COMPLY-Can an algorithm be a fiduciary?
Aug 14 (Reuters) - A surge of interest in automated investment management firms has sparked debate about who or what can best meet the high ethical standards required by regulators for recommending investments - software or flesh-and-blood advisers?
New so-called robo-advisers, like Wealthfront Inc and Betterment LLC, are Web-based firms that use computer algorithms to select investments and tweak portfolios over time, for annual fees that are as low as 0.15 percent of a client's assets under management, compared with 1 percent or more by fee-only independent advisers.
But they are regulated just like independent advisers who set up offices and meet clients on a regular basis - they typically register with the U.S. Securities and Exchange Commission and are deemed "fiduciaries" who must put their clients' interests above their own. (Securities brokers who earn commissions are regulated differently.)
Robo-advisers currently control a small percentage of the assets managed by the advisory industry, but are rapidly moving to the mainstream. Charles Schwab Corp has said it plans to roll out its own automated investment service. Wealthfront just announced a deal with the National Football League's San Francisco 49ers under which some retired football players and team employees will receive the automated investment advice. Other robo-advisers include Hedgeable Inc and Personal Capital Corp.
Compliance professionals and lawyers say they are already fielding questions about whether robo-advisers can live up to regulators' expectations. Critics are skeptical, but lawyers say that robo-advisers can indeed be fiduciaries. Securities regulations do not require human contact in an investment adviser's client relationships, said Michael Koffler, a lawyer in New York who counsels investment advisory firms on regulatory issues.
AN ELECTRONIC FIDUCIARY
To put the client first, both automated and human advisers must discern clients' risk tolerance. To do that, both often use questionnaires, though the human adviser will typically also have a face-to-face meeting with a new client.
At Betterment, for example, new customers fill out questionnaires covering everything from home-buying plans to retirement goals, said Jon Stein, Betterment's co-founder and chief executive. The firm's algorithms then pull from a short list of exchange-traded funds (ETFs) to create a portfolio that fits the client's risk tolerance and time horizon. The firm, based in New York, manages $755 million in assets.
Of course those algorithms are programmed by humans who are informed by an investment policy committee that includes economists and financial analysts. Clients can also speak to customer service representatives, but their role is limited.
While Betterment has given personalized advice to a handful of clients with complex situations, the company's business is primarily automated, Stein said. Critics say automation may overlook details that come up in person, such as a sick parent's expenses.
Automation can actually reduce the risk of human error, said Mike Alfred, head of San Diego-based research firm BrightScope. Robo-advisers are betting "they will know almost as much as a human adviser and what they don't, they'll make up for in not making dumb mistakes," Alfred said.
Steering clear of programming mistakes is part of a robo-adviser's fiduciary obligation. Algorithms must manage clients' accounts in accordance with their goals and work as a firm promises, said Koffler, the lawyer, who notes that disclosures should also be clear. "Clients should know what they're getting and not getting."
For example, at Wealthfront, which manages $1.3 billion in assets, clients know they are not getting budgeting and estate planning advice, said Adam Nash, the firm's president and chief executive. The predominantly young clients sign up because the advice is automated, Nash said, adding, "Most of my clients would pay me to never call them." (Reporting by Suzanne Barlyn in New York; editing by Linda Stern and Matthew Lewis)