Guidance on Non-Face-to-Face Distribution of Investment and Insurance Products

Annex Guidance on non-F2F distribution of investment and insurance products (A) Appropriateness of non-F2F service mode for certain customers While online platforms in general can provide convenience to customers, AIs should consider whether providing investment services or long term insurance products through such channels would be appropriate for certain customers, such as VCs who may not be able to make independent investment or insurance decisions. There are various ways to achieve this objective. Box 1: Practical examples for reference, not prescribed standards: Some possible factors for considering whether online platforms are appropriate for certain customers: • Certain customers may not be capable of subscribing to investment services or purchasing long term insurance products via online platforms (e.g. persons with visual or intellectual impairment, or persons who are illiterate) unless sufficient support is provided (e.g. online platforms may not necessarily be inappropriate to persons who are illiterate if voice navigation is provided to facilitate the use) • Certain customers whose ability to understand complex products, high-risk products or long term insurance products through online platforms may be in doubt • Certain customers lacking technology literacy may not be ready to use online platforms unless relevant consumer education or guidance on the use of online platforms is provided Possible attributes of a customer’s technology literacy may include the customer’s prior experience in using online platforms (say, online banking or shopping platforms) or other digital products for online services or processes. Some possible ways for identifying customers lacking “technology literacy”: • Assess a customer’s technology literacy with the use of simple online questionnaire • Assess a customer’s technology literacy by manned channel through interaction with the customer

2 (B) Handling of transactions with VCs AIs should have regard to the following protection measures related to VCs in providing investment services and long term insurance products through non-F2F channels. (I) Assessment for VCs According to the HKMA’s Investor Protection Measures Circular1 and relevant guidance, AIs should determine whether a customer is a VC (“VC Assessment”) and provide VCs with additional protection in certain circumstances. Unlike having F2F interaction between AIs’ staff and customers at AIs’ premises, certain non-F2F channels without human interaction may pose challenges to AIs in determining whether a customer is a VC. For example, some customer’s circumstances (e.g. observable impairments) or changes in a customer’s circumstances may not be observable by AIs via online platforms. Under such situations, AIs can introduce measures to ascertain relevant customers’ circumstances and any subsequent changes which indicate that a customer may be a VC or could lead to a change in the VC Assessment result (e.g. from non-VC to VC) but may not be observable through online platforms. Box 2: Practical examples for reference, not prescribed standards: • Enquire into a customer’s circumstances for VC Assessment by customer’s declaration or other reasonable approaches • Monitor and review any irregular risk profiling results and suspicious transactions, so as to help reveal any changes in the customer’s circumstances • Provide the option of VC Assessment by manned channel Following the risk-based approach, it is acceptable for AIs not to conduct VC Assessment for sale of investment products where the suitability obligation is not applicable, e.g. online (as well as offline) transactions on non-complex investment products conducted on an execution-only basis. Along this principle, it is also acceptable for AIs not to conduct VC Assessment for sale of long term insurance 1 HKMA’s Circular dated 25 September 2019 on “Investor Protection Measures in respect of Investment, Insurance and Mandatory Provident Fund Products”, see section (A)(III.2) and section (B)(VIII) of Annex 1.

3 products where the product is exempt from or not subject to the financial needs analysis (“FNA”) requirement under the Guideline on Financial Needs Analysis (“GL30”) issued by the IA and no recommendation is provided. (II) Extra care for handling investment and long term insurance product transactions with VCs Unless otherwise stipulated in relevant guidance (e.g. Companion Requirement, PICOP, and mandatory audio-recording and provision of fact sheet for annuity insurance products for retail banking customers), AIs may apply a risk-based approach in considering the extent and operational arrangements of extra care for handling investment and long term insurance product transactions of VCs, having regard to the circumstances of the customers and the products. Box 3: Practical examples for reference, not prescribed standards: • Show reminder message(s) in the online journey to encourage VCs to dial telephone hotline or approach branch for assistance, if necessary • Provide the option of video-recording by the AI for sale conducted through video-conference tools • Send notifications, e.g. via email or SMS, to the VC and his/her designated contact person upon successful online transactions • Conduct post-sale controls (e.g. sample call back on a risk-based approach to ensure VCs’ understanding on their investment transactions placed through online platforms)

Companion requirement for VCs (applicable to retail banking customers only) As set out in section (A)(II.7) and section (B)(VI) of Annex 1 to the Investor Protection Measures Circular1 , AIs should allow VCs to choose whether he/she would like to (i) bring along a companion to witness the sale process; and/or (ii) have a second front-line staff member to handle the sale (“Companion Requirement”). The Companion Requirement is not mandatory for online investment and long term insurance product transactions where companion is not practicable (e.g. transactions through online platforms without involvement of AI’s staff). Nonetheless, it is feasible and prudent to arrange for companion under certain circumstances. For example, where investment or long term insurance product

4 transactions involve the use of video-conference or telephone conference, AIs should follow the choice of the VC on the Companion Requirement, to the extent applicable and practicable. Box 4: Practical examples for reference, not prescribed standards: • Provide option of witnessing through live chat with co-browsing by appropriate staff of the AI (C) Selling process (I) Suitability assessment for investment transactions In the course of our supervisory processes, it is noted that some registered institutions (“RIs”) have implemented online suitability assessment for complex investment products through pragmatic and user-friendly approaches. Box 5: Practical examples for reference, not prescribed standards: • Make use of available and relevant customer’s information from the AI’s records (e.g. retrieving from customer’s valid risk profiling result), instead of asking for customer’s information from scratch, for conducting suitability assessment (II) Suitability assessment for sale of long term insurance products Subject to GL30 and unless otherwise exempted, AIs are required to conduct FNA with a customer to ensure that the insurance product being recommended is suitable for the customer regardless of the channel of distribution. Given the differences in the modes of operation of different non-F2F channels, the extent and granularity of the information to be collected for FNA can be varied taking into account the particular circumstances of the target customers and the modes of the distribution channels. That said, AIs have an obligation to collect adequate information such that they can perform reasonable assessment before making any insurance recommendation. For products that are exempt from or not subject to the FNA requirement, before recommending the products to a customer, AIs are required to carry out appropriate suitability assessment in relation to the customer’s circumstances in

5 accordance with Standards and Practices 6.1 of the Codes of Conduct issued by the IA 2 regardless of the channel of distribution. The level of suitability assessment should be proportionate and reasonable, taking into account the customer’s circumstances and the nature and complexity of the products concerned, and at the minimum comply with the following standards: (i) AIs should ascertain whether the customer has the need(s) and/or objective(s) which match(es) with the nature and features of the product(s). (ii) AIsshould explain clearly to the customer how the recommended product(s) match(es) with his/her need(s) and/or objective(s). (iii) No insurance product should be recommended where information available to AIs reveals any issue of concern about the customer’s affordability. Notwithstanding the foregoing, when deciding whether non-F2F channels should be used to distribute a particular insurance product, AIs should take into consideration, among other things, the features of the product. For example, a product with complex features may not be suitable for distribution through online platforms where advice or explanation to customers cannot be sufficiently and effectively given during the sale process.

(III) Measures relating to “unsolicited” / “execution-only” / “no recommendation” investment and long term insurance product transactions It is noted that some AIs treat certain online investment transactions as “unsolicited” / “execution-only” transactions, or distribute eligible long term insurance products via online platforms on “no recommendation” basis. AIs are reminded that whether solicitation or recommendation is involved depends on the facts and circumstances of each case. AIs which adopt “unsolicited” / “execution-only” / “no recommendation” transactions model should have proper policy, procedures and controls to ensure that such model will not be abused to circumvent their internal policy and/or regulatory requirements such as suitability assessment or FNA requirement. These include clear guidance to relevant staff; effective review mechanism to test check compliance; appropriate management information system (“MIS”) reports to help identify possible abuse; and proper design of sales target and incentive system. 2 Code of Conduct for Licensed Insurance Agents and Code of Conduct for Licensed Insurance Brokers.

6 Box 6: Practical examples for reference, not prescribed standards: • Sample call back to confirm with customer that those transactions classified as “unsolicited” / “execution-only” / “no recommendation” genuinely did not involve any solicitation or recommendation by AI’s staff • Design financial incentives scheme for sales staff that is commensurate with the level of participation of and support provided by them for “execution￾only” or “no recommendation” transactions For the avoidance of doubts, complex investment products (other than derivatives traded on an exchange where paragraph 6.5 of the SFC’s Guidelines on Online Distribution and Advisory Platforms is applicable) should not be put on online platforms for sale without any suitability assessment, regardless of whether solicitation or recommendation is made. (IV) Controls over transactions with mismatch(es) or exception(s) under online environment As set out in section (A)(II.5) and (B)(V) of Annex 1 to the Investor Protection Measures Circular1 , in suitability assessment, AIs should take extra care in handling transactions with mismatch(es) or exception(s). For retail banking customers, there should be an appropriate level of supervision to review and approve as appropriate a transaction with mismatch(es) or exception(s) before execution. In the context of online transactions, corresponding measures could be appropriate system control for assessing the justification (e.g. could be in the form of some pre-set choices of reasons) for the transaction with mismatch(es) or exception(s). For the avoidance of doubt, merely customer’s acknowledgement for the transaction with mismatch(es) or exception(s) shall not discharge AIs’ suitability obligation where applicable. For investment transactions, AIs may also consider conducting the review and approval before the batch processing or day-end processing, instead of during the selling process, if customers’ orders are executed by batches or at day-end. For sale of long term insurance products, review and approval of transactions with mismatch(es) or exception(s) should not be later than the submission of applications to the insurers.

7 (V) Product disclosure The objective of product disclosure is to enable the customer to understand the investment or insurance product before entering into a transaction. AIs are reminded to make clear and adequate disclosure of relevant material information to customers. In doing so, AIs should ensure that information is communicated in an easily comprehensible manner, and use plain language in disclosure and presentation of information to customers. AIs should also provide customers with channels for making enquiries.

Box 7: Practical examples for reference, not prescribed standards: • Pop ups / video clips that explain complex terms through the journey; hotline or live chat in case customers have any questions; Frequently Asked Questions, etc. (VI) Audio-recording of sale process of investment products As mentioned in the HKMA’s circular of 23 December 2020 3 , where the communication facility does not have appropriate record-keeping function which allows AIs to record, retrieve and monitor communication with customers, AIs should put in place compensating measures and controls to ensure compliance with relevant regulatory requirements. For example, if the investment transaction concerned falls under the circumstances that require audio-recording or phone recording, AIs should put in place measures (such as use of office phone with audio-recording capability) to record the relevant conversations with customers or to record the relevant recap. For the avoidance of doubt, AIs should audio-record the relevant conversations with customers or the relevant recap for investment transactions (regardless of whether solicitations or recommendations are involved; and whether complex products are involved) conducted through video-conference. The recording should cover, where applicable, suitability assessment; product disclosure; and order placement and confirmation. The records should be retained according to applicable requirements. AIs are also reminded to comply with other applicable regulatory requirements, including but not limited to the SFC’s “Code of Conduct for Persons Licensed by or Registered with the Securities and Futures Commission”, in particular 3 HKMA’s circular dated 23 December 2020 on “Frequently Asked Questions on Investor Protection Measures”.

8 paragraph 3.9 about order recording and various provisions about information disclosure and suitability obligation, and the SFC’s circular dated 4 May 2018 on “Receiving client orders through instant messaging”, where applicable. (VII) Audio-recording of sale process of long term insurance products AIs are reminded to comply with the existing requirements on audio-recording4 in all non-F2F distribution channels. For example, if audio-recording of the selling process is not practical in a particular non-F2F channel, AIs should not sell annuity insurance products to VCs, or non-VCs who choose to have audio￾recording arrangement. In addition, when selling any long term insurance products through video￾conference, AIs are reminded to also comply with the end-to-end recording requirements in the IA’s circular “Sandbox application for the distribution of long term insurance policies via video conferencing tools” dated 5 August 2020. (VIII) Pre-investment Cooling-off Period (“PICOP”) (applicable to retail banking customers only) PICOP is to provide additional protection to retail banking customers with less sophistication in respect of certain products. PICOP will allow the customer more time to consider the appropriateness of the investment and consult third parties if necessary. AIs are reminded that PICOP applies to applicable investment transactions5 with solicitation or recommendation conducted online. In other words, PICOP does not apply to online transactions without solicitation or recommendation. (IX) Signature requirements for insurance products Obtaining wet signatures may not be practical in some non-F2F channels. According to the IA’s circular “Sandbox application for the distribution of long term insurance policies via video conferencing tools” of 5 August 2020, AIs may adopt any appropriate technologies or arrangements for obtaining a customer’s signature so long as the signature is legally binding and the customer is fully aware of the documents to be signed. In this connection, for online platforms, AIs should 4 Section (B)(II) of Annex 1 to the Investor Protection Measures Circular. 5 Section (A)(II.6) of Annex 1 to the Investor Protection Measures Circular.

9 put in place proper controls in the platforms to ensure that full content of the document has been displayed to the customer before the request for customer’s signature is made. (D) Monitoring AIs should implement proper and effective monitoring and review mechanisms for identifying, monitoring and reviewing any irregular customer risk profiling results, as well as suspicious investment or long term insurance product transactions or trading patterns conducted through non-F2F channels, especially those conducted by VCs. In case of any identified abnormal risk profiling results, transactions or trading patterns, AIs should take appropriate follow-up action. Box 8: Practical examples for reference, not prescribed standards: Some possible alerts that may warrant attention: • Sudden significant changes in customers’ risk profiling results and/or trading patterns • Frequent update of customer risk profile or log-in to online platforms Examples of post-sale monitoring: • Regular MIS reports on transactions conducted by VCs, such as past trading channels and patterns, number of logins to the online platforms, etc. • Conducting call back to VCs upon detection of irregular customer behaviour (e.g. sudden significant changes in customers’ risk profiling results and/or trading patterns) (E) Provision of assistance to customers for using non-F2F channels To facilitate customers’ use of non-F2F channels, AIs should have arrangements for customers to seek assistance and guidance from the AI for their use of non￾F2F channels to conduct investment or long term insurance product transactions. Box 9: Practical examples for reference, not prescribed standards: • Provide hotline, live chat or the option of co-browsing of online platforms by AI’s staff to see what the customer is browsing in real-time so that assistance and guidance could be provided • Provide educational materials or tools to facilitate customers to understand the usage, risks and benefits associated with non-F2F channels (e.g. video

10 demo on the use of online platforms)