Artificial intelligence (AI) is an increasingly important technology within the investment management industry.1AI has been used in a variety of waysincluding as the newest strategy for attempts to "beat the market" by outperforming passive index funds that are benchmarked against the S&P 500, despite the long-standing finding that index funds consistently win that contest.2
Investment advisers who use AI should consider the unique issues the technology raises in light of an adviser's fiduciary duty to its clients. In this client alert, we provide an overview of how AI is being used by investment advisers, the fiduciary duties applicable to investment advisers, and particular issues advisers should consider in designing AI-based programs, to ensure they are acting in the best interests of their clients.3
How Artificial Intelligence Is Being Adopted by Investment Advisers
AI is currently used by investment advisers in a variety of innovative ways:
Issues Raised by an Investment Adviser's Fiduciary Duties
Under federal law, an investment adviser is a fiduciary to its clients.8An adviser's fiduciary duty involves a duty of care and a duty of loyalty, which, although not defined specifically in the Investment Advisers Act of 1940 (Advisers Act), have been addressed and developed through U.S. Securities and Exchange Commission (SEC) interpretive releases and guidance, as well as case law.9As discussed below, these duties have implications for an adviser's use of AI. The specific obligations required by an adviser's fiduciary duty will depend upon what functions the adviser has agreed to assume for the client.10While the SEC has not provided specific guidance for advisers using AI, current guidance raises unique considerations for advisers to consider.11
Duty of Loyalty
The duty of loyalty requires investment advisers not to place their own interest ahead of their clients' interests.12An adviser must make full and fair disclosure to its client of all material facts relating to the advisory relationship and employ reasonable care to avoid misleading clients. Information provided to clients must be sufficiently specific so that a client is able to understand the investment adviser's business practices and conflicts of interest.
An adviser's duty of loyalty raises, among others, the following issues with respect to AI-based investment management programs:
What facts does an adviser need to disclose about its use of AI?
Advisers should consider disclosing information such as the following:
How should advisers think about the tension between disclosure obligations and confidentiality regarding proprietary technologies?
It is important for advisers to disclose enough information for investors to make an informed decision about engaging, and then managing the relationship with, the investment adviser. Advisers should be careful to not mislead clients, and information provided to clients should be sufficiently specific so that a client is able to understand the investment adviser's business practices. However, highly technical information about the process behind the AI's decisions might not be beneficial to a client's understanding of the adviser's platform.
Does an adviser need to disclose the historical success rate of returns from using artificial intelligence?
Historically, funds that employ AI have not outperformed the S&P 500.13Investment advisers might therefore be expected to provide disclosures indicating that an adviser has not conclusively proven AI's ability to predict securities prices and may not "beat the market."14
Duty of Care
The duty of care requires, among other things, the duty to provide advice appropriate for the client and the duty to monitor a client's investments, and the ongoing suitability of those investments, over the course of the relationship.15An adviser must develop a reasonable understanding of the client's objectives and have a reasonable belief that the advice it provides is in the best interest of the client, based on the client's portfolio and objectives.
An adviser's duty of care raises, among others, the following issues with respect to AI-based investment management programs:
Can an adviser replace traditional suitability assessments with alternative data or other AI-based tools?
If an AI-based system makes investment choices on behalf of clients using deep or machine learning that develops on its own, by tracking client behavior, or by using alternative data, the adviser should pay particularly close attention to how the recommendations generated by those data might differ or be in conflict with a client's explicit preferences and investment objectives. It is possible that an adviser using AI-based tools will make different assessments of what is best or appropriate for the client than if the adviser uses more traditional tools like suitability questionnaires that ask a client about her risk profile, investment objectives, and other characteristics. As a result, an adviser using AI-based systems to generate an investment management system may be at cross-purposes with the client, which would raise issues based on the adviser's duty of care.
How frequently should an investment adviser evaluate its AI program?
Because AI programs create their own rules based on the data they analyze, and autonomously make trading decisions, advisers should develop internal procedures for ensuring their programs are operating correctly.16For example, advisers should adequately test their AI before and periodically after it is integrated into the investment platform. In addition, advisers should develop strategies for procedures they can implement to adjust their AI programs if they do not produce favorable results. Advisers should also monitor for possible cybersecurity threats.
How should an adviser review investment decisions directed by AI to ensure the decisions still fit within a client's investment goals?
Advisers using AI should adopt and implement procedures that will periodically review the performance of their AI, to ensure that performance is within expected parameters and that decisions are not being made to the detriment of clients' investment goals. Ultimately, the adviser is responsible for all decisions made by its AI-based program and therefore cannot let an AI-based program simply run without the adviser's active monitoring.17
[1] For example, on May 26, 2020, HSBC announced the launch of its AI Powered US Equity Index family, the first equity index product powered by artificial intelligence and big data, using data insights from IBM Watson to guide equity trading decisions. HSBC Launches First Equity Index Products Powered by AI and Big Data, Business Wire (May 26, 2020), https://www.businesswire.com/news/home/20200526005556/en/HSBC-Launches-Equity-Index-Products-Powered-AI.
[2] In his 1972 book A Random Walk Down Wall Street, Burton Malkiel maintains that share prices evolve similarly to a random walk and have no relationship with historic values or other variables. Because future share prices are not based on any pattern, they cannot be predicted, and it is unlikely any random fund adviser will be able to out-perform a passive investment fund. Burton Malkiel, A Random Walk Down Wall Street: The Time-Tested Strategy for Successful Investing. W.W. Norton, 1972. Since the 1972 publication, research continues to show that beating the market is unlikely. For example, a study released at the end of 2019 indicated that over the past 10 years, 89 percent of large-cap funds, 84 percent of mid-cap funds, and 89 percent of small-cap funds failed to outperform their S&P 500 benchmark on a relative basis. SPIVA U.S. Year-End 2019 Scorecard, S&P Dow Jones Indices (2019), https://us.spindices.com/indexology/core/spiva-us-year-end-2019.
[3] Of course, the federal securities laws raise issues that are important for investment advisers who employ AI to consider even outside their fiduciary duties. For example, advisers using AI may want to rely on Rule 3a-4 under the Investment Company Act of 1940, which provides a safe harbor from registration as an investment company for advisory programs that are reasonably customized to particular investors, with respect to their investment programs. If an AI-based program provides similar advice to a variety of clients, however, the adviser may not be able to rely on Rule 3a-4. Thus, advisers who manage client accounts based on AI may need to consider the specific contours of their AI programs in light of the requirements of Rule 3a-4. This alert by no means covers the range of issues AI models raise for investment advisers.
[4] Adam Satariano and Nishant Kumar, The Massive Hedge Fund Betting on AI, Bloomberg (Sept. 27, 2017), https://www.bloomberg.com/news/features/2017-09-27/the-massive-hedge-fund-betting-on-ai.
[5] Adam Satariano and Nishant Kumar, The Massive Hedge Fund Betting on AI, Bloomberg (Sept. 27, 2017), https://www.bloomberg.com/news/features/2017-09-27/the-massive-hedge-fund-betting-on-ai.
[6] Robo-advisers are online platforms, and typically registered investment advisers, that provide discretionary asset management services to their clients through online algorithmic-based programs. See Division of Investment Management, Robo Advisers, IM Guidance Update No. 2017-02 (Feb. 2017).
[7] See, e.g., Ryan W. Neal, Wealthfront Turns to Artificial Intelligence to Improve Robo Advice (Mar. 31, 2016), https://www.wealthmanagement.com/technology/wealthfront-turns-artificial-intelligence-improve-robo-advice.
[8] SEC v. Capital Gains Research Bureau, Inc., 375 U.S. 180, 194 (1963). An investment advisers fiduciary duty is imposed under the Advisers Act in recognition of the relationship of trust between an investment adviser and a client and is made enforceable by the antifraud provisions of Section 206 of the Advisers Act.
[9] See, e.g., Capital Gains 375 U.S. at 194; SEC v. Tambone, 550 F.3d 106, 146 (1st Cir. 2008); SEC v. Moran, 944 F. Supp. 286, 297 (S.D.N.Y 1996); Transamerica Mortgage Advisors, Inc. v. Lewis, 444 U.S. 11, 17 (1979); Commission Interpretation Regarding Standard of Conduct for Investment Advisers, Investment Advisers Act Release No. 5248 (July 12, 2019); IM Guidance Update No. 2017-02 (Feb. 2017); Proxy Voting by Investment Advisers, Investment Advisers Act Release No. 2106 (Jan. 31, 2003). See also SEC Commissioner Kara M. Stein, Surfing the Way: Technology, Innovation, and Competition, Remarks at Harvard Law Schools Fidelity Guest Lecture Series (No. 9, 2015), available at https://www.sec.gov/news/speech/surfing-wave-technology-innovation-and-competition-remarks-harvard-law-schools-fidelity (former SEC Commissioner Stein discussing the application of fiduciary duties to digital advice provided by robo-advisers).
[10] See, e.g., Investment Advisers Act Release No. 5248) (in interpretive guidance related to investment advisers fiduciary duties, explaining that while all investment advisers owe their clients a fiduciary duty, that fiduciary duty must be viewed in the context of the agreed-upon scope of the relationship between the adviser and the client; the fiduciary duty itself may not be waived, but the exact responsibilities of the adviser in managing the clients account may be limited).
[11] Although not directly on point, guidance released by the SECs Division of Investment Management with respect to robo-advisers provides some light on the types of considerations advisers should address based on their fiduciary duties to clients when adopting non-traditional methods of investment advisory services. IM Guidance Update No. 2017-02 (Feb. 2017). Additionally, former SEC personnel have spoken on the SECs own use of AI. See, e.g., SEC Commissioner Kara M. Stein, From the Data Rush to the Data Wars: A Data Revolution in Financial Markets (Sept. 27, 2018), available at https://www.sec.gov/news/speech/speech-stein-092718; Scott W. Bauguess, Acting Director and Acting Chief Economist, SEC Division of Economic and Risk Analysis, The Role of Big Data, Machine Learning, and AI in Assessing Risks: a Regulatory Perspective (June 21, 2017), available at https:// http://www.sec.gov/news/speech/bauguess-big-data-ai.
[12] Investment Advisers Act Release No. 5248.
[13] Mark Hulbert, Using AI for Picking Stocks? Not So Fast, Wall Street Journal (Jan 5, 2020), https://www.wsj.com/articles/use-ai-for-picking-stocks-not-so-fast-11578279960.
[14] In a response to a request for a no action letter, the staff of the SEC explained to an investment adviser purporting to rely on abilities as a psychic medium that he would be required to disclose that the predictive value of his methods had not been scientifically established. John Anthony, SEC Staff No Action Letter (Mar. 19, 1975). Similarly, because the predictive value of AI has not been conclusively established, investment advisers should consider the extent to which the SEC would expect them to disclose the lack of proof that AI-based algorithms are better than or even equal to the success of more traditional tools.
[15] Investment Advisers Act Release No. 5248.
[16] The Financial Industry Regulatory Authority (FINRA) published guidance in 2016 related to the governance and supervision framework advisers should have in place to adopt client-facing digital investment advice tools. While not directly related to AI, the guidance emphasizes the importance of supervising algorithms used in digital-advice tools and periodically assessing whether [the] algorithm is consistent with the firms investment and analytical approaches. See FINRA Report on Digital Investment Advice (Mar. 2016), available at https://www.finra.org/sites/default/ files/digital-investment-advice-report.pdf.
The SEC has previously brought enforcement actions against investment advisers for failing to supervise those in charge of programming and monitoring algorithmic or other automated trading strategies. In one enforcement action, the founder of a hedge fund allowed his co-founder complete control of operating and monitoring an algorithm to make trade decisions. He was made aware that the algorithm was not working as expected and made no efforts to follow-up with the co-founder or inform investors or prospective investors of any issues related to the algorithm. In the Matter of Timothy S. Dembski, Investment Advisers Act Release No. 4671 (March 24, 2017).
[17] For example, robo-adviser Wealthfront was the subject of an SEC enforcement action for providing a false statement that it monitored all client accounts to avoid transactions that might result in wash sale, which would potentially reduce investor returns, when it did not actually do so. See In the Matter of Wealthfront Advisers, Investment Advisers Act Release No. 5086 (Dec. 21, 2018).
See the rest here:
- Classic reasoning systems like Loom and PowerLoom vs. more modern systems based on probalistic networks - November 8th, 2009 [November 8th, 2009]
- Using Amazon's cloud service for computationally expensive calculations - November 8th, 2009 [November 8th, 2009]
- Software environments for working on AI projects - November 8th, 2009 [November 8th, 2009]
- New version of my NLP toolkit - November 8th, 2009 [November 8th, 2009]
- Semantic Web: through the back door with HTML and CSS - November 8th, 2009 [November 8th, 2009]
- Java FastTag part of speech tagger is now released under the LGPL - November 8th, 2009 [November 8th, 2009]
- Defining AI and Knowledge Engineering - November 8th, 2009 [November 8th, 2009]
- Great Overview of Knowledge Representation - November 8th, 2009 [November 8th, 2009]
- Something like Google page rank for semantic web URIs - November 8th, 2009 [November 8th, 2009]
- My experiences writing AI software for vehicle control in games and virtual reality systems - November 8th, 2009 [November 8th, 2009]
- The URL for this blog has changed - November 8th, 2009 [November 8th, 2009]
- I have a new page on Knowledge Management - November 8th, 2009 [November 8th, 2009]
- N-GRAM analysis using Ruby - November 8th, 2009 [November 8th, 2009]
- Good video: Knowledge Representation and the Semantic Web - November 8th, 2009 [November 8th, 2009]
- Using the PowerLoom reasoning system with JRuby - November 8th, 2009 [November 8th, 2009]
- Machines Like Us - November 8th, 2009 [November 8th, 2009]
- RapidMiner machine learning, data mining, and visualization tool - November 8th, 2009 [November 8th, 2009]
- texai.org - November 8th, 2009 [November 8th, 2009]
- NLTK: The Natural Language Toolkit - November 8th, 2009 [November 8th, 2009]
- My OpenCalais Ruby client library - November 8th, 2009 [November 8th, 2009]
- Ruby API for accessing Freebase/Metaweb structured data - November 8th, 2009 [November 8th, 2009]
- Protégé OWL Ontology Editor - November 8th, 2009 [November 8th, 2009]
- New version of Numenta software is available - November 8th, 2009 [November 8th, 2009]
- Very nice: Elsevier IJCAI AI Journal articles now available for free as PDFs - November 8th, 2009 [November 8th, 2009]
- Verison 2.0 of OpenCyc is available - November 8th, 2009 [November 8th, 2009]
- What’s Your Biggest Question about Artificial Intelligence? [Article] - November 8th, 2009 [November 8th, 2009]
- Minimax Search [Knowledge] - November 8th, 2009 [November 8th, 2009]
- Decision Tree [Knowledge] - November 8th, 2009 [November 8th, 2009]
- More AI Content & Format Preference Poll [Article] - November 8th, 2009 [November 8th, 2009]
- New Planners Solve Rescue Missions [News] - November 8th, 2009 [November 8th, 2009]
- Neural Network Learns to Bluff at Poker [News] - November 8th, 2009 [November 8th, 2009]
- Pushing the Limits of Game AI Technology [News] - November 8th, 2009 [November 8th, 2009]
- Mining Data for the Netflix Prize [News] - November 8th, 2009 [November 8th, 2009]
- Interview with Peter Denning on the Principles of Computing [News] - November 8th, 2009 [November 8th, 2009]
- Decision Making for Medical Support [News] - November 8th, 2009 [November 8th, 2009]
- Neural Network Creates Music CD [News] - November 8th, 2009 [November 8th, 2009]
- jKilavuz - a guide in the polygon soup [News] - November 8th, 2009 [November 8th, 2009]
- Artificial General Intelligence: Now Is the Time [News] - November 8th, 2009 [November 8th, 2009]
- Apply AI 2007 Roundtable Report [News] - November 8th, 2009 [November 8th, 2009]
- What Would You do With 80 Cores? [News] - November 8th, 2009 [November 8th, 2009]
- Software Finds Learning Language Child's Play [News] - November 8th, 2009 [November 8th, 2009]
- Artificial Intelligence in Games [Article] - November 8th, 2009 [November 8th, 2009]
- Artificial Intelligence Resources - November 8th, 2009 [November 8th, 2009]
- Alan Turing: Mathematical Biologist? - April 25th, 2012 [April 25th, 2012]
- BBC Horizon: The Hunt for AI ( Artificial Intelligence ) - Video - April 30th, 2012 [April 30th, 2012]
- Can computers have true artificial intelligence" Masonic handshake" 3rd-April-2012 - Video - April 30th, 2012 [April 30th, 2012]
- Kevin B. Korb - Interview - Artificial Intelligence and the Singularity p3 - Video - April 30th, 2012 [April 30th, 2012]
- Artificial Intelligence - 6 Month Anniversary - Video - April 30th, 2012 [April 30th, 2012]
- Science Breakthroughs - April 30th, 2012 [April 30th, 2012]
- Hitman: Blood Money - Part 49 - Stupid Artificial Intelligence! - Video - April 30th, 2012 [April 30th, 2012]
- Research Members Turned Off By HAARP Artificial Intelligence - Video - April 30th, 2012 [April 30th, 2012]
- Artificial Intelligence Lecture No. 5 - Video - April 30th, 2012 [April 30th, 2012]
- The Artificial Intelligence Laboratory, 2012 - Video - April 30th, 2012 [April 30th, 2012]
- Charlie Rose - Artificial Intelligence - Video - April 30th, 2012 [April 30th, 2012]
- Expert on artificial intelligence to speak at EPIIC Nights dinner - May 4th, 2012 [May 4th, 2012]
- Filipino software engineers complete and best thousands on Stanford’s Artificial Intelligence Course - May 4th, 2012 [May 4th, 2012]
- Vodafone xone™ Hackathon Challenges Developers and Entrepreneurs to Build a New Generation of Artificial Intelligence ... - May 4th, 2012 [May 4th, 2012]
- Rocket Fuel Packages Up CPG Booster - May 4th, 2012 [May 4th, 2012]
- 2 Filipinos finishes among top in Stanford’s Artificial Intelligence course - May 5th, 2012 [May 5th, 2012]
- Why Your Brain Isn't A Computer - May 5th, 2012 [May 5th, 2012]
- 2 Pinoy software engineers complete Stanford's AI course - May 7th, 2012 [May 7th, 2012]
- Percipio Media, LLC Proudly Accepts Partnership With MIT's Prestigious Computer Science And Artificial Intelligence ... - May 10th, 2012 [May 10th, 2012]
- Google Driverless Car Ok'd by Nevada - May 10th, 2012 [May 10th, 2012]
- Moving Beyond the Marketing Funnel: Rocket Fuel and Forrester Research Announce Free Webinar - May 10th, 2012 [May 10th, 2012]
- Rocket Fuel Wins 2012 San Francisco Business Times Tech & Innovation Award - May 13th, 2012 [May 13th, 2012]
- Internet Week 2012: Rocket Fuel to Speak at OMMA RTB - May 16th, 2012 [May 16th, 2012]
- How to Get the Most Out of Your Facebook Ads -- Rocket Fuel's VP of Products, Eshwar Belani, to Lead MarketingProfs ... - May 16th, 2012 [May 16th, 2012]
- The Digital Disruptor To Banking Has Just Gone International - May 16th, 2012 [May 16th, 2012]
- Moving Beyond the Marketing Funnel: Rocket Fuel Announce Free Webinar Featuring an Independent Research Firm - May 23rd, 2012 [May 23rd, 2012]
- MASA Showcases Latest Version of MASA SWORD for Homeland Security Markets - May 23rd, 2012 [May 23rd, 2012]
- Bluesky Launches Drones for Aerial Surveying - May 23rd, 2012 [May 23rd, 2012]
- Artificial Intelligence: What happened to the hunt for thinking machines? - May 25th, 2012 [May 25th, 2012]
- Bubble Robots Move Using Lasers [VIDEO] - May 25th, 2012 [May 25th, 2012]
- UHV assistant professors receive $10,000 summer research grants - May 27th, 2012 [May 27th, 2012]
- Artificial intelligence: science fiction or simply science? - May 28th, 2012 [May 28th, 2012]
- Exetel taps artificial intelligence - May 29th, 2012 [May 29th, 2012]
- Software offers brain on the rain - May 29th, 2012 [May 29th, 2012]
- New Dean of Science has high hopes for his faculty - May 30th, 2012 [May 30th, 2012]
- Cognitive Code Announces "Silvia For Android" App - May 31st, 2012 [May 31st, 2012]
- A Rat is Smarter Than Google - June 5th, 2012 [June 5th, 2012]