As 2018 wound down, the Securities and Exchange Commission for the first time charged two robo-advisers for violations of federal securities laws. These cases represent a trend that has been a long time in the making and will culminate in more such actions in 2019 and beyond: the clash between algorithms and the SEC.
One aspect of the trend is investment advisers' increased use of algorithms — essentially procedures or sets of rules telling a computer how to execute, much like a recipe with steps a person might follow to achieve a certain dish. Robo-advisers invest based on input from investors followed by the output of algorithms, rather than recommendations of human advisers.
The underlying technology is not new — traditional financial advisers have had similar tools available to them for decades. In 2014, one Hong Kong venture capital fund famously appointed an algorithm to its board of directors. With robo-advisers, however, the technology is available directly to retail clients, much like a self-checkout line in a grocery store.
As accounts typically have low minimums and carry minimal fees, robo-advisers are popular with beginner and lower-net-worth investors, for whom traditional financial advisers may be out of reach. By some estimates, robo-advisers collectively manage over $200 billion in assets.
In 2016, the SEC said it would begin auditing robo-advisers for compliance with securities laws, and in February 2017, the SEC issued an investor bulletin and guidance on the topic. Now, in two separate cases, the SEC has made the point that using an algorithm does not relieve investment advisers of their obligations under the federal securities laws.
On Dec. 21, the SEC announced a settlement with Wealthfront Advisers, an online robo-adviser that provides software-based portfolio management, including a tax-loss harvesting program for clients' taxable accounts. In connection with that program, the SEC alleged that Wealthfront falsely represented to clients it would monitor their accounts to avoid transactions that might trigger a wash sale. In fact, the SEC alleged, Wealthfront failed to conduct such monitoring, rendering its representations misleading.
In a separate action against another robo-adviser, Hedgeable Inc., the SEC alleged that the adviser misleadingly compared its results to performances of other robo-advisers. According to the SEC, Hedgeable calculated its returns based on a small subset of client accounts and miscalculated its competitors' trading model returns.
While groundbreaking because they involved robo-advisers, the Wealthfront and Hedgeable actions allege misconduct by humans rather than as a result of malfunctioning algorithms. The SEC has a long history of cases against investment advisers for misrepresenting to clients, so it comes as no surprise that advisers who make such misrepresentations in connection with an algorithmic trading platform will meet a similar fate.
However, another aspect of the trend toward more frequent clashes between algorithms and the SEC: algorithms that misfire.
The impact of automated trading first became clear after a five-minute span on May 6, 2010, in which the Dow Jones Industrial Average dropped by 1,000 points—the so-called flash crash, an event largely attributed, correctly or incorrectly, to algorithmic trading. More recently, a Dec. 25 Wall Street Journal article attributed a "market swoon" to what it called "The Herdlike Behavior of Computerized Trading." Increasingly, algorithms are being faulted for their impact on the market.
With such potential for large market impacts, algorithmic misbehavior will be shown little tolerance by regulators. Indeed, in October 2013, the SEC brought an action against a securities broker after the broker's automated trading system experienced a significant error. According to the SEC, as a result of erroneously placed computer code, the firm, during a 45-minute span, errantly "traded more than 397 million shares, acquired several billion dollars in unwanted positions, and eventually suffered a loss of more than $460 million." Although the error was inadvertent, the SEC found that the firm did not reasonably manage the risk of its market access.
The SEC's message has been clear: Although the federal securities laws were enacted long before anyone imagined trusting a robot with investment decisions, they will be applied to robo-advisers in full force. Investment advisers and others relying on automated technology to serve their clients should thoughtfully review their operations to ensure compliance with securities laws.
Misrepresenting the human oversight of algorithms, misrepresenting the performance of algorithms and failing to properly control the risk of algorithms not performing as intended will increasingly be the subject of SEC enforcement actions in 2019 and beyond.
Nicolas Morgan is a partner at Paul Hastings and former senior trial counsel at the SEC. Lily Lysle is an associate at Paul Hastings.