A financial robo-advisor understands the investment objectives, risk profile and investment horizon of customers through online questionnaires, arrives at the most suitable investment recommendations for the customers through computer algorithms, provides services that cope with market changes through balancing, completely avoids subjective human influences through computer algorithms and maintains the reasonableness and discipline of investing. (Note: 1)
The emergence of financial robo-advisors is quite influential to the promotion of the “financial inclusion” rationale. The core tenet of “financial inclusion” is that everyone has opportunities to participate in economic activities and enjoy financial services. (Note: 2) In contrast to traditional financial recommendations, they require the customers to have a certain scale of assets and to pay a certain consultancy fee in order to receive services from financial advisors and thus exclude a large portion of people from the services, resulting in the phenomenon of so-called “financial exclusion.”The advantages of financial robo-advisors lie in low management fees and low entry barriers, making professional services of financial advisors accessible to almost anyone.
Although financial robo-advisors have the advantages of low cost, automated management and stable investment, they are still subject to their inherent constraints. For example, there are no sufficient customized products. As the source of customer recommendations by financial robo-advisors is information provided by customers via online questionnaires, the contents of online questionnaires cannot necessarily take into account large short-term expenditures or tax, pension and real estate planning. In addition, financial robo-advisors are potentially vulnerable to human interferences, algorithm configuration and provision of market information and are still dependent on background operation of a team of experts. Therefore, whether they can entirely avoid subjective human influences remains an important issue to be observed in the future.
Currently, there are several financial institutions and relevant operators that have actively engaged in such financial robo-advisor services, including O-Bank, Nomura Asset Management, Tarobo Investment Advisors, Business Intelligence, Fu Hua Securities Investment Trust, Fubon Securities, UBS, CTBC Bank, etc. Among them, the services of O-Bank and Tarobo Investment Advisors have been officially launched. It should be noted, however, that in spite of the name “financial robo-advisors,” currently they are still part of smart wealth management, still require a team of experts and are not completely driven by machine self-learning to automatically generate investment portfolios for customers. Therefore, some operators refer to such service as “smart wealth management.” (Note: 3)
With respect to such market changes and demands, the Financial Supervisory Commission (hereinafter, the “FSC”) approved the Operating Guidelines of the Securities Investment & Consulting Association of the ROC for Providing Robo-Advisor Services by Securities Investment Consulting Enterprises through Automated Tools (hereinafter, the “Guidelines”) in June 2017. The Guidelines reference recommendations or guidance for automated investment tools released by the Securities and Exchange Commission (SEC) and the Financial Industry Regulatory Authority (FINRA) of the US and in the European Union, the UK, Australia, Canada, etc. Relevant regulatory highlights are provided as follows:
(1) Notice to customers before the use of such services
Point 8 of the Guidelines stipulates that when engaging in automated investment consulting services, a securities investment consulting enterprise shall, before the first use of the services, communicate to customers and ensure that they are aware of or have learned the following:(i) Before such services are used, the contents of services and other relevant publicly disclosed information should be carefully reviewed; and (ii) the investment tools have their inherent constraints and potential gaps from the reality, including gaps in the basic assumptions of the system or programs and the scope of product offering; investment portfolios recommended by the automated investment consulting services are directly tied to information provided by customers; and system output may not necessarily meet the financial needs or targets of individual customers.
(2) Supervisory mechanisms for algorithms and programs
Point 3 of the Guidelines provides that an operator shall conduct effective supervision and management of the algorithms used by the system and shall set up internal supervisory mechanisms for initial review and regular review. The initial review mechanisms shall include: (i) whether the algorithm used by the system can achieve the expected results should be evaluated by understanding the methodology used by the algorithm and the assumptions, biases and preferences of the system; and (ii) the data input into the system should be understood; (iii) output testing should be conducted to ensure that the output meets the expectation. The regular review mechanisms shall include: (i) whether the model used by the system can still be used appropriately in case of changes in market situations or economic conditions should be evaluated; (ii) system output shall be regularly tested to ensure that the output meets the original expectation; and (iii) personnel shall be assigned to supervise the system.
(3) Supervisory mechanisms for investment portfolios and conflicts of interest
Point 5 of the Guidelines provides that to faithfully observe principles which require priority to customer’s interest, avoidance of conflicts of interest and unjust enrichment, and fair dealings, it is necessary to prevent the establishment and selection of investment portfolios from conflicting the interest of customers by ensuring that the system can fairly and objectively execute functions, including: (i) deciding the parameters for investment portfolios, such as returns, extent of diversification, credit risks or liquidity risks; (ii) establishing the selection criteria for including securities in an investment portfolio, such as trading costs, liquidity risks and credit risks; (iii) selecting appropriate securities for inclusion in an investment portfolio and reviewing the algorithm if the securities are selected by the algorithm; and (iv) reviewing the appropriateness of the system default investment portfolio recommendations to customers concerned in terms of their the risk profile and tolerance.
(4) Requirements under the know-your-customer/ appropriateness principle
Point 4 of the Guidelines provides that before investment portfolio recommendations are provided by an automated investment consulting service, the profile of the customer should be established to perform the know-your-customer operation. The objectives and duration of a customer’s investment should be understood. In addition, matters including but not limited to the customer’s investment knowledge, investment experience, financial status and risk tolerance should be sufficiently learned and evaluated.
(5) Investment portfolio rebalancing
Point 6 of the Guidelines provides that when it is necessary to readjust ratios of the assets as previously recommended due to different return performance of the recommended assets in an investment portfolio, if an automated investment consulting service system has an automatic rebalancing function for the embedded asset portfolios to meet the risk profile of the customers or to maintain the originally determined ratios to reduce investment portfolio risks, relevant rebalancing transaction details about the investment portfolios of the automated investment consulting service should be agreed with the customers in advance, and the customers shall be informed of: (i) the provision of the investment portfolio rebalancing service; (ii) the ways in which the portfolio rebalancing works; and (iii) all kinds of costs and other potential constraints which may arise from the investment portfolio rebalancing. In addition, the system should set up a policy and procedure for the automated investment consulting service system to deal with major market changes.
In addition, the FSC further announced that under certain conditions, an investment trust enterprise is allowed to execute rebalancing transactions for customers automatically via its computer system in August 2017. In order to maintain the investment vehicles and investment ratios originally agreed with customers, an investment trust enterprise and its customers may agree in a contract to rebalancing transactions which are automatically executed by a computer system when the profit or loss of an individual investment vehicle or the overall investment portfolio reaches a predetermined threshold or when the deviation of the originally agreed-upon investment ratios reaches the predetermined threshold, and the results of the transaction execution will be communicated to the customers in real time.
(6) Supervision by a professional committee
Point 7 of the Guidelines provides that a dedicated committee shall be set up within an enterprise or its group to take charge of the design and contents of internal customer questionnaires, the development and adjustment of algorithms, alignment of customers’ portfolio recommendations with their risk profiles, the supervision and management of the fair and objective execution and portfolio rebalancing by the automated investment consulting service system, or the participation in the review and due diligence of external software developers and suppliers to evaluate the appropriateness of the system. The dedicated committee shall also ensure that such enterprise has set up comprehensive preventive, detection and handling network security measures.
While financial robo-advisor services are currently in full swing in Taiwan and the FSC has also approved the relevant requirements mentioned above, still consumer protection is not adequate. How can operators and consumers share their risks in the event of any defect in the design of the algorithms or programs per se or of any hacking? Are mechanisms in place which allow consumers to seek compensation? These are all major issues to be discussed in the future.
Note 1: Trading Valley (2017), Financial Robo-Advisors – An Introduction, https://www.stockfeel.com.tw/機器人理財-引言/ (last date of review: Oct. 6, 2017).
Note 2: Innovative Operation Strategies of the Financial Sector under Key Emerging Trends, Yang-cheng Lu, Shih-chieh Lin, Kai-chun, Chang, Chien-yu Lai, Ming-tai Chung, Shun-feng Hsieh, Chun-yi Yeh, Yuan-teh Huang, Juo-yu Huang, the Taiwan Academy of Banking and Finance, the Bankers of the Republic of China (2015), Pages 25-26 and 82, Taipei: the Taiwan Academy of Banking and Finance
Note 3: CTBC Bank’s Complete Reliance on Artificial Intelligence-based Autonomous Investing by Financial Robo-advisors Still under Experiment, reported by Chung-ying Sun (2017) for the United Daily News, https://udn.com/news/story/7239/2558383?from=udn-ch1_breaknews-1-cate6-news