Removal of References to Credit Ratings From Regulation M

1 17 CFR 242.100 through 242.105. Regulation M is also adopted under the Securities Act of 1933 (‘‘Securities Act’’) and under the Investment Company Act of 1940. 2See Joint Explanatory Statement of the Committee of Conference, Conference Committee Report No. 111–517, to accompany H.R. 4173, 864– 79, 870 (June 29, 2010). 3See Public Law 111–203, sec. 939A(b), 124 Stat. 1376, 1872–90 (2010). Section 939A of the Dodd￾Frank Act also requires the Commission to ‘‘review any regulation issued by [the Commission] that requires the use of an assessment of the credit￾worthiness of a security or money market instrument and any references to or requirements in such regulations regarding credit ratings.’’ Public Law 111–203, sec. 939A(a). The Commission must transmit a report to Congress upon the conclusion of the review required in section 939A(a). Public Law 111–203, sec. 939A(c); see U.S. Securities and Exchange Commission Staff, Report on Review of Reliance on Credit Ratings: As Required by Section 939A(C) of the Dodd-Frank Wall Street Reform and Consumer Protection Act (2011), available at https://www.sec.gov/news/studies/2011/ 939astudy.pdf. Staff reports, Investor Bulletins, and other staff documents (including those cited herein) represent the views of Commission staff and are not a rule, regulation, or statement of the Commission. The Commission has neither approved nor disapproved the content of these documents and, like all staff statements, they have no legal force or effect, do not alter or amend applicable law, and create no new or additional obligations for any person. 4Public Law 111–203, sec. 939A(b). SECURITIES AND EXCHANGE COMMISSION 17 CFR Parts 240 and 242 [Release No. 34–97657; File No. S7–11–22] RIN 3235–AL14 Removal of References to Credit Ratings From Regulation M AGENCY: Securities and Exchange Commission. ACTION: Final rule. SUMMARY: The Securities and Exchange Commission (‘‘Commission’’) is adopting rule amendments to implement section 939A(b) of the Dodd￾Frank Wall Street Reform and Consumer Protection Act of 2010 (‘‘Dodd-Frank Act’’), which requires, among other things, that the Commission remove from its regulations any references to credit ratings and substitute in their place alternative standards of creditworthiness. The amendments remove certain existing rule exceptions that reference credit ratings for nonconvertible debt securities, nonconvertible preferred securities, and asset-backed securities and substitute in their place new exceptions that are based on alternative standards of creditworthiness. These substitutes include exceptions for nonconvertible debt securities and nonconvertible preferred securities (together, ‘‘Nonconvertible Securities’’) of issuers who meet a specified probability of default threshold, as well as exceptions for asset-backed securities that are offered pursuant to an effective shelf registration statement filed on a certain form that is tailored to asset-backed securities offerings. The
Commission is also adopting an amendment to a recordkeeping rule applicable to broker￾dealers in connection with their reliance on an exception involving probability of default determinations. DATES: Effective date: The final rules are effective on August 21, 2023. FOR FURTHER INFORMATION CONTACT:
Jessica Kloss, Attorney-Adviser, Laura Weber, Branch Chief, Josephine Tao, Assistant Director, Office of Trading Practices, or Carol McGee, Associate Director, Office of Derivatives Policy and Trading Practices, at (202) 551– 5777, Division of Trading and Markets, Securities and Exchange Commission, 100 F Street NE, Washington, DC 20549–7010. SUPPLEMENTARY INFORMATION: The Commission is amending the following rules adopted under the Securities Exchange Act of 1934 (‘‘Exchange Act’’):
Commission reference CFR citation Rule 17a–4 ............................... 17 CFR 240.17a–4 Regulation M: 1 Rule 100 ............................... 17 CFR 242.100 Rule 101 ............................... 17 CFR 242.101 Rule 102 ............................... 17 CFR 242.102
Table of Contents
I. Introduction
II. Discussion of the Final Rule Amendments
A. Rule 101(c)(2) of Regulation M:
Implementing Section 939A(b) in Certain Exceptions for Distribution Participants
1. Rule 101(c)(2)(i): Nonconvertible
Securities of Issuers Who Meet a Specified Probability of Default Threshold
2. Rule 101(c)(2)(ii): Asset-Backed
Securities Offered Pursuant to an Effective Shelf Registration Statement Filed on Form SF–3 B. Rule 102(d)(2) of Regulation M:
Implementing Section 939A(b) in Certain Exceptions for Issuers and Selling Security Holders
1. Rule 102(d)(2)(i): Nonconvertible
Securities of Issuers Who Meet a Specified Probability of Default Threshold
2. Rule 102(d)(2)(ii): Asset-Backed
Securities Offered Pursuant to an Effective Shelf Registration Statement Filed on Form SF–3
C. Exchange Act Rule 17a–4(b)(17): Adding
a Record Preservation Requirement for Broker-Dealers in Connection With Probability of Default Determinations
III. Other Issues
IV. Other Matters
V. Economic Analysis
A. Baseline
1. The Investment Grade Fixed Income
Market
2. The Investment Grade Exception
B. Benefits of the Amendments
C. Costs of the Amendments
1. Costs Associated With Obtaining the
Estimate of the Probability of Default
2. Costs Associated With Maintaining
Records Related to the Probability of Default Estimation
3. Costs Associated With Structural Credit
Risk Model Based Probability of Default Being an Imperfect Proxy for Creditworthiness
4. Costs Associated With Asset-Backed
Securities’ Amendments
5. Indirect and Other Costs of the
Amendments
D. Efficiency, Competition, and Capital Formation E. Reasonable Alternatives
1. Alternative Threshold for Probability of
Default
2. Exception Based on Security
Characteristics
3. Exception Based on Issuer
Characteristics
4. Exception Based on Issuer and Issue
Characteristics
5. Elimination of the Investment Grade
Exception From Rule 101
6. Alternative for Asset-Backed Securities
7. Alternatives for Rule 102 Exception
8. Alternative for the Record Preservation
Requirement
VI. Paperwork Reduction Act
A. Respondents
B. Use of Information
C. Collection of Information
1. Burden and Cost Estimates Related to
the Rule 101 Amendments
2. Burden and Cost Estimates Related to
the Rule 17a–4 Amendments D. Collection of Information Is Mandatory E. Confidentiality of Responses to Collection of Information F. Retention Period for Record Preservation Requirement
VII. Regulatory Flexibility Act Certification
Statutory Authority
I. Introduction
To reduce reliance on credit ratings,2
section 939A(b) of the Dodd-Frank Act
requires the Commission, among other things, to ‘‘remove any reference to or requirement of reliance on credit ratings’’ and ‘‘substitute in such regulations such standard of credit￾worthiness’’ as the Commission determines to be appropriate for those regulations.3 In making such a determination, the Commission must seek to establish, to the extent feasible, uniform standards of creditworthiness for use by the Commission, taking into account the entities it regulates and the purposes for which those entities would rely on such standards of creditworthiness.4 Regulation M, which is a set of prophylactic anti-manipulation rules

5See 17 CFR 242.101(c)(2) (‘‘Rule 101(c)(2)’’), 17 CFR 242.102(d)(2) (‘‘Rule 102(d)(2)’’). Both of these rules except Nonconvertible Securities and asset￾backed securities that are rated by at least one nationally recognized statistical rating organization, as that term is used in 17 CFR 240.15c3–1 (‘‘Rule 15c3–1’’), in one of its generic rating categories that signifies investment grade. Throughout this release, each exception in Rule 101(c)(2) or Rule 102(d)(2) that references credit ratings is referred to as an ‘‘Investment Grade Exception,’’ and, together, those exceptions are referred to as the ‘‘Investment Grade Exceptions,’’ as applicable. 6See 17 CFR 242.100(b) (‘‘Rule 100(b)’’) (defining ‘‘distribution’’ as ‘‘an offering of securities, whether or not subject to registration under the Securities Act, that is distinguished from ordinary trading transactions by the magnitude of the offering and the presence of special selling efforts and selling methods’’). 7See 17 CFR 242.100(b) (defining ‘‘covered security’’ as any security that is the subject of a distribution or any reference security, and ‘‘reference security’’ as a security into which a security that is the subject of a distribution may be converted, exchanged, or exercised or which, under the terms of the subject security, may in whole or in significant part determine the value of the subject security’’). 8See 17 CFR 242.100(b) (defining ‘‘distribution participant’’ as any ‘‘underwriter, prospective underwriter, broker, dealer, or other person who has agreed to participate or is participating in a distribution’’). 9See 17 CFR 242.100(b) (defining ‘‘covered security’’ as any security that is the subject of a distribution or any reference security, and ‘‘reference security’’ as a security into which a security that is the subject of a distribution may be converted, exchanged, or exercised or which, under the terms of the subject security, may in whole or in significant part determine the value of the subject security). 10See 17 CFR 242.100(b). 11Anti-Manipulation Rules Concerning Securities Offerings, Release No. 34–38067 (Dec. 20, 1996) [62 FR 520 (Jan. 3, 1997)] (‘‘Regulation M Adopting Release’’), 62 FR 521. Rule 101’s prohibitions apply to distribution participants and their affiliated purchasers, while Rule 102’s prohibitions apply to issuers, selling security holders, and their affiliated purchasers. 12See Trading Practices Rules Concerning Securities Offerings, Release No. 34–37094 (Apr. 11,
1996) [61 FR 17108 (Apr. 18, 1996)] (‘‘Regulation
M Proposing Release’’), 61 FR17111, 17120. 13See Regulation M Adopting Release, 62 FR 527; see also infra note 38 (discussing how the ability to substitute similar securities in the market for the security in distribution limits the potential impact a covered person might attempt to exert on the market and distribution of such security). The Investment Grade Exceptions trace back to a 1975 Commission staff no-action position regarding Exchange Act Rule 10b–6, the predecessor to Rules 101 and 102 of Regulation M. See Letter from Robert C. Lewis, Assoc. Dir., Div. Mkt. Reg., SEC, to Donald M. Feuerstein, Gen. Partner & Counsel, Salomon Bros. (Mar. 4, 1975) (emphasizing the following representations from the lead underwriter-requestor in taking its position: (1) ‘‘because the non-convertible bonds of particular issuers are not considered unique and because of the concept of relative value, it is simply not possible to manipulate the price of a corporate bond that has broad investor interest,’’ and (2) purchasing activities in such securities generally are ‘‘unlikely to materially affect the price of [a nonconvertible debt security being offered] because of the availability of large amounts of securities of other issuers which have comparable quality yield [spreads]’’). For a further discussion of the history of the Investment Grade Exceptions, see Removal of References to Credit Ratings From Regulation M, Release No. 34–94499 (Mar. 23, 2022) [87 FR 18312 (Mar. 30, 2022)] (‘‘Proposal’’), 87 FR 18315. 14Regulation M Proposing Release, 61 FR 17112. 15See Proposal, 87 FR 18316–24. The Commission previously proposed two alternatives to the Investment Grade Exceptions. See Removal of Certain References to Credit Ratings Under the Securities Exchange Act of 1934, Release No. 34– 64352 (Apr. 27, 2011) [76 FR 26550 (May 6, 2011)]; References to Ratings of Nationally Recognized Statistical Rating Organizations, Release No. 34– 58070 (July 1, 2008) [73 FR 40088 (July 11, 2008)] (‘‘2008 Proposing Release’’), 73 FR 40095–97. The Commission did not adopt any rule amendments with regard to the Investment Grade Exceptions based on either of these proposals. 16See Proposal, 87 FR 18317–19. 17See Proposal, 87 FR 18321–22. 18See Proposal, 87 FR 18323–24. 19See Proposal, 87 FR 18324–25. 20Comments received in response to the Proposal are contained in File No. S7–11–2022, available at https://www.sec.gov/comments/s7-11-22/ s71122.htm. 21See, e.g., Letter from Chris Carr (May 19, 2022); Letter from Daniel Kuo (May 19, 2022); Letter from Fred Carter (May 19, 2022); Letter from Biren Patel (May 19, 2022); Letter from Robert Tso (May 23, 2022); Letter from Stephen W. Hall, Legal Dir. & Secs. Specialist, Better Mkts, Inc., to Vanessa A. Countryman, Sec’y, SEC (May 23, 2022) (‘‘Better Markets Letter’’), at 3. 22See, e.g., Letter from Alexandra Merz (May 18,
2022) (‘‘Merz Letter’’); Letter from Gerhard Krohmer
(May 19, 2022); Andriy Granovsky (May 19, 2022); Letter from Jason Smith (May 20, 2022); Letter from Craig Faison (May 20, 2022); Letter from Jaymin Patel (May 20, 2022); Letter from Paul K. Sacco (May 21, 2022); Letter from David Navari (May 22,
2022) (‘‘Navari Letter’’); Letter from Jim Protsenko
(May 24, 2022); Letter from John Hall (May 26, 2022); Letter from Andrew Macafee (May 30, 2022). The Commission also received two anonymous comments on May 19, 2022, both of which stated that credit rating agencies ‘‘have become obsolete.’’ 23See, e.g., Merz Letter; Letter from Robert Long (May 19, 2022); Letter from James R. Brown (May 19, 2022); see also Letter from William Desavigny (May 19, 2022); Letter from Kevin Price (May 19, 2022); Letter from Jason MacKenzie (May 19, 2022); Letter from James Zarbock (May 19, 2022); Letter from Carsten Hensch (May 19, 2022); Letter from Thomas Sutton (May 19, 2022); Letter from Harold VanPatten (May 19, 2022); Letter from Aaron Grimshaw (May 19, 2022); Letter from Andre M (May 19, 2022); Letter from Andrew Oshea (May 19, 2022); Letter from Steven Calvino (May 19, 2022); Continued that is designed to preserve the integrity of the securities trading markets as independent pricing mechanisms by prohibiting activities that could artificially influence the market for an offered security, contains references to credit ratings in identical exceptions under 17 CFR 242.101 (‘‘Rule 101’’) and
242.102 (‘‘Rule 102’’) for investment
grade Nonconvertible Securities and asset-backed securities.5 The Investment Grade Exceptions are two of several exceptions to Rule 101’s and Rule 102’s general prohibitions: in connection with a distribution 6 of covered securities,7 distribution participants,8 issuers, selling security holders, and their affiliated purchasers are prohibited from, directly or indirectly, bidding for, purchasing, or attempting to induce any person to bid for or purchase, a covered security 9 during the applicable ‘‘restricted period.’’ 10 These prohibitions exist to protect the integrity of the offering process by precluding activities that could artificially influence the market for the offered security.11 In adopting the Investment Grade Exceptions, the Commission stated that certain securities and activities should be excepted to allow for activities that are necessary for the distribution to occur; to limit adverse effects to the trading market that could result from these prohibitions absent such exceptions; and to permit conduct that is not likely to have a manipulative impact.12 The Investment Grade Exceptions were premised on the principle that investment grade Nonconvertible Securities and asset￾backed securities are less likely to be subject to the type of manipulation that Regulation M seeks to address because they are largely fungible and trade primarily on the basis of yield and creditworthiness (traditionally measured by credit ratings),13 rather than the identity of the particular issuer.14 In accordance with section 939A(b)’s requirements, in 2022, the Commission proposed rule amendments to remove the Investment Grade Exceptions and substitute them with new exceptions that are based on alternative standards of creditworthiness.15 The Commission proposed to except from Rule 101 (1) Nonconvertible Securities of issuers who meet a specified probability of default threshold,16 and (2) asset-backed securities that are offered pursuant to an effective shelf registration statement filed on Form SF–3.17 The Commission proposed to eliminate, without replacing, the Investment Grade Exception from Rule 102.18 The Commission also proposed to amend 17 CFR 240.17a (‘‘Rule 17a–4’’), specifically paragraph (b) of Rule 17a– 4 (‘‘Rule 17a–4(b)’’), to require broker￾dealers to preserve written probability of default determinations pursuant to Rule 101.19 The Commission received comments from an industry group, a data provider, nonprofit organizations, and individuals.20 Commenters broadly recognized and acknowledged the objectives of the Proposal. Some commenters, including individual commenters, provided general support for the Proposal 21 and stated that reliance on credit ratings is outdated 22 and can be harmful to investors or the markets.23 Another commenter

Letter from Dennis Smith (May 19, 2022); Letter from Devin Dasbach (May 19, 2022); Letter from Mark A. Fritzke (May 19, 2022); Letter from Cameron Beebe (May 19, 2022); Letter from Nick Parasiris (May 19, 2022). 24See Letter from Jacob Rajan (May 19, 2022). 25See, e.g., Letter from Joseph Corcoran, Managing Dir. & Assoc. Gen. Counsel, Secs. Indus. & Fin. Mkts. Ass’n, to Vanessa Countryman, Sec’y, SEC (May 23, 2022) (‘‘SIFMA Letter 1’’); Better Markets Letter. 26See 17 CFR 242.101(c)(2)(i), as amended, 242.102(d)(2)(i), as amended (requiring, for reliance by issuers, selling security holders, and their affiliated purchasers, that the distribution participant acting as the lead manager (or in a similar capacity) of a distribution have made the probability of default determination, as applicable to the subject security, pursuant to Rule 101(c)(2)(i), as amended). 27See infra Part II.A.1. 28See infra Part II.A.1. 29See infra note 126. 30See infra Part II.A.1. 31See 17 CFR 242.100(b). 32See 17 CFR 240.17a–4(b)(17), as amended. 33See 17 CFR 242.101(c)(2)(ii), as amended, 242.102(d)(2)(ii), as amended. 34Regulation M Adopting Release, 62 FR 524. 35See Regulation M Adopting Release, 62 FR 524. To illustrate with a simple example, absent an exception, a broker-dealer who is participating in a distribution of XYZ Corp.’s 3% bonds maturing 12/ 31/2029 would be prohibited from making a market in bonds with those terms prior to completing the distribution. The broker-dealer would not, however, be prohibited from making a market in XYZ Corp.’s 3% bonds maturing 12/31/2030 because the date of maturity, a term of the bond, is different from that of the security in distribution. 36See supra note 15. 37Proposal, 61 FR 18316. In addition, the Commission also stated in the Proposal that another example provided by a commenter is a ‘‘best￾efforts’’ offering. Proposal, 61 FR 18316. One commenter on the Proposal stated that firms rely on the Investment Grade Exceptions in the context of ‘‘sticky deals’’ and ‘‘re-openings’’ of debt issuances. See SIFMA Letter 1, at 4. As discussed below, in
Part V.E.5, any offering can become a sticky
offering. In such case, it may become challenging for the issue to trade based solely on its yield and maturity, notwithstanding the issuer’s creditworthiness. Therefore, a sticky offering does not necessarily indicate a lack of creditworthiness on the part of the issuer. In the Proposal, the Commission asked if sticky offerings of creditworthy issuers disprove the underlying premise for excepting certain Nonconvertible Securities. See Proposal, 87 FR 18320. The Commission also asked if the Investment Grade Exception should be removed from Rule 101, without a replacement, because whether an offering will become sticky is unknown at the beginning of the Regulation M restricted period. See Proposal, 87 FR 18320. One commenter stated that it is unaware of any manipulative issues associated with reliance on the Investment Grade Exceptions in connection with sticky offerings. See SIFMA Letter 1, at 2. supported the Proposal and stated that its adoption will lead to increased market competition.24 Some commenters opposed or expressed concerns about the Proposal, and offered certain recommendations with regard to particular aspects of the proposed rule amendments,25 which are addressed below, in Parts II.A through
C. After reviewing and carefully
considering the public comments and recommendations, and in accordance with the requirements of section 939A(b), the Commission is adopting final rule amendments, with targeted modifications to address comments received and to streamline and clarify the rule text from the Proposal. As discussed below in Parts II.A and II.B, the Commission believes that its original basis for excepting investment grade Nonconvertible Securities and asset-backed securities from Rules 101 and 102 continues to apply to the securities that are captured by the amendments’ substitute standards of creditworthiness.
II. Discussion of the Final Rule
Amendments
The amendments remove the existing Investment Grade Exceptions from both Rule 101 and Rule 102 of Regulation M. For distributions of Nonconvertible Securities, the Commission is adopting two new exceptions—one in 17 CFR 242.101(c)(2)(i) (‘‘Rule 101(c)(2)(i)’’), for reliance by distribution participants and their affiliated purchasers, and one in 17 CFR 242.102(d)(2)(i) (‘‘Rule 102(d)(2)(i)’’), for reliance by issuers, selling security holders, and their affiliated purchasers. Both exceptions are based on the requirements more fully described in Parts II.A.1 and B.1 that relate to the determination of an issuer’s probability of default as derived from a structural credit risk model.26 As discussed below, in Part II.A.1, final Rule 101(c)(c)(i) differs from the Proposal with regard to the exception’s conditions involving who is eligible to make probability of default determinations pursuant to Rule 101(c)(2)(i) and when such probability of default determination must be made to rely on the exception. While the Proposal would have allowed any distribution participant to make the probability of default determination in meeting the conditions of Rule 101(c)(2)(i), the final amendments require the probability of default determination to be made by the distribution participant acting as the lead manager (or in a similar capacity) of a distribution.27 In addition, final Rule 101(c)(2)(i) will require five additional business days before the price determination date from what was proposed to make the probability of default determination in satisfying the exception’s conditions.28 Finally, the Commission is making some technical, non-substantive changes from the Proposal with regard to the wording of the standard,29 as well as some clarifying changes to the proposed definition of ‘‘structural credit risk model.’’ 30 Because the term ‘‘structural credit risk model’’ is used identically in both Rule 101(c)(2)(i) and Rule 102(d)(2)(i), as amended, the Commission is adding a definition for the term ‘‘structural credit risk model’’ in Rule 100(b) of Regulation M.31 In addition, the Commission is adopting new paragraph (b)(17) of Rule 17a–4 (‘‘Rule 17a– 4(b)(17)’’) requiring the preservation of the written probability of default determination, relied upon by a broker￾dealer, pursuant to new Rule 101(c)(2)(i) or new Rule 102(d)(2)(i), as applicable, to facilitate Commission staff examinations.32 For distributions of asset-backed securities, the Commission is adopting identical, new exceptions—one in 17 CFR 242.101(c)(2)(ii) (‘‘Rule 101(c)(2)(ii)’’), for reliance by distribution participants and their affiliated purchasers, and one in 17 CFR 242.102(d)(2)(ii) (‘‘Rule 102(d)(2)(ii)’’), for reliance by issuers, selling security holders, and their affiliated purchasers—requiring that such securities be offered pursuant to an effective shelf registration statement filed on Form SF–3.33 A. Rule 101(c)(2) of Regulation M:
Implementing Section 939A(b) in Certain Exceptions for Distribution Participants The application of Rule 101’s prohibitions to distributions of Nonconvertible Securities and asset￾backed securities generally is limited because, under Regulation M, bids for and purchases of outstanding Nonconvertible Securities are not restricted unless the security being purchased is identical in all of its terms to the security being distributed.34 For example, Rule 101’s restrictions do not apply for a security if there is a single basis point difference in coupon rates or a single day’s difference in maturity dates from the security in distribution.35 In addition, as stated in the Proposal, commenters on the Commission’s previously proposed alternatives to the Investment Grade Exception 36 stated that reliance on the Investment Grade Exceptions largely is limited to two situations: re-openings (e.g., when an issuer may want to make a series of offerings of its fixed-income securities via a re-opening to match its funding needs or the desires of its target investor class, or when a foreign sovereign issuer may conduct a re-opening for public financing purposes) and sticky offerings.37 The securities that meet the

However, no commenter suggested that the Commission should remove the Investment Grade Exception from Rule 101, without a replacement. 38See Regulation M Adopting Release, 62 FR 527; see also Regulation M Proposing Release, 61 FR
17112. For purposes of Regulation M, securities of
issuers of a certain credit quality trade on the basis of their yield and creditworthiness (traditionally measured by credit ratings) and are less susceptible to manipulation because other similar Nonconvertible Securities or asset-backed securities are available to investors as an alternative. If the pricing of an offering is inconsistent with pricing in the overall secondary market for similar Nonconvertible Securities or asset-backed securities, an investor may purchase alternative securities that have a better yield, yet are of comparable creditworthiness, in relation to the security being distributed. Accordingly, the ability to substitute similar Nonconvertible Securities or asset-backed securities for the security in distribution limits the ability of a distribution participant to impact the market and distribution of such security. 39Proposal, 87 FR 18338. The Commission stated that, as discussed in that release, based on an analysis of the probability of default and investment grade ratings of a sample of Nonconvertible Securities available on the market as of Oct. 22, 2021, this was an appropriate substitute standard of creditworthiness in place of the reference to credit ratings in the Investment Grade Exception for Nonconvertible Securities. See Proposal, 87 FR 18318. 40Proposal, 87 FR 18338. 41See generally Proposal, 87 FR 18316–17 (providing a history and overview of structural credit risk models). Generally, these models assume that owners of a company’s equity will continue to pay the company’s liabilities if the company’s value exceeds its liabilities. Equivalently, if the equity owners were considered to own a call option on the value of the company with a strike price equivalent to the liabilities owed, the equity owners would exercise the call on the value of the company. If, however, the company’s liabilities exceed the company’s value, the models assume that the equity owners will choose to default on the company’s liabilities, or equivalently, the equity owners would not exercise the call on the value of the company. Accordingly, these structural credit risk models provide a method, based on the Black-Scholes option pricing model, to estimate the probability that a company might default on its liabilities. See Proposal, 87 FR 18317. 42The Default Point frequently is calculated as all short-term liabilities plus half of the long-term liabilities. See Mario Bondioli et al., The Bloomberg Corporate Default Risk Model (DRSK) for Public Firms (Mar. 2021), available at https://ssrn.com/ abstract=3911300. 43These models calculate the probability of default based on inputs from an issuer’s balance sheet. Transactions in equity securities frequently are used as a proxy to determine the value of the firm and the overall volatility of the issuer’s assets. Even the absence of a market for an issuer’s equities alone does not preclude the ability of a distribution participant to use certain structural credit risk models because the
issuer’s balance sheet will include the liabilities, assets, and equity, which, with further analysis, can be used to determine the inputs for the models. Distribution participants, based on their activities as an underwriter, broker￾dealer, or other person who has agreed to participate in a distribution, can access an issuer’s balance sheet to calculate the issuer’s probability of default. 44Letter from Gregory Babyak, Global Head of Reg. Affairs, Bloomberg L.P., to Vanessa A. Countryman, Sec’y, SEC (May 23, 2022) (‘‘Bloomberg L.P. Letter’’), at 1; Letter from Robert E. Bishop, Fellow, Ctr. Law & Bus., UC Berkeley School of Law, & Frank Partnoy, Adrian A. Kragen Professor of Law, UC Berkeley School of Law, to Vanessa Countryman, Sec’y, SEC (May 23, 2022) (‘‘IILF Letter’’), at 2; Better Markets Letter, at 2. 45Bloomberg L.P. Letter, at 1. 46 IILF Letter, at 2. 47 IILF Letter, at 5. 48 IILF Letter, at 6. 49Proposal, 87 FR 18320. Such a requirement could present operational challenges in connection with deriving an issuer’s probability of default from Continued requirements of Rule 101’s Investment Grade Exception are less likely to be subject to the type of manipulation that Rule 101 seeks to prevent because these securities trade on the basis of their yield and creditworthiness (traditionally measured by credit ratings), rather than the identity of the particular issuer, and are largely fungible.38

1. Rule 101(c)(2)(i): Nonconvertible
   Securities of Issuers Who Meet a Specified Probability of Default Threshold The Commission proposed to except the Nonconvertible Securities of issuers for which the probability of default, estimated as of the day of the determination of the offering pricing and over the horizon of 12 calendar months from such day, is less than 0.055%, as determined and documented in writing by the distribution participant as derived from a structural credit risk model.39 The Commission included a definition for the term ‘‘structural credit risk model’’ as a proviso in proposed Rule 101(c)(2)(i) to mean ‘‘any commercially or publicly available model that calculates the probability that the value of the issuer may fall below a threshold based on an issuer’s balance sheet.’’ 40 Accordingly, as proposed, a distribution participant’s (or its affiliated purchaser’s) reliance on proposed Rule 101(c)(2)(i) would have been conditioned on a probability of default determination that was made by use of any commercially or publicly available model that calculates the probability that the value of the issuer may fall below a threshold based on an issuer’s balance sheet. As discussed in the Proposal, since 1974, structural credit risk models, such as the model first proposed by Robert C. Merton and its successor models, have become widely relied upon to determine the probability that an issuer will default on its loan obligations.41 Many commercial data providers, as part of software suites that allow users to analyze securities, employ certain structural credit risk models that are based on the Black-Scholes option pricing model as a way to measure the creditworthiness of companies. These types of structural credit risk models typically use measures from company accounting statements and company￾specific and aggregate market prices and require input variables to calculate an estimated probability of default for a specified horizon, including the market value and volatility of the assets, as well as assumptions regarding the threshold for company asset values, below which the equity owner would default on its obligations (‘‘Default Point’’).42 In addition, these structural credit risk models provide the probability that a company’s assets will fall below the Default Point at or by the expiration of a defined period. Generally, the following variables are needed to derive an issuer’s probability of default from a typical structural credit risk model: (1) the issuer’s value, which can be based on observed market prices of an issuer’s equity security or estimated based on an issuer’s balance sheet; (2) the volatility of the issuer’s equity or assets, which also can be based on market observations or estimated based on an issuer’s balance sheet; (3) the risk-free rate; (4) a time horizon; and (5) the Default Point. A structural credit risk model’s application may be limited in the absence of a market for an issuer’s equity securities if the market price of the issuer’s assets,
   which, as discussed above, is required to calculate the probability of default, is difficult to determine.43 Some commenters supported the proposed exception for Nonconvertible Securities that is based on an issuer’s probability of default.44 One commenter stated that the proposed application of a structural credit risk model requirement will provide additional transparency for investors and other market participants.45 Another commenter agreed with the Commission that the estimated probability of default of a debt security ‘‘is and should be a central component of the analysis of the credit risk’’ 46 and that the ‘‘expected probability of default can be independently determined by structural credit risk models based on observable market events and information available on a firm’s balance sheet,’’ without having to rely on an investment grade rating.47 However, this commenter also stated that balance sheet measures frequently are inaccurate and that comparisons of market values to book values are subject to concerns about ‘‘garbage in, garbage out.’’ 48 While the Commission requested comments regarding whether the exception proposed in Rule 101(c)(2)(i) should require the issuer’s balance sheet to be audited,49 it did not receive any

a structural credit risk model. As discussed below, in this Part, the determinations must be ‘‘estimated as of the sixth business day immediately preceding the determination of the offering price,’’ which helps to ensure that timely information regarding the issuer is used as a model input. A lead manager may encounter difficulties in obtaining model input information from an issuer’s audited balance sheet each time it needs to determine an issuer’s probability of default for purposes of reliance on the exception for Nonconvertible Securities if such a determination were being made in between audits. 50See infra Part V.E.3. 51 17 CFR 239.13. 52 17 CFR 239.25. 53 17 CFR 239.33. 54 17 CFR 239.34. 55 17 CFR 239.40. 56SIFMA Letter 1, at 5–8; see Letter from Joseph Corcoran, Managing Dir. & Assoc. Gen. Counsel, Secs. Indus. & Fin. Mkts. Ass’n, to Vanessa Countryman, Sec’y, SEC (Oct. 28, 2022) (‘‘SIFMA Letter 2’’). 57SIFMA Letter 1, at 3–4 (stating that this standard would provide a straightforward, uniform standard; align with how the Commission addressed the Dodd-Frank Act-related removal of references to credit ratings from the eligibility criteria for use of certain provisions under the Securities Act and related forms; promote the conduct of offerings on a registered basis by limiting the exception to qualifying registered offerings; afford predictability; avoid complex calculations that could lead to errors or differing results, depending on the particular structural credit risk model used; allow the availability of the exception to be readily and independently verified through a review of the issuer’s Electronic Data Gathering, Analysis, and Retrieval (EDGAR) database filings, which would minimize additional regulatory burdens and obviate the need for any additional broker-dealer recordkeeping obligations; ease the burden on all involved, including for regulators; and provide greater legal certainty to the affected issuers and any selling shareholders). 58SIFMA Letter 2, at 1; SIFMA Letter 1, at 5–7. 59SIFMA Letter 2, at 3. 60Security Ratings, Release No. 33–9245 (July 27,
2011) [76 FR 46603 (Aug. 3, 2011)] (‘‘Form S–3 and
Form F–3 Release’’), 76 FR 46607. When the Commission revised the eligibility criteria for use of Form S–3 and Form F–3 to remove any references to credit ratings, it noted that none of the criteria are a standard of creditworthiness. Form S– 3 and Form F–3 Release, 76 FR 46607 n.60. The Commission stated that ‘‘any alternative standard for Forms S–3 and F–3 eligibility that does not refer to credit ratings should preserve the forms and access to the shelf registration process for issuers who have a wide following in the marketplace.’’ Form S–3 and Form F–3 Release, 76 FR 46607. Form SF–3, the shelf registration statement form for asset-backed securities that is discussed below, in
Part II.A.2, differs from these other forms raised by
the commenter that rely on the eligibility criteria for use of Form S–3 or Form F–3. Whereas the eligibility criteria included in Forms S–3 and F–3 focus exclusively on whether the issuer has a wide following in the marketplace to identify issuers who should be eligible for short-form registration and faster access to capital markets through the shelf registration process, the eligibility criteria and offering requirements included in Form SF–3 help to ensure that asset-backed securities issued in shelf offerings are designed to help ensure that that the securitization is designed to produce expected cash flows that are sufficient to service payments or distributions in accordance with their terms; that obligated parties more carefully consider the characteristics and quality of the assets that are included in the pool; and that asset-backed securities shelf offerings have transactional safeguards and features that make those certain securities appropriate to be issued without prior Commission staff review. See Asset-Backed Securities Disclosure and Registration, Release No. 34–72982 (Sept. 4, 2014) [79 FR 57184 (Sept. 24, 2014)] (‘‘Regulation AB II Adopting Release’’), 79 FR 57267, 57278, 57283. 61See Regulation M Adopting Release, 62 FR 524; supra note 13. For these reasons discussed above, as related to the eligibility criteria of Forms S–3 and F–3, the commenter’s suggestion of excepting Nonconvertible Securities that are offered pursuant to an effective registration statement filed on Form S–4 or F–4 would not be an appropriate substitute standard of creditworthiness in place of the reference to credit ratings in the Investment Grade Exception pursuant to section 939A(b) because Forms S–4 and F–4 include the Forms S–3 and F– 3 eligibility criteria by allowing registrants that meet the registrant eligibility requirements of Form S–3 or F–3 and that are offering investment grade securities to incorporate by reference certain information. See Form S–3 and Form F–3 Release, 76 FR 46611 (citing General Instruction B.1 of Forms S–4 and F–4). Similarly, the commenter’s suggestion of excepting Nonconvertible Securities that are offered pursuant to an effective registration statement filed on Form F–10, provided that the offering also meets the transactional requirements of General Instruction I.B.2. of Form F–3, would not be an appropriate substitute standard of creditworthiness in place of the reference to credit ratings in the Investment Grade Exception pursuant to section 939A(b) because that measure references transactional requirements that have a distinct purpose from the Commission’s original basis for adopting the Investment Grade Exception. As discussed in this Part, the probability of default is an appropriate measure to identify low manipulation risk of Nonconvertible Securities because it allows for the selection of issuers whose securities trade on the basis of yield and creditworthiness. comments in response to this question. In
addition, the Commission considered including in the exception models that may not necessarily rely on an issuer’s balance sheet to determine a firm’s creditworthiness, such as reduced-form models. Reduced-form models, however, do not necessarily predict future defaults better than structural credit risk models do, and they suffer from a lack of theoretical foundation of the assumed relationships, or the intuitive interpretation of the model dependencies and why the defaults occur.50 For these reasons, and as discussed throughout this Part, the Commission is adopting a standard that is based on the use of a structural credit risk model as it is appropriately designed to measure creditworthiness of Nonconvertible Securities in new Rule 101(c)(2)(i), in accordance with the requirements of section 939A(b). One commenter suggested an exception based on alternative standards of creditworthiness that do not utilize structural credit risk models. First, this commenter suggested that the Commission adopt an exception for Nonconvertible Securities that are offered pursuant to an effective registration statement filed on any of the following forms: (1) Form S–3; 51 (2) Form S–4; 52 (3) Form F–3; 53 (4) Form F–4; 54 or (5) Form F–10,55 provided that, for an offering registered on Form F–10, the offering also meets the transactional requirements of General Instruction I.B.2 of Form F–3.56 The commenter stated several reasons for which the use of this standard would be desirable,57 and that allowing some amount of high yield issuers to be eligible for the exception would be an acceptable compromise in light of the benefits the commenter’s proposed standard otherwise provides.58 In addition, this commenter stated that the ‘‘consistently very high percentage of registered nonconvertible debt tranches that were investment grade demonstrates that Securities Act registration alone serves as a reliable proxy for identifying offerings of nonconvertible debt securities that trade primarily based upon their yield and creditworthiness.’’ 59 The Commission acknowledges that the commenter’s suggested standard may capture a consistently very high percentage of registered nonconvertible debt tranches that were rated investment grade and may seem operationally easier to determine whether the new exception in Rule 101(c)(2)(i) for Nonconvertible Securities is available, may help to promote the conduct of offerings on a registered basis, and allow for the use of an exception that can be verified with publicly available information, among other things. The commenter’s suggested standard, however, would not be appropriate because it does not sufficiently focus on creditworthiness. When the Commission revised the eligibility criteria for use of Forms S–3 and F–3 to remove any references to credit ratings, it specifically stated that the eligibility criteria included in those forms did not distinguish among issuers by the quality of their credit but rather focused exclusively on
whether the issuer has a wide following in the marketplace to identify issuers who should be eligible for short-form registration and faster access to capital markets through the shelf registration process.60 This is distinct from the basis for adopting the Investment Grade Exceptions, including under Rule 101(c)(2), which was that Nonconvertible Securities are appropriate to except from Regulation M’s requirements because they are fungible and traded on the basis of their yield and creditworthiness, and therefore are less likely to be manipulated.61 Specifically, securities of issuers of a certain credit quality trade based on

62Bonds trade among investors and dealers in secondary markets at prices that depend on economy-wide interest rates, as well as on market perceptions regarding the likelihood that the issuing company will make the promised payments. Hendrik Bessembinder & William Maxwell, Markets: Transparency and the Corporate Bond Market, 22 J. ECON. PERSP. 217, 220 (2008). 63See infra Part V.E.3. 64SIFMA Letter 1, at 8. 65See, e.g., SIFMA Letter 1, at 8. 66Form S–3 and Form F–3 Release, 76 FR 46607. 67 In 2008, prior to the enactment of the Dodd￾Frank Act, the Commission proposed to substitute credit ratings references in Rules 101 and 102 with a standard for Nonconvertible Securities that was based primarily on the WKSI concept from 17 CFR
230.405 (‘‘Rule 405’’), as well as a standard for
asset-backed securities that were registered on Form S–3. See 2008 Proposing Release, 73 FR 40095–97. The WKSI-based approach, consistent with the definition of WKSI under Rule 405, would have excepted the Nonconvertible Securities of companies that have issued at least $1 billion aggregate principal amount of nonconvertible securities, other than common equity, in primary offerings for cash, not exchange, registered under the Securities Act. See 17 CFR 230.405, paragraph (1)(i)(B)(1) of the definition of WKSI; see also 2008 Proposing Release, 73 FR 40096. 68See SIFMA Letter 1, at 10. This commenter stated that its suggested exception based on Forms S–3 and F–3 is more appropriate than this WKSI￾based approach because the Form S–3/F–3-based approach ‘‘recognizes several different means of qualifying under the transactional requirement, only one of which is based upon the aggregate principal amount of non-convertible securities issued over the preceding three years.’’ SIFMA Letter 1, at 9–10. For the reasons discussed in this
Part, neither of these approaches is an appropriate
standard of creditworthiness in place of the reference to credit ratings in the Investment Grade Exception. 69Proposal, 87 FR 18334–35. 70See infra Part V.E.3. 71See Proposal, 87 FR 18317. Similar to how securities covered by the existing Investment Grade Exception are excepted from Rule 101’s prohibitions, Nonconvertible Securities that trade based on their yield and creditworthiness would be excepted under Rule 101 as amended to include the probability of default-based standard. yield and creditworthiness 62 and are less susceptible to manipulation because other similar Nonconvertible Securities are available to investors as an alternative to the security in distribution. If pricing of a Nonconvertible Security offering is inconsistent with pricing in the overall secondary market for similar Nonconvertible Securities, an investor may purchase alternative Nonconvertible Securities that have a better yield, yet are of comparable creditworthiness, than the security being distributed. Accordingly, the ability to substitute similar Nonconvertible Securities in the market for the security in distribution limits the potential impact that a distribution participant might attempt to exert on the market and distribution of such security. In addition, when debt has a very low probability of default, the cash flows are close to risk-free. Thus, the price of the debt is mainly subject to fluctuations based on aggregate interest rates rather than issuer-specific or security-specific news. The probability of default is an appropriate measure to identify low manipulation risk of such securities, as it allows for the selection of issuers whose securities trade on the basis of yield and creditworthiness (traditionally measured by credit ratings). For issuers with sound creditworthiness, the pricing of securities is unrelated to other risks associated with the identity of the issuer, greatly reducing their uncertainty and manipulation risk. A standard based on a criterion such as being widely followed in the market does not allow for such a clear distinction because such a standard does not differentiate securities that are traded solely on their yield and creditworthiness from securities that trade solely on the issuer’s identity and thus could present a high manipulation risk.63 Accordingly, it is appropriate to implement the section 939A(b) mandate by adopting an exception that is based on a standard that is likewise premised specifically on creditworthiness rather than on whether a particular issuer has a wide following in the marketplace. Second, the commenter suggested that its recommended standard described above, which is based on the Forms S– 3 and F–3 eligibility criteria, could be modified by prohibiting reliance on the exception for Nonconvertible Securities that include both a ‘‘limitation on restricted payments covenant’’ and a ‘‘limitation on sales of assets and subsidiary stock covenant,’’ which the commenter stated are two covenants that
typically are associated with non￾investment grade debt securities and are almost never used in investment grade debt securities.64 However, despite the commenter’s suggestions, the covenant restrictions are features of current market practices 65 but are not necessarily inherent characteristics of the securities related to their creditworthiness. Conditioning the exception on the absence of certain covenants poses the risk that, should market practice change, the exception would quickly become outdated. Therefore, even with the commenter’s two suggested modifications, a standard that is focused on the Form S–3 or Form F–3 eligibility criteria and is premised exclusively on whether an issuer is widely followed,66 rather than on an issuer’s creditworthiness, is not an appropriate substitute standard of creditworthiness in place of the references to credit ratings in Rule 101’s Investment Grade Exception for Nonconvertible Securities to sufficiently respond to the requirements of section 939A(b). Finally, this commenter suggested that the Commission adopt a modified version of the 2008 Proposing Release involving a well-known seasoned issuer (‘‘WKSI’’) standard 67 to except: (1) a WKSI that, as of a date within 60 days of the applicable determination date, has a worldwide market value of its outstanding voting and non-voting common equity held by non-affiliates of $700 million or more; and (2) a non￾WKSI to the extent it is carved out of the WKSI definition solely by virtue of the application of paragraph (1)(v), (vi), or (ix) of the definition of ‘‘ineligible issuer’’ under Rule 405 under the Securities Act.68 The Commission recognizes the advantage a WKSI-based standard might have in terms of its simplicity and straightforward calculation. As noted in the Proposal, however, a WKSI-based standard as proposed in 2008 was criticized for allowing many risky, high-yield issues to be excepted and preventing issues by smaller but otherwise credit-worthy issuers from being eligible for the exception.69 This WKSI-based standard, however, unlike the probability-of￾default-based standard, would fail to capture the pricing point where the sound creditworthiness of the issuer eliminates other risks associated with the issuer identity. The Nonconvertible Securities of such issuers trade solely based on their yields and creditworthiness and not on issuer characteristics, where pricing uncertainty and manipulation risk are at their minimum.70 The WKSI-based standard, therefore, is a less effective measure of manipulation risk as compared to the probability of default measure. For these reasons, and pursuant to the requirements of section 939A(b), the probability of default measure is a more appropriate substitute of creditworthiness for the reference to credit ratings in the existing Investment Grade Exception than is the commenter’s suggested WKSI-based standard.71 As discussed above, in this
Part, when debt has a very low
probability of default, its price fluctuations are mainly based on aggregate interest rates rather than on company-specific or security-specific news. The probability of default measure, in contrast to the commenter’s suggested WKSI-based standard, continues to rely on the premise underlying the Investment Grade Exception: Nonconvertible Securities

72See SIFMA Letter 1, at 10; Better Markets Letter, at 5. 73See SIFMA Letter 1, at 5; Better Markets Letter, at 4. 74Better Markets Letter, at 4 (stating, in part, that the use of structural credit risk models will create a lack of uniformity that conflicts with the mandate in section 939A(b) for the Commission to establish, to the extent feasible, uniform standards of creditworthiness). But see Bloomberg L.P. Letter, at 2 (stating that, although the application of a particular threshold across multiple models may have some unintended consequences (e.g., different point-in-time probability of default models may produce different results for the same issuance), the proposed exception provides an alternative measure of creditworthiness that is practical, appropriately based on objective factors, and can be consistently applied by market participants). The mandate in
section 939A(b) to seek to establish uniform
standards of creditworthiness is limited ‘‘to the extent [that it is] feasible.’’ Public Law 111–203, sec. 939A(b). As discussed in this Part, the use of structural credit risk models to derive an issuer’s probability of default is an appropriate standard of creditworthiness in accordance with section 939A(b)’s requirements. 75Better Markets Letter, at 4. 76Better Markets Letter, at 4. 77See IILF Letter, at 6 (stating that requiring a particular type of model could potentially distort the behavior of market participants in their estimations of probability of default and discouraging further and alternative inquiries into the probability of default). The use of structural credit risk models is required only for purposes of deriving an issuer’s probability of default pursuant to new Rule 101(c)(2)(i). Distribution participants, as well as other market participants, may use other types of models in evaluating the creditworthiness of an issuer outside of making a Rule 101(c)(2)(i) probability of default determination. 78See IILF Letter, at 6. 79Also, as discussed below, in Part II.C, because probability of default estimates may be subjective to some extent and not comparable across different issuers or for the same issuer across different issues if estimates are based on different models, or done by different researchers or vendors, the requirement associated with reliance on new Rule 101(c)(2)(i) to preserve written probability of default determinations is designed to facilitate the Commission’s examinations of broker-dealers who rely on the exception in new Rule 101(c)(2)(i) or new Rule 102(d)(2)(i). 80See infra Part V.C.3. 81See infra Part V.C.3; see also Proposal, 87 FR 18334. 82Proposal, 87 FR 18332. 83SIFMA Letter 1, at 5. 84See infra Part V.B (discussing how structural credit risk models, as defined in Rule 100(b), are designed to measure creditworthiness, and creditworthiness itself is considered to be a good measure of manipulation risk). 85See 17 CFR 240.17a–4(b)(17), as amended; infra
Part II.C.
86Proposal, 87 FR 18320. that trade primarily based on their yield and creditworthiness are less susceptible to the type of manipulation that Rule 101 seeks to prevent. Some commenters stated that the Commission should specify a particular structural credit risk model to be used by all parties in making probability of default calculations.72 These commenters stated their concerns regarding the potential for inconsistent outcomes resulting from the discretion to choose what structural credit risk models to apply.73 One commenter stated that the adoption of the proposed model-based standard would create the risk that the new standard would be manipulated because firms would have a wide variety of models from which to select.74 This commenter stated that, while the proposed probability-of￾default-based standard is a reasonable alternative standard of creditworthiness, the use of structural credit risk models would create challenges for the Commission, with regard to implementing the probability of default standard, and for investors, with regard to confidence in the consistency and reliability of determinations made under the new standard.75 In addition, this commenter stated that setting no minimum standards for the models and allowing market participants the discretion to choose among a wide range of models threatens to create a ‘‘race to the bottom’’ as market participants seek to avoid competitive disadvantages that will arise from having an appropriately rigorous risk of default evaluation.76 Another commenter, however, stated that it would be ‘‘risky’’ for the Commission to engage in ‘‘model preferencing.’’ 77 The Commission agrees with the comment against requiring the use of a specific structural credit risk model.78 On balance, the use of a structural credit risk model to derive an issuer’s probability of default pursuant to Rule 101(c)(2)(i), as amended, provides an appropriate degree of flexibility in terms of model selection while also providing certainty to distribution participants as to the standards for the structural credit risk model required for purposes of compliance in making probability of default determinations. The ability to use an unrestricted universe of models for purposes of meeting the conditions of new Rule 101(c)(2)(i) could provide distribution participants with the opportunity to choose model specifications that enable abuse of the exception for Nonconvertible Securities. In this regard, the definition of ‘‘structural credit risk model,’’ as discussed below in this Part, sets minimum standards for the structural credit risk models that may be used to derive an issuer’s probability of default to meet the conditions of new Rule 101(c)(2)(i). These minimum standards include that the model be a commercially or publicly available model and that it calculate, based on an issuer’s balance sheet, the probability that the value of the issuer will fall below the Default Point, at or by the expiration of a defined period. As
discussed below, in Part V.C.3, the standard’s use of structural credit risk models could incentivize lead managers to select models and estimation specifics in such a way to ensure the resulted estimates are below the threshold, thus allowing securities of issuers with low creditworthiness and high manipulation risk to be eligible for the exception. The public availability of alternative estimates for investors, however, should mitigate this concern.79 Specifically, the limitation that the structural credit risk model must be a commercially or publicly available model would limit a distribution participant’s ability to develop models for the purpose of abusing the exception.80 In this regard, use of a structural credit risk model that is not commercially or publicly available, or one that does not calculate, based on an issuer’s balance sheet, the probability that the value of the issuer will fall below the Default Point, at or by the expiration of a defined period, would not be permissible in meeting the conditions of the exception. While a standard that relies on the use of a structural credit risk model retains a certain level of subjectivity,81 this standard also leaves room for improvement if the market adopts more accurate structural credit risk models in the future. As the Commission stated in the Proposal, the use of any model to estimate creditworthiness necessarily provides an imperfect measure.82 This flexibility in selection may result in an outcome-oriented selection of structural credit risk models, as one commenter suggested.83 However, a selection that meets the definition of ‘‘structural credit risk model,’’ as provided in Rule 100(b), as well as the requirements of new Rule 101(c)(2)(i), would be consistent with the aims of section 939A as well as those of Regulation M.84 In addition, the new record preservation requirement set forth in Rule 17a–4(b)(17), as discussed below in Part II.C, is designed to aid Commission examinations of broker￾dealers who rely on the exception in Rule 101(c)(2)(i) or Rule 102(d)(2)(i), as amended, and can help deter improper adjusting of the estimation to meet the conditions of either of the exceptions.85 The Commission requested comment on whether there are ‘‘any reasons why the Rule should not permit a distribution participant to perform its own calculation (subject to recordkeeping requirements, as proposed).’’ 86 To address concerns that the proposed flexibility in structural

87SIFMA Letter 1, at 5.
88 Distribution participants who act as the ‘‘lead manager’’ of a distribution for purposes of the exception in Rule 101 may, as a practical matter, also use or be known by different titles, such as ‘‘lead underwriter,’’ ‘‘managing lead underwriter,’’ ‘‘syndicate manager,’’ ‘‘stabilizing manager,’’ ‘‘lead bookrunner,’’ or ‘‘co-managing underwriter.’’ The parenthetical ‘‘(or in a similar capacity)’’ is included in FINRA’s underwriting-related rules, such as FINRA Rule 5110, to recognize this common industry practice, as well as to prevent evasion by persons attempting to avoid regulatory responsibility under a particular provision by using a different title or term to refer to themselves, even though they perform the same or similar function. 89See 17 CFR 242.101(c)(2)(i), as amended. 90As discussed below, in Part II.C, broker-dealers who rely on the new exception for Nonconvertible Securities in new Rule 101(c)(2)(i) or new Rule 102(d)(2)(i), as applicable, must preserve certain records pursuant to Rule 17a–4 under the Exchange Act. New paragraph (b)(17) of Rule 17a–4 requires broker-dealers to preserve the written probability of default determination, relied upon pursuant to the exception for Nonconvertible Securities. Accordingly, broker-dealers relying on the exception for Nonconvertible Securities are required to preserve for a period of not less than three years, the first two years in an easily accessible place, the written probability of default determination. 91The term ‘‘lead manager’’ under new Rule 101(c)(2)(i) is consistent with how the term ‘‘manager’’ is applied, for recordkeeping purposes, in 17 CFR 240.17a–2(b)(1) with respect to any person who acts as a manager of a distribution for its sole account or for the account of a syndicate or group in which it is a participant with respect to keeping records of any syndicate covering transactions, penalty bids, and all related stabilizing activity, all three of which are governed under 17 CFR 242.104, which cross-references the recordkeeping requirement in 17 CFR 240.17a–2, as well as the ‘‘managing underwriter’’ in connection with TRACE-reporting of eligible fixed-income securities, or FINRA Rule 5131’s requirement that the lead managing underwriter of a distribution disclose indications of interest and final allocation information to the issuer’s pricing committee, or notify the issuer of any impending release or waiver of lock-ups. Thus, similar to the traditional role played by the lead or managing underwriter in firm commitment offerings—which generally include overseeing the offering process to ensure that the marketing, pricing, and allocation processes all go smoothly; providing critical advice on the structure, size, timing, and price of the offering; and advising on how to best present the issuer’s business in the prospectus or other offering documents—the distribution participant acting as the lead manager (or in a similar capacity) of a distribution is the only market
participant who is eligible to derive the issuer’s probability of default for purposes of meeting the conditions of new Rule 101(c)(2)(i), in recognition that it is in the best position to do so. There may be distributions with more than one distribution participant acting as the lead manager (or in a similar capacity). In such a distribution, because the rule text refers to ‘‘the distribution participant acting as the lead manager,’’ only one of the distribution participants acting as the lead manager would be permitted to make the probability of default determination for the particular distribution. 92See, e.g., infra Part V.C.3 (discussing that the requirement related to the lead manager’s probability of default determination should mitigate the subjectivity (of the analysis involved in probability of default estimation, as well as of the selection of the model and data sample specifics) and the concerns regarding non-uniform probability of default estimates for the same issue—and to some degree across issues for the same issuer to the extent the same parties are engaged by the issuer for different issues). 93See, e.g., supra notes 88, 91. 94See, e.g., Regulation M Adopting Release, 62 FR 534–35. 95See, e.g., supra notes 88, 91; SIFMA, Model Form of Master Agreement Among Underwriters (Dec. 10, 2018), https://www.sifma.org/wp-content/ uploads/2017/08/SIFMA-Model-MAAU.pdf https:// www.sifma.org/wp-content/uploads/2017/08/ SIFMA-Model-MAAU.pdf. 96See Regulation M Adopting Release, 62 FR 522–23 (discussing Rule 100’s definition of ‘‘completion of participation in a distribution’’ when underwriters in a syndicate are involved). 97The Commission asked in the Proposal whether distribution participants should be required to post or make the probability of default public on their website to rely on the exception. See Proposal, 87 FR 18320. As discussed below, in Part II.C, one commenter stated that the Commission also could publish, or require publication of, default probability estimates that market participants derive from various models, along with default probabilities implied by both market prices and credit default swap spreads. See IILF Letter, at 8. However, the Commission did not receive any comment suggesting that distribution participants making probability of default determinations should be required to post or make the probability of default determinations public on their website in order to rely on the exception. Nor did the Commission receive any comment suggesting that the sharing of probability of default determinations among other covered persons should be included as a condition to the exception. 98Rule 100 of Regulation M defines the term ‘‘distribution’’ as ‘‘an offering of securities, whether or not subject to registration under the Securities Act that is distinguished from ordinary trading transactions by the magnitude of the offering and the presence of special selling efforts and selling methods.’’ With regard to shelf
offerings, each takedown of a shelf is to be individually examined to determine whether such offering constitutes a ‘‘distribution’’ (i.e., whether it satisfies the ‘‘magnitude’’ of the offering and ‘‘special selling efforts and selling methods’’ criteria of a distribution). Regulation M Adopting Release, 62 FR 526. In those situations where a broker-dealer sells shares on behalf of an issuer or selling security holder in ordinary trading transactions into an independent market (i.e., without any special selling efforts), the offering will not be considered a distribution, and the broker-dealer will not be subject to Rule 101. Regulation M Adopting Release, 62 FR 526. credit risk model selection could lead to different underwriters coming to different conclusions on the availability of an exception from Regulation M based on which structural credit risk model they use, as well that this flexibility could contribute to inefficiencies, confusion, and dissension among distribution participants,87 the final amendments limit the universe of those who are eligible to determine an issuer’s probability of default under new Rule 101(c)(2)(i) to include only the distribution participant who is acting as the lead manager (or in a similar capacity) 88 of a distribution.89 This limitation will help to ensure consistent reliance on the exception across all distribution participants for the same distribution through use of the same written probability of default determinations and through the new record preservation requirements under Rule 17a–4(b)(17).90 The lead manager’s role and responsibilities in overseeing the distribution process 91 should, for these same reasons, help alleviate concerns regarding the consistency and reliability of the determinations within any particular distribution.92 The lead-manager requirement is intended to broadly reflect current market practices.93 Lead managers will be incentivized to share their probability of default determinations with other distribution participants and their affiliated purchasers (as well as with the issuer, selling security holders, and their affiliated purchasers) in order to rely on the exception for Nonconvertible Securities given their primary role and responsibilities in overseeing the distribution process, which can include providing liquidity and facilitating an orderly distribution and aftermarket in connection with the offering.94 While Regulation M does not require the lead manager to coordinate the activities of the other syndicate members, lead managers are, as a practical matter, concerned that the other underwriters in the syndicate are, among other things, complying with Regulation M’s trading prohibitions 95 so as not to extend the Regulation M restricted period.96 However, Rule 101(c)(2)(i), as amended, does not require that the lead manager making the probability of default determination share the determination with other distribution participants or their affiliated purchasers in order
for those parties to rely on the exception.97 Therefore, non-lead manager distribution participants and their affiliated purchasers (as well as issuers, selling security holders, and their affiliated purchasers, as discussed below, in Part II.B.1) may not be able to rely on the exception for Nonconvertible Securities if the lead manager does not share the probability of default determination or there is no distribution participant to act as the lead manager for the distribution, such as with self￾underwritten offerings, at-the-market offerings, or other shelf offerings, to the extent such an offering meets the definition of a ‘‘distribution’’ under Rule 100(b) of Regulation M. 98 This potential impact is mitigated, however, if the syndicate adjusts the way it interacts with lead managers. For example, the syndicate could decide, by contract (e.g., in an agreement among underwriters) and for the same reason of consistency as discussed above, to specifically authorize the lead manager

99See Proposal, 87 FR 18330. 100See Better Markets Letter, at 4. 101However, as discussed below, in Parts V.C.1 and VI.C.1, some lead managers may rely on third party vendors rather than internally calculate the probability of default. 102As discussed below, in Part II.C, the record preservation requirements provided in new Rule 17a–4(b)(17) are sufficient to help the Commission’s examinations of broker-dealers relying on the new probability-of-default-based standard. 103See, e.g., supra notes 13, 38. 104SIFMA Letter 1, at 5. 105SIFMA Letter 1, at 5. 106 IILF Letter, at 2 (stating that the inability to reliably estimate the probability of default on a debt security using any of a variety of statistical models and market measures is strong evidence that the security should not fall within an exception to Regulation M and could also be evidence that the market participant is not in a position to trade or hold that specific security). As discussed in this
Part, new Rule 101(c)(2)(i) requires that the
probability of default determination be derived from a structural credit risk model. Distribution participants and their affiliated purchasers may not avail themselves of the exception in new Rule 101(c)(2)(i) with regard to any security that does not meet the requirements of that exception. 107 IILF Letter, at 2. 108 IILF Letter, at 2, 6–8 (stating its concerns about the use of balance sheets and suggested that the Commission reference more flexible alternatives, such as the probability of default threshold could vary annually on an ongoing basis depending on a similar analysis of more recent data going forward or peg the annual probability of default threshold based on an analysis of a sample of securities from the previous year). The Commission has considered this comment and, on balance, concerns about the use of an issuer’s balance sheet should be addressed by the rigorous theoretical justification as well as by the economic interpretation of the resulting relationships between the inputs that are embedded in such structural credit risk models. See, e.g., infra note 243. However, allowing a more flexible threshold, as would be done under the commenter’s suggestion, would result in increased subjectivity and non-uniformity of the application of the exception. 109 IILF Letter, at 2. 110See Proposal, 87 FR 18332. 111See Proposal, 87 FR 18332. 112Bloomberg L.P. Letter, at 2. to share its probability of default determination with other distribution participants. The syndicate could also make the decision not to allow the lead manager to share its probability of default determination with unrelated distribution participants in order to keep a tighter control of the distribution. Because the facts and circumstances vary across issues, it is reasonable to let the syndicate decide how widely the probability of distribution determination is shared by the lead manager. If such information is shared, it must be used consistently across the syndicate to rely on the new exception in Rule 101(c)(2)(i) because such reliance is conditioned on the lead manager’s probability of default determination. The lead-manager requirement, however, could affect current market practices by resulting in fewer Nonconvertible Securities being excepted under the new standard in comparison to those currently excepted under the Investment Grade Exception if, as a practical matter, only Nonconvertible Securities that are subject to underwritten offerings become eligible for the new exception in Rule 101(c)(2)(i). With regard to the types of distributions covered, as a result of the condition requiring that the lead manager perform the probability of default determinations in order for reliance on the exception, this exception will be available to a subset of distributions of Nonconvertible Securities covered under the existing Investment Grade Exception and to a subset of the distributions that would have been captured under the Proposal.99 Accordingly, any potential
challenges, as a commenter suggested,100 are likely to be faced in carrying out obligations in order to rely on the new exception in Rule 101(c)(2)(i). The estimated costs associated with the requirement related to the lead manager making the probability of default determination are included below, in Part V.C.1.101 The bright-line requirements of Rule 101(c)(2)(i), as amended, and the corresponding record preservation requirements of new Rule 17a–4(b)(17) also help to promote confidence in the consistency and reliability of the lead manager’s determinations by limiting the degree to which there are variances between probability of default calculations within any one distribution as well as by deterring any improper tweaking of model inputs.102 The bright￾line threshold of 0.055%, as well as the pre-determined time horizon, are model inputs that are uniform and predictable and, thus, should provide the necessary clarity as to what is expected in evaluation and documentation. In addition, the bright-line threshold of 0.055% will help to ensure that only those Nonconvertible Securities that trade on the basis of yield and creditworthiness, and are fungible,103 will meet the exception, regardless of the model picked. Accordingly, the probability-of-default-based standard articulates an appropriate alternative measure of creditworthiness that is practical and is appropriately based on objective factors. One commenter stated that commercially or publicly available structural credit risk models are not used by all firms in the context of evaluating whether to underwrite a security and that such a decision, instead, is focused on the adequacy and accuracy of disclosures.104 The Commission acknowledges that, when a firm evaluates whether to underwrite a distribution of securities, it typically focuses on the adequacy and accuracy of an issuer’s disclosures. However, the adequacy and accuracy of disclosures are a separate question from the creditworthiness of securities for purposes of the exception for Nonconvertible Securities. It is possible that an accurate disclosure statement did not reveal the issuer’s low creditworthiness, making the offered securities inappropriate to be eligible for the exception for Nonconvertible Securities that trade on the basis of their yield and creditworthiness. One commenter stated that the proposed probability of default calculations would create a heightened risk for errors.105 Another commenter stated that market participants can reliably estimate the probability of default, not only by using the proposed ‘‘structural credit risk models’’ but also by using other statistical models,106 market measures of credit risk, and other credit risk measures.107 This commenter encouraged the Commission to adopt a final rule that references market measures of credit risk as part of the estimation of, or as an alternative to, the probability of default.108 The Commission agrees with the comments that market participants can
reliably estimate the probability of default derived from a structural credit risk model.109 The Commission also acknowledges that probability of default estimates are not free of subjectivity and can vary across structural credit risk models, researchers, or vendors.110 The use of any model or market measure to estimate issuer creditworthiness is imperfect.111 The new exception’s bright-line probability of default threshold and time horizon, however, provides predictability and allows for the exception to be applied consistently by distribution participants who are eligible to make probability of default determinations.112 As discussed below, in Part V.E.3, the Commission has considered other types of models, such as reduced-form models, which would generally provide less stable predictions than structural credit risk models do because they can be so flexible that they suffer from a lack of theoretical foundation and a lack of intuitive interpretation of why the defaults occur. Also, unrestricted use of these models might also provide more opportunity to choose a reduced-form model specification to enable use of the exception for Nonconvertible Securities. The Commission therefore is adopting

113SIFMA Letter 1, at 10. The Commission estimated in the Proposal that, while the proposed threshold of 0.055% would capture approximately 90% of the investment grade securities in its sample of nonconvertible fixed income securities, a threshold of 0.5% would capture about 98.6% of investment grade securities. See Proposal, 87 FR 18330, 18334. 114See Bloomberg L.P. Letter, at 2; IILF Letter, at 6. 115Bloomberg L.P. Letter, at 2. Another commenter stated that it is not necessary to state a precise bright-line measure. IILF Letter, at 6. The exception’s use of a bright-line threshold, by imposing specific and clear requirements, helps to ensure that the exception captures only those Nonconvertible Securities that trade on the basis of yield and creditworthiness. It also helps to ensure that the exception is based on objective factors and can be consistently applied by market participants. See Bloomberg L.P. Letter, at 2. 116See infra Part V.B; Proposal, 87 FR 18319, 18330. 117See also Proposal, 87 FR 18334. Based on an analysis of the available data as of Mar. 2023, a 0.5% threshold would make the new exception less restrictive and would result in 124 additional non￾investment grade securities being captured by the standard, from 64 non-investment grade issues under the 0.055% threshold to 188 issues under the 0.5% threshold. See infra Part V.B. These figures differ from those included in the Proposal because they are based on an analysis of the available data as of Mar. 2023, whereas the Proposal’s figures were based on an analysis of the data available as of Oct.
2021. See infra Part V.B.
118See Regulation M Adopting Release, 62 FR 527; supra note 38. 119See infra Part V.E.1. 120See Proposal, 87 FR 18330; see also Proposal, 87 FR 18319, 18332. 121SIFMA Letter 1, at 10. 122See infra Part VI.C.1. For example, because commonly available spreadsheet software can be used to calculate the probability of default, lead managers would not need 10 calendar days to derive an issuer’s probability of default. See, e.g., Proposal, 87 FR 18319. an exception that is based on the use of a structural credit risk model as this model is appropriately designed to measure creditworthiness of Nonconvertible Securities in Rule 101(c)(2)(i), as amended, in accordance with the requirements of section 939A(b). One commenter recommended that the proposed probability of default threshold should be increased to 0.5% in order to capture the maximum amount of issuers who, currently, are eligible under the existing exception (i.e., the proposed threshold of 0.055% is too restrictive with regard to scoping in securities that currently are excepted).113 Other commenters supported the proposed probability of default threshold of 0.055% as reasonable.114 One commenter stated that the bright-line threshold of 0.055% will provide clarity as to what is expected in evaluation and documentation.115 The Commission has considered these comments and concluded that the 0.055% threshold appropriately calibrates the probability of default to determine the creditworthiness of an issuer whose Nonconvertible Securities trade based on their yield and creditworthiness.116 While the higher threshold of 0.5% captures a larger set of securities of creditworthy issuers whose securities are eligible for the existing Investment Grade Exception, it also allows for an exception that captures a larger set of securities that could be prone to manipulation risk in comparison to the 0.055% threshold (i.e., non-investment grade securities).117 Because the commenter’s suggested 0.5% threshold, in comparison to the 0.055% threshold, risks capturing a majority of the securities that are not traded on the basis of their yield and creditworthiness in the same way that Nonconvertible Securities excepted under the existing Investment Grade Exception are traded, too many distributions of these types of securities would be included, which reflects that a 0.5% threshold may not be an appropriate replacement standard of creditworthiness, in accordance with of section 939A(b). Further, even if the 0.055% threshold does not capture the exact same set of Nonconvertible Securities captured by the Investment Grade Exception, the 0.055% threshold nevertheless identifies Nonconvertible Securities that are less susceptible to the manipulation that Rule 101 is designed to prevent because they trade based on their yield and creditworthiness, As discussed above, Regulation M seeks to protect the offering price of a security during a distribution, when there are heightened incentives on the part of those who are
involved in the offering process to influence the subject security’s price. Because these Nonconvertible Securities are traded on the basis of their yield and creditworthiness, and are largely fungible, they are less susceptible to manipulation.118 In other words, the ability of distribution participants and their affiliated purchasers to bid up the price of a Nonconvertible Security of an issuer that meets the 0.055% probability of default threshold is limited by investors’ ability to substitute the security with other securities that are similar and of comparable creditworthiness. In contrast, a non￾investment grade security that has a much higher probability of default tends to have idiosyncratic risks that make them less substitutable and hence more susceptible to manipulation. The threshold of 0.5% would capture more than the majority of non-investment grade securities (approximately 69.9% of non-investment grade securities),119 which indicates that it may not be an appropriate measure of creditworthiness to replace the reference to credit ratings in Rule 101’s Investment Grade Exception. Accordingly, the 0.055% threshold appropriately calibrates the probability of default to determine the creditworthiness of an issuer whose Nonconvertible Securities should trade based on yield and creditworthiness and is an appropriate substitute standard of creditworthiness to replace the credit ratings reference in the Investment Grade Exception pursuant to section 939A(b). While the probability of default measure uses a threshold of 0.055%, as was proposed, the final rule text is changed from the proposed rule text of ‘‘less than 0.055%’’ to state ‘‘0.055% or less.’’ The Commission is clarifying that a determination of a 0.055% probability of default is eligible for the exception, so long as all other conditions of the exception are met. This change is consistent with the estimates included in the Proposal, including with how the Commission calibrated the probability of default threshold in the Proposal.120 The same commenter also suggested that the probability of default calculations should be permitted to be made within a specified duration of time in advance of pricing (rather than as of the day of determining the offering price), for example, within 10 calendar days prior to pricing of the offering, similar to the approach taken with respect to average daily trading volume (‘‘ADTV’’) calculations, to afford distribution participants adequate time to adjust their market activities as necessary.121 The Commission acknowledges that lead managers who make probability of default determinations pursuant to Rule 101(c)(2)(i) may need additional time prior to the pricing of an offering to make the required calculations. However, in light of the comments received, the 10-calendar-day period suggested by the commenter would be unnecessarily long for the lead manager to determine and document in writing the issuer’s probability of default because the
determination is likely to be highly automated.122 In addition, the suggested 10-calendar-day period may not encourage as timely of information about the issuer as possible if the model inputs are taken farther away from the day of the determination of the offering price, as proposed. The Commission is, therefore, modifying the proposed time horizon of ‘‘the day of the determination of the

123SIFMA Letter 1, at 11.
124See, e.g., infra Part V.B (discussing how probabilities of default implied by structural credit risk models generally use current estimates of equity valuation and volatility based on the recent trading activity, and hence incorporate more recent news affecting the valuation and perceived volatility of the firm). 125See, e.g., Proposal, 87 FR 18330. 126The final amendments make the technical change of deleting the word ‘‘the’’ from the beginning of the exception in order to mirror the beginning of new Rule 101(c)(2)(ii). The final rule amendments also make an edit to use the words ‘‘as derived from’’ instead of ‘‘by using,’’ as proposed, to clarify that a structural credit risk model must be the only method of determining an issuer’s probability of default, as opposed to one method among others. This change also conforms the standard to how it was discussed in the Proposal. See, e.g., Proposal, 87 FR 18318–19. The final amendments also make a conforming edit from the proposed rule text to add the word ‘‘full’’ preceding the time horizon of 12 calendar months to make the phrasing of the exception’s time horizon consistent with Regulation M’s other time horizons. See, e.g., 17 CFR 242.100(b) (defining the terms ‘‘ADTV,’’ which uses a time horizon of ‘‘two full calendar months,’’ and ‘‘principal market,’’ which uses the time horizon of ‘‘12 full calendar months’’). 127See infra Parts V.A through E. 128The Commission is making clarifying changes to the proposed definition of the term ‘‘structural credit risk model’’ that conform it to its description in the Proposal. As discussed above, in Part II.A.1, the Proposal stated that a structural credit risk model ‘‘provide[s] a probability that a firm’s assets will fall below the Default Point at or by the expiration of a defined period of time.’’ Proposal, 87 FR 18317. The final rule amendments conform the proposed definition to that description and clarify that the ‘‘threshold’’ referenced in the proposed definition is the Default Point. 129The Commission is also making a non￾substantive, technical change from the proposed definition of the term ‘‘structural credit risk model.’’ As discussed below, because the final rule amendments include the term ‘‘structural credit risk model’’ in both new Rule 101(c)(2)(i) and new Rule 102(d)(2)(i), the Commission is adding to Rule 100(b) a definition for the term ‘‘structural credit risk model.’’ The addition of this definition in Rule 100(b) does not change the definition of the term ‘‘structural credit risk model’’ but rather simplifies the final text of new Rules 101(c)(2)(i) and 102(d)(2)(i) by obviating the need for a proviso containing a definition in each of those rules. Accordingly, use of the term ‘‘structural credit risk model’’ is identical across new Rules 101(c)(2)(i) and 102(d)(2)(i). 130 17 CFR 242.100(b). 131See Regulation AB II Adopting Release. offering pricing’’ to allow the lead manager to make its probability of default determination as of
‘‘the sixth business day immediately preceding the determination of the offering price’’ for purposes of the new exception. This change from the proposed time horizon of ‘‘the day of the determination of the offering price’’ is being made in response to comment that additional time is needed because ‘‘[i]t would be very damaging to the issuer to launch a re-opening, subsequently determine that there is no exception under the probability of default calculation, and then have to extend the pricing of the offering by at least one (or five) business days.’’ 123 The Commission agrees with the commenter that more time would be useful to address the potential of an offering by at least one to five business days but is concerned that the model inputs supporting the probability of default determination may become stale with additional time beyond that.124 Accordingly, the Commission is extending the time horizon to allow for the potential of an offering being extended up to five business days. Further, this will allow the determination to be made before Regulation M’s otherwise applicable five-business-day restricted period (i.e., preceding the determination of the offering price) and should provide a sufficient amount of additional time for the lead manager to account for any relevant market activities and timely information regarding the issuer as a model input in determining the probability of default.125 This essentially allows these distribution participants, in relation to the proposed ‘‘day of the determination of the offering price’’ requirement, five additional business days, as defined in Rule 100(b), before the actual pricing and launch of the offering to make the probability of default determination. In view of the nature and trading characteristics of Nonconvertible Securities, the impact, if any, of a corporate or market event in the intervening five business days would be unlikely to result in the manipulation that Regulation M seeks to prevent. Nonconvertible Securities are priced and traded differently than equity securities in that the focus (with Nonconvertible Securities) is placed on receiving periodic interest payments during the life of the instrument rather than on any potential equity upside or increase in the current trading or offering price. Therefore, trading activity in Nonconvertible Securities at or around the time of the distribution is unlikely to influence the pricing or trading of such securities, particularly during Regulation M’s (otherwise applicable) restricted period. Accordingly, the Commission is adopting, with targeted modifications in consideration of the comments received, as discussed in this Part, as well as certain technical changes,126 a new exception in Rule 101(c)(2)(i) for Nonconvertible Securities of issuers for which the probability of default, estimated as of the sixth business day immediately preceding the determination of the offering price and over the horizon of 12 full calendar months from such
day, is 0.055% or less, as determined and documented, in writing, by the distribution participant acting as the lead manager (or in a similar capacity) of a distribution, as derived from a structural credit risk model. For the reasons discussed above, in this Part, and as supported by an analysis of the probability of default and investment grade credit ratings of a sample of Nonconvertible Securities available on the market,127 the standard used in the new exception is an appropriate substitute standard of creditworthiness to replace the credit ratings reference in the Investment Grade Exception for Nonconvertible Securities. Finally, the Commission is adopting, substantially as proposed, with certain clarifying 128 and technical 129 changes, a definition under Rule 100(b) of Regulation M for the term ‘‘structural credit risk model’’ that means ‘‘any commercially or publicly available model that calculates, based on an issuer’s balance sheet, the probability that the value of the issuer will fall below the threshold at which the issuer would fail to make scheduled debt payments, at or by the expiration of a defined period.’’ 130 Accordingly, a covered person’s reliance on Rule 101(c)(2)(i) or Rule 102(d)(2)(i), as amended, is conditioned on a probability of default determination that was derived from any commercially or publicly available structural credit risk model that calculates, based on an issuer’s balance sheet, the probability that the value of the issuer will fall below the threshold at which the issuer would fail to make scheduled debt payments, at or by the expiration of a defined period.
2. Rule 101(c)(2)(ii): Asset-Backed
Securities Offered Pursuant to an Effective Shelf Registration Statement Filed on Form SF–3 The Commission proposed an amendment to add a new exception in Rule 101(c)(2)(ii) for asset-backed securities that are offered pursuant to an effective shelf registration statement filed on Form SF–3. In 2014, the Commission adopted shelf eligibility criteria for asset-backed securities offerings registered on new Form SF–3 in part to implement section 939A(b) of the Dodd-Frank Act.131 The Commission designed the shelf eligibility requirements to help ensure a

132See Regulation AB II Adopting Release, 79 FR 57189. 133See Regulation AB II Adopting Release, 79 FR 57267. 134See Regulation AB II Adopting Release, 79 FR 57278. 135See Regulation AB II Adopting Release, 79 FR 57283. 136Regulation AB II Adopting Release, 79 FR 57265, 57285. 137See SIFMA Letter 1, at 11. 138 IILF Letter, at 6. 139 IILF Letter, at 6–7 (stating that asset-backed securities are widely traded and have frequently quoted prices and credit spreads and that it is straightforward to calculate the probability of default based on these market measures). For the reasons discussed above, in Part II.A.1, the Commission is requiring that an issuer’s probability of default be derived from a structural credit risk model, and not from other market measures. 140Proposal, 87 FR 18323 (requesting comment on whether a probability-of-default-based standard would be appropriate for the exception for asset￾backed securities; whether there are models that are used to calculate a probability of default threshold for asset-backed securities that would be relevant to consider based on the type of security involved and, if so, what the threshold should be; what benefits this approach would provide; what other concerns this approach could raise; and how this approach would address potential conflicts of interest involving the distribution participant or affiliated purchaser making the determination). 141 IILF Letter, at 7 (citing Flannery, Houston & Partnoy, Credit Default Swap Spreads, 158 U. PA.
L. REV. 2085, 2087 (2010)). The available data
referenced in the commenter’s statement, as well as the article the commenter cited, pertains to debt and credit default swaps. It does not appear from the cited article that the analysis performed related to credit default swaps was performed with regard to asset-backed securities. Further, this commenter did not provide similar information about asset￾backed securities. 142See infra Part VI.B (discussing model inputs). 143See infra Part V.E.6. 144 IILF Letter, at 8. 145 IILF Letter, at 8. 146See Regulation M Adopting Release, 62 FR 527; see also Prohibitions Against Trading by Persons Interested in a Distribution, Release No. 34–19565 (Mar. 4, 1983) [48 FR 10628, 10631 (Mar. 14, 1983)] (stating that the ‘‘fungibility’’ of certain types of securities makes manipulation of their price very difficult). 147See Regulation M Adopting Release, 62 FR
527. 148See Regulation M Adopting Release, 62 FR
527. 149See Regulation M Adopting Release, 62 FR 527
(citations omitted). The Commission stated that such rationale also applies to the existing identical exception in Rule 102(d)(2) of Regulation M. Regulation M Adopting Release, 62 FR 531. certain ‘‘quality and character’’ in light of the requirement to reduce regulatory reliance on credit ratings.132 The shelf eligibility requirements included in Form SF–3 are designed to help ensure that the securitization is designed to produce expected cash flows that are sufficient to service payments or distributions in accordance with their terms; 133 that obligated parties more carefully consider the characteristics and quality of the assets that are included in the pool; 134 that asset￾backed securities shelf offerings have transactional safeguards and features that make those certain securities appropriate to be issued without prior Commission staff review; 135 and that issuers design and prepare asset-backed securities offerings with greater oversight and care.136 The asset-backed securities offered pursuant to an effective shelf registration statement filed on Form SF–3 should trade primarily on the basis of yield and creditworthiness, rather than on the identity of a particular issuer and its idiosyncratic risk. One commenter supported the adoption of the proposed exception for asset-backed securities that are offered pursuant to an effective shelf registration statement filed on Form SF–3 and stated that it appreciated the straightforward nature of the standard, which allows all interested parties to easily determine whether the exception is available.137 Another commenter disagreed with the Commission’s statement that an exception for asset-backed securities that is based on a probability of default threshold may be unfeasible.138 This commenter stated that market participants are able to estimate the probability of default for these securities, not only using statistical models but also based on market measures such as credit spreads.139 In response to Request for Comment (‘‘RFC’’) 29,140 this commenter stated that a probability of default standard based on market measures would have indicated that the exception to Regulation M no longer applied for certain debt securities during the months leading up to the collapse of Lehman Brothers, and available data shows that, if such a market measure￾based standard had been implemented, the exceptions for Regulation M would not have been available for debt securities as early as fall 2007.141 Despite this example, the Commission has considered this comment and determined that an exception for asset￾backed securities that is based on a structural credit risk model to derive an issuer’s probability of default would be unfeasible because distribution participants (including those acting as the lead managers) may not be able to collect all of the information required to calculate the probability of default, such as the value and volatility of the equity.142 In other words, practical challenges of obtaining
reliable fundamental information about the equity would make a probability of default measure unfeasible for an exception for asset-backed securities that trade on the basis of their yield and creditworthiness. The Commission has also determined that a measure based on credit spreads or the use of other models, such as reduced-form models, would not be appropriate to use due to their flexible or unstructured nature, which could result in a standard that can be used to abuse the exception.143 The same commenter stated that ‘‘any final rules governing asset-backed securities also could reference expected recovery in the event of default and default correlation.’’ 144 The commenter suggested to require under the exception that the applicable market participant determine and document a conclusion that the credit risk associated with a security was ‘‘minimal’’ (or some other similar standard) based on these variables, without any requirement that they use a particular model.145 The commenter’s suggested exception, by conditioning reliance on a list of variables and a judgment of ‘‘minimal’’ credit risk, without any bright-line requirements to help deter abuse of the exception through self-serving conclusions, would not be sufficiently objective. The Commission continues to believe that its original basis for excepting securities of a certain quality and character is appropriate and that such securities are less at risk of the manipulation that Regulation M addresses.146 As discussed above, the Commission excepted investment grade asset-backed securities from Rule 101 because such securities trade primarily on the basis of yield and creditworthiness (traditionally measured by credit ratings).147 In providing this rationale, the Commission stated that the principal focus of investors in the asset-backed securities market is on the structure of a class of securities and the nature of the assets pooled to serve as collateral for those securities rather than on the identity of a particular issuer.148 The Commission also stated that Rule 101(c)(2) excepts investment grade securities that are ‘‘primarily serviced by the cashflows of a discrete pool of receivables or other financial assets, either fixed or revolving, that by their terms convert into cash within a finite time period plus any rights or other assets designed to assure the servicing or timely distribution of proceeds to the security holders.’’ 149 As discussed above, in this Part, the practical challenge of obtaining reliable fundamental information about the equity makes a probability of default determination difficult or infeasible. The Commission believes that an appropriate and pragmatic approach is

150See, e.g., supra notes 60–61 (contrasting the focus of creditworthiness in the eligibility criteria and offering requirements included in Form SF–3 with that of other Commission Forms, such as Form S–3 and Form F–3). 151See Proposal, 87 FR 18321. 152See supra note 149. 153See 17 CFR 239.45(b)(v), (vi); Form SF–3, General Instruction I.B.1(e) and (f). 154Asset-Backed Securities, Release No. 33–8518 (Dec. 22, 2004) [70 FR 1506, 1517 (Jan. 7, 2005)] (‘‘Regulation AB Release’’). In adopting the 20% delinquency concentration level, the Commission codified a staff position that an asset-backed security will not fail to meet the definition of ‘‘asset-backed security’’ solely because such a security is supported by assets having total delinquencies of up to 20% at the time of the proposed offering. See Regulation AB Release, 70 FR 1517 (citing Bond Mkt. Ass’n, SEC Staff No￾Action Letter, 1997 WL 634124 (Oct. 8, 1997) (‘‘BMA NAL’’)). This threshold was the same threshold that was applied to certain other matters affecting registration and disclosure requirements for asset-backed securities (e.g., non-recourse commercial mortgage securitizations, pooling of corporate debt securities, and securitizations involving third-party credit enhancement). See BMA NAL, 1997 WL 634124, at * 3. The staff position was based on the premise that such a threshold for total delinquency concentration would, by itself, not present a materially greater risk of asset non-performance or default at the security level. See BMA NAL, 1997 WL 634124, at * 4. 155See Regulation AB Release, 70 FR 1517. 156See Regulation M Adopting Release, 62 FR 527. 157Form SF–3, General Instruction I.B.1(a)–(c). 158See Regulation AB II Adopting Release, 79 FR 57283. 159See Regulation AB II Adopting Release, 79 FR 57283. 160See supra notes 132–136. 161See Regulation AB II Adopting Release, 79 FR 57278. 162See, e.g., Regulation AB II Adopting Release, 79 FR 57277–78. 163See Regulation M Adopting Release, 62 FR 527. 164Public Law 111–203, sec. 939A(b) (requiring agencies to ‘‘seek to establish, to the extent feasible, uniform standards of credit-worthiness for use by each such agency, taking into account the entities regulated by each such agency and the purposes for which such entities would rely on such standards of credit-worthiness’’). to add an exception based on Form SF–3, as proposed, because it sufficiently focuses on creditworthiness.150 In addition, as stated in the Proposal, a standard that relies on Form SF–3 with respect to Nonconvertible Securities would not be appropriate because the transaction requirements included in Form SF–3 are relevant only to asset-backed securities and thus would not be a sufficient measure of creditworthiness for securities that are not subject to the Form SF–3 transaction requirements.151 The transaction requirements included in Form SF–3 allow for shelf offerings of only those asset-backed securities that share the qualities and characteristics of the
investment grade asset-backed securities currently excepted in Rule 101(c)(2): with respect to either set of securities, the principal focus of investors is the structure of a class of securities and the nature of the assets pooled to serve as collateral for those securities, rather than on the identity of a particular issuer.152 First, eligibility for offering securities pursuant to a Form SF–3 is limited, in
part, by the percentage of delinquent
assets and, for certain lease-backed securitizations, by the portion of the pool attributable to the residual value of the physical property underlying the leases.153 For an asset-backed securities offering with an effective Form SF–3, delinquent assets cannot constitute 20% or more of the asset pool. Delinquent assets may not convert into cash within a finite period of time, as required by the definition of ‘‘asset-backed security,’’ because they are not performing in accordance with their terms and management or that other action may be needed to convert the assets into cash. However, as the Commission stated at the time it adopted the 20% delinquency limitation for shelf eligibility, in principle, asset￾backed securities should be primarily dependent on the pool of assets self￾liquidating instead of on the ability of the entity performing collection services.154 The application of the limitation on delinquent assets was designed to ensure that attention is focused on the ability of collateral of the underlying asset pool to generate cash flow rather than on the identity of the issuer and its ability to convert those assets into cash,155 consistent with the Commission’s original basis for excepting investment grade asset-backed securities from Rule 101.156 Second, Form SF–3 includes certain transaction requirements with respect to the structure of the asset-backed security being offered. Such structural requirements include: (1) a certification by the depositor’s chief executive officer that, among other things, the securitization structure provides a reasonable basis to conclude that the expected cash flows are sufficient to service payments or distributions in accordance with their terms; (2) a review of the asset-backed security’s pool of assets upon the occurrence of certain triggering events, including delinquencies, by a person that is unaffiliated with certain transaction parties, such as the sponsor, depositor, servicer, trustee, or any of their affiliates; and (3) a dispute resolution provision, contained in the underlying transaction documents, for any repurchase request.157 When adopting the transaction requirements included in Form SF–3, the Commission stated that sponsors may have an increased incentive to carefully consider the characteristics of the assets underlying the securitization and accurately disclose these characteristics at the time of offering.158 The Commission also stated that investors should benefit from the reduced losses associated with nonperforming assets because, as a result of this new shelf requirement, sponsors will have less of an incentive to include nonperforming assets in the pool.159 Because the transactional safeguards included in Form SF–3 provide incentives for obligated parties to, among other things,160 more carefully consider the characteristics and quality of the assets that are included in the pool,161 asset-backed securities that are offered pursuant to an effective Form SF–3 should trade
based on their yield and creditworthiness rather than on the identity of a particular issuer.162 The requirement regarding an effective shelf registration statement filed on Form SF–3 is an appropriate substitute for the reference to credit ratings in the Investment Grade Exception because the standard is designed to limit eligibility for that exception to only those asset-backed securities that should trade based on their yield and creditworthiness due to their particular qualities and characteristics. Because the ability of distribution participants and their affiliated purchasers to bid up the price of an asset-backed security offered pursuant to an effective Form SF–3, during a distribution, is limited by a market participant’s ability to substitute the security with other securities that are similar and of comparable creditworthiness,163 such a security is less susceptible to the types of manipulation that Regulation M seeks to prevent. The application of the transaction requirements included in the Commission’s Form SF–3, therefore, should result in the offering of asset￾backed securities that have similar qualities and characteristics to the investment grade asset-backed securities currently excepted under the existing provision in Rule 101(c)(2). In addition, the exception for asset-backed securities that are offered pursuant to an effective shelf registration statement filed on Form SF–3 carries over the standard of creditworthiness included in the Commission’s Form SF–3 and helps to implement the mandate that, to the extent feasible, uniform standards of creditworthiness be used.164

165The final rule amendments make a non￾substantive, technical change that replaces the proposed reference to ‘‘17 CFR 239.45’’ with a reference to ‘‘§ 239.45 of this chapter’’ when referencing Form SF–3. 166See Proposal, 87 FR 18323–24. 167See Better Markets Letter, at 3–4 (stating that the proposal to eliminate, without replacing, the exception in Rule 102 for certain investment grade securities is appropriate because issuers and selling security holders have comparatively strong incentives to manipulate the price of the distributed security); IILF Letter, at 7. As discussed below, in
Part II.B.1, the new exception for Nonconvertible
Securities takes account of this consideration. 168 IILF Letter, at 7. 169SIFMA Letter 1, at 12–13. This commenter stated that the Investment Grade Exception in Rule 102 is relied upon in the context of sticky offerings and re-openings of debt issuances. See SIFMA Letter 1, at 4. 170See SIFMA Letter 1, at 12. With regard to replacing the Investment Grade Exception in Rule 102 pursuant to section 939A(b)’s requirements, the Commission requested comment on whether it should adopt an exception based on either the probability-of-default-based standard for Nonconvertible Securities or the standard for asset￾backed securities that are offered pursuant to an effective shelf registration statement filed on Form SF–3 instead of removing the Investment Grade Exception, without substituting an alternative, and whether it should adopt an exception in Rule 102 if a distribution participant determines that a security is an excepted security pursuant to Rule 101(c)(2). Proposal, 87 FR 18324 (request for comment (RFC) 35). One commenter replied, in response to RFC 35, that, to the extent the Commission receives comments that market participants on their own cannot make decisions and judgments about credit risk related to Rule 102, an exception based on probability of default would be a viable alternative. See IILF Letter, at 7. The Commission did not receive any such comment in response to the Proposal. 171See Regulation M Adopting Release, 62 FR 527; see also Prohibitions Against Trading by Persons Interested in a Distribution, Release, No. 34–19565 (Mar. 4, 1983) [48 FR 10628, 10631 (Mar. 14, 1983)]. 172See Proposal, 87 FR 18333. 173Regulation M Adopting Release, 62 FR 531. 174Regulation M Adopting Release, 62 FR 527. Accordingly, the Commission is adopting, substantially as proposed, with a technical change,165 an amendment to add a new exception in Rule 101(c)(2)(ii) for asset-backed securities that are offered pursuant to an effective shelf registration statement filed on Form SF–3. B. Rule 102(d)(2) of Regulation M:
Implementing Section 939A(b) in Certain Exceptions for Issuers and Selling Security Holders The Commission proposed to remove, without replacing, the Investment Grade Exception in Rule 102(d)(2) of Regulation M. The Commission stated in the Proposal that this removal without replacement was appropriate given that the retention of an exception for creditworthy Nonconvertible Securities and asset-backed securities would not likely be necessary to facilitate orderly distributions or limit disruptions in the trading market in light of issuers’ limited market access needs and the apparent limited reliance on Rule 102’s Investment Grade Exception, coupled with the incentive for issuers, selling security holders, and their affiliated purchasers to manipulate the market for the distributed security, regardless of the security’s credit quality.166 Two commenters supported the proposed elimination, without replacement, of Rule 102(d)(2)’s Investment Grade Exception.167 One commenter stated that it is ‘‘appropriate in contexts where deference arguably should be made to independent decisions and judgment by market participants, without the crutch of reliance on credit ratings.’’ 168 One commenter objected to the proposed elimination, without replacement, of the Investment Grade Exception in Rule 102(d)(2) and stated that the ‘‘continued availability of such an exception would be important to broker-dealers who are affiliated with an issuer but are not, for whatever reasons, serving as an underwriter or other participant in connection with the distribution.’’ 169 This commenter stated that the Commission, instead, should substitute the existing standard in Rule 102(d)(2) with ‘‘the same standards as used for purposes of the exceptions under Rule 101(c)(2).’’ 170 For reasons explained below, the Commission agrees with this commenter and is replacing the Investment Grade Exception in Rule 102, rather than eliminating the exception, without replacement, as proposed. The Commission continues to believe that its original basis for excepting Nonconvertible Securities and asset￾backed securities of a certain quality and character from Rule 102’s prophylactic prohibitions is appropriate and that the substitute standards discussed below are appropriate to ensure that those securities are less at risk of the manipulation that Regulation M addresses.171 The Commission is adopting rule amendments that remove the Investment Grade Exception from Rule 102 of Regulation M and substitute in its place exceptions based on alternative standards of creditworthiness to except Nonconvertible Securities. The reasons for adding these new exceptions are discussed below with regard to Nonconvertible Securities, in Part II.B.1, and asset-backed securities, in Part II.B.2.

1. Rule 102(d)(2)(i): Nonconvertible
   Securities of Issuers Who Meet a Specified Probability of Default Threshold The Commission acknowledges that eliminating the exception, without replacement, may impact entities, such as broker-dealers who are not distribution participants (and are not eligible to quality for an exception under Rule 101) but may qualify for a comparable exception under Rule 102 as a result of being an affiliate of an issuer or selling security holder and meeting the exception’s conditions. The continued availability of an exception for the Nonconvertible Securities will also provide issuers and selling security holders with more flexibility during distributions as compared to the Proposal. The elimination, without replacement, of the Investment Grade Exception from Rule 102 for issuers and selling security holders could increase issuance costs or deter market participants from issuing Nonconvertible Securities with low manipulation risk.172 In adopting the Investment Grade Exception, the Commission stated that it determined to include the Investment Grade Exception in Rule 102 based, in
   part, on the rationales indicated for an
   identical exception to Rule 101.173 The Commission excepted investment grade Nonconvertible Securities from Rule 101 ‘‘based on the premise that these securities traded on the basis of their yield and credit ratings, are largely fungible and, therefore, are less likely to be subject to manipulation.’’ 174 As discussed above, in Part II.A.1, the new exception in Rule 101(c)(2)(i) for Nonconvertible Securities of issuers for which the probability of default, estimated as of the sixth business day immediately preceding the determination of the offering price and over the horizon of 12 full calendar months from such day, is 0.055% or less, as determined and documented, in writing, by the distribution participant acting as the lead manager (or in a similar capacity) of a distribution, as derived from a structural credit risk model, is an appropriate substitute standard of creditworthiness in place of the reference to credit ratings in the Investment Grade Exception for Nonconvertible Securities. The standard of creditworthiness, which was the basis of the Investment Grade Exception for Nonconvertible Securities in Rule 102, is still appropriate to use as the basis of an

175See supra note 13 and accompanying text. 176See Regulation M Adopting Release, 62 FR 530. 177See Proposal, 87 FR 18323. 178See supra note 88. 179See Regulation M Adopting Release, 62 FR 530. 180See, e.g., Proposal, 87 FR 18330. As discussed above, in Part II.A.1, this may be the case in, for example, self-underwritten offerings, at-the-market offerings, or other shelf offerings, to the extent such an offering meets the definition of a ‘‘distribution’’ under Rule 100(b) of Regulation M. With regard to shelf offerings, each takedown is to be individually examined to determine whether such offering constitutes a ‘‘distribution.’’ Regulation M Adopting Release, 62 FR 526. If, as a result of the amendments, the exception for Nonconvertible Securities is no longer available in connection with a distribution, and if no other exception is available, Rule 102’s prohibitions would apply. Accordingly, an issuer and all of its affiliated purchasers would be subject to the applicable restricted period of Rule 102 when sales off a shelf by an issuer, or by any affiliated purchaser, constitute a distribution of securities. Similarly, when a selling security holder sells off the shelf and such sales constitute a distribution, all other shelf security holders who are affiliated purchasers of the selling security holder would be subject to the applicable restricted period of Rule 102. See Regulation M Adopting Release, 62 FR 531. 181See Proposal, 87 FR 18333. 182See Proposal, 87 FR 18323. exception to Rule 102 for Nonconvertible Securities.175 In addition, the standard of creditworthiness used in the exception in Rule 101(c)(2)(i), as amended, is an appropriate standard of creditworthiness to use in place of the reference to credit ratings in the Investment Grade Exception in Rule 102 pursuant to the requirements of section 939A(b). That standard, which is based on an issuer’s probability of default, is designed to identify Nonconvertible Securities that are less susceptible to the manipulation that Regulation M is designed to prevent because they trade based on their yield and creditworthiness, as determined by the current financial condition of the issuer. However, given that issuers and selling security holders have the greatest interest in an offering’s outcome,176 regardless of the credit quality of the security,177 it would not be appropriate for the exception to permit those parties to make their own probability of default determinations (by their own or a third party calculation) in order to meet the conditions of the exception. Therefore, Rule 102(d)(2)(i), as amended, uses the same bright-line test for excepting Nonconvertible Securities. For issuers, selling security holders, and their affiliated purchasers to use the exception in Rule 102(d)(2)(i), as amended, however, they must rely on the probability of default determination made by the distribution participant acting as the lead manager (or in a similar capacity) 178 of the distribution and
documented in writing pursuant to Rule 101(c)(2)(i), as amended. Rule 102(d)(2)(i), as amended, does not permit reliance on the exception if issuers, selling security holders, or their affiliated purchasers make the required probability of default determinations themselves, or rely on a determination made by a non-lead manager or any other third party. This condition to the exception means that issuers, selling security holders, and their affiliated purchasers would not be able to rely on the new Rule 102(d)(2)(i) exception when selling securities directly, unless a lead manager is involved in the distribution and had made (and documented) the qualifying probability of default determination. This condition provides for the continued availability of an exception under Rule 102 for creditworthy Nonconvertible Securities for broker-dealers who are affiliated with an issuer but are not serving as an underwriter or other participant in connection with the distribution. At the same time, this condition is designed to prevent abuse of the exception by issuers, selling security holders, and their affiliated purchasers by taking into account that these market participants have the greatest interest in an offering’s outcome and generally do not have the same market access needs as underwriters.179 In addition, the condition regarding lead-manager probability of default determinations in new Rule 102(d)(2)(i) is consistent with the condition regarding lead-manager probability of default determinations in new Rule 101(c)(2)(i). Even though probability of default determinations made by or directly for issuers or selling security holders or affiliated purchasers cannot be used in order for such parties to rely on the new exception in Rule 102(d)(2)(i) for Nonconvertible Securities, these parties would, however, be able to avail themselves of the exception in reliance on a probability of default determination made by the distribution participant acting as the lead manager (or in a similar capacity) of the distribution pursuant to Rule 101(c)(2)(i), as amended. Similar to how the lead or managing underwriter in a firm commitment offering communicates certain pricing, allocation, and other distribution￾related information to the issuer or selling security holder in connection with that particular distribution, the lead managing underwriter’s communications regarding its probability of default determination may vary based on the parties and their prior course of conduct as to the frequency and manner or mode of such communication. However, Rule 101(c)(2)(i), as amended, does not require that the lead manager making the probability of default determination share the determination with the issuer, selling security holders, or their affiliated purchasers in order for those parties to rely on the exception. Therefore, issuers, selling security holders, and their affiliated purchasers will not be able to rely on the exception for Nonconvertible Securities if the lead
manager does not share the probability of default determination or there is no distribution participant to act as the lead manager for the distribution. With regard to the types of distributions covered, as a result of the condition related to lead-manager probability of default determinations, the exception for Nonconvertible Securities is available to a subset of distributions covered under the existing Investment Grade Exception 180 but more distributions than what was covered under the Proposal given that the Proposal would have removed, and not replaced, the Investment Grade Exception in Rule 102.181 The estimated costs associated with the condition related to the lead manager making the probability of default determination are included below, in Part V.C.1. As discussed below, in Part II.C, broker-dealers who rely on the new exception in Rule 102(d)(2)(i) are required to preserve the written probability of default determination made pursuant to Rule 101(c)(2)(i), as amended. This record preservation requirement could help facilitate Commission examinations of broker￾dealers who rely on Rule 102(d)(2)(ii), as amended, and could deter their misuse of the exception through relying on determinations that do not meet the conditions of the exception or through relying on the exception when no determination has been made.
2. Rule 102(d)(2)(ii): Asset-Backed
Securities Offered Pursuant to an Effective Shelf Registration Statement Filed on Form SF–3 The Commission is adopting in new Rule 102(d)(2)(ii) an exception for asset￾backed securities that are offered pursuant to an effective shelf registration statement filed on Form SF–
3. The Commission stated in the
Proposal that the incentive for issuers, selling shareholders, and their affiliated purchasers to manipulate the market of a distributed security exists regardless of the credit quality of the security.182

183See supra notes 153–161 and accompanying text. 184Regulation M Adopting Release, 62 FR 531. 185See Regulation M Adopting Release, 62 FR 527. 186See Regulation M Adopting Release, 62 FR 527. 187See Regulation M Adopting Release, 62 FR 527 (citations omitted). 188See Regulation M Adopting Release, 62 FR 527. 189See supra note 149. 190See Regulation M Adopting Release, 62 FR 527. 191 In the Proposal, this requirement was described as a ‘‘recordkeeping’’ requirement. For clarity and consistency with the title of Rule 17a– 4 (‘‘Records to be preserved by certain exchange members, brokers, and dealers’’), this requirement is referred to throughout this release as a record preservation requirement. 192Proposal, 87 FR 18324. While this incentive may exist, transaction requirements included in Form SF–3 allow for shelf offerings of only those asset-backed securities that share the qualities and characteristics of the investment grade asset-backed securities that meet the Investment Grade Exception 183 and thus are less likely to be subject to the type of manipulation that Regulation M seeks to address. In addition, in contrast to how Nonconvertible Securities would be excepted, whether an asset-backed security is rated investment grade is an objective, observable fact, as is whether an asset-backed security is offered pursuant to an effective shelf registration statement filed on Form SF–
3. Reliance on the new exception in
Rule 102(d)(2)(i) does not require issuers, selling security holders, or their affiliated purchasers to make a calculation in determining whether the subject asset-backed security meets the conditions of that exception (i.e., that the asset-backed security is offered pursuant to an effective shelf registration statement filed on Form SF– 3), in contrast to how reliance on the new exception in Rule 101(c)(2)(i) or Rule 102(d)(2)(i) for Nonconvertible Securities is conditioned on a calculation determining whether the issuer’s probability of default meets the specified threshold. Because the Investment Grade Exception for asset￾backed securities does not focus on the potential interests of those covered persons seeking to rely on the exception but rather the particular qualities of the securities themselves (i.e., that the asset￾backed securities are appropriate to except from Regulation M because they trade on the basis of their yield and creditworthiness, traditionally measured by credit ratings, and are largely fungible), the measure based on the Form SF–3 shelf eligibility requirements, which similarly focuses on the particular qualities of the asset￾backed securities, is an appropriate substitute standard of creditworthiness to replace the reference to credit ratings in the existing Investment Grade Exception in accordance with section 939A(b)’s requirements, without having to restrict or place any further conditions on who may rely on the exception under Rule 102(d)(2)(ii), as amended. The Commission stated that it determined to include the Investment Grade Exception in Rule 102 based, in
part, on the rationales indicated for an
identical exception to Rule 101.184 As discussed above, in Part II.A.2, the Commission excepted investment grade asset-backed securities from Rule 101 because such securities trade primarily on the basis of yield and credit rating.185 When the Commission adopted the Investment Grade Exception in Rule 101, it stated that the principal focus of investors in the asset-backed securities market is on the structure of a class of securities and the nature of the assets pooled to serve as collateral for those securities rather than on the identity of a particular issuer.186 The Commission also stated that the Investment Grade Exception is for securities that are ‘‘primarily serviced by the cashflows of a discrete pool of receivables or other financial assets, either fixed or revolving, that by their terms convert into cash within a finite time period plus any rights or other assets designed to assure the servicing or timely distribution of proceeds to the security holders.’’ 187 The standard in new Rule 102(d)(2)(i) that relies on the Form SF–3 eligibility requirements continues to be derived from the premise that certain asset￾backed securities are traded based on factors such as their yield and creditworthiness.188 As discussed above, in Part II.A.2, the transaction requirements included in Form SF–3 allow for shelf offerings of only those asset-backed securities that share the qualities and characteristics of the investment grade asset-backed securities that meet the Investment Grade Exception: with respect to either set of securities, the principal focus of investors is the structure of a class of securities and the nature of the assets pooled to serve as collateral for those securities, rather than on the identity of a particular issuer.189 The application of the transaction requirements included in the Commission’s Form SF–3, therefore, should result in the offering of asset￾backed securities that have similar qualities and characteristics to the asset￾backed securities currently excepted under Rule 102’s Investment Grade Exception. Because the ability of issuers, selling security holders, and their affiliated purchasers to bid up the price of an asset-backed security offered pursuant to an effective Form SF–3, during a distribution, is limited by a market participant’s ability to substitute the security with other securities that are similar and of comparable creditworthiness,190 such a security is less susceptible to the types of manipulation that Regulation M seeks to prevent. In accordance with section 939A(b), it is appropriate to continue to except in Rule 102(d)(2) asset-backed securities that trade on the basis of their yield and creditworthiness. For these reasons, the Commission is adopting in new Rule 102(d)(2)(ii) an exception for asset-backed securities that are offered pursuant to an effective shelf registration statement filed on Form SF–3. The new exception in Rule 102(d)(2)(i) may be relied upon by issuers, selling security
holders, and their affiliated purchasers if all conditions of the exception are met.
C. Exchange Act Rule 17a–4(b)(17):
Adding a Record Preservation Requirement for Broker-Dealers in Connection With Probability of Default Determinations The Commission proposed a new record preservation 191 requirement that broker-dealers who are distribution participants or affiliated purchasers must preserve certain records pursuant to Rule 17a–4 under the Exchange Act, the Commission’s broker-dealer record retention rule. Proposed paragraph (b)(17) of Rule 17a–4 would have required broker-dealers relying on the exception for Nonconvertible Securities to preserve the written probability of default determination made pursuant to proposed paragraph (c)(2)(i) of Rule 101. Accordingly, those broker-dealers would be required to preserve for a period of not less than three years, the first two years in an easily accessible place, the written probability of default determination made pursuant to proposed paragraph (c)(2)(i) of Rule 101.192 One commenter stated that the proposed record preservation requirement for certain broker-dealers is ‘‘plainly appropriate as a means of facilitating the Commission in its examination and oversight of broker￾dealers who rely on the exception in Rule 101 and would be required to

193Better Markets Letter, at 4. 194SIFMA Letter 1, at 12; see also SIFMA Letter 2, at 2. 195 17 CFR 242.101(c)(1). 196 17 CFR 242.101(c)(3). 197 17 CFR 242.101(c)(4). 198See infra Part V.C.3; see also Proposal, 87 FR 18334. 199 IILF Letter, at 8. 200Proposal, 87 FR 18324. 201See, e.g., 17 CFR 242.101(c)(2)(i), as amended, 242.102(d)(2)(i), as amended. Both of the new exceptions for Nonconvertible Securities in Rules 101 and 102 require that the issuer’s probability of default be documented and determined, in writing, without necessarily requiring the other information included in the commenter’s suggestion. 202See Proposal, 87 FR 18329; infra Part V.A.2. 203See, e.g., infra Part V.C. 204Proposal, 87 FR 18324. 205New Rule 17a–4(b)(17) requires broker-dealers to preserve the written probability of default determination, relied upon by such broker-dealer, pursuant to § 242.101(c)(2)(i) or § 242.102(d)(2)(i) (Rule 101 or Rule 102 of Regulation M), as applicable. conduct the new probability of default determination.’’ 193 Another commenter stated that the Commission should not adopt the proposed record preservation requirement for broker-dealers relying on new Rule 101(c)(2)(i)’s exception for Nonconvertible Securities because ‘‘firms are already subject to extensive recordkeeping requirements [and] should continue to have flexibility in determining the precise nature and types of records they make and retain for such purpose, just as they do for purposes of the various other exceptions to Rule 101 of Regulation M.’’ 194 However, unlike the other securities￾based exceptions in Rule 101, which apply to ‘‘actively-traded securities,’’ 195 ‘‘exempted securities,’’ 196 and ‘‘face￾amount certificates or securities issued by an open-end management investment company or unit investment trust,’’ 197 and which are based on standards that rely on the use of publicly available information that can be verified, this exception is subject to the specific requirements in section 939A(b) to use ‘‘standards of credit-worthiness.’’ As discussed above, in Parts II.A.1 and B.1, the probability of default, as derived from structural credit risk models, is an appropriate substitute standard of creditworthiness to replace the reference to credit ratings in the existing Investment Grade Exceptions in accordance with section 939A(b)’s requirements. Due to the number of variations among structural credit risk models and their estimated inputs, the probability of default estimates may be subjective to some extent.198 As discussed below, in Part V.A.2, creditworthiness is an appropriate standard to reflect manipulation risk because securities issued by firms with sound creditworthiness trade primarily on yield and creditworthiness (traditionally measured by credit ratings) and have low pricing uncertainty and manipulation risk. Reliance on the new exception in Rule 101(c)(2)(i) or Rule 102(d)(2)(i) for issuers of Nonconvertible Securities is conditioned on the use of a written
probability of default calculation that has been determined and documented, in writing, by the distribution participant acting as the lead manager. This exception is in contrast to the other Regulation M exceptions that require the use of publicly available information that can be verified. Accordingly, requiring a record of the written probability of default determination to be preserved will help facilitate the Commission’s examinations of broker￾dealers relying on the new exception in Rule 101(c)(2)(i) or Rule 102(d)(2)(i). The record preservation requirement, therefore, is appropriate to help deter improper adjusting of the estimation to meet the conditions of either of the exceptions. Further, because probability of default estimates may be subjective to some extent and not comparable across different issuers or for the same issuer across different issues if estimates are based on different models, or done by different researchers or vendors, the requirement associated with reliance on new Rule 101(c)(2)(i) to preserve written probability of default determinations is designed to facilitate the Commission’s examinations of broker-dealers. Another commenter stated that the Commission could publish, or require publication of, point estimates at a particular time and as rolling averages of default probabilities during a specified period, default probability estimates that market participants derive from various models, along with default probabilities implied by both market prices and credit default swap spreads (to the extent those are traded for a particular issuance).199 While access to such information could be informative for certain market participants and investors, the record preservation requirement set forth in new Rule 17a–4(b)(17), as stated in the Proposal, was designed to aid the Commission in its examinations of broker-dealers relying on the exception in Rule 101(c)(2)(i), as amended, by requiring such broker-dealers to retain the written probability of default determination supporting their reliance on the exception.200 As such, a requirement for these entities to publish the information from the commenter’s suggestion would not serve this purpose because such a requirement may not necessarily involve the type of information needed to meet the conditions new Rule 101(c)(2)(i) or new Rule 102(d)(2)(i) 201 and, therefore, would not facilitate the Commission’s examinations of broker-dealers relying on those exceptions. In addition, the cost burden of doing so on a regular basis could be disproportionate to the infrequent usage of the exception, as these entities could incur other burdens associated with disclosing such information.202 Accordingly, it could discourage some entities from participating in certain issues, which could increase the costs of the affected issues.203 Similarly, for the Commission to publish this information, such that parties could rely on the information, would also not be appropriate because this approach
would not facilitate its examinations of broker-dealers relying on the exception. After reviewing the comments, the Commission is adopting Rule 17a–4(b) under the Exchange Act largely as proposed, by adding new paragraph (17), which requires broker-dealers to preserve the written probability of default determination, relied upon pursuant to the new exception in Rule 101(c)(2)(i) or Rule 102(d)(2)(i), as applicable. As discussed above in Part II.B.1, the new exception in Rule 102(d)(2)(i) was not originally proposed. However, proposed Rule 17a–4(b)(17) was intended to capture ‘‘broker-dealers relying on the exception for Nonconvertible Securities’’ 204 and, therefore, the Commission believes it is appropriate to apply it to both exceptions for Nonconvertible Securities that are being adopted. Moreover, broker-dealers relying on either exception should be in a position to comply with the requirements of new Rule 17a–4(b)(17) because both exceptions require that the lead manager of a distribution make and document in writing the probability of default determination pursuant to Rule 101(c)(2)(i), as amended.205 Accordingly, the final rule adds ‘‘or § 242.102(d)(2)(i) . . . or Rule 102 . . . as applicable’’ in light of the addition of the new exception for Nonconvertible Securities in Rule 102(d)(2)(i) and that broker-dealers may be relying on the new exception either in Rule 101(c)(2)(i) or in Rule 102(d)(2)(i), depending on whether they are a covered person under Rule 101 or Rule 102. In addition, the final rule adds the text ‘‘, relied upon by such broker-dealer,’’ to clarify that the written probability of default determination must be preserved in connection with a broker-dealer’s

206See Proposal, 87 FR 18325.
207 17 CFR 240.17a–4(b).
208Letter from Brian (Mar. 25, 2022). 209Letter from Patrick Lawson (Mar. 26, 2022). 210SIFMA Letter 2, at 4–5. 211Letter from Jason Wallace (May 19, 2022). 212See, e.g., Letter from Anthony Frattin (May 19, 2022); Kern Letter; Wang Letter; Ferguson Letter; Navari Letter. 213See Better Markets Letter, at 2. 214See Letter from Senator Thom Tillis to Vanessa Countryman, Sec’y, SEC (Nov. 4, 2022). 215Letter from The Delois Albert Brassell Estate and the Robert James Brassell Estate (June 1, 2022). 216 5 U.S.C. 801 et seq. 217 15 U.S.C. 78c(f). 218 15 U.S.C. 78w(a)(2). 219The term ‘‘fixed-income securities’’ in the Economic Analysis section refers to nonconvertible debt securities, nonconvertible preferred securities, and asset-backed securities. reliance on the new exception in Rule 101(c)(2)(i) or in Rule 102(d)(2)(i), as applicable. New paragraph (b)(17) of Rule 17a–4 would affect the existing practices of broker-dealers by imposing new record preservation requirements when relying on the exception in new Rule 101(c)(2)(i) or new Rule 102(d)(2)(i). A broker-dealer who is a distribution participant acting as the lead manager (or in a similar capacity) of a distribution and uses a vendor to determine the probability of default could satisfy this record preservation requirement by maintaining documentation of the assumptions used in the vendor model, as well as the output provided by the vendor supporting the probability of default determination. Such a broker-dealer calculating the probability of default on its own could satisfy the record preservation requirement by maintaining documentation of the value of each variable in deriving the probability of default, along with a record identifying the specific source(s) of such information for each variable. Other broker-dealers, namely those that rely on the written probability of default determination of another broker-dealer acting as the lead manager (or in a similar capacity), could satisfy the record preservation requirement by maintaining a copy of the documentation described above, or by retaining a written notice it received of the probability of default determination. The requirement to preserve, pursuant to Rule 17a–4(b), the written probability of default determination is consistent with other record retention obligations that Exchange Act rules impose on broker-dealers.206 Exchange members and broker-dealers currently are required to comply with the three-year preservation period in Rule 17a–4(b) for other records and should have in place procedures to satisfy such preservation requirements.207
III. Other Issues
Certain commenters urged the Commission to take additional or different regulatory and non-regulatory actions than the approaches that were proposed, including actions that the Commission did not propose. These suggestions covered a variety of areas, including use of the term ‘‘investors,’’ 208 SEC enforcement actions,209 other provisions of Regulation M,210 insurance company ratings,211 individual securities,212 credit ratings industry reforms,213 agency operations,214 and nondisclosures.215 These issues are outside the scope of the Proposal and that the final amendments to Rules 100(b), 101(c)(2), 102(d)(2), and 17a– 4(b) appropriately further the Commission’s objectives of promoting investor protection, enhancing market efficiency, and facilitating capital formation by implementing the requirements of section 939A(b) of the Dodd-Frank Act and facilitating the Commission during examinations of broker-dealers.
IV. Other Matters
If any of the provisions of these rules, or the application thereof to any person or circumstance, is held to be invalid, such invalidity shall not affect other provisions or application of such provisions to other persons or circumstances that can be given effect without the invalid provision or application. Pursuant to the Congressional Review Act,216 the Office of Information and Regulatory Affairs has designated these rules as not a ‘‘major’’ rule as defined by 5 U.S.C. 804(2).
V. Economic Analysis
The Commission is sensitive to the economic consequences and effects, including costs and benefits, of its rules. Some of these costs and benefits stem from statutory mandates, while others are affected by the discretion exercised in implementing the mandates. Section 3(f) of the Exchange Act 217 provides that whenever the Commission is engaged in rulemaking pursuant to the Exchange Act and is required to consider or determine whether an action is necessary or appropriate in the public interest, the Commission shall also consider, in addition to the protection of investors, whether the action will promote efficiency, competition, and capital formation. Additionally, section 23(a)(2) of the Exchange Act 218 requires the Commission, when making rules under the Exchange Act, to consider the impact such rules would have on competition. Section 23(a)(2) also provides that the Commission shall not adopt any rule which would impose a burden on competition that is not necessary or appropriate in furtherance of the purposes of the Exchange Act. The analysis below addresses the likely economic effects of the amendments, including the anticipated benefits and costs of the amendments, and their likely effects on efficiency, competition, and capital formation. The Commission also discusses the potential economic effects of certain alternatives to the approach taken by these amendments. Some of the benefits and costs discussed below are impracticable to quantify. For example, sticky offerings are generally not identified in the available data and may be difficult to trace in the appropriate records of the distribution participants. Therefore, much of the discussion of economic effects is qualitative. A. Baseline

1. The Investment Grade Fixed Income
   Market
   To assess the economic effects of the amendments, the Commission is using as the baseline the nonconvertible debt, nonconvertible preferred, and asset￾backed securities markets as they exist at the time of this release, including applicable rules that the Commission has already adopted. The affected parties include Nonconvertible Securities and asset￾backed securities (collectively ‘‘fixed￾income securities’’) 219 distribution and other market participants, such as issuers, selling security holders, underwriters, banks, broker-dealers, and their affiliated purchasers; fixed-income security investors, such as retail investors, mutual funds, exchange traded funds, and separate investment accounts; vendors of the relevant market data; and nationally recognized statistical rating organizations (‘‘NRSROs’’). Currently a majority of the distribution participants in the relevant markets are subscribed to a major vendor of the market data necessary to evaluate various aspects of the distribution. Further, a rating by an NRSRO is necessary in order for distribution participants to rely on the Investment Grade Exception. Today there are ten credit rating agencies registered with the Commission as

220See Current NRSROs, U.S. Sec. & Exch. Comm’n, available at https://www.sec.gov/about/ divisions-offices/office-credit-ratings/current-nrsros. 221See Staff Report on Nationally Recognized Statistical Rating Organizations (Feb. 2023) at 23, available at https://www.sec.gov/files/2023-ocr￾staff-report.pdf. 222The nonconvertible debt securities also include preferred securities. 223The statistics are based on the data from Mergent. Some agents were reported as performing two or more functions, for example as an underwriter and as a lead underwriter. 224The information is based on EDGAR data for public offerings of asset-backed securities. It should be noted that prospectuses may contain multiple tranches, including non-offered tranches excluded from the public offering of asset-backed securities. 225See supra Part I. 226See 17 CFR 242.101(a), 242.102(a); see, e.g., 17 CFR 242.101(c)(2), 242.102(d)(2). 227See Regulation M Adopting Release, 62 FR 527. 228There are other metrics that could serve as a proxy for manipulation risk in Rules 101 and 102, such as security public float or visibility to other market participants. One commenter for instance proposed Form S–3 and F–3 standard and a WKSI￾based standard to measure manipulation risk (SIFMA Letter 1). However, unlike measures of creditworthiness, such criteria fail to capture the pricing point where the security is trading solely based on its yield and maturity and thus has low pricing uncertainty and low manipulation risk. Therefore, measures of creditworthiness are a better proxy for manipulation risk. 229See Proposal, 87 FR 18316. Note, however, that not every foreign sovereign issue is conducted in the form of re-opening. 230See Proposal, 87 FR 18316. In a ‘best-effort’ offering, the underwriters are not required to sell any specific number or dollar amount of securities but will use their best efforts to sell the securities offered. See Plain English Disclosure, Release No. 34–38164, (Jan. 14, 1997) [62 FR 3152 (Jan. 21, 1997)]. Note, however, that not every best-effort offering will become sticky, where the underwriter is unable to sell all of the securities in the distribution. 231See Proposal, 87 FR 18329. Note that the Commission received no comments on the types of issues that typically rely on the exception. 232The estimate is obtained using Mergent data for relevant securities during 2021. 233See Letter from Kenneth E. Bensten, Jr., Executive Vice President, Public Policy and Advocacy, SIFMA to Elizabeth M. Murphy, Secretary (July 5, 2011) at 6; John Berkery & Remmelt Reigersman, Re-openings: Issuing Additional Debt Securities of an Outstanding Series, Mayer Brown 1–2 (2020), available at https://www.mayerbrown.com/-/media/files/ perspectives-events/publications/2020/05/ reopenings_-issuing-additional-debt-securities-of￾an-outstanding-series.pdf. See also Proposal, 87 FR 18329. 234See Proposal, 87 FR 18329. 235We note, however, that not all sticky offerings are issued by an issuer
with a low creditworthiness and have a high manipulation risk. It is thus important to have a standard of creditworthiness that is able to capture most recent available information on issuer creditworthiness, such as probability of default, and account for the cases of possible sudden declines in creditworthiness. 236See supra Part I for a relevant discussion. 237SIFMA Letter 1, at 12–13. NRSROs.220 Three large NRSROs (S&P Global Ratings, Moody’s Investors Service, Inc., and Fitch Ratings, Inc.) have historically accounted for most of the market share in this market. As of December 31, 2021, these three market participants accounted for 94.4% of all of the NRSRO credit ratings outstanding.221 The affected securities are nonconvertible debt, nonconvertible preferred, and asset-backed securities. In 2021, there were 33,798 issues of nonconvertible debt securities,222 with 687 issuers and 301 agents involved (266 reported as participating underwriters, of which 201 were the lead underwriters; 39—as trustees, and 10—as fiscal agents).223 Additionally, in 2021, there were 114 filed prospectuses for public offerings of asset-backed securities.224
2. The Investment Grade Exception
Regulation M is designed to prevent manipulative activities that could artificially influence the demand and pricing of covered securities.225 In particular, Rules 101 and 102 of Regulation M prohibit distribution and certain other market participants from bidding for or purchasing a covered security, in connection with a distribution of securities unless an exception, such as the Investment Grade Exception, applies.226 At the time the exception was included, the investment grade securities, that is securities characterized by sound creditworthiness, as measured by credit rating, were considered to be traded primarily on yield and credit ratings, and to be largely fungible.227 Therefore, sound creditworthiness was considered to be a good proxy for low manipulation risk. Investment Grade issues were presumed to have low probability of default and were thus considered to have low pricing uncertainty and low manipulation risk, which formed the basis for the exception. For purposes of these amendments sound creditworthiness is a good proxy for low manipulation risk since securities issued by firms with sound creditworthiness trade primarily on yield and creditworthiness (traditionally measured by credit ratings).228 Further, none of the commenters on the Proposal raised concern that creditworthiness would not be an adequate proxy for manipulation risk. The application of the Investment Grade Exception to Rules 101 and 102 is primarily limited to two cases: re￾openings (an offering of an additional principal amount of securities that are identical to the securities already outstanding, for example, when an issuer wishes to make a series of offerings via a re-opening to match its funding needs or when some foreign sovereign issuers conduct a re-opening for public finance purposes 229) and sticky offerings (an offering where a lack of demand results in an underwriter being unable to sell all of the securities in a distribution, for example, when an investor failed to honor a previously expressed indication of interest; also, as stated in the Proposal, another example a commenter provided is in a best￾efforts offering 230).231 Re-openings are used infrequently and constitute about 0.3% of the relevant securities’ markets’ issuance volume.232 Sticky offerings are not identified in the relevant databases, making it difficult to assess their relative magnitude. Re-openings are used in situations when such financing method offers the benefit of cost-effectiveness. For example, it may be cheaper for an issuer to offer a series of small offerings as opposed to one large offering, as the latter could result in a lower offering price due to the supply pressure. Further, since a re-opening issue is fungible with securities already in circulation and can be traded interchangeably with these securities in the secondary market, it provides additional liquidity benefits to the investors.233 As discussed above, sticky offerings typically result when a large investor
fails to fulfill its expressed purchase interest in the issue, which could be due to a negative factor that transpired about the issue or issuer.234 Any offering of the relevant security thus can become a sticky offering. In such cases it may become challenging to trade the issue based solely on the yield and maturity (otherwise it would have become possible to find another purchaser in a timely manner). This may give rise in some cases to a heightened risk of manipulation in connection with a distribution of securities even if the security is rated as investment grade.235 Rule 102 provides that, in connection with a distribution of securities effected by or on behalf of an issuer or selling security holder, it shall be unlawful for such person, or any affiliated purchaser of such person, directly or indirectly, to bid for, purchase, or attempt to induce any person to bid for or purchase, a covered security during the applicable restricted period.236 Issuers and selling security holders generally do not have the same market access needs as underwriters and are not expected to buy the securities they are issuing. However, as pointed out by one of the commenters, their affiliated broker￾dealers, which do not serve as an underwriter, may seek to rely on Rule 102 exception.237

238See supra Part I.
239We note that the SEC staff took a similar position in the COVID–19 Market Monitoring Group, Credit Ratings, Procyclicality and Related Financial Stability Issues: Select Observations, SEC Staff (July 15, 2020) (‘‘Cost of debt capital is driven by a wide range of financial and non-financial factors and forces; ratings downgrades are generally lagging indicators of cost of debt capital.’’), available at https://www.sec.gov/news/public￾statement/covid-19-monitoring-group-2020-07-15. 240Public Law 111–203, sec. 939A(a). The Commission has issued several releases concerning the removal of references to credit ratings: Security Ratings, Release No. 34–64975 (July 27, 2011) [76 FR 46603 (Aug. 3, 2011)]; Removal of Certain References to Credit Ratings Under the Securities Exchange Act of 1934, Release No. 34–71194 (Dec. 27, 2013) [79 FR 1522 (Jan. 8, 2014)]; Removal of Certain References to Credit Ratings under the Investment Company Act, Release No. IC–30847 (Dec. 27, 2013) [79 FR 1316 (Jan. 8, 2014)]; Asset￾Backed Securities Disclosure and Registration, Release No. 34–72982 (Sept. 4, 2014) [79 FR 57184 (Sept. 24, 2014)]; Removal of Certain References to Credit Ratings and Amendment to the Issuer Diversification Requirement in the Money Market Fund Rule, Release No. IC–31828 (Sept. 16, 2015) [80 FR 58124 (Sept. 25, 2015)]. 241See, e.g., the seminal model by Robert C. Merton, On the Pricing of Corporate Debt: The Risk Structure of Interest Rates, 29 Journal of Finance 449, 449–70 (1974), along with related successive refinement models such as Fischer Black & John C. Cox, Valuing Corporate Securities: Some Effects of Bond Indenture Provisions, 31 J. Fin. 351, 351–67 (1976); Robert Geske, The Valuation of Corporate Liabilities as Compound Options, 12 J. Fin. & Quantitative Analysis 541, 541–52 (1977); and Oldrich A. Vasicek, Credit Valuation, KMV (Mar. 22, 1984), among others. 242For example, the Merton (1974) Model and the Successor Models are included in the curriculum for such credentials as the Chartered Financial Analyst. See, e.g., Credit Analysis Models, ≤CFA Inst. (2022), available at https:// www.cfainstitute.org/en/membership/professional￾development/refresher-readings/credit-analysis￾models. One commenter, however, suggested that ‘‘most of our member firms do not use them [the credit risk models] for other purposes either, to the extent such models are used at all, they serve merely as a supplement to member firms’ own proprietary credit analysis as part of their decision making on whether to extend a loan or other credit.’’ (SIFMA Letter 1 at 5). See also Part II.A.1 for a relevant discussion. 243An alternative set of models used to derive probability of default are ‘reduced-form models’. The reduced-form models rely on statistical analysis rather than the balance sheet to determine a firm’s creditworthiness. However, compared to structural credit risk models, they lack in rigorous theoretical justification as well as economic interpretation
of the resulted relationships between the model inputs. See, e.g., Edward Altman, Andrea Resti, & Andrea Sironi, Default Recovery Rates in Credit Risk Modeling: A Review of the Literature and Empirical Evidence, 33 Econ. Notes 183 (2004) (discussing the competing models), available at https://onlinelibrary.wiley.com/doi/10.1111/j.0391- 5026.2004.00129.x. 244See, e.g., IILF Letter, Bloomberg L.P. Letter, and Better Markets Letter. 245See SIFMA Letters 1 and 2 and the relevant discussion in Part II.A.1. 246See Bloomberg L.P. Letter, at 2. 247See SIFMA Letter, at 10. The Investment Grade Exception was included in Regulation M as it was considered a good proxy for the likelihood of manipulation risk.238 However, the reference to credit ratings in the Commission’s rules may encourage investors to place undue reliance on the credit ratings. Credit ratings themselves are potentially imprecise and often lagging indicators of creditworthiness.239 B. Benefits of the Amendments As mentioned above, section 939A(b) of the Dodd-Frank Act requires the Commission to ‘‘remove any reference to or requirement of reliance on credit ratings, and to substitute in such regulations such standard of creditworthiness as the Commission determines to be appropriate.’’ 240 In this amendment, the Commission will require distribution participants, issuers, selling security holders and affiliated purchasers, in order to avail themselves of these exceptions from Regulation M, to rely upon the structural credit risk models as a measure of creditworthiness.241 These models have been used to estimate the probability of default of an issuer.242 Structural credit risk models typically take the issuer balance sheet measures of debt obligations as given and estimate a probability of default based on the market value and volatility of the firm’s equity. The value of equity is viewed in these models as the value of a call option on firm assets where the strike price is the total notional value of debt. Since the market value of equity, the volatility of equity, and the notional value of debt can be calculated from the market trading and balance sheet data, under the structural credit risk models the volatility of the value of the assets and the market value of assets, which are not observable, can be estimated. The probability of default can be calculated as the probability that the call option will expire out-of-the￾money, which occurs when the value of the company falls below the book value of the debt. As discussed above, structural credit risk models are based on the structure of the balance sheet.243 Since the future value of the firm is unknown, a structural credit risk model must make assumptions about the probability distribution of possible firm values in different scenarios, some of which may trigger default. These assumptions include the current firm value and the volatility of firm value, for which the observed market value of equity and the volatility of equity is often an
input. Some models include assumptions over the firm’s dividend policy. For purposes of these amendments, the probability of default derived from the structural credit risk models is an appropriate proxy for creditworthiness. As discussed previously in Part V.A.2, creditworthiness is an appropriate standard to reflect manipulation risk since securities issued by firms with sound creditworthiness trade primarily on yield and creditworthiness (traditionally measured by credit ratings) and have low pricing uncertainty and manipulation risk. The Commission received several comments supporting the probability of default as a standard for Rules 101 and 102 exception.244 However, one commenter opposed this option and suggested a standard based on Forms S–3 and F–3 or on WKSI standard.245 However, these alternatives are not good measures of sound creditworthiness as compared to probability of default because they fail to reflect the pricing point where a security is traded solely on its yield and maturity. Thus, the probability of default based on structural credit risk models is a more appropriate proxy for creditworthiness, and thereby for manipulation risk. Consistent with the Proposal, the Commission is adopting a 0.055% probability of default threshold. The Commission requested and received comments on this proposed threshold level. Specifically, one commenter expressed support of the proposed threshold and also noted that at any given date, the composition and population of any selected sample meeting the threshold could change; 246 as such some variation of the estimated percentages of the captured universe of securities eligible for the existing Investment Grade Exception is to be expected. Another commenter expressed that the threshold should be increased to 0.5% because it believes the exceptions as amended should be crafted to capture as many of the securities covered under the existing investment grade exceptions as possible; this commenter did not address the corresponding increase in the percentage of currently ineligible securities or the costs of that increase.247 No other commenters suggested a different or lower threshold and, overall, the commenters did not identify any economic effects of the proposed threshold level that were not considered in the Proposal. The Commission acknowledges that the percentage of investment grade securities that would be captured under a specific threshold fluctuates over time and as conditions change that affects the various inputs into the models. As of March 2023, the 0.055% probability of default threshold captured

248The investment grade status for nonconvertible securities issued between 2018 and 2023 was obtained from Mergent (as of the last available Mergent update through Mar. 2023) while the probability of default estimates were obtained for a cross-section of securities available in Bloomberg (as of Mar. 28, 2023). Please refer to Mario Bondioli, Martin Goldberg, Nan Hu, Chengrui Li, Olfa Maalaoui Chun, & Harvey J. Stein, The Bloomberg Corporate Default Risk Model (DRSK) for Public Firms (working paper Aug. 28, 2021), available at https://ssrn.com/abstract=3911300 (retrieved from SSRN Elsevier database), for methodology description of Bloomberg probability of default measure. 249See Proposal, 87 FR 18330. 250As discussed above, one commenter expressed general support of the proposed 0.055% threshold (see Bloomberg L.P. Letter, at 2) while another commenter suggested increasing the level to 0.5% (See SIFMA Letter 1, at 10). 251As pointed out by one commenter, some variation of the estimates is unavoidable, and ‘‘this highlights the importance of selecting an objective, data driven model that is consistently applied over time and documented by the distribution participant.’’ See Bloomberg Letter, at 2. 252We note that the SEC staff took a similar position in the COVID–19 Market Monitoring Group, Credit Ratings, Procyclicality and Related Financial Stability Issues: Select Observations, SEC Staff (July 15, 2020) (‘‘Cost of debt capital is driven by a wide range of financial and non-financial factors and forces; ratings downgrades are generally lagging indicators of cost of debt capital.’’), available at https://www.sec.gov/news/public￾statement/covid-19-monitoring-group-2020-07-15. Some academic studies find evidence that structural credit risk models may be able to respond to aggregate and firm specific news faster than credit ratings. Also, such models are able pick up on differences in default risk within a credit rating bucket. However, credit ratings do not necessarily imply probabilities of default and thus may not be directly comparable to probability of default estimated using a structural credit risk model. See Jing-zhi Huang & Hao Zhou, Specification Analysis of Structural Credit Risk Models (Fed. Res. Bd., Fin. & Econ. Discussion Series, 2008–552008), available at https://www.federalreserve.gov/pubs/feds/2008/ 200855/200855pap.pdf; Moody’s Analytics, EDF Overview (2011) (outlining the approach by Moody’s KMV), available at https:// www.moodysanalytics.com/-/media/products/EDF￾Expected-Default-Frequency-Overview.pdf; Giuseppe Montesi & Giovanni Papiro, Risk Analysis Probability of Default: A Stochastic Simulation Model, 10 J. Credit Risk 29 (2014). 253Some of the costs associated with this option are discussed in the Costs Section of the Economic Analysis. 254However, this will not be the case for other distribution participants who are not considered the lead manager of the distribution, which may deter such participants from relying on the
exception. Further, this may result in lead managers’ selecting a model that allows them to rely on the exception but is not necessarily the best model of the securities’ creditworthiness and manipulation risk. These issues are discussed in more detail in infra
Part V.C.
255Even though the lead manager would have to use a structural credit risk model, there are many versions of such models available, and the specific model parameters can be selected as well, providing considerable flexibility of the estimates as compared to the specific choices used in the assessments by NRSROs. 256 Vendors offer a number of commercial applications based on structural credit risk models. The probability of default calculated by structural credit risk models, such as the Merton (1974) Model and the Successor Models, can also be calculated by lead managers without the use of a vendor. One commenter, however, suggested that currently firms seldom use probability of default models in connection with issuances of the relevant securities. See SIFMA Letter 1 at 5. approximately 76% of the investment grade securities in the final sample of nonconvertible Fixed-Income Securities used (1996 distinct investment grade issues with probability of default below 0.055% out of 2637 total investment grade rated issues in the sample).248 This threshold also captured approximately 24% of non-investment grade issues (64 out of 269 non￾investment grade issues in the sample). This estimation differs from that in the Proposal. In the Proposal, we observed, using data from October 2021, that the 0.055% threshold captured about 90% of investment grade securities (2436 out of 2710 issues) and about 37% of non-investment grade issues (125 of 341 non-investment grade issues).249 Overall, at the time of the analysis of data as of March 2023, 2060 issues met the proposed exception as compared with the 2637 issues under the current exception. Given the reaction of commenters to the proposed 0.055% threshold 250 and that there is an unavoidable trade-off between capturing securities that are ineligible for the existing Investment Grade Exception and leaving out some securities that are currently eligible, the proposed threshold is intended to strike a reasonable balance between these two statistical realities over time.251 Nonconvertible debt securities and nonconvertible preferred securities of issuers for which the probability of default, estimated as of the sixth business day immediately preceding the determination of the offering price and over the horizon of 12 full calendar months from such day, is 0.055% or less, as determined and documented, in writing, by the distribution participant acting as the lead manager (or in a similar capacity) of a distribution, as derived from a structural credit risk model are to be excepted from Rules 101 and 102. An advantage of using probabilities of default implied by structural credit risk models instead of NRSRO credit ratings is that these model-implied probabilities of default generally use current estimates of equity valuation and volatility based on the recent trading activity, and hence incorporate more recent news affecting the valuation and perceived volatility of the firm. In contrast, credit rating agencies
are generally slower than the market in updating credit ratings and outlooks and thus may reflect less up-to-date information.252 The Proposal did not limit which distribution participants are allowed to produce probability of default estimations for the purposes of the exception. In order to ensure consistency and reliability of the estimates within any particular distribution and reduce the potential subjectivity and non-uniformity of the estimates the amendments specify that only lead managers are responsible for estimating the probability of default for a given distribution.253 Lead managers would have flexibility of either calculating the probability of default internally using structural credit risk models, given the wide availability of software products available on the market that perform such calculations, or obtaining an estimate from a vendor. One of the benefits of the amendment is that the lead managers will have the flexibility of selecting the model they find most appropriate to assess the creditworthiness of issuers for the purposes of using the exception.254 This means the lead managers will not have to rely on a credit rating for the issue in order to determine its eligibility for Rules 101 and 102 exception and will no longer have to rely on an NRSRO’s choice of the model for such purposes.255 Furthermore, multiple vendors currently provide estimates of the probability of default based upon structural credit risk models as a part of default packages that include various market data and metrics.256 Removing and replacing the references to credit ratings from Rules 101 and 102 of Regulation M may also have a benefit of expanding the number of options available to lead managers compared to what they would have under the requirements of the Investment Grade Exception. Specifically, the exceptions’ requirement will no longer rely on a limited number of vendors providing credit ratings, which may reduce possible negative consequences from limited competition. Structural credit risk models as a measure for creditworthiness could therefore serve as a better proxy for manipulation risk than credit ratings because, by prescribing a methodology rather than a metric generated by only a certain category of regulated vendors (that is, NRSROs), distribution lead managers may have more options for either using a vendor-supplied structural credit risk model or using their own proprietary version of a publicly available structural credit risk model. Under the final rule amendments, the structural credit risk models cannot be applied to asset-backed securities due to

257See a relevant discussion in supra Part II.B.2. 258See IILF Letter, at 6. 259SIFMA Letter 1, at 11. 260See supra notes 121–125 and accompanying text. 261One commenter opposed use of SF–3 standard for asset-backed securities and suggested relying on the probability of default instead (IILF Letter, at 7). However, probability of default calculations based on a structural credit risk model are complex for this type of securities due to their complex structure and are not routinely used. Another commenter in fact expressed support of using SF– 3 standard for asset-backed securities (SIFMA Letter 1, at 11). 262Based on EDGAR database filings from 2022. 263See SIFMA letter 1, at 5. 264See infra Part VI. 265The Commission estimates the wage rate based on salary information for the securities industry compiled by SIFMA. See Management & Professional Earnings in the Securities Industry— 2013, SIFMA (Oct. 7, 2013). These estimates are modified by the Commission staff to account for an 1800 hour work-year and multiplied by 5.35 (professionals) or 2.93 (office) to account for bonuses, firm size, employee benefits and overhead. These figures have been adjusted for inflation through Jan. 2023 using data published by the Bureau of Labor Statistics’ Consumer Price Index inflation calculator, available at https:// www.bls.gov/data/inflation_calculator.htm. 201 lead managers × 3 hours × $363 hour for a compliance manager = $218,889. 266See infra Part VI.C.1. 267Cost estimated is based on the sum of 33,798 offerings multiplied by 1 burden hour multiplied by $363, for a compliance manager. See Management & Professional Earnings in the Securities Industry— 2013, SIFMA (Oct. 7, 2013). These estimates are modified by the Commission staff to account for an 1800 hour work-year and multiplied by 5.35 (professionals) or 2.93 (office) to account for bonuses, firm size, employee benefits and overhead. These figures have been adjusted for inflation through Jan. 2023 using data published by the Bureau of Labor Statistics’ Consumer Price Index inflation calculator, available at https:// www.bls.gov/data/inflation_calculator.htm. 268See infra note 256. One commenter suggested that firms rarely use probability of default models in connection with issuances of the relevant securities. However, probability of default estimates are typically provided by the vendors in a package with other data firms are often subscribed to. the complexity of the structure of such instruments.257 Even though one commenter suggested that probability of default can be estimated for asset￾backed securities 258 such estimation based on structural credit risk models is not routinely used due to the complexity of the structure of these securities and the corresponding complex application of such models. Further, another commenter supported proposed Form SF–3 standard for the Investment Grade Exception with respect to asset-backed securities.259 The final amendments provide that securities that are
offered pursuant to an effective shelf registration statement filed on Form SF–3 should also be excepted from Rules 101 and 102. The Form SF–3 shelf eligibility requirements provide objective criteria that can also ensure that the securities are consistent with the Commission’s original basis for the Investment Grade Exceptions. Asset￾backed securities that are offered pursuant to an effective shelf registration statement filed on Form SF– 3 are less at risk of the manipulation that Regulation M addresses. Specifically, the Form SF–3 shelf eligibility requirements limit the number of nonperforming assets in the asset-backed security pool, require review of the pool assets if certain conditions are met, and require certification by the chief executive officer, among other things. As the Commission noted when adopting Form SF–3, the Form incentivizes sponsors to carefully review and disclose the underlying assets’ characteristics, reducing the overall uncertainty about the asset￾backed security 260 and, with respect to these final amendments, the risk of manipulation. The Commission received no comments that suggest otherwise. Asset-backed securities that are offered pursuant to an effective shelf registration statement filed on Form SF– 3 have similar qualities and characteristics to the investment-grade asset-backed securities currently excepted in Rule 101(c)(2).261 A review of recent EDGAR database filings confirms that almost all asset-backed securities issued pursuant to an effective shelf registration statement filed on Form SF–3 have investment grade ratings.262
C. Costs of the Amendments
The Commission recognizes that some of the affected underwriters, their affiliates, as well as issuers, selling security holders and affiliated purchasers may bear costs from the amendments. The amendments may alter the universe of securities that are eligible for the new exceptions. If some distribution participants decide not to participate in certain issues because of the rule amendments, the costs of the affected issues may increase. For example, when fewer banks or broker￾dealers are available, the underwriters may be able to charge higher fees. Additionally, as the result of the amendments, fewer issues may take place or issuers may rely more on private markets,263 potentially limiting issuers’ ability to raise capital and affecting investors in the relevant securities as the available security selection and liquidity may be reduced. There are several types of costs that could arise: (1) costs associated with calculations or obtaining the probability of default estimate; (2) costs associated with preserving records related to the probability of default estimation; (3) costs due to the probability of default being an imperfect proxy for creditworthiness, (4) asset-backed securities’ costs associated with the amendments, (5) indirect and other costs of the amendments. We discuss these costs in detail below.

1. Costs Associated With Obtaining the
   Estimate of the Probability of Default Lead managers may incur costs related to determining the probability of default. Consistent with the Paperwork Reduction Act (‘‘PRA’’) section,264 the Commission estimates that it will take a lead manager 3 hours to establish a system to gather the data serving as the inputs and then perform the analysis necessary to calculate the probability of default of the issuer whose securities are the subject of the distribution, for an aggregate cost of $218,889 265 Consistent with the PRA section,266 the Commission also estimates that it will take a lead manager one hour to gather the inputs required to calculate probability of default each time it participates in a distribution of Nonconvertible Securities. There were 33,798 offerings of Nonconvertible Securities in 2021. Therefore, it is estimated that annually lead managers will spend maximum of $12,268,674 267 in the aggregate complying with this requirement if all lead managers choose to estimate the probability of default internally. However, some lead managers may rely on third party vendors rather than internally calculate the probability of default. Any costs associated with using a vendor to obtain probability of default estimate, however, should be small, as the vendors typically already have subscriptions available to provide calculations regarding the probability of default based on structural credit risk models.268 Furthermore, lead managers, in particular those that choose to determine the probability of default estimate internally, are likely to already have the computational resources necessary to conduct such analysis internally. Therefore, the total costs for the lead managers of complying with the requirement should be below $12,268,674. Further, since the rule amendments specify that only the lead manager can supply the estimate of the probability of default for the purposes of relying on the exception, some issues where there is no distribution participant to act as the lead manager for the distribution, such as with self-underwritten offerings, at-the-market offerings, or other shelf offerings, may not be able to rely on the

269Such costs, however, cannot be quantified due to lack of available data. 270See supra Part VI.C.2. 271 301 distribution participants × 25 hours × $363 hour for a compliance manager = $2,731,575. See Management & Professional Earnings in the Securities Industry—2013, SIFMA (Oct. 7, 2013). These estimates are modified by the Commission staff to account for an 1800 hour work-year and multiplied by 5.35 (professionals) or 2.93 (office) to account for bonuses, firm size, employee benefits and overhead. These figures have been adjusted for inflation through Jan. 2023 using data published by the Bureau of Labor Statistics’ Consumer Price Index inflation calculator, available at https:// www.bls.gov/data/inflation_calculator.htm. 272See supra Part VI.C.2. 273Cost estimated based on the sum of 301 distribution participants multiplied by 10 burden hours multiplied by $363, for a compliance manager. See Management & Professional Earnings in the Securities Industry—2013, SIFMA (Oct. 7, 2013). These estimates are modified by the Commission staff to account for an 1800-hour work￾year and multiplied by 5.35 (professionals) or 2.93 (office) to account for bonuses, firm size, employee benefits and overhead. These figures have been adjusted for inflation through Jan. 2023 using data published by the Bureau of Labor Statistics’ Consumer Price Index inflation calculator, available at https://www.bls.gov/data/inflation_ calculator.htm. 274See, e.g., Rothwell, at 2 and ABA Letter, at 15 –17. 275See, e.g., John Y. Campbell, Jens Hilscher, & Jan Szilagyi, In Search of Distress Risk, 63 J. Fin. 2899 (2008), available at https://scholar. harvard.edu/files/campbell/files/ campbellhilscherszilagyi_jf2008.pdf. 276We note that the SEC staff took a similar position the COVID–19 Market Monitoring Group, Credit Ratings, Procyclicality and Related Financial Stability Issues: Select Observations, SEC (July 15,
2020) (‘‘Cost of debt capital is driven by a wide
range of financial and non-financial factors and forces; ratings downgrades are generally lagging indicators of cost of debt capital.’’), available at https://www.sec.gov/news/public-statement/covid￾19-monitoring-group-2020-07-15. 277See supra Part II.B.1. exception. These issues may therefore be subject to Regulation M restrictions and may have to rely on private markets and may face potentially higher issuing costs or not take place.269
2. Costs Associated With Maintaining
Records Related to the Probability of Default Estimation Broker-dealers relying on the new exception in Rule 101(c)(2)(i) or Rule 102(d)(2)(i) must preserve the written probability of default determination made pursuant to Rule 101(c)(2)(i), as amended. Consistent with the PRA
section,270 the Commission estimates
that it will take a distribution participant 25 hours to update the applicable policies and systems required to account for capturing the records made pursuant to new Rule 101(c)(2)(i), for an aggregate cost of $2,731,575.271 Consistent with the PRA
section,272 the Commission also
estimates that it will take a distribution participant 10 hours to maintain such records as well as to make additional updates to the applicable record preservation policies and systems to account for the rules. Therefore, it is estimated that annually broker-dealers will spend $1,092,630 273 in the aggregate complying with this requirement.
3. Costs Associated With Structural
Credit Risk Model Based Probability of Default Being an Imperfect Proxy for Creditworthiness As discussed previously, the structural credit risk models are designed to measure creditworthiness, and creditworthiness itself is considered a good measure of manipulation risk. There are costs that are currently present in the relevant markets associated with creditworthiness being an imperfect proxy for manipulation risk. However, in the absence of a better proxy for manipulation risk, creditworthiness has continued to successfully serve the purpose of measuring such risk for many years. This is also supported by the comments stating that the investment grade standard has been successfully used in Rules 101 and 102 exception.274 The final rule amendments are not expected to alter those costs and the discussion that follows focuses instead on the costs associated with the structural credit risk models as a proxy for creditworthiness. The use of any model to estimate creditworthiness necessarily provides an imperfect measure. Structural credit risk models are no exception. We note, however, that models such as structural credit risk models often are a part of the analysis involved in obtaining a credit rating.275 Some ways to implement structural credit risk models make use of historical trading data to produce a reliable estimate of the model input parameters. These data may not be available for certain infrequently traded securities. In some circumstances, the market for a security has not yet been established and sufficient trading data are unavailable, making it difficult to apply the exception. Additionally, structural credit risk models rely on a number of parameter estimates such as firm market value and volatility, which could be difficult to assess as these values change with market conditions and business fluctuations. A changing term structure of interest rates and noise trading in the market can further distort the probability of default estimates. Incorrect parameter estimates may result in the incorrect estimates of default probability and allow distribution participants to rely on the exception for risky issues or prevent distribution participants from relying on the exception for safe issues. Implied probabilities of default are sensitive to market prices and estimates of market volatility and consequently tend to be counter cyclical, increasing during market downturns, which are often also periods of increased uncertainty. A constant threshold which is not time￾varying will potentially result in fewer firms qualifying for the exception during market downturns, which may result in more issuances during this period not qualifying or firms choosing not to issue, hence increasing their cost of capital or limiting their access to capital. While credit rating downgrades are also countercyclical occurring more frequently during market downturns, they tend to be slow in incorporating updates.276 Thus, the impact of the counter cyclicality of
default probabilities implied by structural credit risk models could be stronger relative to using credit ratings: during periods of distress, using these probabilities of default will likely result in fewer firms with an investment grade credit rating falling below the threshold, and thus fewer firms qualifying for the exception relative to using credit ratings. Lead managers who make probability of default determinations pursuant to new Rule 101(c)(2)(i) could make reasonable adjustments to model parameters and inputs to recalculate the probability of default as market conditions change, mitigating the costs discussed above. Due to the number of variations among structural credit risk models and their estimated inputs, the probability of default estimates may be subjective to some extent and not comparable across different issuers or for the same issuer across different issues if estimates are based on different models, or done by different researchers or vendors. The latter may affect market participants’ ability to effectively rely on the estimates to make comparative assessments across multiple securities. However, this is also true of the credit ratings that often rely on similar models, which mitigates these costs of the amendments relative to the market baseline. Further, as a result of the Rules 101 and 102 amendments, all underwriters as well as issuers, selling security holders and affiliated purchasers will rely on the lead manager’s assessment of the probability of default in order to use the exception.277 This should mitigate the subjectivity and non-uniformity of the estimation concerns for the same

278The definition of structural credit risk models for purposes of new Rule 101(c)(2)(i) is limited to commercially or publicly available models, which would limit a distribution participant’s ability to develop its own models to achieve favorable results. 279See Bloomberg L.P. Letter. at 2, which provides analysis supporting the proposed probability of default threshold. Additionally, IILF Letter, at 6 suggests that the proposed threshold is in a reasonable range. 280See SIFMA Letter 1 at 10. 281See IILF Letter, at 6. 282We note that the SEC staff took a similar position in the COVID–19 Market Monitoring Group, Credit Ratings, Procyclicality and Related Financial Stability Issues: Select Observations, SEC Staff (July 15, 2020) (‘‘Cost of debt capital is driven by a wide range of financial and non-financial factors and forces; ratings downgrades are generally lagging indicators of cost of debt capital.’’), available at https://www.sec.gov/news/public￾statement/covid-19-monitoring-group-2020-07-15. 283Such changes in fees or changes in size cannot be reasonably quantified due to lack of available data on the respective changes (before and after an occurrence) in the relevant values. 284These costs are estimated as $363 per participant per distribution if estimates are obtained internally. Consistent with the PRA, the Commission estimates that it would take one hour per issue to calculate probability of default. Cost estimated is based on 1 burden hour multiplied by $363, for a compliance manager. See Management & Professional Earnings in the Securities Industry— 2013, SIFMA (Oct. 7, 2013). These estimates are modified by the Commission staff to account for an 1800 hour work-year and multiplied by 5.35 (professionals) or 2.93 (office) to account for bonuses, firm size, employee benefits and overhead. These figures have been adjusted for inflation through Jan. 2023 using data published by the Bureau of Labor Statistics’ Consumer Price Index inflation calculator, available at https:// www.bls.gov/data/inflation_calculator.htm. issue and to some degree across issues for the same issuer to the extent the same parties are engaged by the issuer for different issues. This requirement allows the lead manager to perform estimations which determine if the resulted probability of default falls below the threshold for all the distribution participants and their affiliates and thus the availability of the exception. Some of these participants may decide to withdraw if the exception is not available. However, the lead manager is interested in the best outcome of the distribution and therefore has strong incentives to encourage the participation of these entities in the distribution, mitigating the above concern. This may, on the other hand, incentivize lead managers to select models and estimation specifics in such a way to ensure the resulted estimates are below the threshold, potentially allowing issues of issuers with low creditworthiness and high
manipulation risk to rely on the exception. The public availability of alternative probability of default estimates available for the investors through multiple vendors, however, should mitigate this concern. In addition, as discussed above in reference to the selected threshold, the proposed amendment may expand the universe of issuers of nonconvertible securities that qualify for the exception and include issuers that did not receive an investment grade credit rating, but have a structural credit model implied probability of default that falls below the threshold. The debt prices of these firms may be prone to manipulation if the price of their debt is relatively more sensitive to the idiosyncratic risks of the issuers. Additionally, this amendment may create potential opportunities for new products offered by the vendors designed specifically for a given issue or issuer. A custom designed estimate paid for by a party with an interest in the outcome of the distribution may lead to potential conflicts of interest since the vendor is incentivized in this case to produce an estimate which will allow the issuer, their affiliates and selling security holders, and other distribution participants to rely on the exception. However, the existing major vendors supplying probability of default estimates have numerous clients currently using this information for business purposes other than the Rules 101 and 102 exception. Therefore, given the reputational concerns it is unlikely that these vendors will produce a product to cater specifically to the use of these estimates for purposes of relying on the Rules 101 and 102 exception. Additionally, the model input estimates or assumptions may be selected by the lead manager in such a way as to produce the desired estimation result if the model is estimated internally and may result in lead managers’ selecting the models so as to be able to rely on the exception.278 This may result in an additional cost of adding some manipulation risk to the relevant markets if manipulation prone issues are allowed to rely on the exception as a result. Finally, the threshold of 0.055% for the exception is based on model assumptions and available data. Some commenters expressed support for the proposed threshold level selection,279 while one commenter suggested a higher level.280 Future market evolution may result in this threshold becoming either too large or too small, allowing risky issues to rely on the exception or preventing less risky issues from using it. One commenter expressed a similar concern about a set-level threshold specification in the rules.281 The threshold may vary by industry, with the threshold being more restrictive in some industries relative to the original NRSRO investment grade designation. Moreover, probabilities of default as implied by structural credit risk models tend to be counter-cyclical and can spike in periods of crisis due to decreases in market valuation and increases in equity volatility.
Consequently, during such periods, fewer investment grade firms generally fall below the threshold. Credit ratings by NRSROs are also countercyclical but tend to be slow-moving, since credit rating changes often lag updates to firm conditions that will impact cost of capital.282
4. Costs Associated With Asset-Backed
Securities’ Amendments
The amendments may render some asset-backed securities ineligible to rely on the exception from the Regulation M. This may increase issuance costs for the underwriters as well as issuers, selling security holders and affiliated purchasers. For instance, broker-dealers may reduce an offering’s size or increase fees if the exception to Regulation M is no longer available.283 Additionally, issuers may need to establish new business relationships due to Regulation M restrictions. Furthermore, some issuers may decide not to issue the affected securities if the exceptions to Regulation M are no longer available. As a result, some asset-backed securities’ issues may not take place, which could affect issuers’ ability to raise capital and could affect investors in the relevant markets by potentially reducing the selection of the available asset-backed securities.
5. Indirect and Other Costs of the
Amendments
Besides the direct effects on the distribution participants and affected securities discussed above the final rule amendments may also generate indirect effects including on investors in these securities and NRSROs. For instance, distribution participants other than lead managers may want to verify the estimates provided by the lead manager by either obtaining the estimate from a vendor or making the calculations internally, which will result in additional costs for these participants.284 Additionally, the lead managers, although not required, may need to expend resources in terms of their staff time and resources in order to notify other distribution participants, their affiliated purchasers, issuers, selling security holders, and their affiliated purchasers of their probability of default

285 33,798 issues times 0.25 hours.
286 8,450 hours * $205 hour for a junior business analyst wage = $1,732,250 See Management & Professional Earnings in the Securities Industry— 2013, SIFMA (Oct. 7, 2013). These estimates are modified by the Commission staff to account for an 1800 hour work-year and multiplied by 5.35 (professionals) or 2.93 (office) to account for bonuses, firm size, employee benefits and overhead. These figures have been adjusted for inflation through January 2023 using data published by the Bureau of Labor Statistics’ Consumer Price Index inflation calculator, available at https:// www.bls.gov/data/inflation_calculator.htm. 287See supra Part V.C.1. determinations that were estimated pursuant to Rule 101(c)(2)(i). The Commission estimates that it will take 0.25 hours per lead manager per issue (8,450 hours annually) 285 to notify other distribution participants of the probability of default estimates. Therefore, the total estimated cost for the lead managers associated with notifying other distribution participants is estimated as $1,732,250.286 Further, if issuer participation in the relevant security issues, for example in the case of re-openings or issues that are more likely to become sticky offerings, becomes limited, some issues may not take place that otherwise would. Investors may additionally face a more limited choice of investment instruments as a result. This may also affect liquidity of their portfolios in the case of re-openings, since re-openings can offer additional liquidity benefits as the securities offered in re-openings are interchangeable with the existing issues. However, as already discussed in the case of re-openings, these costs are expected to be minimal as re-openings are used infrequently. The rule amendments do not rely on an NRSRO rating in order to determine if an issue is eligible for the exception. This may diminish NRSROs’ clientele to the extent NRSROs choose not to provide structural credit risk model￾based estimates of the probability of default for their existing clients opting to rely on the exception. However, the amendment may increase the clientele of the vendors that supply relevant data and metrics to the lead managers or other distribution participants who wish to verify the lead manager estimates, if such vendors already supply probability of default estimates or choose to offer this estimate as a part of their services. In addition, if firms do not solicit credit rating services from NRSROs beyond the estimate of a probability of default implied by a structural credit risk model, investors will not be able to benefit from the information provided by a credit rating report and ongoing coverage of the firm that otherwise will be provided through the distribution participant. D. Efficiency, Competition, and Capital Formation As discussed previously, lead managers will have flexibility in selecting the structural credit risk model to access creditworthiness as a measure of manipulation risk for the business. This may
encourage issuers to issue securities in relevant markets, as well as participation of other distribution participants, such as selling security holders and affiliated purchasers. As a result, this could improve competition between issuers for investors as well as competition between lead managers for underwriting business. Further, widely available estimates of the probability of default as well as an option of internal model estimation could lead to a more competitive environment in the provision of models as the requirement to rely on proprietary credit risk models of a small number of NRSROs is removed. The improved competition, market participation and efficiency ultimately should lead to more efficient capital formation as the access to and functioning of the relevant fixed income markets improves. However, it is possible that a new business model could emerge in the relevant markets that leads to conflicts of interest and neutralizes the effects discussed above. For instance, lead managers could contract with a vendor or a credit rating agency directly to create a custom estimate of the probability of default. This could result in a business model where an interested party pays for the supplied estimate and where vendors may be incentivized to produce an estimate designed to fit the desired estimation result. Thus issuers that otherwise will not be able to rely on the exception could end up being excepted potentially increasing the manipulation risk in the relevant markets, which in turn could negatively affect competition and capital formation. The reputational concerns, however, would generally prevent vendors from generating estimates specifically designed for the needs of a small number or a single customer. Additionally, the positive effects discussed above could be offset by the fact that only lead managers can obtain an estimate of the probability of default for the distribution. Some issues where there is no distribution participant to act as the lead manager for the distribution, such as with self-underwritten offerings, at-the-market offerings, or other shelf offerings’’ may not have the exceptions available.287 This may deter participants from such distributions and in some cases result in securities being issued in private markets or issues not taking place. This may negatively affect the competition and capital formation in the relevant market. Some issuers may also face higher costs or no longer be able to use the exception, for example, due to imperfect model estimates because of market fluctuations or changing market. High costs of issuance or inability to rely on the exception may deter participants from issuing the affected securities, which could affect competition and capital formation in the relevant markets. Further, potential negative effects of non-uniform estimates and subjectivity additionally reduce these benefits. As discussed previously, variations in model assumptions, parameters, or data sample used
necessarily introduce an element of subjectivity in the final estimates and leads to differences in the estimates across different issues or issuers. Finally, potentially increased issuance costs due to some asset-backed securities being ineligible for the exception may also negatively affect market participation and competition of the relevant markets. E. Reasonable Alternatives Alternative 1 discussed below deals with the probability of default threshold, alternatives 2–4 discuss alternative approaches to using structural credit risk models as a standard of creditworthiness to measure manipulation risk. Alternative 5 discusses elimination of the exception from Rule 101, alternative 6 deals with asset-backed securities, alternative 7 discusses Rule 102 options, while the last alternative discussed the record preservation requirement.

1. Alternative Threshold for Probability
   of Default
   The Probability of Default threshold of 0.055% was chosen in an effort to maximize investment grade securities captured and minimize the non￾investment grade securities captured. However, a different threshold could be used in the Rule exception, which would capture different proportions of investment and non-investment grade securities. For example, based on data as of March 2023, a higher threshold of 0.5% is estimated to capture about 97% of investment grade securities (2550 out of 2637 investment grade issues) and about 70% of non-investment grade issues (188 out of 269 non-investment grade issues). A lower threshold of 0.03% is estimated to capture about 64% of investment grade securities (1675 out of 2637 investment grade

288Based on the data as of Oct. 2021. Based on Mar. 2023 data, 0.055% threshold scopes in about 76% of investment grade issues (1996 out of 2637 issues) and about 24% or non-investment grade issues (64 out of 269 issues). 289SIFMA Letter 1, at 7 and 10, see also a relevant discussion in Part II.B.1. 290See Bloomberg L.P. Letter, at 2. 291See IILF Letter, at 7. 292See SIFMA Letter 1, at 5, Better Markets Letter, at 4, and the relevant discussion in Part II.D. 293Empirical duration is bond duration calculated based on historical data rather than a formula. Typically, it is estimated using a regression analysis of the relationship between market bond prices and Treasury yields. 294 2011 Proposing Release, 76 FR 26557–64. issues) and 11% of non-investment grade issues (29 out of 269 non￾investment grade issues). The advantage of a higher threshold is that it captures a larger set of investment grade securities, but at the expense of also capturing an additional set of non￾investment grade securities, which could be prone to manipulation risk. Increasing the threshold would allow more investment grade securities to rely on the exception at expense of a potentially higher manipulation risk; on the other hand, decreasing the threshold would limit the ability of some of the investment grade securities to use the exception, but would also limit the number of non-investment grade securities allowed to rely on the exception and, as a result, also limit manipulation risk. The Commission proposed 0.055% threshold level, which scoped in about 90% of investment grade issues and about 37% of non-investment grade issues.288 One of the commenters suggested increasing the threshold in order to capture a larger percentage of the previously eligible investment grade issues,289 another commenter suggested that the proposed threshold level is appropriate,290 while none of the commenters suggested decreasing the threshold. Furthermore, at any given date, the proportion of currently eligible securities that would be captured varies. Manipulation risk remains the primary concern of Regulation M. Because the originally proposed threshold of 0.055% remains appropriate for these purposes, and acknowledging the variation in eligible securities that would be captured over time, increasing (or decreasing) this threshold for the primary aim of capturing more (or fewer) of currently eligible securities does not justify changing this threshold. Rather than providing a specific number as a threshold, a method for distribution participants to use in calculating such a threshold could be specified instead. For example, such method could involve calculating a set of probability of default estimates for a sample of Nonconvertible Securities with characteristics such as yield and maturity similar to the distribution participant’s securities issued over a specified time interval and comparing it to a specified standard of creditworthiness. A longer time interval of the data sample would
capture more issues and improve statistical accuracy at expense of having market conditions potentially changing and generating incorrect estimates. A shorter time interval of the sample ensures the market conditions have not changed but includes fewer issues resulting in a smaller sample and lower statistical accuracy. One of the commenters expressed similar ideas advocating for an estimation method rather a fixed threshold level, which would result in a more flexible threshold level.291 The main advantage of specifying a method as opposed to a number for the threshold is its flexibility with respect to changing market conditions. The main disadvantage of this alternative is subjectivity of the analysis involved, which may lead to non-uniform application of the Regulation M exceptions across issues or issuers if the estimated threshold differs considerably across issues or issuers; or incentivize market participants to adjust the threshold estimation to be able to rely on the exception. Some commenters expressed a concern for the estimates’ subjectivity and non-uniformity as discussed previously. This alternative could introduce additional subjectivity and non-uniformity and thus is sub-par to the originally proposed option.292
2. Exception Based on Security
Characteristics
As an alternative replacement for the reference to investment grade securities, the Commission considered analysis that could be based on security characteristics, such as (1) total amount of issue outstanding (public float); (2) yield to maturity of the security during a past trading period; or (3) empirical duration.293 Other relevant security characteristics that could be used are outlined in the 2011 Proposal.294 Such analysis could be performed internally or externally and could be additionally verified by a third party. All of these alternatives were included in the Proposal and the Commission received no comments in regards to these alternatives. Below we discuss public float, yield to maturity and empirical duration criteria in more detail.

• Exception Based on the Total
  Amount of Issue Outstanding (Public Float). To the extent that it is more difficult to manipulate price of a larger issue, public float could be used as an alternative criterion to reflect manipulation risk. This criterion has the advantage of being straightforward and easy to evaluate. Due to its simplicity, it lacks the estimation issues associated with other measures such as the probability of default. However, determination of a threshold for public float to select securities for the exception is complicated due to its considerable variation across issuers or industries. A specific threshold selection could potentially disadvantage smaller issuers—especially during periods of market downturns when valuations are low. Additionally, public float is not inherently an indication of low credit risk since a distressed firm can have a large amount of debt.
• Exception Based on Yield to
  Maturity.
  Securities that are traded primarily on yield and maturity have low manipulation risk, as discussed before, since their pricing does not reflect issuer specific risks. Yield to maturity, therefore, can be used as an alternative criterion to evaluate manipulation risk. However, using yield to maturity as a criterion for securities eligible for the exception is also problematic. Even though this criterion is similarly easy to obtain and lacks any major estimation issues, selecting a threshold is not straightforward. For instance, yield to maturity differs considerably by industry. Selecting a fixed threshold may result in some industries being under-represented and others over￾represented in the pool of eligible issues. Moreover, yield to maturity often moves with risk-free rates; thus fewer firms would be excepted during periods of high interest rates. The default-free component of yield to maturity makes this measure a very noisy proxy of credit worthiness.
• Exception Based on Empirical
  Duration.
  Empirical duration is another alternative proxy that could be used to evaluate Nonconvertible Securities for an exception from Regulation M. Negative empirical duration might be an indication that a Nonconvertible Security or its issuer is of low creditworthiness. A Nonconvertible Security with negative empirical duration is less affected by changes in interest rates than Nonconvertible Securities of creditworthy issuers and trades similar to equity securities. Although negative empirical duration may demonstrate that a particular issuer or security is not creditworthy, it has some limitations that affect the viability

295 2008 Proposing Release, 73 FR 40095–97. 296See SIFMA Letter 1 at 9. 297ABA Letter, at 15–17 and Letter from Deborah A. Cunningham and Boyce I. Greer, Co-chairs, Securities Industry and Financial Markets Association (‘‘SIFMA’’) Credit Rating Agency Task Force, to Florence E. Harmon, Acting Secretary (Sep. 4, 2008) at 13. 298See SIFMA Letter 1 at 9. 299See SIFMA Letter 1 at 5. 300See SIFMA Letter 1, at 5–8 as well as the related discussion in Part II.B.1. 301Form S–3 and Form F–3 Release, 76 FR 46607. 302The reduced-form credit risk models are discussed, for example, in Robert Litterman & Thomas Iben, Corporate Bond Valuation and the Term Structure of Credit Spreads, 17 (3) Fin. Analysts J. 52, 52–64 (1991); Robert A. Jarrow & Stuart M. Turnbull, Pricing Derivatives on Financial Securities Subject to Default Risk, 50 J. Fin. 53, 53– 86 (1995); Robert A. Jarrow, David Lando, & Stuart M. Turnbull, A Markov Model for the Term Structure of Credit Risk Spreads, 10 Rev. Fin. Stud. 481, 481–523 (1997); Darrell Duffie & Kenneth J. Singleton, Modeling the Term Structures of Defaultable Bonds, 12 Rev. Fin. Stud. 687, 687–720 (1999). 303 IILF Letter, at 2. 304See SIFMA Letter 1, at 5, Better Markets Letter, at 4. of negative empirical duration as a substitute for the reference to credit ratings in the Investment Grade Exception. In particular, this measure relies heavily on statistical analysis, requires the Nonconvertible Security to be traded, and may lack intuitive interpretation, which renders empirical duration a poor proxy for the type of manipulation that Regulation M is designed to prevent.
3. Exception Based on Issuer
Characteristics
The Commission also considered an exception based on issuer characteristics, for example, the interest coverage ratio, the WKSI standard, as suggested in the 2008 Proposing Release,295 a Form S–3/F–3-based standard, or a criterion based on a reduced-form credit risk model, as an alternative to the structural credit risk models. We discuss these alternatives below.

• Exception Based on the WKSI
  Standard.
  The Commission could adopt a standard based on the amount of the issuer’s total securities outstanding or based on the WKSI standard as a criterion to determine eligibility for the exception. The issuers that fall under the WKSI definition or with sufficient amounts of total securities issued or outstanding are large and established firms that typically have sound creditworthiness. The Commission included this alternative in the Proposal. One commenter expressed some support for this alternative.296 The advantage of this characteristic is its simplicity, uniformity, and the lack of subjectivity of the analysis. However, the WKSI standard as discussed in the 2008 Proposing Release, for example, was heavily criticized for allowing risky high-yield issues to be eligible for the exception and preventing issues by smaller but otherwise creditworthy issuers from relying on the exception, which remains a considerable concern.297 Even though one of the commenters suggested a standard based on the WKSI standard due to its simplicity, uniformity and lack of subjectivity,298 such a standard would fail to capture the pricing point where securities trade solely based on their yields and maturity and not on the issuer characteristics, where pricing uncertainty and manipulation risk are at their minimum. Thus, such a standard would be a sub-par measure of manipulation risk as compared to the probability of default.
• Exception Based on Forms S–3 and
  F–3.
  One commenter stated that the complexity of the proposed probability of default calculations would impose additional regulatory burdens that could be avoided if the exception, instead, relied on a standard based on readily verifiable and publicly available information.299 This commenter proposed using Form S–3 or Form F–3 as a standard for the exception given the uniformity, simplicity and a lack of subjectivity of such a standard.300 The Form S–3 or Form F–3 eligibility criteria are intended to access whether an issuer is widely followed,301 rather than an issuer’s creditworthiness. A widely followed issuer may be more likely to have a low manipulations risk, making this a reasonable alternative criterion to consider for the Investment Grade Exception. However, such a standard does not differentiate securities that are traded solely on their yield and creditworthiness from securities that trade also on the issuer identity and thus have a high manipulation risk. Therefore, probability of default is a preferred standard to rely upon in the assessment of manipulation risk for the purposes of the Investment Grade Exception.
• Exception Based on the Interest
  Coverage Ratio.
  Another possible issuer-based criterion for exception eligibility is the interest coverage ratio. This alternative was included in the Proposing Release and no commenters expressed a view on this option. A high interest coverage ratio typically indicates the issuer’s ability to repay debt and can be used as a criterion to reflect creditworthiness. It has the advantage of being a simple and easy to calculate value. However, the interest coverage ratio is an accounting measure that can result in inconsistent outcomes as it is based on the reported earnings rather than cash flows. Reported earnings may differ based on accounting practices of the firm. Structural credit risk models have an advantage over interest coverage ratio since they are not dependent on reported earnings, which are heavily influenced by accounting practices.
• Exception Based on Reduced-Form
  Credit Risk Model.
  An alternative to using structural credit risk models is reduced-form credit risk models.302 The latter models could be a good measure of creditworthiness and of manipulation risk to the extent that creditworthiness is a good proxy for manipulation risk. This alternative was discussed in the Proposal. One of the commenters proposed a similar alternative relying on debt security prices, yields, or credit spreads instead of using a structural credit risk model for the probability of default estimation.303 Unlike structural models, reduced-form models do not assume default occurs when firm value falls below a threshold. The default is instead assumed to follow an unobserved process and the default model can be fitted to the market data. The advantage of these models is they do away with some of the unrealistic requirements of structural credit risk models, for example when the firm value, its volatility or other required parameters are unobserved. Even though such models can be considered more flexible and may provide better fit for the observed default events, their ability to predict future defaults may not necessarily exceed that of the structural models. In addition, unlike structural models, they suffer from a lack of theoretical background of the assumed relationships, or the intuitive interpretation of the model dependencies and why the defaults occur. Unrestricted use of these models might also provide more opportunity to choose a reduced-form model specification which enables use of the exception. Further, some commenters expressed a concern for a lack of consistency and uniformity across issues or issuers in using probability of default standard for the exception.304 Since reduced-form models are more flexible and less structured than structural credit risk models, such concerns would be more pronounced in

305See Removal of Certain References to Credit Ratings Under the Securities Exchange Act of 1934, Release No. 34–71194 (Dec. 27, 2013) [79 FR 1522, 1527–28 (Jan. 8, 2014)]. 306This is unlike the structural credit risk model based probability of default that would imply the same costs for all the participants who obtain the estimated values. 307See, e.g., SIFMA Letter 1, at 2; Bloomberg L.P. Letter, at 1. a standard that is based on the reduced￾form models.
4. Exception Based on Issuer and Issue
Characteristics
The Commission considered, as another alternative, an analysis based on both security and issuer characteristics; for example, characteristics outlined in Exchange Act Rule 15c3–1. This alternative was discussed in the Proposal and the Commission received no comments in regard to this option. Rule 15c3–1 specifies a set of factors to determine a minimum amount of credit risk broker-dealers can use to determine if a security can qualify for lower haircuts: (1) credit spreads; (2) securities-related research; (3) internal or external credit assessments; (4) default statistics; (5) inclusion in an index; (6) enhancements and priorities; (7) price, yield and/or volume; or (8) asset-class specific factors.305 Some of these factors, such as default statistics or credit assessments, measure issuer creditworthiness, while others, such as price, yield, or volume, measure the manipulation risk present in each specific issue, providing a good overall assessment of manipulation risk. The advantage of this alternative is that it would align the exception with already existing standards that broker￾dealers might apply to determine whether a security has a minimal amount of credit risk. The standard in Rule 15c3–1 was adopted in 2013 as a replacement for a reference to investment grade securities pursuant to
section 939A of the Dodd-Frank Act.
Such test could have minimum additional costs for broker-dealers who already have all the necessary procedures in place for its application. The Rule 15c3–1 standard is commonly used for seasoned securities and, therefore, includes a longer time period to obtain information about issues that may not be available for the new issuances and for seasoned (actively traded) distributions that may have only a one-day restricted period also subject to Regulation M. Moreover, the Rule 15c3–1’s minimal credit risk standard is based on a set of eight different factors, some of which include price or volume, with respect to each specific issue. Depending on these other participants’ systems and regulatory obligations, it may be costly for them to replace the investment grade standard with the minimal credit risk standard. This could result in a situation where different distribution participants are facing different costs,306 possibly deterring some market participants.
5. Elimination of the Investment Grade
Exception From Rule 101
The Commission also considered eliminating the Investment Grade Exception for Fixed-Income Securities from Rule 101. Elimination of the exception was discussed as an alternative in the Proposal and the Commission did not receive any direct comments on this option. However, as discussed in Part II, commenters broadly supported the Commission’s efforts to find an alternative standard of creditworthiness in place of the references to credit ratings in Rule 101’s Investment Grade Exception (as opposed to removing the Investment Grade Exception, without a replacement).307 The advantage of this alternative is eliminating the situations when manipulation-prone securities fall under the exception due to limitations of proxies used to select the securities to be excepted. For instance, as discussed above, there are various limitations of the structural credit risk models’ applications, which may limit the ability of certain issuers to rely on the exception or allow issuers with a higher risk of having their securities manipulated to avoid Regulation M. If the exception is eliminated, any limitations of such a proxy for manipulation risk are eliminated as well. In addition, this approach could ultimately relieve lead managers from the need to spend time or costs to implement, understand, and calibrate any standard such as a structural credit risk model. However, this approach raises a number of concerns. Specifically, eliminating the exception could make some offerings in the excepted securities considerably more costly. For example, with respect to re-openings, broker￾dealers who might otherwise elect to re￾open a bond offering may determine not to do so to avoid restrictions of Regulation M that could arise during such a re-opening if it becomes a sticky offering. This could increase the cost of the issue that has to rely on the next￾best alternative structure. Further, an alternative transaction structure, if selected, may decrease the liquidity of the securities being issued because they would not be fungible with the previously issued securities. This may also result in some distribution participants, such as broker-dealers, deciding not to participate. This could limit the number of available broker￾dealers, potentially increasing fees faced by the issuers. Further, if certain issues do not take place under the amendments, it could reduce the selection of available securities for the investors in the relevant markets and may limit issuers’ ability to raise capital. However, these costs might be mitigated because a party subject to the prohibitions of Rule 101 could structure its buying activity before or after the applicable restricted period so as not to incur any costs associated with relying on the exception. The above arguments apply to all currently excepted investment grade securities because any such issue can become a sticky offering and the underwriters have to account and adjust for this possibility ex-ante. In a scenario where an
underwriter is unable to sell its allotted securities to the public on or promptly after the pricing date, there is no exception on which to rely, the underwriter/broker-dealer would likely ex-ante adjust the cost of issuance to reflect this added risk. Broker-dealers could be more cautious in structuring potentially sticky offerings if they know they will be required to comply with Rule 101 (and have no exceptions available), by reducing an offering’s size or increasing fees as a risk premium. This could potentially raise the cost of investment grade offerings. However, this could also decrease the probability of an offering to become sticky, potentially reducing manipulation risk in the relevant markets. The removal, without replacement, of the Investment Grade Exception could also affect the liquidity of the Fixed￾Income Securities if re-openings of issues already in circulation are more costly, potentially reducing issuers’ reliance on this financing structure, which negatively affects the investors in the relevant markets. This alternative could also disrupt some established business relationships. In certain circumstances new relationships may need to be established. For example, if an offering becomes sticky, absent Investment Grade Exception to rely on some broker￾dealers may be limited in their ability to trade relevant securities and decide to withdraw, in which case the issuer may need to seek a different broker-dealer. This would increase costs of the affected security offerings, including the new broker dealer fees or the search costs, especially when the market has a limited number of available broker￾dealers.

308See IILF Letter, at 6.
309See SIFMA Letter 1, at 11.
310See SIFMA Letter 1, at 12.
311See SIFMA Letter 1, at 11.
312 44 U.S.C. 3501 et seq. The burdens associated with the information collection requirements are referred to as ‘‘PRA burdens.’’ 313See 44 U.S.C. 3507; 5 CFR 1320.11. 314 In the Proposal, the Commission proposed to eliminate the Investment Grade Exception under Rule 102 of Regulation M, without proposing an alternative standard in its place. However, as discussed above, in Part II.C, the Commission is adopting an exception that is based on an issuer’s probability of default in both Rule 102 and Rule 101.
6. Alternative for Asset-Backed
Securities
As an alternative for asset-backed securities the Commission considered using a standard based on the value at risk. This alternative was included in the Proposal and no commenter expressed any view on this standard. Value at risk measures the percentage loss of the security in the worst case scenarios over a specified time period. It can be estimated by performing a simulation over the underlying securities’ pool and determining the cash flows available to the asset-backed security in each scenario. A number of commercially available options can be used to perform this analysis. Value at risk can be a good indicator of manipulation risk since low value at risk indicates that the majority of the cash flows are sufficiently assured. The price of the asset-backed security in this case is more certain and is less subject to manipulation risk. However, value at risk is by construction estimated for a specified time period and thus only accounts for the potential losses during such period, while losses may also occur after this time period. In this case the price of the asset-backed security may depend on issue-specific factors and be prone to manipulation despite the estimated value at risk over the specified time period being low. This may allow securities with high manipulation risk to rely on the exception. One of the commenters proposed as an alternative to use probability of default-based standard for asset-backed securities calculated using prices and credit spreads.308 However, probability of default is typically not used for these securities due to the complexity of their structure and corresponding complexity of the calculations. Further, another commenter supported proposed Form SF–3 standard to use for the asset￾backed securities due to its uniformity and simplicity.309
7. Alternatives for Rule 102 Exception
The Commission also considered and proposed eliminating, without replacing, the Investment Grade Exception in Rule 102. Disruption to the trading market may be limited because distribution participants will still be able to rely on the exception from Rule 101 if they meet the requirements of the proposed rules. However, one of the commenters pointed out that eliminating the exception from Rule 102 may affect issuer-affiliated broker￾dealers that do not act as an underwriter and may need to rely on the Rule 102 exception.310 Eliminating the exception from Rule 102 may increase issuance costs or deter market participants from issuing such securities. Therefore, elimination of the exception from Rule 102 was not the best option in comparison to the alternative selected.
8. Alternative for the Record
Preservation Requirement
The Commission considered not adding the record preservation requirement. The option of not adding the record preservation requirement for broker-dealers was suggested by one of the commenters due to the additional burdens it creates for the broker￾dealers.311 However, the record preservation requirement may help ensure an estimate of the probability of default is produced for all the distribution participants to rely upon to determine eligibility of the issue for the exception. The record preservation requirement also helps address concerns about the existence of some subjectivity involved in the selection of a particular structural credit risk model and data sample specifics by the lead managers, and the possibility of lead managers selecting these specifics so as to generate a probability of default estimate below the threshold level. The potential consequences of not including a record preservation requirement, therefore, could be that issues with high manipulation risk are allowed to rely on the exceptions from Regulation M. It is also intended to aid the Commission staff in examinations of the broker￾dealers in evaluation of the specific model and data used to determine the probability of default of the issue in addition to exception eligibility.
VI. Paperwork Reduction Act
Certain provisions of the final amendments contain ‘‘collection of information’’ requirements within the meaning of the Paperwork Reduction Act of 1995 (‘‘PRA’’).312 The hours and costs associated with determining whether a Nonconvertible Security qualifies for the new exception in Rule 101(c)(2)(i) and preserving the corresponding records under Rule 17a– 4(b)(17) constitute PRA burdens. In accordance with the PRA, the Commission is submitting the final amendments to the rules to the Office of Management and Budget (‘‘OMB’’) for review.313 The Commission published a notice requesting comment on these collections of information requirements in the Proposal and submitted these requirements to the OMB for review in accordance with the PRA. An agency may not conduct or sponsor, and a person is not required to respond to, a collection of information unless it displays a currently valid control number. The titles and control numbers for these collections of information are as follows:
Rule Title OMB control no.
Rule 101 ...................................................................................... Rule 101, 17 CFR 242.101 (Activities by Distribution Partici￾pants). 3235–0464 Rule 17a–4 .................................................................................. Records to be Preserved by Certain Brokers and Dealers ....... 3235–0806 These PRA burdens are distinct from the existing OMB-approved collection of information burden estimates under Rules 101, 102, and 17a–4 because the Commission has not estimated that respondents incur PRA burdens when determining whether a security qualifies for the Investment Grade Exception, nor did Rule 17a–4 include a recordkeeping requirement in connection with reliance on the Investment Grade Exception.314 The Commission did not receive any comments on the Proposal’s PRA analysis. While one commenter did reference the potential burden of the

315SIFMA Letter 1, at 5.
316 17 CFR 242.101.
317 17 CFR 242.102.
318 17 CFR 240.17a–4.
319 17 CFR 242.101(c)(2)(i), as amended, 242.102(d)(2)(i), as amended. 320 17 CFR 240.17a–4(b)(17), as amended. 321 17 CFR 242.101(c)(2)(ii), as amended, 242.102(d)(2)(ii), as amended. 322See supra Part V.A.1. 323 [201 lead manager broker-dealers] + [100 non￾lead manager broker-dealers] = 301 respondents under new Rule 17a–4(b)(17). proposed amendments generally,315 no commenters specifically addressed the Commission’s estimates of burdens and costs in the Proposal’s PRA analysis. In addition, the Commission’s estimates of the collection of information for the amendments, as adopted, have been updated from the estimates included in the Proposal, as appropriate, with the updated estimates based on the modifications in the adopted rule and based on more recent data. The Commission is adopting in Rules 101 and 102 the proposed exception that is based on an issuer’s probability of default, as described above in Parts II.A and B, to replace the Investment Grade Exceptions. The Commission is also adopting a corresponding record preservation requirement in Rule 17a– 4(b), which requires broker-dealers to preserve the written probability of default determination, relied upon pursuant to the new exception in Rule 101(c)(2)(i) or Rule 102(d)(2)(i), as applicable. As discussed above, Regulation M is designed to preserve the integrity of the securities trading market as an independent pricing mechanism by prohibiting activities that could artificially influence the market for an offered security. Subject to exceptions, Rule 101 prohibits distribution participants and their affiliated purchasers,316 and Rule 102 prohibits issuers, selling security holders, and their affiliated purchasers, from directly or indirectly bidding for, purchasing, or attempting to induce another person to bid for or purchase a covered security during a restricted period.317 Rule 17a– 4 requires a broker-dealer to preserve certain records if it makes or receives them.318 In accordance with the requirements of section 939A(b), the Commission is adopting amendments to Rules 101 and 102 of Regulation M that remove the Investment Grade Exceptions and add, in their place, new exceptions for Nonconvertible Securities for which the issuer’s probability of default, estimated as of the sixth business day immediately preceding the determination of the offering price and over the horizon of 12 full calendar months from such day, is 0.055% or less, as determined and documented, in writing, by the distribution participant acting as the lead manager (or in a similar capacity) of a distribution, as derived from a structural credit risk model.319 The Commission is also adopting Rule 17a– 4(b)(17), which requires broker-dealers to preserve the written probability of default determination, relied upon pursuant to the new exception in Rule 101(c)(2)(i) or Rule 102(d)(2)(i), as applicable, for a period of not less than three years, the first two years in an easily accessible place.320 The Commission is
also adopting identical new exceptions in Rules 101(c)(2)(ii) and 102(d)(2)(ii) for asset￾backed securities that are offered pursuant to an effective shelf registration statement filed on Form SF–
3.321 The discussion of estimates that
follows is limited to the new information collection requirements that result from the final amendments related to the probability of default determinations in Rule 101(c)(2)(i), as amended, as well as the record preservation thereof in reliance on the new exceptions provided in Rule 101(c)(2)(i) or Rule 102(d)(2)(i) pursuant to Rule 17a–4(b)(17). The Commission is not estimating that the new exception for asset-backed securities that are offered pursuant to an effective shelf registration statement filed on Form SF– 3 in Rules 101(c)(2)(ii) and 102(d)(2)(ii) will increase or decrease the existing approved information collections because whether an asset-backed security is offered pursuant to an effective shelf registration statement filed on Form SF–3 is an objective, observable fact that would not incur any PRA burden. A. Respondents The respondents under the amended rules are lead managers who choose to make a probability of default determination in order to rely on the exception for Nonconvertible Securities and other broker-dealers who use the lead manager’s probability of default determination in relying on an exception for Nonconvertible Securities. As noted in Part V.A.1, there were 201 lead managing underwriters and 100 other non-lead manager broker-dealers of Nonconvertible Securities in 2021.322 The Commission assumes that, on balance, these numbers will remain consistent given the capital, expertise, and relationships needed to serve as the lead underwriter of a Nonconvertible Securities offering. The Commission, therefore, is estimating that 301 respondents will be subject to PRA burdens under the amendments. The respondents under the amendments to Rule 101(c)(2)(i) are lead managers who make probability of default determinations. The Commission, therefore, is estimating that 201 respondents will be subject to PRA burdens under Rule 101(c)(2)(i), as amended. The respondents under the amendments to Rule 17a–4(b)(17) are broker-dealers who rely on the new exception in Rule 101(c)(2)(i) or Rule 102(d)(2)(i). The Commission, therefore, is estimating that 301 respondents will be subject to PRA burdens under new Rule 17a–4(b)(17).323 B. Use of Information The information collected under the amendments ensures that the Nonconvertible Securities that are least likely to be subject to the type of manipulation that Regulation M seeks to address are excepted from Rules 101 and 102. Further, the Commission believes that the information contained in the records required to be preserved pursuant to Rule 17a–4(b)(17) will facilitate the Commission in conducting examinations of broker-dealers who rely on the new exceptions in Rule 101(c)(2)(i) or Rule 102(d)(2)(i).
C. Collection of Information
As discussed below, the Commission believes that respondents will incur PRA burdens under the amendments to Rule 101(c)(2)(i) because distribution participants who are acting as the lead manager (or in a similar capacity) of a distribution and make a probability of default determination are required for each distribution of Nonconvertible Securities to determine the subject issuer’s probability of default in order to rely on the exception. These respondents may also incur PRA burdens in their probability of default determinations. Respondents who are broker-dealers and rely on the new exception in Rule 101(c)(2)(i) or Rule 102(d)(2)(i) will incur PRA burdens under the requirements set forth in new Rule 17a–4(b)(17) because they are required to preserve records of the written probability of default determination.

1. Burden and Cost Estimates Related to
   the Rule 101 Amendments
   Rule 101(c)(2)(i), as amended, permits lead managers to gather the data serving as the inputs and then perform the analysis necessary to calculate the probability of default of the issuer

324The Commission recognizes that some respondents may choose to utilize the probability of default estimates that are calculated and made available by a third-party vendor rather than make the determination themselves. In the Proposal, the Commission noted that the Commission’s burden estimates for the adopted amendments to Rule 101 are based on respondents gathering the required data and calculating the probability of default, internally, without the use of third-party vendors, because the Commission lacks granular information from which to base an estimate of the proportion of respondents that would use vendors. The Commission requested comment on the extent to which respondents may use third-party vendors, as well as the costs and time burdens of using such services. See Proposal, 87 FR 18326 n.129. However, the Commission did not receive comments in response to this request. For purposes of estimating the PRA burdens under the final rules as amended, the Commission continues to assume that all respondents will make the probability of default determination internally with data they have gathered, rather than use third party vendors. As discussed above, in Part II.A.1, there may be distributions with more than one distribution participant acting as the lead manager (or in a similar capacity), but only one of the distribution participants acting as the lead manager would be permitted to make the probability of default determination for the particular distribution. See supra note 91. Therefore, for purposes of the PRA estimations in this release, only one lead manager on any distribution for purposes of these calculations is assumed. 325Further, respondents who choose to utilize probability of default estimates that are calculated and made available by a third-party vendor will already have access to the vendor’s software and systems containing these estimates, typically as part of an existing subscription, so they will not need to procure further services or subscriptions from these vendors to access any such determinations. However, as noted above, for purposes of estimating these PRA burdens, the Commission assumes all respondents would make their own calculations and not use third party vendors. This assumption is being made to provide an estimate reflecting for the more costly of the two approaches. 326 [201 lead managers] × [3 hours] = 603 hours. The Proposal included 237 respondents, which was taken from available data from 2020. The number included herein reflects the number from the available data from 2021, as discussed above, in
Part V.A.1. In addition, under the Proposal, the 237
figure included non-lead manager broker-dealers
who would have been eligible, under the proposed Rule 101(c)(2)(i), to make probability of default determinations in order to meet the Nonconvertible Securities exception’s conditions. 327This number was obtained from Mergent, a financial data provider. 328 [603 hours (initial burden)] + [33,798 hours (ongoing annual burden)] = 34,401 hours. 329See 17 CFR 239.45. 330 [301 respondents] × [25 hours] = 7,525 hours. 331 [301 respondents] × [10 hours] = 3,010 hours. 332 [7,525 hours (initial burden)] + [3,010 hours (ongoing annual burden)] = 10,535 hours. whose securities are the subject of the distribution to meet the conditions of the exception.324 This requirement will result in respondents incurring a PRA recordkeeping burden. This process will likely be highly automated, and that respondents will initially comply with this requirement by reprograming systems to create a means to calculate electronically the probability of default based on manually gathered and entered inputs for financial modeling. The respondents who make probability of default determinations will be broker￾dealers serving as lead managers and are likely to have experience in using their own proprietary version of a publicly available structural credit risk model. Accordingly, the initial configuration of systems will be handled internally and take 3 hours per respondent. The Commission also assumes that broker￾dealers serving as lead managers already have the software and systems in place required to make the calculations.325 The Commission therefore estimates that the total industry-wide initial burden for configuring systems to make and probability of default estimates is 603 hours.326 An issuer’s probability of default is forward-looking and changes over time, so the Commission believes that respondents will manually gather the inputs required to calculate an issuer’s probability of default each time it participates in a distribution of Nonconvertible Securities. There were 33,798 offerings of Nonconvertible Securities in 2021.327 Because financial modeling generally, and the probability of default calculation more specifically, is well-known by industry participants, the Commission believes that respondents have employees who are familiar with how to gather the required model inputs. The Commission, therefore, estimates that it will take lead-manager respondents roughly one hour per distribution of Nonconvertible Securities to determine and document, in writing, the probability of default determinations. Accordingly, the Commission estimates that calculating the probability of default pursuant to Rule 101(c)(2)(i), as amended, will result in an aggregate annual ongoing industry-wide burden of 33,798 hours. The Commission estimates that the total PRA burden resulting from the final amendments to Rule 101 is 34,401 hours in the first year 328 and 33,798 hours thereafter. The Commission does not believe that the amendments to Rule 101(c)(2)(ii) excepting asset-backed
securities that are offered pursuant to an effective shelf registration statement filed on Form SF– 3 will result in respondents incurring PRA burdens because whether an asset￾backed security has an effective shelf registration statement filed on Form SF– 3 is an objective, observable fact.329 Further, there is no corresponding record preservation requirement for respondents documenting reliance on the exception for asset-backed securities under Rule 101(c)(2)(ii), as amended.
2. Burden and Cost Estimates Related to
the Rule 17a–4 Amendments New Rule 17a–4(b)(17) requires broker-dealers to preserve the written probability of default determination, relied upon pursuant to the new exception in Rule 101(c)(2)(i) or Rule 102(d)(2)(i), as applicable, for a period of not less than three years, the first two years in an easily accessible place. The Commission estimates that this record preservation requirement imposes an initial burden of 25 hours per respondent for updating the applicable policies and systems required to account for preserving the records made pursuant to Rule 101(c)(2)(i), as amended. Accordingly, the Commission estimates that the total industry-wide initial burden for this requirement is 7,525 hours.330 The Commission also estimates that respondents will incur an ongoing annual burden of 10 hours per firm for maintaining such records, as well as to make additional updates to the applicable record preservation policies and systems to account for preserving the records pursuant to new Rule 17a– 4(b)(17), resulting in a total ongoing industry-wide burden of 3,010 hours.331 The Commission, therefore, estimates that the total PRA burden resulting from the amendment to Rule 17a–4 is 10,535 hours in the first year 332 and 3,010 hours per year thereafter. PRA SUMMARY TABLE Industry-wide burden due to amendments to Initial burden hours Ongoing annual burden hours/year (after first year) Total PRA burden hours in first year Rule 101 ...................................................................................................................................... 603 33,798 34,401

333 5 U.S.C. 601 et seq.
334 5 U.S.C. 605(b).
335Although section 601(b) of the RFA defines the term ‘‘small business,’’ the statute permits agencies to formulate their own definitions. The Commission has adopted definitions for the term ‘‘small business’’ for the purposes of Commission rulemaking in accordance with the RFA. Those definitions, as relevant to this rulemaking, are set forth in 17 CFR 240.0–10 (‘‘Rule 0–10’’). Rule 0–10 also provides that the Commission may, if warranted by the circumstances, use a different definition for particular rulemakings. See 17 CFR 240.0–10. 336 5 U.S.C. 603(a). 337 5 U.S.C. 551 et seq. 338 5 U.S.C. 605(b). 339See 17 CFR 240.17a–5(d). 340See 17 CFR 240.0–10(c). 341 17 CFR 242.0–10(a). 342See Proposal, 87 FR 18337. 343As discussed above, in Part II.B, broker-dealers who are affiliated with the issuer and do not meet the definition of ‘‘distribution participant’’ under Rule 100(b) of Regulation M may be covered persons under Rule 102. Even if those broker￾dealers had net capital over $500,000, they would not be small entities under Rule 0–10 because they are affiliated with an issuer (of investment grade securities) that is not a small entity. PRA SUMMARY TABLE—Continued Industry-wide burden due to amendments to Initial burden hours Ongoing annual burden hours/year (after first year) Total PRA burden hours in first year Rule 17a–4 .................................................................................................................................. 7,525 3,010 10,535 D. Collection of Information Is Mandatory The information collections for making probability of default determinations under the amendments to Rule 101 are mandatory for reliance on exceptions in Rule 101(c)(2)(i) or Rule 102(d)(2)(i). In addition, the information collections involving the preservation of written probability of default determinations under the amendments to Rule 17a–4 are mandatory if a broker-dealer relies on the new exception in Rule 101(c)(2)(i) or Rule 102(d)(2)(i). E. Confidentiality of Responses to Collection of Information The Commission would not typically receive confidential information as a result of these collections of information. To the extent that the Commission receives—through its examination and oversight program, through an investigation, or by some other means—records or disclosures from a distribution participant regarding the probability of default determination, such information would be kept confidential, subject to the provisions of applicable law. F. Retention Period for Record Preservation Requirement Pursuant to new Rule 17a–4(b)(17), a broker-dealer is required to preserve the written probability of default determination, relied upon pursuant to the new exception in Rule 101(c)(2)(i) or Rule 102(d)(2)(i), as applicable, for a period of not less than three years, the first two years in an easily accessible place.
VII. Regulatory Flexibility Act
Certification
The Regulatory Flexibility Act (‘‘RFA’’) 333 requires Federal agencies, in promulgating rules, to consider the impact of those rules on ‘‘small entities,’’ 334 a term that includes ‘‘small businesses.’’ 335 Section 603(a) 336 of the Administrative Procedure Act,337 as amended by the section 604(a) of the RFA requires the Commission to undertake a final regulatory flexibility analysis of rules it is adopting, unless the Commission certifies that the rules would not have a significant impact on a substantial number of small entities.338 Small entities include broker-dealers with total capital (net worth plus subordinated liabilities) of less than $500,000 on the date in the prior fiscal year as of which its audited financial statements were prepared pursuant to Rule 17a–5(d) under the Exchange Act,339 or, if not required to file such statements, a broker-dealer who had total capital (net worth plus subordinated liabilities) of less than $500,000 on the last day of the preceding fiscal year (or in the time it has been in business, if shorter), and is not affiliated with any person (other than a natural person) who is not a small business or small organization.340 A small business or small organization, for purposes of ‘‘issuers’’ or ‘‘person’’ other than an investment company, is defined as a person who, on the last day of its most recent fiscal year, had total assets of $5 million or less.341 In the Proposal, the Commission certified, pursuant to section 605(b) of the RFA, that the proposed amendments to Rules 101 and 102 would not have a significant economic impact on a substantial number of small entities.342 The Commission requested but did not receive any comments on the certification as it related to the entities impacted by Rule 101 or Rule 102 of Regulation M, or by Rule 17a–4 under the Exchange Act. Based on the Commission’s analysis of the existing information relating to broker-dealers who are subject to Rules 101, 102,343 and 17a–4, it is unlikely that any broker-dealer categorized as a ‘‘small business’’ or ‘‘small organization’’ under Rule 0–10 could serve as an underwriter or other distribution participant, as they would almost certainly have insufficient capital to participate in underwriting activities. In addition, the Commission continues to believe that none of the various persons affected by the amendments would qualify as a small entity under the Rule 0–10 definition as it is unlikely that any issuer of that size had investment grade securities that were eligible for the Investment Grade Exception. Accordingly, the Commission believes it is unlikely that, in the future, a small entity may become impacted by the amendments because broker-dealers who enter this market are likely to have at least $500,000 in total capital, as described above, or to be affiliated with a person who is not a small business or small organization as defined under Rule 0–10, and because issuers of securities that are eligible for the new exceptions
provided in Rules 101(c)(2) and 102(d)(2) are likely to have total assets greater than $5 million. For the foregoing reasons, the Commission certifies, pursuant to
section 605(b) of Title 5 of the U.S.
Code, that the amendments to Rules 100, 101, 102, and 17a–4 will not have a significant economic impact on a substantial number of small entities. Statutory Authority The final amendments contained in this release are being adopted under the authority set forth in sections 939 and 939A of the Dodd-Frank Act and

sections 3(b), 15, 23(a), and 36 of the Exchange Act. List of Subjects in 17 CFR Parts 240 and 242 Broker-dealers, Fraud, Issuers, Reporting and recordkeeping requirements, Securities. Text of Rule Amendments For the reasons set out in the preamble, the Commission is amending 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, 78j–4, 78k, 78k–1, 78l, 78m, 78n, 78n–1, 78o, 78o–4, 78o–10, 78p, 78q, 78q–1, 78s, 78u–5, 78w, 78x, 78dd, 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. 1376 (2010); and Pub. L. 112–106, sec. 503 and 602, 126 Stat. 326 (2012), unless otherwise noted.

---

Section 240.17a–4 also issued under secs.
2, 17, 23(a), 48 Stat. 897, as amended; 15 U.S.C. 78a, 78d–1, 78d–2; sec. 14, Pub. L. 94– 29, 89 Stat. 137 (15 U.S.C. 78a); sec. 18, Pub.
L. 94–29, 89 Stat. 155 (15 U.S.C. 78w);

---

• 2. Amend § 240.17a–4 by adding
     paragraph (b)(17) to read as follows:
     § 240.17a–4 Records to be preserved by certain exchange members, brokers and dealers.

---

(b) * * *
(17) The written probability of default determination, relied upon by such broker or dealer, pursuant to § 242.101(c)(2)(i) or § 242.102(d)(2)(i) of this chapter (Rule 101 or Rule 102 of Regulation M), as applicable.

---

PART 242—REGULATIONS M, SHO,
ATS, AC, NMS, AND SBSR AND CUSTOMER MARGIN REQUIREMENTS FOR SECURITY FUTURES

• 3. The authority citation for part 242
  continues to read as follows:
  Authority: 15 U.S.C. 77g, 77q(a), 77s(a), 78b, 78c, 78g(c)(2), 78i(a), 78j, 78k–1(c), 78l, 78m, 78n, 78o(b), 78o(c), 78o(g), 78q(a), 78q(b), 78q(h), 78w(a), 78dd–1, 78mm, 80a– 23, 80a–29, and 80a–37.
• 4. Amend § 242.100 in paragraph (b)
     by adding in alphabetical order a definition for ‘‘Structural credit risk model’’ to read as follows:
     § 242.100 Preliminary note; definitions.

---

(b) * * *
Structural credit risk model means any commercially or publicly available model that calculates, based on an issuer’s balance sheet, the probability that the value of the issuer will fall below the threshold at which the issuer would fail to make scheduled debt payments, at or by the expiration of a defined period.

---

• 5. Amend § 242.101 by revising
     paragraph (c)(2) to read as follows:
     § 242.101 Activities by distribution participants.

---

(c) * * *
(2) Certain nonconvertible and asset￾backed securities. (i) Nonconvertible debt securities and nonconvertible preferred securities of issuers for which the probability of default, estimated as of the sixth business day immediately preceding the determination of the offering price and over the horizon of 12 full calendar months from such day, is 0.055% or less, as determined and documented, in writing, by the distribution participant acting as the lead manager (or in a similar capacity) of a distribution, as derived from a structural credit risk model; or (ii) Asset-backed securities that are offered pursuant to an effective shelf registration statement filed on Form SF– 3 (§ 239.45 of this chapter); or

---

• 6. Amend § 242.102 by revising
     paragraph (d)(2) to read as follows:
     § 242.102 Activities by issuers and selling security holders during a distribution.

---

(d) * * *
(2) Certain nonconvertible and asset￾backed securities. (i) Nonconvertible debt securities and nonconvertible preferred securities of issuers for which the probability of default, estimated as of the sixth business day immediately preceding the determination of the offering price and over the horizon of 12 full calendar months from such day, is 0.055% or less, as determined and documented, in writing, by the distribution participant acting as the lead manager (or in a similar capacity) of a distribution, as derived from a structural credit risk model, pursuant to § 242.101(c)(2)(i); or (ii) Asset-backed securities that are offered pursuant to an effective shelf registration statement filed on Form SF– 3 (§ 239.45 of this chapter); or

---

By the Commission.
Dated: June 7, 2023.
Vanessa A. Countryman,
Secretary.
[FR Doc. 2023–12591 Filed 6–16–23; 8:45 am] BILLING CODE 8011–01–P