On the heels of the Financial Industry Regulatory Authority issuing a Report on Digital Investment Advice last month, on April 7 the Securities and Exchange Commission approved a FINRA rule amendment that requires registration of persons that are “primarily responsible for the design, development or significant modification of algorithmic trading strategies, or who are responsible for the day-to-day supervision or direction of such activities.”
The fundamental proposition is that all algorithms are conceived and built by humans, and said humans should be both qualified and readily identifiable so they may be held accountable for any harm their creations cause.
The specific rule amendment is to NASD Rule 1032(f). This rule calls for the registration of any “Securities Trader,” which as you likely guessed is a representative that is engaged in securities trading on behalf of the firm. One becomes a registered Securities Trader by passing the Series 57 exam; supervision of Securities Traders requires both the Series 57 and Series 24 (with some exceptions).
Algorithmic trading designers and developers are now swept under this same regime, and therefore must pass the Series 57 and fulfil requisite continuing education like any other registered Securities Trader. FINRA justifies the rule expansion given the widespread use of algorithmic trading strategies used by member firms, as well as their significance to the market as a whole.
So what exactly is an algorithmic trading strategy? To quote the SEC’s Notice of Filing: “An automated system that generates or routes orders or order-related messages such as routes or cancellations, but does not include an automated system that solely routes orders received in their entirety to a market center.” Such systems generate orders or execute trades “without material intervention by any person.”
Importantly, FINRA notes that the scope of what will be considered an algorithmic trading strategy will continue to evolve as new systems and strategies are developed. It is essentially an admission that regulators can’t write or modify rules fast enough to keep up with technological innovation.
Equally as important, especially for robo-advisors, is that algorithms that solely generate a recommended asset allocation are specifically carved out: “An automated investment service that constructs portfolio recommendations, but that is not equipped to automatically generate orders and order-related messages to effectuate such trading ideas into the market” would not trigger registration requirements under Rule 1032(f).
The designers and developers of algorithms that merely analyze clients’ risk tolerance questionnaire answers and spit out a recommended portfolio of securities are safe. If the algorithm goes a step further to also cause the recommended portfolio of securities to be purchased, rebalanced or sold in a client’s account via automated order-flow, however, registration would indeed appear to be required.
This latest rule approval is but a microcosm of the larger regulatory trend toward increased oversight of the humans that write the codes that drive the financial services industry.
Both FINRA and the SEC, through a combination of public speeches, rulemaking, guidance updates and enforcement actions, are becoming more and more curious about fintech’s architects. Engineers, coders and developers will be thrust into the spotlight as time goes on, whether they like it or not. How this ultimately affects the fiduciary accountability puzzle is the next chapter that has yet to be written (I’m far from the first person to query how a line of source code can be fiduciary – a very human construct).
“Pay no attention to that man behind the curtain,” said the Wizard of Oz at the end of the 1939 fairytale. Dorothy and her gang knew better than to be satisfied with blind trust, and the same can be said of our friends in Washington.