note 115 above. 197See 2009 SEC Staff Study, note 123 above, and Cindy Alexander, Scott Bauguess, Gennaro Bernile, Alex Lee, & Jennifer Marietta-Westberg, The Economic Effects of SOX Section 404 Compliance:
A Corporate Insider Perspective, 56 J. of Acct and Econ. 267 (2013) (‘‘Alexander et al. 2013 Study’’). 198These estimates are based on staff analysis of data from the 2008–09 Survey. The analysis considers responses pertaining to the most recent year for which a given respondent provided a response. We note that the rate of responses to the question about net benefits was lower than for other questions. See the 2009 SEC Staff Study, note 123 above, and Alexander et al. 2013 Study, note 197 above, for details on the survey and analysis methodology. 199See, e.g., William Kinney & Marcy Shepardson (2011), Do Control Effectiveness Disclosures Require SOX 404(b) Internal Control Audits? A Natural Experiment with Small U.S. Public Companies, 49(2) J. of Acct. Res. 413 (‘‘Kinney and Shepardson 2011 Study’’) (considering those accelerated filers that have newly crossed the $75 million public float threshold in a given year); Iliev 2010 Study, note 186 above (considering those accelerated filers with between $75 million and $100 million in public float); Michael Ettredge, Matthew Sherwood, & Lili Sun (2017), Effects of SOX 404(b) Implementation on Audit Fees by SEC Filer Size Category, 37 (1) J. of Acct. and Pub. Pol’y 21 (considering accelerated filers as a category, as opposed to large accelerated filers, but also finding a contemporaneous 42.7% increase in audit fees for non-accelerated filers even though were not subject to the independent auditor attestation requirement); and Susan Elridge & Burch Kealey, SOX Costs: Auditor Attestation under
Section 404, Working Paper (2005), available at
http://papers.ssrn.com/sol3/papers.cfm?abstract_ id=743285 (considering accelerated filers in the lowest quintile of total assets). 200See, e.g., Alexander et al. 2013 Study, note 197 above. focus on studies that consider events that allow the effects of the ICFR auditor attestation requirement to be isolated from those of the other requirements that were imposed by SOX, as many early studies did not isolate the effects of the ICFR auditor attestation requirement from other changes required by the same legislation, such as the audit committee requirements of SOX Section 301191 and the certifications required pursuant to SOX
Section 302. Regardless, the results of
the studies we focus on have been mixed, perhaps due, in part, to changes over time in how the ICFR auditor attestation requirement has been implemented. For example, a study analyzing the response to announcements of initial delays in the application of the requirements to some issuers found that the ICFR auditor attestation requirement was associated with a net reduction in stock market valuation for foreign issuers.192 On the other hand, a study of the response to the later permanent exemption from the ICFR auditor attestation requirement for some issuers found that this requirement was associated with a net increase in stock market valuation for smaller reporting issuers.193 This finding is consistent with studies that conclude that the requirement is valueenhancing based on a negative stock market reaction to issuers excluding acquired operations from management’s assessment of ICFR and the ICFR auditor attestation, though these studies do not determine the extent to which this effect is attributable to the ICFR auditor attestation.194 Similarly, a study of smaller reporting issuers that switched regimes over time found that being subject to the ICFR auditor attestation requirement was associated with an increase in stock market valuation for these issuers.195 The rate of voluntary compliance with the ICFR auditor attestation requirement among exempt issuers has generally been low,196 which may indicate that exempt issuers, when considering their own net cost or benefit of compliance, have typically deemed it to be more beneficial to expend these resources on other uses. Finally, when considering the net tradeoff between costs and benefits for accelerated filers with low revenues in particular, we also reexamined data from the SEC-sponsored survey of financial executives conducted during December 2008 and January 2009 (‘‘2008–09 Survey’’).197 While the results of this survey might not be directly applicable a decade later, particularly given the changes over time discussed in Section III.B.1 above, they provide some suggestive evidence that low-revenue issuers are more likely than other accelerated filers to believe that the costs of complying with SOX
Section 404 substantially outweigh the
benefits. In particular, when asked about the net costs or benefits of complying with SOX Section 404, 30% of respondents at an accelerated filer with revenues below $100 million indicated that the costs far outweighed the benefits, in contrast to 14% of respondents at an accelerated filer with greater revenues.198 b. Potential Reduction in Audit Fees While issuers disclose their total audit fees, they are not required to disclose the portion of these fees that is attributable to the ICFR auditor attestation requirement. Studies of the initial implementation of the ICFR auditor attestation requirement found that it was associated with a roughly 100% increase in audit fees for small accelerated filers.199 However, these early estimates likely include some initial start-up costs, which were found to diminish over time.200 Further, these estimates do not incorporate the effect of later developments such as the adoption of AS 2201 (previously AS No. 5), which was expected to reduce compliance costs for smaller issuers, and the adoption of the new risk assessment auditing standards, which may reduce the incremental cost of an integrated audit over a financialstatement only audit. We therefore begin by considering current audit fees for accelerated filers that are subject to the ICFR auditor attestation requirement and have revenues of less than $100 million, as well as issuers in our comparison populations (non-accelerated filers, other than EGCs, and EGCs, neither of which is required to comply with the ICFR auditor attestation requirement) that also have revenues of less than $100 million. Table 12 presents the average total audit fees for these categories of filers.

201The estimates in the table are based on staff analysis of data from Ives Group Audit Analytics and include issuers in this revenue category and of each filer type with revenue data and a SOX Section 404(a) management report available in the Ives Group Audit Analytics database. See note 116 above for details on the identification of filer type. 202For EGCs, the average difference is $437,917 minus $317,360, or $120,557, which is about 27.5% of $437,917. For non-accelerated filers other than EGCs, the average difference is $437,917 minus $173,881, or $264036, which is about 60.3% of $437,917. 203 It is also possible that these estimates may be inflated due to the cost in recent years of transitioning to the 2013 COSO framework for evaluating ICFR. See note 111 above. 204See 2009 SEC Staff Study, note 123 above. See also 2013 GAO Study, note 115 above (finding, based on a survey conducted in December 2012 through February 2013, that 29% of audit fees for companies with a market capitalization of less than $10 billion and that obtained an auditor attestation in 2012 was attributable to these attestations). 205See Ge et al. 2017 Study, note 177 above (stating this difference as an increase of about 36% over the total audit fees of non-accelerated filers, which represents 0.36 divided by 1.36 or about 26% of the total audit fees of the small accelerated filers). 206See Ge et al. 2017 Study, note 177 above (finding an increase in audit fees of about 35%, representing 0.35 divided by 1.35 or about 26% of the total audit fees as a new accelerated filer). 207See Jia et al. 2014 Study, note 195 above (performing a regression analysis of total audit fees, including control variables for company size, auditor type, company and audit complexity, company performance, company operational risk, and financial risk). 208See letters from Acorda et al., Calithera, and CONNECT. These estimates are also generally consistent with the estimate set forth by a presenter at a recent Advisory Committee on Small and Emerging Companies (‘‘ACSEC’’) meeting. The presenter stated that some biotechnology companies that anticipate losing their status as EGCs in the next few years ‘‘believe they will incur somewhere between $150,000 to $350,000 in additional audit fees, $50,000 to $150,000 in other consulting costs and either $40,000 or as much as $200,000 for internal labor.’’ See William Newell, Presentation at ACSEC Meeting 49 to 54 (Sept. 13, 2017) (‘‘William Newell 2017 Presentation Transcript’’), available at https://www.sec.gov/info/smallbus/acsec/acsectranscript-091317.pdf. See also William J. Newell, Sarbanes-Oxley Section 404(b): Costs of Compliance and Proposed Reforms, Presentation at ACSEC Meeting (Sept. 13, 2017) available at https:// www.sec.gov/info/smallbus/acsec/william-newellacsec-091317.pdf.
TABLE 12—AVERAGE TOTAL AUDIT FEES IN DOLLARS BY FILER TYPE 201
Issuers with revenues <$100 million Accelerated (ex. EGCs) Non-Accelerated (ex. EGCs) EGC 2014 ........................................................................................................................... $424,019 $179,925 $199,744 2015 ........................................................................................................................... 436,190 183,077 463,403 2016 ........................................................................................................................... 446,381 167,214 317,433 2017 ........................................................................................................................... 445,079 165,307 288,860 Average/year .............................................................................................................. 437,917 173,881 317,360 For these low-revenue issuers, the difference between the average annual audit fees for accelerated filers and the comparison populations represents, as a percentage of the total audit fees for accelerated filers, roughly 25 to 60% of those total audit fees.202 Some part of this 25 to 60% difference is likely attributable to the ICFR auditor attestation requirement. However, as discussed in Section III.C.2, audit fees have been found in general to increase with total assets and other measures of issuer size, and the median issuer in the comparison populations is substantially smaller than the median affected issuer (in terms of total assets, number of employees, or public float). To account for the fact that some portion of the 25 to 60% difference in audit fees across these groups may be attributable to their difference in size,203 we select an estimate at the low end of the range, resulting in a percentage estimate of 25% of total audit fees that would be saved by issuers newly exempted from the ICFR auditor attestation requirement. This estimate is generally consistent with a range of estimates from other sources that use data from after the 2007 change in the ICFR auditing standard, but that are not focused on low-revenue issuers. These other estimates, which range from approximately five to 35% of total audit fees, are based on a variety of samples and methodologies. For example, the 2008–09 Survey asked respondents what portion of their audit fees were attributable to the ICFR auditor attestation. The average reported percentage for the fiscal year in progress at the time of the survey was 34% for issuers with public float between $75 million and $700 million.204 One study considered the difference in the change in audit fees from 2003 through 2014 for non-accelerated filers versus smaller accelerated filers (i.e., those with market capitalization less than $300 million) and concluded that about 26% of the total audit fees for smaller accelerated filers was attributable to the ICFR auditor attestation requirement.205 This study also found a similar percentage effect when considering the change in audit
fees for issuers that newly entered accelerated filer status.206 A different study that controls for additional factors that could be associated with total audit fees finds a more modest effect, estimating that, on average, a five percent increase in audit fees was attributable to transitioning to accelerated filer status over the period from 2007 to 2013 (compared to an average increase of 59.52% for the period from 2002 to 2006, before the 2007 change in the ICFR auditing standard).207 We note that these studies do not separately consider the audit fees of low-revenue issuers and may not fully incorporate the effects of recent developments, such as the increased focus of PCAOB inspections on ICFR auditor attestations beginning around 2010 and the new risk assessment auditing standards. Given the average total audit fees of about $440,000 per year for accelerated filers with revenues of less than $100 million, we preliminarily estimate that about 25% of these fees, or about $110,000 per year, would be saved on average by the affected issuers as a result of the proposed amendments. The audit fee savings are expected to vary across the affected issuers, with some experiencing smaller savings and some experiencing much larger savings depending on their individual circumstances. For example, a few of the commenters to the SRC Proposing Release cited costs of $400,000 to over $1 million associated with the ICFR auditor attestation requirement (though it is possible that these estimates include costs other than audit fees, which are discussed below).208 Further, we note that some issuers may voluntarily choose to continue to make these expenditures if they deem the benefits of the ICFR auditor attestation to exceed the cost, and that the extent of savings may be affected if auditors continue to test the operating effectiveness of some controls as part of their financial statement audit. Our estimate is subject to significant uncertainty, given the lack of a perfect comparison group, as discussed above, and the fact that it is difficult to isolate the recurring cost of the ICFR auditor attestation requirement from the effects of other key factors that may affect audit fees in our sample, such as the recent

209See, e.g., Leuz and Wysocki 2016 Study, note 193 above. 210See, e.g., Protiviti 2018 Report, note 135 above (finding, for example, total internal costs associated with all aspects of SOX compliance to be $282,900 for 2018 for respondents with less than $100 million in revenues) and SOX & Internal Controls Professionals Group, Moss Adams LLP, and Workiva (2017), ‘‘2017 State of the SOX/Internal Controls Market Survey’’ (‘‘2017 SICPG Survey Report’’), available at www.mossadams.com/ landingpages/2017-sox-and-internal-controlsmarket-survey. 211For example, a presenter at a recent ACSEC meeting provided four examples of biotechnology companies with actual or expected costs other than audit fees attributable to audits of ICFR of $190,000 (Example A), $135,000 (Example B), greater than $110,000 (Example C), and $175,000 (Example D), including the costs of outside vendors, consultants and internal labor. The presenter also cited discussions with other companies that are currently EGCs but ‘‘believe they will incur . . . $50,000 to $150,000 in other consulting costs and either $40,000 or as much as $200,000 for internal labor.’’ See William Newell 2017 Presentation Transcript, note 208 above. See also BIO White Paper, Science or Compliance: Will Section 404(b) Compliance Impede Innovation by Emerging Growth Companies in the Biotech Industry? (February 2019) (‘‘BIO Study’’), available at https://www.bio.org/sites/ default/files/BIO_EGC_White_Paper_02_11_2019_ FINAL.pdf (finding that five biotechnology companies incurred an average cost of outside vendors and consultants related to SOX Section 404(b) compliance of $192,200 and an average cost of associated internal labor of $163,000, for a total of $355,200, based on the responses of these companies, which may or may not overlap with the companies cited in the presentation to ACSEC, to a survey). 212See letter from BIO (supporting allowing ‘‘issuers and their investors the flexibility to determine for themselves whether Section 404(b) is relevant to their business’’). 213For example, one commenter indicated that ‘‘pre-revenue small businesses utilize only investment dollars to fund their work’’ and that any cost savings thus ‘‘could lead to funding for a new life-saving medicine.’’ See letter from BIO. 214For example, in a survey of issuers in the biotech industry, among 11 biotech EGCs that responded to a question regarding how an extension of the exemption from the independent auditor attestation requirement would affect them given the costs associated with the requirement, eight out of the 11 issuers indicated that they expected a positive impact on investments in research and development and six out of the 11 issuers indicated that they expected a positive impact on hiring employees. See BIO Study, note 211 above. 215See, e.g., letter from ICBA. changes in accounting standards discussed above. Also, the costs of obtaining an ICFR auditor attestation may decline over time with the adoption of more
automated controls testing and process automation.
c. Additional Potential Compliance Cost
Savings
The ICFR auditor attestation requirement is associated with substantial other compliance costs beyond audit fees, including outside vendor costs and internal labor costs.209 However, these costs are difficult to measure because they are not required to be reported. Practitioner studies based on surveys of issuers often report non-audit costs of the internal control assessment and reporting requirements of SOX Section 404 in particular or of SOX in general, but the costs attributable to the ICFR auditor attestation requirement versus the SOX
Section 404(a) management ICFR
reporting requirements or other requirements are generally not broken out separately.210 The 2008–09 Survey asked respondents to report their non-audit costs of SOX Section 404 in general, such as their outside vendor costs, labor, and non-labor costs (such as software, hardware and travel costs), as well as the percentage of the outside vendor costs and labor hours that were attributable to the ICFR auditor attestation requirement. For the fiscal year in progress at the time of the survey, the mean (median) annual costs for issuers with between $75 million and $700 million in public float were $134,691 ($50,000) for outside vendors, $489,302 ($242,000) for internal labor costs, and $79,348 ($20,000) for nonlabor costs. Respondents indicated that, on average, ten percent of the outside vendor costs and 25% of the internal labor costs were attributable to the ICFR auditor attestation requirement. A breakdown was not provided for the non-labor costs, which we believe are primarily attributable to management’s ICFR responsibilities under SOX
Section 404(a) rather than the ICFR
auditor attestation.
The average non-audit costs attributable to the ICFR auditor attestation requirement at the time of the survey were thus approximately $125,000 per year ($134,691 times ten percent, plus $489,302 times 25%). In more recent years, the adoption of the new risk assessment auditing standards may have increased the non-audit costs of a financial statement only audit, and thus reduced the incremental costs attributable to the ICFR auditor attestation requirement. We therefore adjust the historical cost downward slightly and estimate that the average non-audit costs attributable to the ICFR auditor attestation requirement are approximately $100,000 per year. This estimate is subject to uncertainty because it is unclear exactly how the current costs may differ from the survey responses a decade ago, and the costs may be different for low-revenue issuers. As in the case of audit fees, some of the affected issuers are expected to experience lower cost savings while others would experience greater savings, depending on their individual circumstances. For example, some issuers have reported potential cost savings other than audit fees ranging from about $110,000 to about $350,000.211 d. Implications of the Cost Savings While we estimate the average compliance cost associated with the ICFR auditor attestation requirement for the affected issuers, it is more difficult to discern whether incurring the costs of this requirement represents the most effective use of funds for these issuers. As discussed in Section III.C.4.c below, issuers for whom the requirement is eliminated may determine that it is worthwhile to use these funds to voluntarily undergo an audit of ICFR.212 Alternatively, some of these issuers could directly invest the compliance cost savings in their control systems, or in improving their operations and prospects for growth. In total, we estimate an average cost savings of $210,000 per issuer per year, with some of the affected issuers experiencing lesser or greater savings. This represents a significant cost savings for issuers with less than $100 million in revenue and may thus have beneficial economic effects on competition and capital formation. In particular, because of the fixed costs component of compliance costs, smaller issuers generally bear proportionately higher compliance costs than larger issuers. For example, we estimate that total audit fees for the past three years have represented about 22% of revenues on average for accelerated filers with less than $100 million in revenues, versus 0.5% of revenue for those above $100 million in revenues. Reducing the affected issuers’ costs would reduce their overhead expenses and may enhance their ability to compete with larger issuers. Importantly, low-revenue issuers are likely to face financing constraints because they do not have access to internally-generated capital.213 Resources saved by the affected parties therefore may be likely to be put to productive use,214 such as towards
capital investments, which would enhance capital formation. The alleviation of these costs could be a positive factor in the decision of additional companies to enter public markets,215 particularly in the case of companies that expect low levels of revenue to persist for many years into the future. That is, if future compliance costs associated with ICFR auditor attestations weigh against these companies becoming publicly traded,

216There is some evidence of a decreased rate of initial public offerings and an increased rate of going private transactions and deregistrations in the United States after SOX. However, it is unclear to what extent these changes can be attributed to SOX (or to the auditor attestation requirement in particular) versus other factors, and to what extent these changes are a cause for concern. See e.g., Coates and Srinivasan 2014 Study, note 181 above, at 636–640 (summarizing a number of studies in this area). 217See note 88 above regarding the exemption of EGCs from the auditor attestation requirement. 218See BIO Study, note 211 above (finding that biotechnology EGCs have lower restatement frequencies than other issuers, after controlling for other factors, and attributing this to their ‘‘absence of product revenue’’ based on findings that revenue recognition is one of the most frequent drivers of financial restatements). 219See notes 228 through 232 below and accompanying text. 220See 2011 SEC Staff Study, note 49 above, at 97–99 and 102–104. See also Coates and Srinivasan 2014 Study, note 181 above. 221See, e.g., Sarah Rice & David Weber, How Effective is Internal Control Reporting under SOX 404? Determinants of the (Non-)Disclosure of Existing Material Weaknesses, 50(3) J. OF ACCT. RES. 811 (2012); William Kinney, Roger Martin, & Marcy Shepardson, Reflections on a Decade of SOX 404(b) Audit Production and Alternatives, 27(4) Acct. Horizons 799 (2013); and Daniel Aobdia, Preeti Choudhary, & Gil Sadka, Do Auditors Correctly Identify and Assess Internal Control Deficiencies? Evidence from the PCAOB Data, Working Paper (2018), available at http:// papers.ssrn.com/sol3/papers.cfm?abstract_ id=2838896. See also Kinney and Shepardson 2011 Study, note 199 above. 222See, e.g., 2011 SEC Staff Study, note 49 above, at 86 (citing evidence that while both issuers Continued reducing these expected future costs may enhance capital formation in the public markets and the efficient allocation of capital at the market level. However, research investigating the link between SOX and companies exiting or choosing not to enter public markets has been inconclusive.216 Further, newly public issuers can already avail themselves of an exemption from the ICFR auditor attestation requirement for at least one and generally up to five years after their initial public offering.217 To the extent that companies may be more focused on costs during those first five years or other factors associated with the decision to go public, the impact of the proposed amendments on the number of publicly traded companies may be limited.
4. Potential Costs of Eliminating the
ICFR Auditor Attestation Requirement for Affected Issuers Exempting the affected issuers from the ICFR auditor attestation requirement may result, over time, in management at this category of issuers being less likely to maintain effective ICFR, which in turn may result in less reliable financial statements, on average, for these issuers. The discussion below explores this potential effect and its implications in detail. We also consider two mitigating factors that could be associated with the affected issuers on average, though they may not apply equally to all of the affected issuers. First, low-revenue issuers may be less susceptible to the risk of certain kinds of misstatements, such as errors associated with revenue recognition.218 Second, in many cases, the market value of such issuers may be driven to a greater degree by their future prospects than by the current period’s financial statements, which may affect how, on average, investors use these issuers’ financial statements. Exempting the affected issuers from the ICFR auditor attestation requirement could also reduce the information available to investors for gauging the reliability of these issuers’ financial statements. In this regard, we discuss below the potential effects related to the identification and disclosure of material weaknesses in ICFR at the affected issuers. However, given the recent findings discussed in Section III.C.4.a below on how ICFR auditor attestations may provide limited information about the risk of future restatements,219 we preliminarily believe that any such effect would not meaningfully affect investors’ overall ability to make informed investment decisions. a. Considerations and Evidence Regarding the Effects of ICFR Auditor Attestations on Financial Reporting This section summarizes a number of broad economic considerations related to the possible effects of an ICFR auditor attestation requirement on financial reporting in order to provide context for the more detailed analysis of the costs of exempting the affected parties from this requirement that follows. As discussed below, the anticipated effects of changes to the population of issuers subject to the ICFR auditor attestation requirement will depend on the characteristics of the specific group of issuers that would be affected. In this regard we note that prior research has not focused on the effects of the ICFR auditor attestation requirement on lowrevenue issuers in particular. As discussed in Section III.B.1, there also have been significant changes over time in the implementation of the ICFR auditor attestation requirement, the standards applying to a financial statement audit even in the absence of an audit of ICFR, and the execution of audits of financial statements and of ICFR, which may have had the effect of reducing both the incremental costs and incremental benefits of an ICFR auditor attestation since the periods studied in much of the existing research. We therefore acknowledge that,
while we believe that consideration of the past research is an important part of our analysis, these factors limit our ability to rely on the findings of past research to predict how the proposed amendments would affect the particular class of issuers implicated by this rulemaking. ICFR auditor attestations can have two primary types of benefits. First, the ICFR auditor attestation reports can provide incremental information to investors about the reliability of the financial statements. Second, the reliability of the financial statements can itself be enhanced. That is, the expectation of, or process involved in, the ICFR auditor attestation could lead issuers to maintain better controls, which could lead to more reliable financial reporting. Importantly for our evaluation of these possible benefits, however, we do not directly observe the effectiveness of ICFR and the reliability of financial statements, but only the associated disclosures by issuers. For example, while restatements may indicate that controls have failed, such restatements are often predicated on the underlying misstatements being detected. Given such limitations with the available data, the analysis in existing studies and in this release is necessarily less than definitive. Regarding the first possible benefit of ICFR auditor attestations, academic research provides some evidence that ICFR auditor attestation reports contain information about the reliability of financial statements, but also demonstrates that the incremental information provided by these reports may be limited. The 2011 SEC Staff Study summarizes evidence that ICFR auditor attestations generally resulted in the identification and disclosure of material weaknesses that were not previously identified or whose severity was misclassified when identified by management in its assessment of ICFR, and that investor risk assessments and investment decisions were associated with the findings in auditor attestation reports.220 However, more recent studies have found that auditor identification of material weaknesses in ICFR tends to be concurrent with the disclosure of restatements, rather than providing advance warning of the potential for restatements.221 While these findings do not imply that ICFR auditor attestation reports fail to provide any useful information about the risk of future restatements,222 they demonstrate that

subject to SOX Section 404(b) as well as those only subject to SOX Section 404(a) often report restatements despite previously reporting that their ICFR was effective, such restatements were 46% higher among those filing only SOX Section 404(a) reports). See also PCAOB Investor Advisory Group, Report from the Working Group on the Investor Survey (2015), available at https://pcaobus.org/ News/Events/Documents/09092015_IAGMeeting/ Investor_Survey_Slides.pdf (reporting survey findings that 72% of institutional investors indicated that they relied on independent auditor attestations of ICFR either ‘‘extensively’’ or ‘‘a good bit’’). 223See, e.g., Ge et al. 2017 Study, note 177 above. 224See Sarah Rice, David Weber, & Biyu Wu, Does SOX 404 Have Teeth? Consequences of the Failure to Report Existing Internal Control Weaknesses, 90(3) Acct. Rev. 1169 (2015). We note that auditors have a duty to follow auditing standards and, if they do not, face associated enforcement, inspection, reputation, and litigation risks that provide a countervailing incentive. 225See, e.g., Defond and Lennox 2017 Study, note 126 above (finding that PCAOB inspections may increase auditors’ issuance of adverse internal control opinions to clients with later restatements). 226See Coates and Srinivasan 2014 Study, note 181 above, and Leuz and Wysocki 2016 Study, note 193 above (both articles discussing the limited ability to make causal attribution based on research on the effects of the provisions of SOX, but also highlighting the specific studies that can more plausibly make causal claims). See also Report to Congress: Access to Capital and Market Liquidity, August 2017 SEC Staff study 24–27 (discussing similar limitations, in a different context, in the ability to make causal inferences about the effects of regulation because of data and experimental design issues), available at https://www.sec.gov/ files/access-to-capital-and-market-liquidity-studydera-2017.pdf. 227See Section III.B.1 above. 228See 2011 SEC Staff Study, note 49 above, at 41 and 86–87. The rate of ineffective ICFR is based on the findings of management reports on ICFR pursuant to SOX Section 404(a). Because auditor attestations of ICFR are associated with an increased detection and disclosure of material weaknesses, as discussed above, the rate of ineffective ICFR reported by issuers not subject to the auditor attestation requirement may be understated, which would result in this difference also being understated. 229See, e.g., Audit Analytics, SOX 404 Disclosures: A Fourteen Year Review (Sept. 2018) (‘‘2018 Audit Analytics Study’’), available at www.auditanalytics.com/blog/sox-404-disclosuresa-fourteen-year-review/. 230See Schroeder and Shepardson 2016 Study, note 124 above (using quarterly accruals quality, measured by the level of quarterly discretionary working capital accruals and the quarterly accrual estimation error, as a proxy for internal control quality based on the argument that internal control
improvements should be exhibited in unaudited financial reports). 231 Id. 232See Vishal Munsif & Meghna Singhvi, Internal Control Reporting and Audit Fees of NonAccelerated Filers, 15(4) J. of Acct., Ethics & Pub. Pol’y 902 at 915 (2014) (finding that 49 out of 160, or 30%, of non-accelerated filers that disclosed a material weakness in 2008 reported no material weaknesses in 2009, in contrast to 64 out of 83, or 77%, of accelerated filers in a similar situation). See also Jacqueline Hammersley, Linda Myers, & Jian Zhou, The Failure to Remediate Previously Disclosed Material Weaknesses in Internal Controls, 31(2) Auditing: A J. of Prac. & Theory 73 (2012); and Karla Johnstone, Chan Li, & Kathleen Rupley, Changes in Corporate Governance Associated with the Revelation of Internal Control Material Weaknesses and their Subsequent Remediation, 28(1) Contemp. Acct. Res. 331 (2011) (both finding a similar rate of remediation for accelerated filers for an earlier sample period). 233See Coates and Srinivasan 2014 Study, note 181 above, at 649–650. 234See Dain Donelson, Matthew Ege, & John McInnis, Internal Control Weaknesses and Financial Reporting Fraud, 36(3) Auditing: A J. of Prac. and Theory 45 (2017) (finding that issuers with a material weakness in ICFR are 1.24 percentage points more likely to have a fraud revelation within the next three years compared to issuers without a material weakness, relative to a 1.60% unconditional probability of fraud). 235See Hollis Asbhaugh-Skaife, David Veenman, & Daniel Wangerin, Internal Control over Financial Reporting and Managerial Rent Extraction:
Evidence from the Profitability of Insider Trading, 55(1) J. of Acct. and Econ. 91 (2013). 236See, e.g., Sarah Clinton, Arianna Pinello, & Hollis Skaife, The Implications of Ineffective Internal Control and SOX 404 Reporting for Financial Analysts, 33(4) J. of Acct. and Pub. Pol’y 303 (2014). 237See 2011 SEC Staff Study, note 49 above, at 98–100. For more recent evidence, see, e.g., Yuping Zhao, Jean Bedard, & Rani Hoitash, SOX 404, Auditor Effort, and the Prevention of Financial Report Misstatements, 151 (2017); and Lucy Chen, Jayanthi Krishnan, Heibatollah Sami, & Haiyan Zhou, Auditor Attestation under SOX Section 404 and Earnings Informativeness, 32(1) Auditing: A J. of Prac. & Theory 61 (2013). 238See Bhaskar et al. 2018 Study, note 124 above (finding that, among companies with less than $150 million in market capitalization, those providing this information may be limited. Further, researchers have been able to predict the identification by auditors of material weaknesses in ICFR beyond those identified by management, to some extent, by using otherwise available information about issuers beyond current restatements, such as their institutional ownership, aggregate losses, past restatements, and late filings.223 The limited incremental information provided by ICFR auditor attestation reports about the risk of future restatements may result from disincentives, such as the increased risk of litigation and greater likelihood of management and auditor turnover that have been associated with earlier material weakness disclosures, for issuers and their auditors to disclose material weaknesses in the absence of restatements.224 It may also result from issues with the quality of the audit of ICFR. In this regard, researchers have found that PCAOB scrutiny of these audits has been associated with a slightly higher rate of identification of material weaknesses in ICFR prior to a later restatement.225 A further reason why ICFR auditor attestation reports may provide only a weak warning about future restatements is that the audit of ICFR may contribute to the avoidance of misstatements, leading us to observe only the residual restatements where the misstatement risk was not foreseen or a misstatement was not detected for reasons unrelated to internal controls. Thus, the second possible benefit we consider is that the audit of ICFR may encourage management to maintain more effective controls and thereby deter accounting errors and fraud. The academic research discussed below documents substantial evidence that would be consistent with such effects, though, as is common in financial economics, it is difficult to determine whether the documented differences can be causally linked to the audit of ICFR.226 In particular, while issuers are subject to a number of requirements discussed above that are intended to help to ensure adequate internal controls and reliable financial statements,227 studies have documented a significant association between audits of
ICFR and the maintenance of better internal controls. The 2011 SEC Staff Study provides analysis and summarizes research indicating that issuers that were not required to obtain an ICFR auditor attestation disclosed ineffective ICFR at a greater rate than those that were subject to such requirements,228 and newer studies demonstrate that this difference has remained consistent in recent years.229 Further, a recent paper finds that the ICFR auditor attestation requirement, but not management ICFR reporting requirements alone, are associated with enhanced quarterly earnings accrual quality, and argues that this is an indication of the improved quality of internal controls.230 We note, however, that this study finds that the improvements for issuers subject to the ICFR auditor attestation requirement are attenuated after the 2007 change in the ICFR auditing standard discussed in
Section III.B.1 above.231 The ICFR
auditor attestation requirement has also been associated with a higher rate of remediation of material weaknesses after they are disclosed.232 To the extent that the ICFR auditor attestation requirement leads to more effective ICFR, this requirement may thereby lead to more reliable financial statements. Some studies have found that a failure to maintain effective ICFR has been associated with a higher rate of future restatements and lower earnings quality,233 a higher rate of future fraud revelations,234 more profitable insider trading,235 and less accurate analyst forecasts.236 Generally, ICFR auditor attestations also have been found to be directly associated with financial statements that are more reliable than in the absence of these attestations.237 However, one study finds conflicting evidence using data from 2007 through 2013,238 consistent

auditor attestations of ICFR, whether voluntarily or because they are accelerated filers, had a higher rate of material misstatements and lower earnings quality than others in this category in the period from 2007 through 2013). 239See, e.g., Douglas Diamond & Robert Verrecchia, Disclosure, Liquidity, and the Cost of Capital, 46(4) J. of Fin. 1325 (1991) (‘‘Diamond and Verrecchia 1991 Study’’); David Easley & Maureen O’Hara, ‘Information and the Cost of Capital,’ 59(4) J. of Fin. 1553 (2004); Richard Lambert, Christian Leuz, & Robert Verrecchia, Accounting Information, Disclosure, and the Cost of Capital,’’ 45(2) J. OF ACCT. RES. 385 (2007); and Christopher Armstrong, John Core, Daniel Taylor, & Robert Verrecchia, When Does Information Asymmetry Affect the Cost of Capital? 49(1) J. OF ACCT. RES. 1 (2011). We note that these articles also detail limited theoretical circumstances under which more reliable disclosures could lead to a higher cost of capital, such as in the case where improved disclosure is sufficient to reduce incentives for market making. 240See, e.g., Dragon Tang, Feng Tian, & Hong Yan, Internal Control Quality and Credit Default Swap Spreads, 29(3) Acct. Horizons 603 (2015); Lawrence Gordon & Amanda Wilford, An Analysis of Multiple Consecutive Years of Material Weaknesses in Internal Control, 87(6) Acct. Rev. 2027 (2012) (‘‘Gordon and Wilford 2012 Study’’); and H. Ashbaugh-Skaife, D. Collins, W. Kinney, & R. LaFond, The Effect of SOX Internal Control Deficiencies on Firm Risk and Cost of Equity, 47(1) J. of Acct. Res. 1 (2009) (‘‘Ashbaugh-Skaife et al. 2009 Study’’). We note that earlier work did not detect an association between SOX Section 404 material weaknesses and the equity cost of capital. See, e.g., M. Ogneva, K. R. Subramanyam, & K. Rachunandan, Internal Control Weakness and Cost of Equity: Evidence from SOX Section 404 Disclosures, 82(5) Acct. Rev. 1255 (2007) (‘‘Ogneva et al. 2007 Study’’). See also 2011 SEC Staff Study, note 49 above, at 101–102. 241See, e.g., Paul Hribar & Nicole Jenkins, The Effect of Accounting Restatements on Earnings Revisions and the Estimated Cost of Capital, 9 Rev. of Acct. Stud. 337 (2004) (‘‘Hribar and Jenkins 2004 Study’’). 242See, e.g., Jennifer Francis, Ryan LaFond, Per M. Olsson, & Katherine Schipper, Cost of Equity and Earnings Attributes, 79(4) Acct. Rev. 967 (2004) (‘‘Francis et al. 2004 Study’’). 243We note that empirical studies of the cost of equity capital face particular challenges in accurately measuring the cost of equity capital, which can reduce their reliability, but that this is mitigated in studies that look at changes over time (Gordon and Wilford 2012 Study, note 240 above, Ashbaugh-Skaife et al. 2009 Study, note 240 above, and Hribar and Jenkins 2004 Study, note 241 above) rather than in the cross-section (Ogneva et al. 2007 Study, note 240 above, and Francis et al. 2004 Study, note 242 above). See, e.g., Stephannie Larocque & Matthew R. Lyle, Implied Cost of Equity
Capital Estimates as Predictors of Accounting Returns and Stock Returns, 2(1) J. of Fin. Rep. 69 (2017) (discussing concerns about measures of the cost of equity capital); and Charles M. C. Lee, Eric
C. So, & Charles C. Y. Wang, Evaluating Firm-Level
Expected-Return Proxies, Harvard Business School Working Paper 15–022 (2017) (finding that ‘‘in the vast majority of research settings, biases in [equity cost of capital measures] are irrelevant’’ and that the cost of equity capital measures used in the accounting literature ‘‘are particularly useful in tracking time-series variations in expected returns’’). 244See, e.g., Ge et al. 2017 Study at 359 (arguing that internal control misreporting leads to lower operating performance due to the non-remediation of ineffective controls, and estimating the degree of such underperformance based on the improvement shown by issuers that are non-accelerated filers after disclosing and remediating material weaknesses, relative to other such issuers that are suspected of having unreported material weaknesses). We note that companies may choose to improve their controls when they are otherwise expecting to enter a period of improved performance, which could lead to a similar association without such improved performance being caused by the changes in internal controls. 245See Mei Cheng, Dan Dhaliwal, & Yuan Zheng, Does Investment Efficiency Improve After the Disclosure of Material Weaknesses in Internal Control over Financial Reporting?, 56(1) J. of Acct. and Econ. 1 (2013). 246See Mei Feng, Chan Li, Sarah McVay, & Hollis Skaife, Does Ineffective Internal Control Over Financial Reporting Affect a Firm’s Operations? Evidence From Firms’ Inventory Management,’’ 90(2) Acct. Rev., 529 (2015). 247See, e.g., Leuz and Wysocki 2016 Study, note 193 above (stating that researchers ‘‘generally lack evidence on market-wide effects and externalities from regulation, yet such evidence is central to the economic justification of regulation’’ and acknowledging that ‘‘the identification of such market-wide effects and externalities is even more difficult than the identification of direct economic consequences on individual firms’’). 248There is also some evidence that more reliable financial disclosures also facilitate a more effective market for corporate control, which can increase overall market discipline and thus enhance the efficiency of production by incentivizing more effective management. See Amir Amel-Zadeh & Yuan Zhang, The Economic Consequences of Financial Restatements: Evidence from the Market for Corporate Control, 90(1) Acct. Rev. 1 (2015). See also Vidhi Chhaochharia, Clemens Otto, & Vikrant Vig, The Unintended Effects of the Sarbanes–Oxley Act, 167(1) J. of Institutional and Theoretical Econ. 149 (2011). 249See, e.g., 2013 GAO Study, note 115 above (finding that 52% of the companies surveyed reported greater confidence in the financial reports of other companies due to the ICFR auditor attestation requirement; in contrast, 30% of the respondents reported that they believed this requirement raised investor confidence in their own company). 250For a further discussion of potential externalities, see Coates and Srinivasan 2014 Study, note 181 above, at 657–659.
251See Ge et al. 2017 Study (regarding the term ‘‘younger,’’ this study defines company age as the number of years a company has been covered in the Compustat database). See also 2011 SEC Staff Continued with concerns discussed in Section III.B.1 above that the quality of audits of ICFR dropped at least temporarily after 2007. To evaluate the economic implications of any effects the ICFR auditor attestation requirement has on ICFR and financial statements, these effects can be tied to their possible effects on factors such as production or investment at the issuer or market level. For example, at the issuer level, more reliable disclosures are generally expected, based on economic theory, to lead investors to demand a lower expected return to hold an issuer’s securities (i.e., a lower cost of capital).239 A lower cost of capital may enhance capital formation by encouraging issuers to issue additional securities in order to make new investments. Empirically, material weaknesses in ICFR,240 restatements,241 and low earnings quality 242 have all been associated with a higher cost of debt or equity 243 capital. More effective ICFR and more reliable financial reporting may also lead to improved efficiency of production if managers themselves thereby have access to more reliable data that facilitates better operating and investing decisions.244 For example, one study finds that the investment efficiency of issuers improves, in that both underinvestment and over-investment are curtailed, after the disclosure and remediation of material weaknesses.245 Another study finds that issuers that remediate material weaknesses in ICFR that are related to inventory tracking thereafter experience higher inventory turnover, together with improvements in sales and profitability.246 That said, it is difficult to generalize the results beyond these samples to determine whether non-remediating issuers or issuers with different types of material weaknesses in ICFR could expect similar operational benefits from remediation. The ICFR auditor attestation requirement may also result in benefits at the market level, though these are more difficult to measure than those at the issuer level.247 The potential for market-level impact is largely driven by network effects (which are associated with the broad adoption of practices) and by other externalities (i.e., spillover effects on issuers or parties beyond the issuer in question). For example, to the extent that the ICFR auditor attestation requirement leads to more reliable financial statements at a large number of issuers, it may lead to a more efficient allocation of capital across different investment opportunities at the market level.248 The ICFR auditor attestation requirement also can enhance capital formation to the extent that it improves overall investor confidence, for which there is some suggestive evidence,249 and thus encourages investment in public markets.250 Importantly, all of these benefits, at both the
issuer and market level, likely vary across issuers of different types. For example, younger, loss-incurring issuers with lower market capitalization and lower institutional ownership, as well as those with more segments, tend to be more likely to newly disclose material weaknesses as they transition into the ICFR auditor attestation requirement.251 However, the market

Study, note 49 above, at 96 (summarizing previous research finding that internal control deficiencies are associated with smaller, complex, riskier, and more financially-distressed issuers). 252See Ashbaugh-Skaife et al. 2009 Study, note 240 above. 253See, e.g., Committee of Sponsoring Organizations of the Treadway Commission, Fraudulent Financial Reporting 1998–2007: An Analysis of U.S. Public Companies (2010) (‘‘COSO 2010 Fraud Study’’), available at http:// www.coso.org/documents/COSO-Fraud-Study2010–001.pdf (finding that companies allegedly engaging in financial disclosure fraud in the period from 1998 through 2007 had median assets and revenue under $100 million and were often lossincurring or close to breakeven) and Characteristics of Financial Restatements and Frauds, CPA J. (Nov. 2017), available at www.cpajournal.com/2017/11/ 20/characteristics-financial-restatements-frauds/ (for more recent evidence). 254See, e.g., Patricia Dechow & Catherine Schrand, Earnings Quality, Research Foundation of CFA Institute 12 (2004) (‘‘Dechow and Schrand 2004 Monograph’’). 255See, e.g., Joel Peress & Lily Fang, Media Coverage and the Cross-Section of Stock Returns, 64(5) J. of Fin. 2023 at 2030 (2009) (finding that ‘‘firm size has an overwhelming effect on media coverage: large firms are much more likely to be covered’’); Armando Gomes, Gary Gorton, & Leonardo Madureira, SEC Regulation Fair Disclosure, Information, and the Cost of Capital, 13 J. of Corp. Fin. 300 at 307 (2007) (stating that ‘‘there is overwhelming evidence that size can explain analyst following’’); and Eliezer Fich, Jarrad Harford, & Anh Tran, Motivated Monitors: The Importance of Institutional Investors’ Portfolio Weights, 118(1) J. of Fin. Econ. 21 (2015) (finding that institutional monitoring is greatest when a company represents a significant allocation of funds in the institution’s portfolio, which is strongly associated with company size). 256See Bhaskar et al. 2018 Study, note 124 above, as discussed in note 238 above. appears to account for the association of material weaknesses with these and other observable issuer characteristics. Thus, issuers with characteristics associated with a higher rate of material weaknesses but that receive an auditor attestation report that does not find such weaknesses are found to have the greatest cost of capital benefit from such a report.252 Small, loss-incurring issuers are also disproportionately represented amongst issuers that have allegedly engaged in financial disclosure frauds, indicating that any benefits in terms of investor protection and investor confidence may be particularly important for this population of issuers.253 On the other hand, marginal changes in the reliability of the financial statements of issuers whose valuation is driven primarily by their future prospects could have limited issuer- and market-level effects to the extent that the current financial statements of these issuers are less critical to assessing their
valuation.254 b. Estimated Effects on ICFR and the Reliability of Financial Statements The academic literature discussed in
Section III.C.4.a above suggests that the
scrutiny associated with the ICFR auditor attestation may lead issuers that are required to obtain this attestation to maintain more effective ICFR and to remediate material weaknesses in ICFR more quickly, leading to more reliable financial statements. Further, as discussed above, studies have highlighted that smaller reporting issuers are disproportionately represented in populations of issuers with ineffective ICFR and financial statements that require material restatement. In addition, smaller issuers are less likely to have significant external scrutiny in the form of analyst and media coverage and monitoring by institutional owners,255 which could otherwise provide another source of discipline to maintain the reliability of financial statements. However, one study cited above finds that the ICFR auditor attestation requirement was associated with less reliable financial statements for lower market capitalization issuers from 2007 through 2013,256 and the existing studies in general may not be directly applicable to current circumstances given the 2010 change in risk assessment auditing standards, the 2007 change in the ICFR auditing standard and other recent changes discussed in
Section III.B.1 above. Importantly, the
existing literature also does not directly examine low revenue issuers. This section therefore provides an analysis of low-revenue issuers using recent data to complement the existing studies and better inform our consideration of the possible costs of the proposed amendments. However, some uncertainty will remain due to the challenges discussed above in measurement and in ascribing causality in any such analysis, the limited sample sizes that result when restricting the analysis to recent years, and the general difficulty of predicting how the parties involved would react to the proposed changes. As discussed in Section III.C.2 above, our analysis includes an examination of two comparison populations of issuers that are not subject to the ICFR auditor attestation requirement but that otherwise have similar responsibilities with respect to ICFR (i.e., non-accelerated filers, other than EGCs, and EGCs), with consideration given to the ways in which these issuers differ from the affected issuers. We first consider possible effects related to the effectiveness of the affected issuers’ ICFR. Because the issuers in our comparison groups are not required to obtain an ICFR auditor attestation, we focus on the findings of SOX Section 404(a) management reports on ICFR, with the caveat that management may not report as many material weaknesses in the absence of an audit of ICFR. The percentage of issuers reporting ineffective ICFR in their management report by issuer type and revenue category for each of the last four years is presented in Table 13.
TABLE 13—PERCENTAGE OF ISSUERS REPORTING INEFFECTIVE ICFR IN MANAGEMENT REPORT 257
Ineffective ICFR year
Accelerated
(ex. EGCs)
(percent)
Non-Accelerated
(ex. EGCs)
(percent)
EGC
(percent)
Revenue <$100M:
2014 .................................................................................................................... 6.0 27.0 43.7 2015 .................................................................................................................... 6.7 26.5 23.8 2016 .................................................................................................................... 9.0 25.9 33.5 2017 .................................................................................................................... 8.4 28.1 36.1 Average/year ...................................................................................................... 7.5 26.9 34.3 Revenue ≥$100M:
2014 .................................................................................................................... 8.6 11.3 5.4 2015 .................................................................................................................... 9.5 10.1 12.1 2016 .................................................................................................................... 8.9 9.0 9.2 2017 .................................................................................................................... 10.1 7.6 10.3 Average/year ...................................................................................................... 9.2 9.5 9.2

257The estimates in this table are based on staff analysis of Ives Group Audit Analytics data. ICFR effectiveness is based on the last amended management report for the fiscal year. Percentages are computed out of all issuers of a given filer type and revenue category with revenue data and a SOX
Section 404(a) management report available in the
Ives Group Audit Analytics database. The accelerated and non-accelerated categories exclude EGCs. See note 116 above for details on the identification of filer type. 258See 2017 SICPG Survey Report, note 210 above, at 6 (finding that 33% of survey respondents with revenues of $75 million or less reported that they manage no more than 100 total controls, as compared to 13% of those with revenues of $76 to $700 million and zero percent of those with revenues greater than $700 million). 259These estimates are based on staff analysis of Ives Group Audit Analytics data. Material weaknesses are considered to be staffing-related if they are categorized in the database as either ‘‘Segregations of duties/design of controls (personnel)’’ or ‘‘Accounting personnel resources, competency/training.’’ In comparison, roughly 70% of the accelerated filers reporting ineffective ICFR in Table 13, whether in the high- or low-revenue category, reported at least one staffing-related material weakness. See also 2018 Audit Analytics Study, note 229 above, at 6 (stating, ‘‘The fact that staffing shortfalls are a pervasive difficulty for many smaller companies explains why the percentage of smaller companies that must disclose ineffective ICFRs maintains a value of 30% or more since 2007,’’ where those companies that provide only a management assessment of ICFR, and not an ICFR auditor attestation, are considered to be ‘‘smaller’’ companies). 260See Ge et al. 2017 Study, note 177 above, at 372 (finding that 62.5% of companies that reported material weaknesses as non-accelerated filers remediate these upon entering accelerated filer status). The 62.5% remediation rate estimated in this study would imply that an additional 15 percentage points of issuers with ineffective ICFR would be expected without the ICFR auditor Continued
TABLE 13—PERCENTAGE OF ISSUERS REPORTING INEFFECTIVE ICFR IN MANAGEMENT REPORT 257—Continued
Ineffective ICFR year
Accelerated
(ex. EGCs)
(percent)
Non-Accelerated
(ex. EGCs)
(percent)
EGC
(percent)
Difference in average/year .......................................................................... ¥1.7 17.4 25.1 Among accelerated filers, the rates of ineffective ICFR are relatively similar for issuers with revenue below $100 million, which would be newly exempted from the ICFR auditor attestation requirement, and those above $100 million. Because all of these issuers are currently subject to the ICFR auditor attestation requirement, we next examine non-accelerated filers (other than EGCs) and EGCs for insight into whether lower revenue issuers may behave differently than others in the absence of such a requirement. When considering these categories of issuers, there is a clear and consistent pattern:
those with low revenues report ineffective ICFR at much higher rates (roughly 15 to 25% higher) than others. Those with higher revenues report ineffective ICFR at rates that are more similar to those for accelerated filers. Because we must rely on disclosed rates of ineffective ICFR, it is difficult to separate the extent to which these rates are affected by the detection and disclosure of material weaknesses in ICFR as opposed to actual underlying material weaknesses in ICFR. As discussed in Section III.C.4.a above, studies have found that audits of ICFR often result in the identification and disclosure of material weaknesses that were not previously identified or whose severity was misclassified in management’s initial assessment. Thus, extending the exemption from the ICFR auditor attestation requirement to the affected issuers may decrease the likelihood that, when these issuers have underlying material weaknesses in ICFR, these material weaknesses are detected and disclosed. It is possible that low-revenue issuers may be less likely than other issuers to fail to detect and disclose material weaknesses in the absence of an ICFR auditor attestation, perhaps because they have less complex financial systems and controls.258 Consistent with this hypothesis, Table 13 demonstrates that the low-revenue issuers that are not subject to the ICFR auditor attestation requirement report relatively high rates of ineffective ICFR. However, it is unclear whether these issuers, if subject to an ICFR auditor attestation requirement, may have been even more likely to uncover material weaknesses. We consider how those affected issuers whose proclivity to detect and disclose underlying material weaknesses in the absence of an ICFR auditor attestation differs from other affected issuers may be differentially affected by the proposed amendments in Section III.C.4.c. below. Regardless of the extent to which the detection of material weaknesses may be improved by an ICFR auditor attestation, the pattern across the comparison populations in Table 13 suggests that, in the absence of an ICFR auditor attestation requirement, lowrevenue issuers are less likely than higher revenue issuers to have effective ICFR in place or to remediate their material weaknesses in ICFR. This may not be surprising, as certain material weaknesses in ICFR may be corrected by, for example, hiring additional staff, which managers of an issuer that is not currently producing much revenue may prefer to defer to a later time. Indeed, about 80 to 85% of the low-revenue issuers reporting ineffective ICFR in the comparison populations in 2017 reported at least one staffing-related material weakness, though these were generally accompanied by other types of material weaknesses.259 As discussed in Section III.C.2, the issuers in the comparison groups may have higher rates of ineffective ICFR than would a group of issuers that is more comparable to the affected issuers in terms of size and maturity. In
addition, besides having low revenues, the issuers in the comparison groups have lower-valued assets and fewer employees than the corresponding accelerated filers, and may therefore be less inclined to expend resources on remediating their ICFR. However, because the rates of ineffective ICFR are similar for the higher revenue issuers of all types in Table 13, but low-revenue issuers that are not subject to the ICFR auditor attestation requirement report ineffective ICFR at much higher rates than the corresponding higher revenue issuers, it is likely that these differences are due at least in part to the nature of low-revenue issuers rather than being driven solely by the differences between the affected issuers and our comparison populations. We therefore expect that extending the exemption from the ICFR auditor attestation requirement, as proposed, may result over time in a lower number of the affected issuers establishing or maintaining effective ICFR. While lowrevenue issuers in the comparison populations report ineffective ICFR at rates that average 15 to 25% percentage points higher than low-revenue accelerated filers, given the differences in the affected issuers versus the comparison populations, we look to the low end of this range and preliminarily estimate that, over time, an additional 15% of the affected issuers may fail to maintain effective ICFR. This estimate is consistent with the estimated effect on ICFR based on a study of issuers transitioning into the ICFR auditor attestation requirement.260 We do not

attestation when 15 times (1/0.625–1) or nine percent of issuers had ineffective ICFR with the ICFR auditor attestation, which is similar to the rate of ineffective ICFR we find for accelerated filers. 261See Audit Analytics, 2017 Financial Restatements: A Seventeen Year Comparison, (May 2018), available at https://www.auditanalytics.com/ blog/2017-financial-restatements-review/, and COSO 2010 Fraud Study, note 253 above. 262The estimates in this table are based on staff analysis of Ives Group Audit Analytics data. Percentages are computed out of all issuers of a given filer type and revenue category with revenue data and a SOX Section 404(a) management report available in the Ives Group Audit Analytics database. The accelerated and non-accelerated categories exclude EGCs. See note 116 above for details on the identification of filer type. 263This result is consistent with the BIO Study, which finds that biotechnology EGCs have a two to three percentage point lower restatement rate than other non-accelerated or accelerated filers and attribute this to their ‘‘absence of product revenue.’’ See BIO Study, note 211 above (finding a 6.20% restatement rate for biotechnology EGCs compared to rates of 7.98% and 9.25% for other nonaccelerated and accelerated filers respectively). expect the full estimated effect to be experienced immediately upon effectiveness of the proposed amendments. Instead, as discussed in detail at the end of this section, we expect a movement towards this higher rate of ineffective ICFR over time as some of the affected issuers make incremental changes in their investment in ICFR and as additional issuers enter the category of affected issuers. We next consider to the extent to which this possible effect might translate into less reliable financial statements. By definition, material weaknesses represent a reasonable possibility that a material misstatement of the issuer’s financial statements will not be prevented or detected on a timely basis, and as discussed above, existing studies have demonstrated that ineffective ICFR is associated with less reliable financial statements. Thus, our estimated increase in the rate of ineffective ICFR likely would translate into a decrease in the reliability of the financial statements of the affected issuers. However, low-revenue issuers could be less susceptible, on average, to at least certain kinds of misstatements. In particular, ten to 20% of restatements and about 60% of the cases of financial disclosure fraud in recent times have been associated with improper revenue recognition,261 which is less of a risk, for example, for issuers that currently have no revenues. We explore this possibility empirically in Table 14, which presents the percentage of issuers in different categories that eventually restated some of the financial statements that they reported for a given year. We consider financial statements associated with years 2014 through 2016, but we note that the restatement rates
that we observe for 2016 are lower than for previous years (and would be even lower for 2017) because of the lag between the initial reporting of financial statements and the detection and filing of restatements for those disclosures.
TABLE 14—PERCENTAGE OF ISSUERS ISSUING RESTATEMENTS BY YEAR OF RESTATED FINANCIALS, BY REVENUE
CATEGORY 262
Restated year
Accelerated
(ex. EGCs)
(percent)
Non-Accelerated
(ex. EGCs)
(percent)
EGC
(percent)
Revenue <$100M:
2014 .................................................................................................................... 6.2 10.3 14.7 2015 .................................................................................................................... 6.9 8.4 10.9 2016 .................................................................................................................... 5.4 5.7 7.9 Average/year ...................................................................................................... 6.2 8.2 11.2 Revenue ≥$100M 2014 .................................................................................................................... 14.1 15.9 29.7 2015 .................................................................................................................... 13.1 10.6 23.1 2016 .................................................................................................................... 8.2 6.1 8.6 Average/year ...................................................................................................... 11.8 10.9 20.5 Difference in average/year .......................................................................... ¥5.6 ¥2.7 ¥9.3
Table 14 demonstrates that issuers
with revenues of less than $100 million have, on average, restatement rates that are three to nine percentage points lower than those for higher revenue issuers.263 This is the case for all three categories of issuers in the table, including the non-accelerated filers (other than EGCs) and EGCs, neither of which is subject to the ICFR auditor attestation requirement. This result is consistent with low-revenue issuers being less likely to make restatements, even (per Table 13) when they experience high rates of ineffective ICFR, perhaps because they are less susceptible to certain kinds of misstatements (such as those related to revenue recognition). As discussed above, observed restatements reflect misstatements that were detected and may only be a subset of actual misstatements. However, because we see the same pattern in each column of Table 14 when moving from low revenue to higher revenue, including for accelerated filers other than EGCs (which have relatively low rates of ineffective ICFR), we preliminarily believe that the lower restatement rates for low-revenue issuers are not driven by a difference in the ability to detect misstatements among these categories of issuers. Despite the lower restatement rates of low-revenue issuers, we expect that the proposed amendments will have some eventual adverse impact on the restatement rates of the affected issuers.
Table 14 demonstrates that, among lowrevenue issuers, the accelerated filers
other than EGCs have a two percent (relative to non-accelerated filers other than EGCs) or five percent (relative to EGCs) lower restatement rate than the issuers in the comparison populations, which are not subject to the ICFR auditor attestation requirement. However, as discussed in Section III.C.2 above, the issuers in the comparison groups may have higher rates of restatement than would a group of issuers that is more comparable to the affected issuers in terms of size and maturity. We therefore look to the low end of this range and preliminarily estimate that, over time, the rate of restatements among the affected issuers may increase by two percentage points. However, given their lower current rates of restatement, even after such an increase the affected issuers may, on average, restate their financial

264We note that an estimate on the high end of the range also would not lead to an estimated eventual restatement rate for the affected issuers that would exceed the estimated average restatement rate of those that would remain accelerated filers. 265See, e.g., Dechow and Schrand 2004 Monograph, note 254 above. 266See Jennifer Francis & Katherine Schipper, Have Financial Statements Lost Their Relevance?, 37(2) J. of Acct. Res. 319 (1999) (‘‘Francis and Schipper 1999 Study’’). 267The reported statistics are adjusted R-squared statistics based on regression analysis by staff using data from the Standard & Poor’s Compustat and Center for Research in Security Prices databases. Market value and financial variables are measured as of the end of the fiscal year. Earnings is income before extraordinary items. Stock return is the 15- month stock return ending three months after fiscal year-end, to account for reporting lags. For stock return regression, earnings are scaled by the lagged market value of equity, and outliers in one percent tails of variable distributions are dropped to reduce noise. See id. for additional details. 268See, e.g., Francis and Schipper 1999 Study. While that study ends in 1994, before our 20-year horizon, the results are similar. For example, for the most recent ten years in that study, the book values of assets and liabilities explain 54 to 70% of the variation in equity market valuation, the book value of equity and earnings explain 63 to 78% of the variation in equity market valuation, and earnings and the change in earnings explain six to 20% of the variation in stock returns. 269We note that there is a relatively small sample of accelerated filers transitioning to non-accelerated filer status because of changes in their public float, as compared to transitions in the other direction, and that such transitions likely represent special circumstances such as underperformance. Therefore, such transitions are not particularly helpful for predicting the outcomes of accelerated filers transitioning to non-accelerated filer status because of the proposed amendments. statements at a rate that is lower than that of issuers that would remain accelerated filers, and that does not exceed that of non-accelerated filers and EGCs with comparable revenues.264 While we anticipate that the frequency of ineffective ICFR and, to a lesser extent, restatements may increase among the affected issuers as a result of the proposed amendments, the economic effects of these changes may be mitigated by another factor that may apply to many of these issuers. In particular, the usefulness of more reliable financial statements is linked to the degree to which they factor into the decisions of investors,265 for example, with respect to these investors’ valuations of issuers.266 The financial statements of many low-revenue issuers may have relatively lower relevance for market performance if, for example, relative to higher revenue issuers, their valuation
hinges more on their future prospects than on their current financial performance. We explore this possibility empirically in Table 15, which uses the methodology applied in previous studies to calculate, for issuers above and below the $100 million revenue threshold, the extent to which the variation in market performance is related to the variation in financial measures.
TABLE 15—PERCENTAGE OF VARIATION IN MARKET PERFORMANCE EXPLAINED BY VARIATION IN FINANCIAL
PERFORMANCE FOR 1998 THROUGH 2017, BY REVENUE CATEGORY 267 Market variable Explanatory variables Revenue <$100MM (percent) Revenue ≥$100M (percent) Market value of equity ...................................................... Book value of assets, book value of liabilities ................. 29.5 62.3 Market value of equity ...................................................... Book value of equity, earnings ........................................ 30.5 70.0 Stock return ....................................................................... Earnings, change in earnings .......................................... 4.6 7.5 For issuers at or above $100 million in revenue, we find that the financial variables used as explanatory variables in Table 15 explain about 60 to 70% of the variation in equity market capitalization and 7.5% of the variation in stock returns. These results are consistent with the findings of previous studies for all issuers.268 In contrast, for issuers with revenues of less than $100 million, we find that these financial variables explain about 30% of the variation in equity market capitalization and just over 4.5% of the variation in stock returns. Importantly, these results show that financial statements are not irrelevant for low-revenue issuers. Thus, the anticipated reduction in the reliability of financial statement for the affected issuers is expected to have some negative implications. However, the lower empirical relevance of financial statements on average for these issuers may partially mitigate the potential adverse effects of the proposed amendments. Finally, we anticipate that the potential adverse effects of the proposed amendments will develop gradually and are likely to be relatively limited in the short term. The preceding discussion is based on the comparison of steady-state differences across issuers in different categories, and represents an analysis of the eventual effects of the proposed amendments. Because the proposed amendments would allow some current accelerated filers to transition to nonaccelerated filer status, some issuers that have already been subject to an audit of ICFR for one or more years may no longer be required to obtain an ICFR auditor attestation. While other issuers will enter into the affected issuers category without having previously obtained an ICFR auditor attestation, and such issuers are likely to represent a larger fraction of the affected issuers over time, initially issuers with experience with ICFR auditor attestations are expected to represent a substantial fraction of the affected issuers. Nevertheless, we recognize that a delay in realizing some of the associated costs from the proposed amendments would not necessarily mitigate their ultimate effects. Newly exempt issuers may have implemented control improvements that would persist regardless of a transition. For example, they may have made investments in systems, procedures, or training that are unlikely to be reversed.
It is difficult to predict the degree of inertia in ICFR and financial reporting in order to gauge how quickly, if at all, issuers that cease audits of ICFR may evolve such that their ICFR and the reliability of their financial statements is more characteristic of exempt issuers.269 The gradual nature of such an evolution, and the associated halo effect of the last disclosed ICFR auditor attestation, may limit the short-term costs of the proposed amendments. In addition, issuers that believe control improvements are valuable for reporting and certifying results would be free to spend the resources saved on the attestations on such improvements. Affected issuers with experience with audits of ICFR may also be more likely to continue to obtain an ICFR auditor attestation on a voluntary basis than other exempt issuers are to begin voluntary audits of ICFR. This may be due to such issuers having already incurred certain start-up costs or facing demand from their current investors to continue to provide ICFR auditor

270See Section III.
271See note 243 above.
272Studies have associated voluntary compliance with the ICFR auditor attestation requirement with decreased cost of capital and value enhancements. See, e.g., Cory Cassell, Linda Myers, & Jian Zhou, The Effect of Voluntary Internal Control Audits on the Cost of Capital, Working Paper (2013) (Cassell et al. 2013 Study), available at http:// papers.ssrn.com/sol3/papers.cfm?abstract_ id=1734300; Todd Kravet, Sarah McVay, & David Weber, Costs and Benefits of Internal Control Audits: Evidence from M&A Transactions, Rev. of Acct. Stud. (forthcoming 2018), available at http:// papers.ssrn.com/sol3/papers.cfm?abstract_ id=2958318; and Carnes et al. 2019 Study, note 194 above. We note that the latter two studies are not able to differentiate between the effects of the ICFR auditor attestation and of management’s assessment of ICFR under SOX Section 404(a). 273See Brown et al. 2016 Study, note 193 above. 274See Cassell et al. 2013 Study. 275There is substantial literature describing the fact that in certain circumstances the incentives of managers are not perfectly aligned with those of shareholders. See, e.g., Michael Jensen & William Meckling, Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure, 3(4) J. of Fin. Econ. 305 (1976). Also, as discussed in Section III.C.4.a above, the ICFR auditor attestation requirement can have important market-level benefits through network and spillover effects that issuers are unlikely to internalize. That is, issuers are likely to balance the issuer-level benefits against the issuer-level costs of voluntary compliance without considering these externalities. 276See 2013 GAO Study, note 115 above. 277See Daniel Cohen, Aiyesha Dey, & Thomas Lys, Real and Accrual-Based Earnings Management in the Pre- and Post-Sarbanes Oxley Periods, 83(3) Acct. Rev. 757 (2008) (finding that an increase in real earnings management partially offset the decrease in accruals-based earnings management that followed SOX). See also Coates and Srinivasan 2014 Study, note 181 above, at 646–647. 278See Sarah Clinton, Arianna Pinello, & Hollis Ashbaugh-Skaife, The Implications of Ineffective Internal Control and SOX 404 Reporting for Financial Analysts,’’ 33(4) J. of Acct. and Pub. Pol’y 303 (2013) (finding that the disclosure of internal control weaknesses are followed by a decline in analyst coverage). attestations. Some issuers in the groups that we use for comparison, which are not subject to an ICFR auditor attestation requirement, voluntarily obtain an ICFR auditor attestation. Thus, the comparisons made above at least partially account for the fact that some issuers may choose to obtain an ICFR auditor attestation even in the absence of a requirement. However, to the extent the rate of voluntary ICFR auditor attestations would be higher amongst the issuers that would be newly exempt from the ICFR auditor attestation requirement than other exempt issuers, the anticipated costs of the proposed amendments in the near
term may be further reduced.
c. Potential Economic Costs of Effects on
ICFR and Reliability of Financial Statements Per the discussion in Section III.C.4.a above, any impact of the proposed amendments on the effectiveness of ICFR and the reliability of financial statements may have issuer-level implications as well as market-level implications. At the issuer level, the potential increase, on average, in the rate of ineffective ICFR and restatements may lead investors to charge a somewhat higher average cost of capital for the affected issuers. An issuer’s cost of capital, or the expected return that investors demand to hold its securities, determines the price at which it can raise funds. Thus, any such increase may be associated with a reduction in capital formation to the extent that it decreases the rate at which the affected issuers raise new capital towards new investments. Further, the affected issuers may also experience reduced operational efficiency because of the reduced reliability of financial information available to management for the purpose of making operating decisions. These potential effects are supported by a number of studies discussed above.270 The potential issuer-level effects on cost of capital and operating performance are difficult to confirm and to quantify for the affected issuers because the existing studies may not be generalizable to the affected issuers and to the current nature of ICFR auditor attestations (after the 2007 change in the ICFR auditing standard, the 2010 change in risk assessment auditing standards, and recent PCAOB inspections focused on these aspects of audits). Further, some of these studies provide mixed evidence, as discussed in Section III.C.4.a above. Moreover, the methods used in previous studies are difficult to apply to a comparable sample of lowrevenue issuers in more recent years because, for example, there would only be a small sample of such issuers that recently switched filing status and because methods of measuring the implied cost of capital are particularly problematic for such issuers.271 The available evidence supports the qualitative, directional effects noted above. However, the previous section demonstrated that the potential increase in material weaknesses in ICFR that we estimate could occur may translate into a more limited effect on the reliability of disclosures, as measured by the rate of restatements, for the affected issuers. Also, based on our analysis, the financial metrics of these issuers have lower explanatory power for investors’ determination of their value than in the case of other issuers. These two factors may mitigate the potential adverse effects on the affected issuers’ cost of capital and operating performance. Importantly, some of the costs of extending the exemption from the ICFR auditor attestation requirement to additional issuers may be further mitigated by the fact that some issuers, even if exempted, may voluntarily choose to bear the costs of obtaining such an attestation.272 Affected issuers that expect a lower
cost of capital with an ICFR auditor attestation, such as those with effective ICFR,273 and particularly those that will be raising new debt or equity capital,274 are more likely to voluntarily obtain an ICFR auditor attestation. We note that lowrevenue issuers have less access to internally-generated capital, as discussed above, so they may be more reliant on external financing for capital. However, it is probably not the case that voluntary compliance with the ICFR auditor attestation requirement would be undertaken in every case in which the total benefits of doing so would exceed the total costs.275 Further, we note that the benefits of voluntary compliance may be partially constrained by a lack of prominent disclosure of such compliance, in that investors may not be able to readily discern which issuers voluntarily comply,276 although we expect that voluntary compliers may be likely to make investors aware of their compliance through other means. Issuers and other market participants may also adapt to the proposed changes in other ways, which may serve to enhance or mitigate the anticipated costs. However, these actions, and therefore their net effects, are difficult to predict. For example, it has been posited that issuers reacted to the requirements of SOX by reducing accruals-based earnings management and, in its stead, making suboptimal business decisions for the purpose of real earnings management.277 It is therefore possible that newly exempt issuers could, to some extent, reduce real earnings management in favor of accruals-based management. Another possibility is that scrutiny from analysts may provide an alternative source of discipline for some of the affected issuers, although there is evidence that analysts may stop covering issuers whose financial statements are deemed to have become less reliable.278 While the preceding analysis considers the average effects across the affected issuers on the effectiveness of ICFR and the reliability of financial statements, the potential issuer-level costs of the proposed extension of the exemption from the ICFR auditor attestation requirement likely vary

279See, e.g., Ashbaugh-Skaife et al. 2009 Study, note 240 above (finding that an unqualified SOX 404 opinion is associated with a 116 basis point decrease in the cost of capital for companies with the characteristics most associated with having ICFR deficiencies, and no significant change for those with characteristics least associated with such deficiencies). See also Ge et al. 2017 Study, note 177 above, at 372 (finding that 90% of issuers with management reports disclosing effective ICFR that then transition to accelerated filer status receive an auditor attestation that also finds no material weaknesses in ICFR). 280See Ge et al. 2017 Study, note 177 above, at 372 (finding that 62.5% of companies that reported material weaknesses as non-accelerated filers remediate these upon entering accelerated filer status). We note that this rate is significantly higher than the remediation rate for non-accelerated filers in general. We estimate that 10%, 11%, and six percent respectively of the non-accelerated filers reporting material weaknesses in ICFR in 2014, 2015, and 2016 that remain non-accelerated filers in the following year report no such weaknesses in the following year. See note 143 above for detail on the data sources and methodologies underlying this estimate. 281 Id. (finding that about ten percent of issuers reporting effective ICFR in their management reports as non-accelerated filers report ineffective ICFR upon entering accelerated filer status). 282See, e.g., Ashbaugh-Skaife et al. 2009 Study, note 240 above (finding that companies that newly disclose material weaknesses in their ICFR have an increase in their cost of capital, but that this increase is lower for companies with the characteristics most associated with having such material weaknesses, at about 50 basis points, and higher for companies without such characteristics, at about 125 basis points). 283See Ge et al. 2017 Study, note 177 above. See also the evidence summarized in Section III.C.4.a. 284The efficient allocation of capital may be further reduced to the extent that the potential cost of capital effects discussed above operate through a reduction in the liquidity of the market for these issuers’ shares, which increases the costs to investors looking to adjust their investments or redeploy their capital. See Diamond and Verrecchia 1991 Study, note 239 above. 285See, e.g., Colleen Boland, Scott Bronson, & Chris Hogan, Accelerated Filing Deadlines, Internal Controls, and Financial Statement Quality: The Case of Originating Misstatements, 29(3) Acct. Horizons 551 (2015) (‘‘Boland et al. 2015 Study’’); and Lisa Bryant-Kutcher, Emma Yan Peng, & David Weber, Regulating the Timing of Disclosure:
Insights from the Acceleration of 10–K Filing Deadlines, 32(6) J. of Acct. and Pub. Pol’y 475- (2013). 286See Joost Impink, Martien Lubberink, & Bart van Praag, Did Accelerated Filing Requirements and SOX Section 404 Affect the Timeliness of 10–K Filings?, 17(2) Rev. of Acct. Stud. 227 (2012) and Eli Bartov & Yaniv Konchitchki, SEC Filings, Regulatory Deadlines, and Capital Market Consequences, 31(4) Acct. Horizons 109 (2017). across different types of affected issuers. In particular, for issuers without (and that continue not to have) underlying material weaknesses in their ICFR, a lack of an auditor attestation may decrease confidence in the effectiveness of their ICFR and therefore increase their cost of capital, particularly for those with characteristics that might otherwise lead the market to believe that they likely have unreported material weaknesses.279 Issuers without underlying material weaknesses in their ICFR are less likely to experience effects on the reliability of their financial statements or operating performance. Among issuers with (or that develop) material weaknesses in ICFR, some may fully detect and disclose these in their SOX Section 404(a) management reports even in the absence of an ICFR auditor attestation requirement. For such issuers, evidence suggests that the removal of the ICFR auditor attestation requirement may reduce the likelihood that they remediate, or the speed with which they remediate, such material weaknesses.280 For these issuers, an exemption from the ICFR auditor attestation requirement may, over time, result in less reliable financial statements, a higher cost of capital, and some operational underperformance. Other issuers with (or that develop) material weaknesses in ICFR may not detect or disclose all of these material weaknesses in the absence of an ICFR auditor attestation requirement. Those that would, however, report ineffective ICFR when subject to the ICFR auditor attestation requirement 281 may have a temporarily reduced cost of capital if exempted from this requirement, particularly if they have characteristics that do not otherwise lead the market to suspect that their ICFR may be ineffective (such as those without past restatements).282 Any such reduced cost of capital for these under-reporters may be temporary, as such issuers may be less likely to remediate underlying material weaknesses in their ICFR and could thus eventually face a higher cost of capital due to less reliable financial statements and could experience negative effects on their operating performance.283 To the extent that the reliability of financial statements is somewhat reduced on average at the issuer level for the affected issuers, the efficient allocation of capital at the market level may be negatively affected given a diminished ability to reliably evaluate different investment alternatives.284 Further, such effects could negatively impact capital formation through a reduction in investor confidence.
Section III.C.4.a provides additional
discussion of these market-level factors. We anticipate that any such marketlevel effects may be limited by the small percentage of the total value of traded securities that is represented by the affected issuers and the size of the expected effect on the reliability of these issuers’ disclosures.
5. Potential Benefits and Costs Related
to Other Aspects of the Proposed Amendments In this section we consider the potential effects of the proposed amendments with regard to other implications of accelerated filer status, specifically with respect to the timing of filing deadlines, certain required disclosures, and the determination of filer status. We also consider below some incremental effects of the proposed amendments to the thresholds for exiting accelerated and large accelerated filer status. a. Filing Deadlines As discussed in Section III.B.1 above, non-accelerated filers are permitted an additional 15 days and five days, respectively, beyond the deadlines that apply to accelerated filers, to file their annual and quarterly reports. Extending these later deadlines to the affected issuers may provide these issuers with additional flexibility in preparing their disclosures, while modestly decreasing the timeliness of the data for investors.
Table 8 in Section III.B.3
demonstrates that while the filing deadlines are not a binding constraint for most accelerated filers, with 64% filing their annual reports over five days early in recent years, some accelerated filers would benefit from an extended deadline. For example, filing Form NT automatically provides a grace period of an additional 15 days to file an annual report, and over the past four years, about five percent of accelerated filers filed their annual reports within this grace period rather than by the original deadline. A further four percent of accelerated filers filed their annual reports after these additional 15 days had passed. Even affected issuers that would otherwise have filed by the accelerated filer deadline may avail themselves of the additional time provided under the proposed amendments to balance other obligations or to prepare higher quality disclosures. The 2003 acceleration of filing deadlines for accelerated filers from 90 to 75 days was associated, at least initially, with a higher rate of restatements for the affected issuers.285 This finding suggests that a later deadline may allow some issuers to provide more reliable financial disclosures. While these issuers could alternatively file Form NT to receive an automatic extension, studies have found that investors interpret such filings as a negative signal, resulting in a negative stock price reaction.286 Issuers may thus prefer to meet the original deadline if possible. On the other hand, allowing the affected issuers to file according to the

287See, e.g., Dan Givoly & Dan Palmon, Timeliness of Annual Earnings Announcements:
Some Empirical Evidence, 57(3) Acct. Rev. 486 (1982). 288See, e.g., Nils Hakansson, Interim Disclosure and Public Forecasts: An Economic Analysis and a Framework for Choice, 52(2) Acct. Rev. 396 (1977) and Baruch Lev, Toward a Theory of Equitable and Efficient Accounting Policy, 63(1) Acct. Rev. 1 (1988). We note that Regulation FD generally prohibits public companies from disclosing nonpublic, material information to selected parties unless the information is distributed to the public first or simultaneously. See 17 CFR 243.100 to 17 CFR 243.103. 289See Jeffrey Doyle & Matthew Magilke, Decision Usefulness and Accelerated Filing Deadlines, 51(3) J. of Acct. Res. 549 (2013). We note that this study found the reverse to be true for large accelerated filers. 290 Id. 291See, e.g., Boland et al. 2015 Study, note 285 above. 292See, e.g., Patricia Dechow, Alastair Lawrence, & James Ryans, SEC Comment Letters and Insider Sales, 91(2) Acct. Rev. 401 (2015) and Lauren Cunningham, Roy Schmardebeck, & Wei Wang, SEC Comment Letters and Bank Lending, Working Paper (2017), available at http://papers.ssrn.com/sol3/ papers.cfm?abstract_id=2727860. 293Based on staff analysis using the Intelligize database, approximately 20 issuers included Item 1B disclosures in Forms 10–K filed in 2017. later non-accelerated filer deadlines may reduce the timeliness and therefore usefulness of the disclosures to investors. Studies have found a reduction in the market reaction to disclosure when the reporting lag between the end of the period in question and the disclosure date is lengthy, as more of the information becomes available through other public channels.287 Researchers have also questioned whether such lags increase information asymmetries, because some investors are more able to access or process information that could provide indirect insight into an issuer’s financial status or performance through alternative channels.288 One study found that the 2003 acceleration of filing deadlines was associated with a decrease in the market reaction to the disclosure of annual reports for accelerated filers.289 Based on this result and supplementary tests regarding the change in disclosure quality and change in timeliness after the acceleration of deadlines, the authors concluded that the negative effect of the shorter deadline on the quality of disclosure appeared to dominate the beneficial effect on the timeliness of the disclosure for these issuers.290 While this finding might not be directly applicable 15 years later, and there is some evidence that some of these effects were temporary,291 in the absence of other evidence we preliminarily expect the net effect of the extended filing deadlines to be beneficial on average but modest overall. b. Disclosures Required of Accelerated Filers Non-accelerated filers are not required to provide disclosure regarding the availability of their filings under Item 101(e)(4) of Regulation S–K. While some investors may benefit from reduced
search costs due to such disclosures, we do not expect that extending the exemption from these disclosures to the affected issuers would have significant economic effects. Non-accelerated filers are not required to provide disclosure required by Item 1B of Form 10–K or Item 4A of Form 20–F about unresolved staff comments on their periodic and/or current reports. Studies have found that the eventual disclosure of staff comments and related correspondence, as well as interim information about these comments before they are made public, are valuerelevant (in that they affect the pricing of securities) for investors.292 While our understanding is that Items 1B and 4A disclosures are relatively uncommon,293 extending the exemption from the requirement to disclose unresolved staff comments to the affected issuers may, in some circumstances, prevent the timely disclosure of value-relevant information to public market investors. Moreover, because Item 1B of Form 10–K and Item 4A of Form 20–F requires unresolved staff comments to be disclosed if they were made not less than 180 days prior to the end of that fiscal year, issuers no longer subject to this disclosure requirement may have a reduced incentive to resolve comments in a timely manner. This could reduce the efficiency of the review process and could increase the number of unresolved staff comments at any given time, and thus also decrease the quality of reporting for the period over which comments continue to be unresolved.
c. Transition Thresholds
The proposed amendments include revisions to the transition thresholds that address when an accelerated filer or large accelerated filer can transition into a different filer status. The proposed amendments would allow accelerated or large accelerated filers to become nonaccelerated filers if they qualify under the SRC revenue test or meet a revised public float transition threshold. An issuer whose revenues previously exceeded the SRC initial revenue threshold of $100 million will not qualify under the SRC revenue test unless its revenues fall below $80 million. The $80 million transition threshold for the SRC revenue test is 80% of the initial threshold of $100 million in revenue. An issuer whose public float previously exceeded the $75 million initial threshold for accelerated filer status would become a nonaccelerated filer if its public float fell below $60 million, or 80% of that initial threshold, as opposed to the current threshold of $50 million. Finally, the proposed amendments also revise the public float transition threshold for exiting large accelerated filer status and becoming an accelerated filer from $500 million to $560 million in public float, or 80% of the $700 million entry threshold, to align with the transition threshold for entering SRC status after having exceeded $700 million in public float. The filer type exit thresholds in Rule 12b–2 are set below the corresponding entry thresholds to provide some stability in issuer classification given normal variation in public float and revenues. The exact placement of these thresholds involves a tradeoff between the degree of volatility in classification versus the extent to which the categories persistently include issuers that are below the initial entry thresholds. Table 16 illustrates this tradeoff using 20 years of data on the evolution of company year-end market capitalizations and revenues. While market capitalization is different from public float, we expect the volatility of these measures to be similar because changes in stock price represent the dominant source of variation in both measures.
TABLE 16—TRANSITIONS IN EQUITY MARKET CAPITALIZATION AND REVENUE LEVEL, 1998 THROUGH 2017 294
Entry threshold
Exit threshold as percentage of entry threshold 60% 70% 80% 90% 100% Percentage of new entrants that exit and re-enter over next two years:

294The estimates in this table are based on staff analysis of data from Compustat. 295This estimate is based on staff analysis of the number of accelerated filers in 2017 with public float of at least $60 million but less than $250 million and prior fiscal year revenues of at least $100 million and that are eligible to be SRCs (i.e., excluding ABS issuers, RICs, BDCs, and subsidiaries of non-SRCs). Revenue data is sourced from XBRL filings, Compustat, and Calcbench. See note 116 above for details on the identification of the population of accelerated filers. We note that the incremental number of affected issuers could be higher than this estimate because there are approximately 230 issuers, the vast majority of which are foreign issuers, for which filer status and/ or public float data are not available (and revenue data is either unavailable or revenues are at least $100 million).
TABLE 16—TRANSITIONS IN EQUITY MARKET CAPITALIZATION AND REVENUE LEVEL, 1998 THROUGH 2017 294—Continued
Entry threshold
Exit threshold as percentage of entry threshold 60% 70% 80% 90% 100% $700M market cap .................................................................................................................... 3.0 3.5 4.7 6.6 9.5 $250 M market cap ................................................................................................................... 3.1 4.0 5.1 6.9 9.1 $75M market cap ...................................................................................................................... 3.1 4.3 5.6 7.1 8.4 $100 M revenue ........................................................................................................................ 0.9 1.1 1.4 2.3 4.5 Percentage of new entrants that do not exit but are below entry threshold for next two years:
$700 M market cap ................................................................................................................... 5.7 3.4 1.6 0.4 0.0 $250 M market cap ................................................................................................................... 4.6 2.8 1.4 0.5 0.0 $75M market cap ...................................................................................................................... 4.0 2.5 1.3 0.5 0.0 $100 M revenue ........................................................................................................................ 3.6 2.8 1.9 0.6 0.0 Consider an entry threshold of $700 million in market capitalization. The first panel of Table 16 demonstrates potential fluctuations in issuer classification based on this entry threshold. A higher exit threshold is associated with more volatility in classification. For example, an exit threshold of $700 million, or 100% of the entry threshold, would have led almost ten percent of the new entrants to exit the following year and then reenter the year after that. Issuers and investors may be confused as a result of such frequent fluctuations in filer type. They may also bear resulting costs, such as (for issuers) the cost of frequently revising disclosure schedules and the scope of auditing contracts and (for investors) any incremental cost of evaluating the reliability of financial disclosures for an issuer that is not consistently subject to the ICFR auditor attestation requirement. The second panel of Table 16 demonstrates the persistence of classification for issuers that drop below the entry threshold. A lower exit threshold is associated with a greater number of issuers remaining in a particular category despite falling below the entry threshold. For example, in the first row of this panel, an exit threshold of $420 million, or 60% of the $700 million entry threshold, would have prevented almost six percent of the new entrants from exiting despite falling below the entry threshold in the next two years. A low exit threshold can thus risk having a filer status effectively apply to a broader group of issuers than intended.
Table 16 demonstrates that the
balance between limiting filer status volatility while enabling filer status mobility provided by an exit threshold of 80% is similar around a $250 million, $75 million, and $700 million market capitalization. We therefore expect the proposed increase in the thresholds to exit accelerated and large accelerated filer status to $60 and $560 million, or 80% of the entry thresholds, to lead to a similar tradeoff in these factors as the 80% public float threshold to re-enter SRC status. Table 16 also demonstrates that revenue is more stable than market capitalization, so the 80% threshold in the revenue test for exiting accelerated and large accelerated filer status is expected to provide a lower degree of filer status fluctuations for a comparable degree of filer status mobility. Overall, we expect the proposed transition thresholds to provide a tradeoff between filer status mobility and volatility that is consistent with the tradeoff provided by the recently revised SRC transition provisions.
6. Alternatives to the Proposed
Amendments
Below we consider the relative costs and benefits of reasonable alternatives to the implementation choices in the proposed amendments. a. Exclude All SRCs From Accelerated Filer Category We have considered excluding all SRCs from the accelerated filer definition, consistent with the past alignment of the SRC and nonaccelerated filer categories. This alternative would include SRCs that meet the revenue test, as proposed, as well as those that have a public float of less than $250 million when initially determining SRC status. Incremental Benefits of Excluding All SRCs From Accelerated Filer Category This alternative would have several benefits, such as promoting regulatory simplicity and reducing any frictions or confusion caused by issuers having to make multiple determinations of their filer type. It would also expand the benefits of the proposed amendments to additional issuers. We estimate that 357 additional issuers 295 would be nonaccelerated filers rather than accelerated filers under this alternative, of which 68 are EGCs and 289 would newly be exempt from the ICFR auditor attestation requirement under SOX
Section 404(b) (although we estimate
that 13 of these newly exempt filers would still be subject to the FDIC auditor attestation requirement). To estimate the benefits to these additional issuers, we begin by considering the audit fees of lower-float issuers of different types, as we did for low-revenue issuers in Table 12 of
Section III.C.3. These results are
presented in Table 17.

296The estimates in this table are based on staff analysis of data from Ives Group Audit Analytics and public float data from XBRL filings. The accelerated and non-accelerated categories exclude EGCs. See note 116 above for details on the identification of filer type. 297The analyses in Table 18 and 19 that follow exclude non-accelerated filers (other than EGCs) because of a lack of higher-float non-accelerated filers and also include, per year, 436 to 583 higherfloat accelerated filers (other than EGCs) and 89 to 135 higher-float EGCs. The sample size varies across years and is based on issuers of a given filer type with public float data available from XBRL filings and a SOX Section 404(a) management report available in the Ives Group Audit Analytics database. See note 116 above for details on the identification of filer type. 298For non-accelerated filers other than EGCs, the average difference is $788,393 minus $365,542, or $422,851, which is about 53.6% of $788,393. For EGCs, the average difference is $788,393 minus $235,307, or $553,086, which is about 70.2% of $788,393. 299 In the case of low-revenue issuers, the assets and employees of the comparison population were about one-third of what they were for the accelerated filers in the analysis, as discussed in
Section IV.C.2 above. In the case of low float
issuers, the assets and employees of the comparison population are about one-fifth of what they were for the accelerated filers in the analysis.
TABLE 17—AVERAGE TOTAL AUDIT FEES IN DOLLARS BY FILER TYPE 296
Issuers with public float <$250 million Accelerated (ex. EGCs) Non-Accelerated (ex. EGCs) EGC 2014 ........................................................................................................................... $750,550 $294,576 $232,006 2015 ........................................................................................................................... 723,337 309,296 239,374 2016 ........................................................................................................................... 837,010 419,357 225,294 2017 ........................................................................................................................... 842,675 438,939 244,554 Average/year .............................................................................................................. 788,393 365,542 235,307 The analysis includes, per year, 551 to 675 lower-float accelerated filers (other than EGCs), 1,537 to 2,784 lower-float non-accelerated filers (other than EGCs), and 163 to 985 lower-float EGCs.297 For these lower-float issuers, the difference between the average annual audit fees for accelerated filers subject to the ICFR auditor attestation requirement and the comparison populations that are exempt from this requirement represents, as a percentage of the total audit fees for accelerated filers, roughly 50 to 70% of those total audit fees.298 This range of percentages is significantly higher than the estimates of the cost of an ICFR auditor attestation from other sources discussed in Section III.C.3.b above. Also, as discussed in Section III.C.2 above, the lower audit fees for the comparison populations may be partially attributable to their smaller size, and the disparity in size in this case is greater than in the analysis of a revenue threshold.299 We therefore select a lower estimate of 40% for the audit fee savings associated with an exemption of these issuers from the ICFR auditor attestation requirement, which is still significantly higher than the 25% we applied for low-revenue issuers and is higher than the five percent to 35% range of estimates from other sources, resulting in an estimate of 40% of $788,393 or about $315,000 in average savings on audit fees under this alternative. Adding this cost savings to our estimate of additional potential compliance cost savings beyond audit fee savings of $100,000 from Section III.C.3.d above, for which the analysis for lower public float issuers would not differ, we estimate an average cost savings of $415,000 for the additional issuers that would be affected under this alternative, with some of these issuers experiencing lesser or greater savings. This represents a significant cost savings for issuers with less than $250 million in public float and may thus have beneficial economic effects on competition and capital formation. As discussed above, smaller issuers generally bear proportionately higher compliance costs than larger issuers. Reducing these
additional issuers’ costs would reduce their overhead expenses and may enhance their ability to compete with larger issuers. To the extent that the cost savings for the additional affected issuers enable capital investments that would not otherwise be made, this alternative would also lead to additional benefits in capital formation. Incremental Costs of Excluding all SRCs From Accelerated Filer Category This alternative could also impose several costs. Overall, we expect costs of this alternative to be greater than for the proposed amendments, primarily due to the broader application of the exemption from the ICFR auditor attestation requirement and the diminished impact of some of the mitigating factors discussed in Section III.C.4 above on SRCs that meet the public float test rather than the revenue test. To explore these potential costs further, we follow the analysis set forth in Section III.C.4 above. We begin by considering the potential impact of an exemption from the ICFR auditor attestation requirement on the effectiveness of ICFR and reliability of financial statements for these issuers.
Table 18 presents our estimates of the
percentage of issuers with public float below $250 million and those with public float of at least $250 million that report ineffective ICFR in their management report in recent years. We compare accelerated filers (other than EGCs) to EGCs because the latter are not currently subject to the ICFR auditor attestation requirement but may have public float that is greater or less than $250 million (while non-accelerated filers are not suitable for this analysis because they would generally not have public float of greater than $250 million). We omit the year 2014 in the second panel because of an insufficient sample of EGCs with public float greater than $250 million in 2014.
TABLE 18—PERCENTAGE OF ISSUERS REPORTING INEFFECTIVE ICFR IN MANAGEMENT REPORT 300
Ineffective ICFR Year
Accelerated
(ex. EGCs)
(percent)
EGC
(percent)
Public Float <$250M:
2014 ...................................................................................................................................................... 9.0 46.6 2015 ...................................................................................................................................................... 9.5 48.0

300The estimates in this table are based on staff analysis of Ives Group Audit Analytics data and public float data from XBRL filings. ICFR effectiveness is based on the last amended management report for the fiscal year. Percentages are computed out of all issuers of a given type and float category with a SOX Section 404(a) management report available in the Ives Group Audit Analytics database. The accelerated category excludes EGCs. 2014 statistics are omitted in this
table, relative to Table 13, because of an insufficient
sample of EGCs with float greater than $250 million in that year. See note 116 above for details on the identification of filer type. 301The estimates in this table are based on staff analysis of Ives Group Audit Analytics data and public float data from XBRL filings. Percentages are computed out of all issuers of a given filer type and float category with a SOX Section 404(a) management report available in the Ives Group Audit Analytics database. The accelerated category excludes EGCs. 2014 statistics are omitted in this
table, relative to Table 14, because of an insufficient
sample of EGCs with float greater than $250 million in that year. See note 116 above for details on the identification of filer type.
TABLE 18—PERCENTAGE OF ISSUERS REPORTING INEFFECTIVE ICFR IN MANAGEMENT REPORT 300—Continued
Ineffective ICFR Year
Accelerated
(ex. EGCs)
(percent)
EGC
(percent)
2016 ...................................................................................................................................................... 10.9 50.0 2017 ...................................................................................................................................................... 10.5 51.8 Average/year ........................................................................................................................................ 10.0 49.1 Public Float ≥$250M:
2015 ...................................................................................................................................................... 7.7 11.2 2016 ...................................................................................................................................................... 6.3 12.6 2017 ...................................................................................................................................................... 8.0 7.6 Average/year ........................................................................................................................................ 7.3 10.5 Difference in average/year ............................................................................................................ 2.7 38.6 As in the case of EGCs and nonaccelerated filers (other than EGCs) with low revenues, as shown in Table 13,
Table 18 demonstrates that EGCs with
lower public float are significantly more likely to report ineffective ICFR than those with higher public float. In comparison, as in the case of our revenue analysis, there is not a distinct pattern in the rate of ineffective ICFR across this public float threshold for accelerated filers. EGCs with lower public float report ineffective ICFR at a rate that is almost 40 percentage points higher than EGCs with higher public float or accelerated filers (other than EGCs) with lower public float. As in our estimation for low-revenue issuers, we acknowledge the potential inflation of these statistics due to the relation between size and age and rates of material weakness. Because we have a single comparison sample in this case, rather than a range of statistics based on two comparison samples as in our analysis based on revenue, we apply a downward adjustment to account for these differences and preliminarily estimate that extending the exemption from the ICFR auditor attestation requirement to issuers that are eligible to be SRCs based on their public float may result in an average increase in the rate of ineffective ICFR of about 25 percentage points among these issuers, somewhat higher than our estimate for low-revenue issuers. We next look to see whether, as with the low-revenue issuers analyzed in Section III.C.4, there are mitigating factors that could limit the potential adverse effects of extending the exemption from the ICFR auditor attestation requirement.
Table 19 presents the rate of
restatements in recent years by issuers in these categories. As in the case of
Table 18, 2014 is excluded in the
second panel due to an insufficient sample size of high float EGCs.
TABLE 19—PERCENTAGE OF ISSUERS ISSUING RESTATEMENTS BY YEAR OF RESTATED FINANCIALS, BY PUBLIC FLOAT
CATEGORY 301
Restated year
Accelerated
(ex. EGCs)
(percent)
EGC
(percent)
Public Float <$250M:
2014 ...................................................................................................................................................... 10.4 17.2 2015 ...................................................................................................................................................... 12.3 16.2 2016 ...................................................................................................................................................... 7.3 8.8 Average/year ........................................................................................................................................ 10.0 14.1 Public Float ≥$250M:
2015 ...................................................................................................................................................... 10.1 16.9 2016 ...................................................................................................................................................... 8.3 11.9 Average/year ........................................................................................................................................ 9.2 14.4 Difference in average/year ............................................................................................................ 0.8 ¥0.3 In this case, the results are distinct from the results in Table 14, which had analyzed the restatement rates for issuers around the $100 million revenue threshold. As shown in Table 14, low revenue issuers restated their financial statements at rates that were three to nine percentage points lower than for higher revenue issuers, whether or not they were subject to the ICFR auditor attestation requirement. In contrast, as shown in Table 19, restatement rates are quite similar above and below a $250 million public float threshold. We therefore believe that the proposition that low-revenue issuers may, on average, be less susceptible to certain kinds of misstatements may not apply to the same extent to issuers with low public float. Because the lower-float EGCs on average restate their financials

302The reported statistics are adjusted R-squared statistics based on regression analysis by staff using data from the Standard & Poor’s Compustat and Center for Research in Security Prices databases. Market value and financial variables are measured as of the end of the fiscal year. Earnings is income before extraordinary items. Stock return is the 15- month stock return ending three months after fiscal year-end, to account for reporting lags. Earnings are scaled by the lagged market value of equity, and outliers in one percent tails of variable distributions are dropped to reduce noise. See Francis and Schipper 1999 Study for additional details. 303These estimates are based on staff analysis of data from the 2008–09 Survey. The analysis considers responses pertaining to the most recent year for which a given respondent provided a response. We note that the rate of responses to the question about net benefits was lower than for other questions. See 2009 SEC Staff Study, note 123 above, and Alexander et al. 2013 Study, note 197 above, for details on the survey and analysis methodology. 304Our staff used market capitalization valuations as of February 2019 to determine the set of potentially affected BDCs. While this methodology is different than the approach used by Rule 12b– 2, which uses the aggregate worldwide market value of the voting and non-voting common equity held by non-affiliates as of the last business day of the issuer’s most recent second fiscal quarter, we do not believe that it would substantially change our analysis. This analysis did not remove BDCs who may qualify as non-accelerated filers based on their status as EGCs. After identifying the set of potentially affected BDCs, our staff manually reviewed the most recent Form 10–K filed on our EDGAR system for each BDC. at a rate about four percentage points higher than that for lower-float accelerated filers (other than EGCs), which is comparable to the five percentage point difference between the corresponding rates for low-revenue EGCs and low-revenue accelerated filers (other than EGCs) in Table 14, we preliminarily estimate that the increase in restatement rates for the additional affected issuers may be comparable to the two percentage points we estimated for low-revenue issuers. However, in contrast to the results for low-revenue issuers, this effect may result in higher restatement rates for the affected issuers than for the higher public float issuers that would remain accelerated filers. To the extent that extending the exemption from the ICFR auditor attestation requirement may reduce the reliability of financial statements for the affected issuers, Table 15 in Section III.C.4 demonstrates that the potential adverse impact of such a change may be mitigated by the lower empirical relevance of financial statements for the market valuation of these issuers. Therefore, we next consider whether a similar proposition could hold for lower public float issuers. In Table 20, we
consider the extent to which the variation in stock returns can be explained by the variation in earnings and changes in earnings for these lower and higher public float issuers over a 20-year horizon. We use market capitalization as a rough proxy for public float, given the limited availability of public float data over the horizon of this analysis. We cannot reliably apply the relevance analysis using market capitalization that we considered in the first two rows of Table 15 in this setting because dividing the sample by the same variable that is being analyzed in a regression analysis like this one generally results in biasing estimates downward (an ‘‘attenuation bias’’), and we are unable to correct for such a bias. However, the analysis below based on stock returns mirrors the analysis in the third row of Table 15 and should not be subject to such a bias.
TABLE 20—PERCENTAGE OF VARIATION IN MARKET PERFORMANCE EXPLAINED BY VARIATION IN FINANCIAL
PERFORMANCE FOR 1998 THROUGH 2017, BY MARKET CAPITALIZATION CATEGORY 302 Market variable Explanatory variables Market Cap <$250M (percent) Market Cap ≥$250M (percent) Stock return .................................................................. Earnings, change in earnings ....................................... 6.7 6.7 We find that the percentage of the variation in returns that is explained by the explanatory financial variables is similar for issuers with market capitalization of less than $250 million as compared to those with higher market capitalization, at about 6.7%. That is, it does not appear that the market relies on financial statements to a lesser extent for the valuation of issuers with public float less than $250 million (as compared to issuers with a larger public float), and so this further mitigating factor that applies to lowrevenue issuers likely does not apply equally to lower public float issuers. Finally, as in Section III.C.3, we reexamine responses to the 2008–2009 Survey. When asked about the net benefits of complying with SOX Section 404, 16% of respondents at accelerated filers with public float of less than $250 million claimed that the costs far outweighed the benefits, in contrast to, as reported above, 30% of respondents at accelerated filers with revenues of less than $100 million.303 While this survey data is somewhat dated, it provides an indication as to the perception by executives at issuers at that time of the relative costs and benefits of the ICFR auditor attestation requirement. To the extent that this perception is borne out by the actual costs and benefits of the ICFR auditor attestation requirement for issuers that meet the SRC revenue test and for those that would otherwise be SRCs under the public float test, this data may suggest that low-revenue issuers would benefit more from qualifying as non-accelerated filers than would other types of SRCs. We are soliciting comment on our analysis of the benefits and costs of extending non-accelerated filer status to all SRCs and whether there are benefits and/or costs of this alternative that we have overlooked. We particularly invite comment on the methodology used to carry out this analysis and any suggestions for alternative or supplemental methodologies to help inform our analysis. b. Include or Exclude Certain Issuer Types Alternatively, we have considered approaches that would include or exclude additional issuer types. For example, we could extend nonaccelerated filer status to other issuers with between $75 million and $700 million in public float that meet the SRC revenue test but would not be eligible to be SRCs due to other reasons. In particular, BDCs and majority-owned subsidiaries of non-SRCs cannot qualify as SRCs and are not otherwise excluded from the ICFR auditor attestation requirement. We estimate that 28 BDCs and one majority-owned subsidiary of a non-SRC parent would meet the same public float and revenue thresholds as the affected
issuers.304 We estimate that 29 BDCs have a market capitalization between $75 million and $700 million, and of these BDCs, 13 have market capitalizations between $250 million and $700 million and the remainder had market capitalizations between $75 million and $250 million. Given the limited number of issuers that are excluded due to their disqualification from SRC status, we preliminarily expect the aggregate incremental costs and benefits of this alternative relative to the proposed approach to be modest,

305 17 CFR 210.6–01 et seq.
306See, e.g., Suraj Srinivasan, Aida Sijamic Wahid, & Gwen Yu, Admitting Mistakes: Home Country Effect on the Reliability of Restatement Reporting, 90(3) Acct. Rev. 1201 (2015). 307See Final Report of the 2017 SEC Government Business Forum on Small Business Capital Formation (Mar. 2018), available at https:// www.sec.gov/files/gbfor36.pdf; and William J. Newell, Presentation at ACSEC Meeting SarbanesOxley Section 404(b): Costs of Compliance and Proposed Reforms, (Sept. 13, 2017), available at https://www.sec.gov/info/smallbus/acsec/williamnewell-acsec091317.pdf. 308See 2013 GAO Study, note 115 above. as compared to the universe of Form 10–K filers, although they could be significant for any particular issuer and significant for BDCs as a class of Form 10–K filers as we estimate the total number of BDC filers to be 50 (of which six have a market capitalization below $75 million and would be already considered non-accelerated filers). Since BDCs do not report revenue on their financial statements, we examined potential alternative metrics to the SRC revenue test threshold of less than $100 million. Of the 29 BDCs with a market capitalization between $75 million and $700 million, our review found that only one BDC reported investment income in excess of $100 million. No BDC reported changes in net realized and unrealized gains and losses or net increase in net assets resulting from operations in amounts greater than $100 million.
TABLE 21—CHARACTERISTICS OF BDCS WITH MARKET CAPITALIZATION BETWEEN $75 AND $700 MILLION
[In millions]
Market capitalization as of February 2019 Investment income for most recent fiscal year Net realized and unrealized gains and losses for most recent fiscal year Net increase in net assets resulting from operations for most recent fiscal year High .......................................................................................... $507.91 $108.28 $43.12 60.69 Low .......................................................................................... 89.69 1.62 (¥123.33) (¥114.28) Average .................................................................................... 255.30 49.37 (¥11.15) 7.70 Median ..................................................................................... 244.72 47.67 (¥4.44) 13.01 We also considered whether to permit BDCs to provide an independent public accountant’s report on internal controls, similar to the one required by RICs on Form N–CEN, since both RICs and BDCs prepare financial statements under
Article 6 of Regulation S–X,305 in place
of the auditor attestation required by SOX Section 404(b). We considered whether such a substitution should be permitted for all BDCs or only those BDCs that would no longer be required to provide a report under SOX Section 404(b) if BDCs were permitted to be a non-accelerated filer based on a test similar to the SRC revenue test. We do not have any data, however, regarding the potential benefits and costs of using a Form N–CEN report on internal controls as compared to the auditor attestation required by SOX Section 404(b). We have also considered excluding FPIs, which are included in the affected issuers to the extent that they meet the required thresholds and other qualifications, from the proposed amendments. Researchers have found that the restatement rates of foreign issuers may be artificially depressed due to a lower likelihood of detection and disclosure of misstatements for these issuers.306 It is therefore possible that encouraging more effective ICFR through an ICFR auditor attestation requirement may be particularly important for such issuers. However, because of limitations in the availability of data such as filing status or public float for many FPIs, we are unable to reliably measure the potential effects for this subset of issuers. Because lowrevenue FPIs may have similar characteristics to low-revenue domestic issuers, including them in the group of affected issuers may help to maintain an even playing field for competition amongst these issuers and avoid discouraging foreign companies from issuing securities in U.S. public markets.
c. Alternative Threshold or Alternative
Metrics
We have considered alternative levels at which a revenue threshold could be set. A $100 million dollar revenue threshold was recommended, in conjunction with a public float threshold, for the accelerated filer definition as well as the SRC definition by the 2017 Small Business Forum and a participant at the September 2017 meeting of the ACSEC.307 The $100 million threshold is also aligned with the SRC revenue test. Empirically, we find no obvious break in the distribution of revenue or in the results of our analysis. In general, lowering the revenue threshold would reduce the expected benefits of the proposed amendments by reducing the number of issuers that would experience cost savings, while also reducing the expected costs of the proposed amendments by reducing the potential adverse impact on the reliability of financial statements. Increasing the threshold would increase the expected benefits while also increasing the expected costs. d. Disclosure While filer status is reported prominently on the cover page of annual reports for most issuers, there is not similarly prominent disclosure of whether an ICFR auditor attestation is provided. In addition to, or in lieu of, the proposed amendments, we could permit or require such disclosure, as recommended by the GAO.308 This would make it easier for investors to identify issuers that undergo a voluntary ICFR auditor attestation with only minimal additional disclosure expense for registrants. This, in turn, may enhance the value to issuers of pursuing an ICFR auditor attestation even when it is not required. While those issuers that voluntarily obtain an ICFR auditor attestation would bear additional costs to do so, we expect they would voluntarily bear these costs only if they believe that the associated issuer-level benefits (e.g., a reduced cost of capital), which could be enhanced by more prominent disclosure, would more than offset those costs. Voluntary compliance with the ICFR auditor attestation requirement by some of the issuers for which this requirement would be eliminated, as discussed above, could mitigate some of the potential negative effects of the proposed amendments. However, we note that investors can already ascertain whether an ICFR auditor attestation is included by searching an issuer’s annual report, and

that including additional items on the annual report cover page could marginally decrease the salience of each item already reported there. D. Request for Comment Throughout this release, we have discussed the anticipated costs and benefits of the proposed amendments. We request and encourage any interested person to submit comments regarding the proposed amendments and all aspects of our analysis of the potential effects of the amendments. We request comment from the point of view of investors, issuers, and other market participants. With regard to any comments, we note that such comments are particularly helpful to us if accompanied by quantified estimates or other detailed analysis and supporting data regarding the issues addressed in those comments. We also are interested in comments on the alternatives presented in this release, in particular, the alternative of extending nonaccelerated filer status to all SRCs, as well as any additional alternatives to the proposed amendments that should be considered.

What are the costs and benefits of
the proposed amendments for investors and issuers? For example, what are the direct costs associated with an ICFR auditor attestation requirement, such as audit fees, as well as indirect costs, such as those related to managerial time and attention, for the group of SRCs that would be exempted from that requirement under the proposed approach? What would be the effects on potential direct and indirect benefits associated with the ICFR auditor attestation requirement for the group of SRCs that would be exempted from that requirement under the proposed approach? Is it possible to relate the benefits to restatement rates or other measures of financial reporting quality for this group? What would be the effect on these issuers’ cost of capital and investor confidence?
For issuers with revenues of less
than $100 million, how do the costs of ICFR auditor attestations compare with the benefits? Do such issuers have simpler financial statements, less variation in their revenue arrangements, fewer revenue-related records to reconcile, or other characteristics that lead to a lower opportunity for misstatements? Or do such issuers have a greater opportunity for errors, perhaps due to staffing constraints or to lower external scrutiny of their disclosures?
Do investors rely to a lesser extent
on the financial statements of issuers with revenues of less than $100 million than on the financial statements of other types of issuers when making investment decisions? Or is the reliability of the financial statements of such issuers particularly important for valuation because of the sensitivity of future projections to current data?
To what extent is the ability of
investors to gauge the reliability of financial statements likely to be affected by the proposed amendments? To what extent is the actual reliability of financial statements likely to be affected by the proposed amendments?
We request comment on our
estimate of the number of affected issuers, our estimates of the internal and external costs of the ICFR auditor attestation requirement, our estimates of the potential changes in the rates of ineffective ICFR and restatements among the affected issuers, and other estimates made in this release. We also request comment on whether there are additional costs and benefits that we can reliably quantify, and request any data that could allow us to make more precise estimates.
We request comment on our
analysis of existing studies. Are there additional considerations or additional studies that we should consider?
We request comment on the
methodologies used to estimate the internal and external costs of the ICFR auditor attestation requirement, to estimate the potential changes in rates of ineffective ICFR and restatements, and to make other estimates in this release. Is our consideration of the experience of issuers that are not currently subject to the ICFR auditor attestation requirement (non-accelerated filers, other than EGCs, and EGCs) in estimating the potential effects on the affected issuers appropriate? Are our estimates and the related adjustments that we make when comparing accelerated filers with issuers that are not currently subject to the ICFR auditor attestation requirement appropriate given the smaller size and lower age of the issuers in our comparison samples? Are there alternative methodologies that we should consider?
We request comment on our
estimate of the average savings on audit fees that would be associated with the proposed amendments. Is our estimate of audit fee savings of about 25% of total audit fees or about $110,000 per year on average across the issuers that would be newly exempt from the ICFR auditor attestation requirement appropriate, too high, or too low? We request specific estimates of fees paid to auditors by issuers to obtain ICFR auditor attestations, separated to the extent possible from other audit costs and accounting for the risk assessment standards that would apply even to a financial statement only audit. We also request specific estimates of other costs associated with obtaining these attestations, such as the hours of managerial and internal staff time spent to facilitate the audit of ICFR. In addition, we request data that would allow us to better understand how all of these costs vary across issuers of different types.
We request statistics on FPIs that
would allow us to better characterize the anticipated effects on these issuers. Do low-revenue FPIs have similar characteristics as low revenue domestic issuers?
We request statistics and analysis
that would allow us to better understand the externalities that the quality of ICFR at one issuer may have on other issuers and on the market as a whole.
Would issuers or auditors take
actions in response to the proposed amendments that would affect the potential economic effects of the proposed amendments? If so, what actions would they take and why? Do issuers currently take actions to stay below the accelerated filer public float threshold? If so, to what extent would such actions be expected to continue or change under the proposed amendments? Is the pricing of auditing services for all issuers likely to change as a result of the proposed amendments? For example, are auditors likely to change the incremental fees they charge for integrated, rather than financial statement only, audits due to the decrease in the number of companies required to obtain an ICFR auditor attestation?
Are there current or developing
auditing practices or technology that may impact the economic effects of the proposed amendments? What are those practices or technology and what effects are they likely to have? For example, are there anticipated effects of the proposed amendments on the cost or quality of substantive testing in the financial statement audit? Are there any effects of automation technology in auditing that we should consider? Overall, how would accounting for such auditing practices or technology change the analysis of the benefits and costs of the proposed amendments and alternatives in this release?
We request comment on our
analysis of the benefits and costs of the alternative of extending non-accelerated filer status to all SRCs, including the quantitative estimates of the number of additional affected issuers, the cost savings, and the potential impact on the rate of ineffective ICFR and restatements

309 44 U.S.C. 3507(d) and 5 CFR 1320.11. 310 17 CFR 249.308a. 311The only proposed revision to this form would be changing filing deadlines, which would neither increase nor decrease the burden hours necessary to prepare the filing because there would be no change to the amount of information required in the filing. 312 17 CFR 240.12b–1 through 240.12b–37. 313Our estimates for Forms 10–K, 20–F, and 40– F take into account the burden that would be incurred by including the proposed disclosure in the applicable annual report. To avoid a PRA inventory reflecting duplicative burdens, we estimate that the proposed disclosure would not impose an incremental burden related to Regulation 12B. 314See Section III.C.1 above. 315See Section I.A above. 316We estimate that the remaining 181 of the 539 affected issuers are EGCs, which are not required to comply with the ICFR auditor attestation requirement under SOX Section 404(b). See Section III.C.1 above. In addition to the 181 EGCs, we estimate that a further 76 of the 539 affected issuers are currently also subject to the FDIC’s auditor attestation requirement. See Section 18A of
Appendix A to FDIC Rule 363. These issuers would
continue to incur burden hours and costs associated with an auditor attestation requirement even if the proposed amendments were adopted. However, the FDIC’s auditor attestation requirement is not part of our rules. For purposes of considering the PRA effects of the proposed amendments, therefore, we have reduced the burden hours and costs for these 76 issuers as we would for the other affected issuers that are not EGCs. for these additional affected issuers. Are there additional benefits and/or costs of this alternative that we have overlooked? What would be the effects of this alternative on efficiency, competition, and capital formation?
14. We request comment on the
alternative of requiring or permitting prominent disclosure of whether an ICFR auditor attestation is provided, either in addition to, or in lieu of, amendments to the accelerated filer and large accelerated filer definitions. For example, what would be the economic effects of requiring issuers to prominently identify whether they voluntarily comply with the ICFR auditor attestation requirement, such as adding a check box to the cover page of appropriate filings? Would such disclosure result in more voluntary compliance with the ICFR auditor attestation requirement? Could prominent disclosure of whether an ICFR auditor attestation is included have the unintended consequence of confusing investors, such as by leading some investors to incorrectly interpret the cover page disclosure as a sign of effective ICFR even if the more detailed disclosure included in the ICFR auditor attestation report shows otherwise?
15. We request comment on
alternative approaches that would include or exclude additional issuer types. For example, what would be the economic effects of allowing BDCs and/ or subsidiaries of non-SRCs, which are excluded from the definition of an SRC, to be non-accelerated filers if they meet the proposed thresholds? What would be the economic effects of excluding FPIs from the proposed changes? What would be the economic effects of using a different threshold or different metric to identify the additional issuers that would become non-accelerated filers? What would be the economic effects of allowing all BDCs that meet the public float and revenue thresholds in the SRC definition, or those criteria with any alternative metric in lieu of annual revenues, to be non-accelerated filers? For BDCs, what would be the benefits and costs to providing an independent public accountant’s report on internal controls required by Form N–CEN as compared to an auditor attestation under SOX Section 404(b)? What would be the economic effects on BDC investors if a Form N–CEN report on internal controls was provided in place of a SOX Section 404(b) attestation? Does it decrease the efficiency of independent auditors to provide different types of internal control audits for RICs and BDCs, even though both types of issuers provide financial reporting under Article 6 of Regulation S–X? Are there other alternatives we should consider?
16. What effect would the proposed
amendments have on competition?
Would the proposed amendments put any issuers at a significant competitive advantage or disadvantage? If so, what changes to the proposed requirements could mitigate any such impact?
17. What effect would the proposed
amendments have on efficiency? How could the proposed amendments be changed to promote any positive effect or to mitigate any negative effect on efficiency?
18. What effect would the proposed
amendments have on capital formation?
Are there any positive or negative effects of the proposed amendments on capital formation that we have overlooked? How could the proposed amendments be changed to better promote capital formation or to mitigate any negative effect on capital formation resulting from the amendments?
IV. Paperwork Reduction Act
A. Summary of the Collections of Information Certain provisions of our rules and forms that would be affected by the proposed amendments contain ‘‘collection of information’’ requirements within the meaning of the Paperwork Reduction Act (‘‘PRA’’). We are submitting the proposal to the Office of Management and Budget (‘‘OMB’’) for review in accordance with the PRA.309 The hours and costs associated with preparing and filing the forms and reports constitute reporting and cost burdens imposed by each collection of information. An agency may not conduct or sponsor, and a person is not required to respond to, a collection of information requirement unless it displays a currently valid OMB control number. Compliance with the information collections is mandatory. Responses to the information collections are not kept confidential and there is no mandatory retention period for the information disclosed. The titles for the affected collections of information are:

‘‘Form 10–K’’ (OMB Control No.
3235–0063);
‘‘Form 10–Q’’ 310 (OMB Control No.
3235–0070); 311
‘‘Form 20–F’’ (OMB Control No.
3235–0288);
‘‘Form 40–F’’ (OMB Control No.
3235–0381); and
‘‘Regulation 12B’’ 312 (OMB Control
No. 3235–0062); 313
The regulation and forms listed above were adopted under the Exchange Act. The regulation and forms set forth the disclosure requirements for periodic reports filed by registrants to help investors make informed investment decisions. A description of the proposed amendments, including the need for the information and its proposed use, as well as a description of the likely respondents, can be found in Section II above, and a discussion of the economic effects of the proposed amendments can be found in Section III above. B. Burden and Cost Estimates Related to the Proposed Amendments We estimate that the proposed amendments would result in approximately 539 additional issuers being classified as non-accelerated filers.314 Accelerated filers are subject to the ICFR auditor attestation requirement and shorter deadlines for filing their Exchange Act periodic reports.315 Additionally, accelerated filers must provide disclosure regarding the availability of their filings and the disclosure required by Item 1B of Form 10–K and Item 4A of Form 20–F about unresolved staff comments on their periodic and/or current reports.

ICFR Auditor Attestation
Requirement
We believe that eliminating the ICFR auditor attestation requirement would reduce the PRA burden for 358 of the 539 affected issuers.316 An ICFR auditor

317See Sections III.C.3 and III.C.5 above. 318Form 40–F does not require disclosure of filer status or public float, which makes it very difficult to determine filer status. So as not to overestimate the burden hour and cost reduction of the proposed amendments, we estimate that only one MJDS issuer that files on Form 40–F would not be subject to the ICFR auditor attestation requirement. 319As discussed in Section III.C.3, above, in deriving this estimate of the reduction in non-audit costs, we have looked to outside vendor and internal labor costs, and not to non-labor costs, because we believe that those non-labor costs (such as software, hardware, and travel costs) are primarily attributable to management’s ICFR responsibilities under SOX Section 404(a) and thus would continue to be incurred. To the extent elimination of the auditor attestation requirement also results in a reduction in management’s time burden, we believe this reduction generally would be captured by the estimated $100,000 reduction, as this amount reflects an overall reduction in nonaudit costs. attestation is required only in annual reports on Forms 10–K, 20–F, and 40– F. Table 22, below, shows the estimated number of affected issuers that are subject to the ICFR auditor attestation requirement that file on each of these forms and the average estimated auditfee and non-audit costs, as described above,317 to comply with the ICFR auditor attestation requirement.
TABLE 22—ESTIMATED ANNUAL COSTS PER ISSUER OF ICFR AUDITOR ATTESTATION REQUIREMENT FOR SPECIFIED
FORMS
Form type
Number of affected issuers Audit-fee costs per issuer Non-audit costs per Issuer Form 10–K ................................................................................................................................... 322 $110,000 $100,000 Form 20–F ................................................................................................................................... 35 110,000 100,000 Form 40–F 318 .............................................................................................................................. 1 110,000 100,000 Because these issuers would no longer be subject to the ICFR auditor attestation requirement under the proposed amendments, they would no longer incur these costs. For purposes of the PRA, this reduction in total burden is to be allocated between a reduction in internal burden hours and a reduction in outside professional costs. Table 23, below, sets forth the percentage estimates we typically use for the burden allocation for each form.
TABLE 23—STANDARD ESTIMATED BURDEN ALLOCATION FOR SPECIFIED FORMS
Form type Internal
(percent)
Outside professionals
(percent)
Form 10–K ........................................................................................................................................................... 75 25 Form 20–F ........................................................................................................................................................... 25 75 Form 40–F ........................................................................................................................................................... 25 75 For the $100,000 reduction in annual non-audit costs,319 we allocate the burden based on the percentages in
Table 23 above. However, we believe
that 100% of the $110,000 annual burden reduction for audit-fee costs related to the ICFR auditor attestation requirement should be ascribed to outside professional costs because that amount is an estimate of fees paid to the independent auditor conducting the ICFR attestation audit. Table 24, below, shows the resulting estimated reduction in cost per issuer associated with outside professionals.
TABLE 24—ESTIMATED REDUCTION IN OUTSIDE PROFESSIONAL COSTS FROM PROPOSED ELIMINATION OF ICFR AUDITOR
ATTESTATION REQUIREMENT
Issuer type
(form used)
Outside professional costs per issuer (Non-audit) Outside professional costs per issuer (audit fees) Total outside professional costs per issuer (non-audit + audit fees) Number of affected issuers Total proposed reduction in outside professional costs Form 10–K ........................................................................... $25,000 $110,000 $135,000 322 $43,470,000 Form 20–F ........................................................................... 75,000 110,000 185,000 35 6,475,000 Form 40–F ........................................................................... 75,000 110,000 185,000 1 $185,000 For PRA purposes, an issuer’s internal burden is estimated in internal burden hours. We are, therefore, converting the internal portions of the non-audit costs to burden hours. These activities would mostly be performed by a number of different employees with different levels of knowledge, expertise, and responsibility. We believe these internal labor costs will be less than the $400 per hour figure we typically use for outside professionals retained by the issuer. Therefore, we use an average rate of $200 per hour to estimate an issuer’s internal non-audit labor costs. Table 25, below, shows the resulting estimated reduction in internal burden hours from the proposed elimination of the ICFR auditor attestation requirement.

320Revisions to Accelerated Filer Definition and Accelerated Deadlines for Filing Periodic Reports, Release No. 33–8644 (Dec. 21, 2005) [70 FR 76634 (Dec. 27, 2005)]. 321We believe that this one-hour reduction will be solely for an issuer’s internal burden hours.
TABLE 25—ESTIMATED REDUCTION IN INTERNAL BURDEN HOURS FROM PROPOSED ELIMINATION OF ICFR AUDITOR
ATTESTATION REQUIREMENT
Issuer type
(form used)
Internal cost per issuer
(non-audit)
Burden hours per issuer
(internal cost/
$200)
Number of affected issuers Total proposed reduction in internal burden hours Form 10–K ....................................................................................................... $75,000 375 322 120,750 Form 20–F ....................................................................................................... 25,000 125 35 4,375 Form 40–F ....................................................................................................... 25,000 125 1 125
2. Filing Deadlines; Disclosure
Regarding Filing Availability and Unresolved Staff Comments As the Commission has recognized previously, changing filing deadlines neither increases nor decreases the burden hours necessary to prepare the filing because there is no change to the amount of information required in the filing.320 Therefore, we do not believe that the proposed change to the filing deadlines would affect an issuer’s burden hours or costs for PRA purposes. We believe that eliminating the requirements to provide disclosure regarding the availability of their filings and the disclosure required by Item 1B of Form 10–K and Item 4A of Form 20– F about unresolved staff comments on their periodic and/or current reports would reduce their burden hours and costs, but we do not expect that reduction to be significant. As opposed to the burden reduction resulting from the elimination of the ICFR auditor attestation requirement, which would apply only to 358 of the 539 total affected issuers that are not EGCs, the burden reduction from eliminating these disclosure requirements would apply to all the 539 affected issuers, including the 181 affected issuers that are EGCs. Of these 181 affected EGC issuers, 160 file annual reports on Form 10–K, 21 file annual reports on Form 20–F, and none file annual reports on Form 40–F. For purposes of the PRA, we estimate the reduction to be approximately one hour for each of the 539 affected issuers.321 That reduction is allocated by form as shown in Table 26, below.
TABLE 26—ESTIMATED REDUCTION IN INTERNAL BURDEN HOURS PER ISSUER FROM PROPOSED ELIMINATION OF
DISCLOSURE REQUIREMENTS REGARDING FILING AVAILABILITY AND UNRESOLVED STAFF COMMENTS Form type Burden hours per issuer Number of affected issuers Proposed reduction in internal burden hours Form 10–K ................................................................................................................................... 1 482 482 Form 20–F ................................................................................................................................... 1 56 56
Form 40–F ................................................................................................................................... 1
3. Total Burden Reduction
Table 27, below, shows the total
estimated reduction in internal burden hours and outside professional costs for all aspects of the proposed amendments.
TABLE 27—REQUESTED PAPERWORK BURDEN UNDER THE PROPOSED AMENDMENTS
Current burden Proposed burden change Current annual responses Current burden hours Current cost burden Proposed change in company hours from auditor attestation Proposed change in company hours from disclosure requirement elimination Proposed total change in company hours Proposed change in professional costs Proposed burden hours for affected responses Proposed cost burden for affected responses (A) (B) (C) (D) (E) (F) = (D) + (E) (G) (H) = (B) + (F) (I) = (C) + (G) 10–K .................................... 8,137 14,217,344 $1,896,280,869 (120,750) (482) (121,232) ($43,470,000) 14,096,112 $1,852,810,869 20–F .................................... 725 480,226 576,270,600 (4,375) (56) (4,431) (6,475,000) 475,795 569,795,600 40–F .................................... 160 14,187 17,025,360 (125) (1) (126) (185,000) 14,187 16,840,360
C. Request for Comment
Pursuant to 44 U.S.C. 3506(c)(2)(B), we request comment in order to:

Evaluate whether the proposed
collections of information are necessary for the proper performance of the functions of the Commission, including whether the information will have practical utility;
Evaluate the accuracy of our
assumptions and estimates of the burden of the proposed collection of information;
Determine whether there are ways
to enhance the quality, utility, and clarity of the information to be collected;
Evaluate whether there are ways to
minimize the burden of the collection of information on those who respond, including through the use of automated collection techniques or other forms of information technology; and

322Public Law 104–121, tit. II, 110 Stat. 857 (1996). 323 5 U.S.C. 601 et seq. 324 5 U.S.C. 553. 325 5 U.S.C. 601(6). 326See 17 CFR 240.0–10(a) under the Exchange Act. 327This estimate is based on staff analysis of issuers, excluding co-registrants, with EDGAR filings of Form 10–K, 20–F and 40–F, or amendments, filed during the calendar year of January 1, 2018 to December 31, 2018. This analysis is based on data from XBRL filings, Compustat, and Ives Group Audit Analytics. 328 17 CFR 270.0–10(a). 329These estimates are based on staff analysis of Morningstar data and data submitted by investment company registrants in forms filed on EDGAR between April 1, 2018 and June 30, 2018.

Evaluate whether the proposed
amendments would have any effects on any other collection of information not previously identified in this section. Any member of the public may direct to us any comments concerning the accuracy of these burden estimates and any suggestions for reducing these burdens. Persons submitting comments on the collection of information requirements should direct their comments to the Office of Management and Budget, Attention: Desk Officer for the U.S. Securities and Exchange Commission, Office of Information and Regulatory Affairs, Washington, DC 20503, and send a copy to, Vanessa A. Countryman, Acting Secretary, U.S. Securities and Exchange Commission, 100 F Street NE, Washington, DC 20549, with reference to File No. S7–06–19. Requests for materials submitted to OMB by the Commission with regard to the collection of information requirements should be in writing, refer to File No. S7–06–19 and be submitted to the U.S. Securities and Exchange Commission, Office of FOIA Services, 100 F Street NE, Washington, DC 20549. OMB is required to make a decision concerning the collection of information requirements between 30 and 60 days after publication of the proposed amendments. Consequently, a comment to OMB is best assured of having its full effect if the OMB receives it within 30 days of publication.
V. Small Business Regulatory
Enforcement Fairness Act
For purposes of the Small Business Regulatory Enforcement Fairness Act of 1996 (‘‘SBREFA’’),322 the Commission must advise OMB as to whether the proposed amendments constitute a ‘‘major’’ rule. Under SBREFA, a rule is considered ‘‘major’’ where, if adopted, it results or is likely to result in:
An annual effect on the economy of
$100 million or more (either in the form of an increase or a decrease);
A major increase in costs or prices
for consumers or individual industries; or
Significant adverse effects on
competition, investment, or innovation. We request comment on whether our proposed amendments would be a ‘‘major rule’’ for purposes of SBREFA. We solicit comment and empirical data on:
The potential effect on the U.S.
economy on an annual basis;
Any potential increase in costs or
prices for consumers or individual industries; and
Any potential effect on competition,
investment, or innovation.
Commenters are requested to provide empirical data and other factual support for their views to the extent possible.
VI. Initial Regulatory Flexibility Act
Analysis
The Regulatory Flexibility Act (‘‘RFA’’) 323 requires the Commission, in promulgating rules under Section 553 of the Administrative Procedure Act,324 to consider the impact of those rules on small entities. The Commission has prepared this Initial Regulatory Flexibility Analysis (‘‘IRFA’’) in accordance with Section 603 of the RFA. It relates to the proposed amendments to the accelerated filer and large accelerated filer definitions in Rule 12b–2 under the Exchange Act. A. Reasons for, and Objectives of, the Proposing Action The purpose of the proposed amendments to the accelerated filer and large accelerated filer definitions in Rule 12b–2 is to promote capital formation by more appropriately tailoring the types of issuers that are included in the category of accelerated filers and revising the transition thresholds for accelerated and large accelerated filers. The reasons for, and objectives of, the proposed amendments are discussed in more detail in Sections I and II above. B. Legal Basis We are proposing the rule and form amendments contained in this release under the authority set forth in Sections 3(b), 12, 13, 15(d) and 23(a) of the Exchange Act, as amended.
C. Small Entities Subject to the
Proposed Rules
The proposed changes would affect some registrants that are small entities. The RFA defines ‘‘small entity’’ to mean ‘‘small business,’’ ‘‘small organization,’’ or ‘‘small governmental jurisdiction.’’ 325 For purposes of the RFA, under our rules, an issuer, other than an investment company, is a ‘‘small business’’ or ‘‘small organization’’ if it had total assets of $5 million or less on the last day of its most recent fiscal year.326 We estimate that there are 1,171 issuers that file with the Commission, other than investment companies, which may be considered small entities and are potentially subject to the proposed amendments.327 Investment companies, which include BDCs, qualify as small entities if, together with other investment companies in the same group of related investment companies, they have net assets of $50 million or less as of the end of their most recent fiscal year.328 Commission staff estimates that, as of June 2018, approximately 19 BDCs are small entities.329 We believe it is likely that virtually all issuers that would be considered small businesses or small organizations, as defined in our rules, are already non-accelerated filers and would continue to be encompassed within that category if the proposed amendments are adopted. To the extent any such issuers are not already nonaccelerated filers, we believe it is likely that the proposed amendments would capture those entities. D. Projected Reporting, Recordkeeping, and Other Compliance Requirements The proposed amendments would reduce the number of accelerated filers, which would reduce the compliance burden for those issuers, some of which may be small entities, because they would no longer have to satisfy the ICFR auditor attestation requirement, comply with accelerated deadlines for filing their Exchange Act periodic reports, provide disclosure regarding the availability of their filings, or provide disclosure required by Item 1B of Form 10–K and Item 4A of Form 20–F about unresolved staff comments on their periodic and/or current reports. Compliance with certain rules affected by the proposed amendments would require the use of professional skills, including accounting and legal skills. The proposed amendments are discussed in detail in Sections I and II above. We discuss the economic effect including the estimated costs and burdens, of the proposed amendments on all registrants, including small entities, in Section III above. E. Duplicative, Overlapping, or Conflicting Federal Rules We believe that the proposed amendments would not duplicate, overlap, or conflict with other federal rules.

F. Significant Alternatives The RFA directs us to consider alternatives that would accomplish our stated objectives, while minimizing any significant adverse effect on small entities. Accordingly, we considered the following alternatives:

Establishing different compliance or
reporting requirements or timetables that take into account the resources available to small entities;
Clarifying, consolidating or
simplifying compliance and reporting requirements for small entities under our rules as revised by the amendments;
Using performance rather than
design standards; and
Exempting small entities from
coverage of all or part of the amendments. We do not believe that establishing different compliance or reporting obligations in conjunction with the proposed amendments is necessary. The proposed amendments would not impose any significant new compliance obligations. In fact, the proposed amendments would reduce the compliance obligations of affected issuers by increasing the number of issuers, including small entities, that are subject to the different, less burdensome, compliance and reporting obligations for non-accelerated filers. Similarly, because the proposed amendments would reduce the burdens for these issuers, we do not believe it is appropriate to exempt small entities from all or part of the proposed amendments. We believe that some of the issuers that would become eligible to be nonaccelerated filers if the proposed amendments are adopted may be smaller entities. Therefore, to the extent that any small entities would become newly eligible for non-accelerated filer status under the proposed amendments, their compliance and reporting requirements would be further simplified. We note in this regard that the Commission’s existing disclosure requirements provide for scaled disclosure requirements and other accommodations for small entities, and the proposed amendments would not alter these existing accommodations. Finally, with respect to the use of performance rather than design standards, because the proposed amendments are not expected to have any significant adverse effect on small entities (and may, in fact, relieve burdens for some such entities), we do not believe it is necessary to use performance standards in connection with this rulemaking. G. Request for Comment We encourage the submission of comments with respect to any aspect of this IRFA. In particular, we request comments regarding:
How the proposed rule and form
amendments can achieve their objective while lowering the burden on small entities;
The number of small entities that
may be affected by the proposed rule and form amendments;
The existence or nature of the
potential effects of the proposed amendments on small entities discussed in the analysis; and
How to quantify the effects of the
proposed amendments.
Commenters are asked to describe the nature of any effect and provide empirical data supporting the extent of that effect. Comments will be considered in the preparation of the Final Regulatory Flexibility Analysis, if the proposed rules are adopted, and will be placed in the same public file as comments on the proposed rules themselves.
VII. Statutory Authority and Text of
Proposed Rule Amendments
The rule amendments described in this release are being proposed pursuant to Sections 3(b), 12, 13, 15(d) and 23(a) of the Exchange Act, as amended. List of Subjects in 17 CFR Part 240 Reporting and recordkeeping requirements, Securities. For the reasons set out in the preamble, the Commission is proposing to amend title 17, chapter II of the Code of Federal Regulations as follows:
PART 240—GENERAL RULES AND
REGULATIONS, SECURITIES
EXCHANGE ACT OF 1934
1. The authority citation for part 240
  continues to read in part as follows:
  Authority: 15 U.S.C. 77c, 77d, 77g, 77j, 77s, 77z–2, 77z–3, 77eee, 77ggg, 77nnn, 77sss, 77ttt, 78c, 78c–3, 78c–5, 78d, 78e, 78f, 78g, 78i, 78j, 78j–1, 78k, 78k–1, 78l, 78m, 78n, 78n–1, 78o, 78o–4, 78o–10, 78p, 78q, 78q–1, 78s, 78u–5, 78w, 78x, 78ll, 78mm, 80a–20, 80a–23, 80a–29, 80a–37, 80b–3, 80b– 4, 80b–11, 7201 et seq.; and 8302; 7 U.S.C. 2(c)(2)(E); 12 U.S.C. 5221(e)(3); 18 U.S.C. 1350; and Pub. L. 111–203, 939A, 124 Stat. 1887 (2010); and secs. 503 and 602, Pub. L. 112–106, 126 Stat. 326 (2012), unless otherwise noted.

1. Amend § 240.12b–2 by, in the
  definition of ‘‘Accelerated filer and large accelerated filer,’’:
a. Removing ‘‘.’’ at the end of
paragraph (1)(iii) and adding in its place ‘‘; and’’;
b. Adding paragraph (1)(iv);
c. Removing ‘‘.’’ at the end of
paragraph (2)(iii) and adding in its place ‘‘; and’’;
d. Adding paragraph (2)(iv); and
e. Revising paragraphs (3)(ii) and
(3)(iii).
The addition and revisions read as follows:
§ 240.12b–2 Definitions.

Accelerated Filer and large accelerated filer— (1) * * * (iv) The issuer is not eligible to use the requirements for smaller reporting companies under the revenue test in paragraph (2) or (3)(iii)(B) of the ‘‘smaller reporting company’’ definition in this section, as applicable. (2) * * * (iv) The issuer is not eligible to use the requirements for smaller reporting companies under the revenue test in paragraph (2) or (3)(iii)(B) of the ‘‘smaller reporting company’’ definition in this section, as applicable. (3) * * * (ii) Once an issuer becomes an accelerated filer, it will remain an accelerated filer unless: the issuer determines, at the end of a fiscal year, that the aggregate worldwide market value of the voting and non-voting common equity held by its non-affiliates was less than $60 million, as of the last business day of the issuer’s most recently completed second fiscal quarter; or it determines that it is eligible to use the requirements for smaller reporting companies under the revenue test in paragraph (2) or (3)(iii)(B) of the ‘‘smaller reporting company’’ definition in this section, as applicable. An issuer that makes either of these determinations becomes a nonaccelerated filer. The issuer will not become an accelerated filer again unless it subsequently meets the conditions in paragraph (1) of this definition. (iii) Once an issuer becomes a large accelerated filer, it will remain a large accelerated filer unless: it determines, at the end of a fiscal year, that the aggregate worldwide market value of the voting and non-voting common equity held by its non-affiliates (‘‘aggregate worldwide market value’’) was less than $560 million, as of the last business day of the issuer’s most recently completed second fiscal quarter or it determines that it is eligible to use the requirements for smaller reporting companies under the revenue test in paragraph (2) or (3)(iii)(B) of the ‘‘smaller reporting company’’ definition in this section, as applicable. If the issuer’s aggregate worldwide market value was $60 million or more, but less than $560

million, as of the last business day of the issuer’s most recently completed second fiscal quarter, and it is not eligible to use the requirements for smaller reporting companies under the revenue test in paragraph (2) or (3)(iii)(B) of the ‘‘smaller reporting company’’ definition in this section, as applicable, it becomes an accelerated filer. If the issuer’s aggregate worldwide market value was less than $60 million, as of the last business day of the issuer’s most recently completed second fiscal quarter, or it is eligible to use the requirements for smaller reporting companies under the revenue test in paragraph (2) or (3)(iii)(B) of the ‘‘smaller reporting company’’ definition in this section, it becomes a nonaccelerated filer. An issuer will not become a large accelerated filer again unless it subsequently meets the conditions in paragraph (2) of this definition.

By the Commission.
May 9, 2019.
Vanessa A. Countryman,
Acting Secretary.
[FR Doc. 2019–09932 Filed 5–28–19; 8:45 am] BILLING CODE 8011–01–P