MIT Sloan Professor Andrew Lo
From the Wall Street Journal
At a conference last year, I was approached by an audience member after my talk. He thanked me for my observation that it’s unrealistic to expect investors to do nothing in the face of a sharp market-wide selloff, and that pulling out of the market can sometimes be the right thing to do. In fact, this savvy attendee converted all of his equity holdings to cash by the end of October 2008.
He then asked me for some advice: “Is it safe to get back in now?” Seven years after he moved his money into cash, he’s still waiting for just the right time to reinvest; meanwhile, the S&P 500 earned an annualized return of 14% during this period.
Investing is an emotional process. Managing these emotions is probably the greatest open challenge of financial technology. Investing is much more complicated than other chores like driving, which is why driverless cars are already more successful than even the best robo advisers.
Despite the enthusiasm of tech-savvy millennials—the generation of investors now in their 20s and 30s who are just as happy interacting with an app as with warm-blooded humans—robo advisers don’t take into account the limits of human cognition; they don’t make allowances for emotional reactions like fear and greed; and they can’t eliminate blind spots. Robo advisers don’t do emotion. When the stock market roils, investors freak out. They need comfort and encouragement. During last August’s stock-market rout, Vanguard Group told The Wall Street Journal it was “besieged” with calls from jittery investors and had to pull volunteers from across the company to handle the call volume.
But what if a robo adviser could identify the precise moment you freak out and encourage you not to sell by giving you historical context that calms your nerves? Better yet, what if this digital adviser could actively manage the risk of your portfolio so you don’t freak out at all?
Imagine if, like your car’s cruise control, you can set a level of risk that you’re comfortable with and your robo adviser will apply the brakes when you’re going downhill and step on the gas when you’re going uphill so as to maintain that level of risk. And if you do decide to temporarily take over by stepping on the brakes, the robo adviser will remind you from time to time that you need to step on the gas if you want to reach your destination in the time you’ve allotted. Instead of artificial intelligence, we should first conquer artificial emotion—by constructing algorithms that accurately capture human behavior, we can build countermeasures to protect us from ourselves.
Robo advisers have great potential but the technology is still immature; they’re the rotary phones to today’s iPhone.
Marvin Minsky, the recently deceased founding father of artificial intelligence, summarized the ultimate goal of his field by saying that he didn’t just want to build a computer that he could be proud of, he wanted to build a computer that could be proud of him. Wouldn’t it be grand if we built a robo adviser that could be proud of our portfolio?
See the post at WSJ “The Experts”
Andrew W. Lo is the Charles E. and Susan T. Harris Professor at MIT Sloan School of Management, director of the MIT Laboratory for Financial Engineering, principal investigator at MIT Computer Science and Artificial Intelligence Laboratory, and chief investment strategist at AlphaSimplex Group.