The Taiwan Banker

Taiwan fully allowed robo-advisory services in June 2017. Later, the pandemic spurred demand for digital financial management, as well as the development of AI applications. According to the Financial Supervisory Committee (FSC), 17 companies had launched robo-advisory services as of the end of August, managing NT$9.728 million in assets for 200,000 customers. However, the scale has the FSC concerned. It announced that robo-advisory services will be included in the scope of the Securities Investment Advisor Occupational Regulations in the fourth quarter, held to a higher legal standard. New robotic advisors launched by fintech firms will also be similarly regulated.Supervisory requirements for algorithmsIf an AI is your financial planner, then who is responsible for the quality of the advice given? In practice, the world’s major powers have all recognized the demands of regulating robotic advisory services, such as the US, UK, and Singapore. Even though the US currently lacks specific robo-advising regulations, the Securities Exchange Commission (SEC) has issued relevant guidelines requiring companies to supervise their algorithms effectively, and to conduct targeted audits. The UK’s Financial Conduct Authority (FCA) and the Monetary Authority of Singapore (MAS) have set professional qualification standards for robo-advisory services. MAS, meanwhile, has set further regulatory conditions for algorithms.Because the applications of AI and other machine learning algorithms remain opaque and difficult to explain, many legislators and markets are apprehensive. In June, to strengthen supervision, the FSC gave notice to robotic advisors, currently self-regulated, that it will be including robotic advising within the scope of the Securities Investment Advisor Occupational Regulations doctrine, raising its legal status. Since the notice ended on September 2, the FSC has been evaluating external suggestions, and it is expected to release the formal guidelines in Q4. In other words, if fintech companies want to launch automated investment portfolio advisory service, they must comply with the application process by obtaining a license. Moreover, the new system will increase the capital requirement to NT$50 million, and the margin requirement to NT$10 million. Companies who currently do not meet these requirements, currently two, will have one year to make adjustments.New regulations: carrot or stick?So-called “automated investment advisory” refers to the use of technology like AI and algorithms to offer customers automated investment portfolio recommendations. It began in June 2017, when the FSC released the relevant operating essentials. Investors were required to fill in a form before investing.According to FSC statistics, robotic advisory has grown year after year in the number of customers and assets under management (AUM) since 2018, growing to and 175,000 customers and NT$7.792 billion by the end of 2023, up 1267% and 700% over just five years from the end of 2018 from NT$570 million and 21,000 respectively.At the beginning of this year, the FSC further simplified the procedures for automatic rebalancing. Based on a three-way contract signed by the advisor, securities trader, and client, if the advisor obtains the client’s agreement allowing them to complete automatic rebalancing trades,the advisor and securities dealer won’t need customer consent again going forward, allowing for true automation. Industry leaders are now optimistic that this field will break NT$10 billion in 2024.This prediction will indeed be realized in September. According to FSC statistics from the end of August, robo-advisory AUM has grown to NT$9.728 billion. If the growth rate remains unchanged, it will exceed NT$10 billion by September.This is the reason FSC is giving robo-advisory the carrot, while also holding the stick of increased supervision. An FSC Securities and Futures Bureau official interviewed by the Taiwan Banker explained that in January 2018, when domestic firms started to offer automated investment advisory services, only two companies provided it. Currently, in contrast, 17 companies offer this service. With the rapid growth in customers and AUM, increased oversight is a necessity to safeguard the rights of investors. Large consumer banks join robo-advisoryBroadly speaking, most large domestic consumer banks are investing heavily in robo-advisory services.For example, CTBC, the largest domestic bank, launched CTBC Intelligent Investment, which pairs AI with big data. Investors need only to log in to the app, select their preferred investments, and decide on either a single transaction or periodic fixed investment. The app then integrates the customer profile with market trends to develop an “Intelligent Investment Allocation” (iMAP) and “Investment Think Tank” for the customer.Cathay United has similarly launched the “Cathay Intelligent Investment Platform” in collaboration with its own Cathay Securities, offering US stock ETFs. The platform allows customers to customize their investment portfolio based on their personal risk preferences. It also has a rebalancing mechanism, which monitored 24 hours a day. If the portfolio ratio drifts from the original allocation goal, the system proactively issues a rebalancing notification.According to FSC statistics, the top three service providers are Cathay United, First Bank, and Mega Bank with NT$2 billion, NT$1.5 billion, and NT$1 billion in AUM respectively. In the investment banking industry, Fuhua Trust and Nomura Investment Bank each have NT$200 million. Many large consumer banks have added robo-advisory sertvices in turn, and more Taiwanese people are becoming accustomed to using AI services. Therefore, it is likely that the growth of this market will only accelerate, but because the current algorithms are self-regulated, it is difficult to enforce accountability. Therefore, officials have signaled that the FSC is resolved to upgrade the supervision of automated investment services from self-regulation to legal enforcement. At the same time, it is strengthening external investigation and industry self-regulation mechanisms. The FSC is planning to release these regulations in Q4, but will continue to examine the laws and adjust them as needed.Providers must obtain operating licensesWill all domestic companies specializing in robo-advisory, be they banks or non-banks, begin to be regulated in the fourth quarter? A Securities and Futures Bureau official responded that the decision on what companies to include in the newest regulations will depend on whether the services conform with the definition established by law or by self-regulation – that is, whether are completely internet-based, with little to no human input on investment portfolio advice.The official indicates that if the company meets the definition “automated investment advisory services,” it must first obtain a FSC-issued investment advisory license to operate. The 17 companies that have already launched robo-advisors all must still meet the aforementioned requirements, but no application will be required.However, if fintech companies only provide programming or support software development but do not provide portfolio advice, they do not meet the law’s definition. The official says that in such cases, there is no requirement to apply to the FSC.The official added that as of mid-September, no fintech companies have inquired about this application process. After the regulations are implemented, fintech companies who desire to launch robo-advisory services must first receive investment advisory licenses, and comply with the new system’s capital and operating margin standards. As the scope of robo-advisory grows, the big will get bigger, while the weaker withdraw. According to FSC statistics from August, there were 17 automated investment advisory companies, one fewer than six months prior. Officials confirm that the company withdrew from the market because their robo-advising business didn’t reach its intended scope. With little business, it decided to discontinue its services.Finance professionals generally believe that those who will dominate the robo-advisory market must be able to acquire a large customer base and asset management scale, reducing the fixed cost pressures on model development.