BOT SERIOUSLY

Fear of algorithm trading is really just the fear of the unknown

Will the machines replace us all?
Will the machines replace us all?
Image: AP Photo/Henny Ray Abrams
By
We may earn a commission from links on this page.

It’s all the machines’ fault.

That’s the conclusion of traders and hedge fund managers interviewed by the Financial Times in a feature (paywall) in which many attributed market volatility to the increased use of computerized trading models. They’re “wreaking havoc on markets and rendering obsolete old-fashioned analysis and common sense,” Robin Wigglesworth reports.

But there’s another take, as Bloomberg’s Matt Levine lays out in his Money Stuff newsletter: Some members of the financial old-guard simply “personally find algorithms confusing.” Levine, a former investment banker himself, has little sympathy. “Hedge fund managers don’t—or aren’t supposed to—get tenure; there is nothing unnatural about someone making a lot of money in the olden days and then being out-competed by the next generation.”

That confusion may be one reason why  the “humbled one-time masters of the universe,” in the words of  Wigglesworth, are so roiled by algorithm trading—they don’t understand it, and that’s very scary.

The funny thing is, we’ve always been quite bad at knowing how to attribute market volatility, which long predates algorithm trading. The start of a much-shared satirical Wall Street Journal article from the 1990s sums it up: “The market rallied early this morning for reasons nobody understands and nobody predicted. CNBC analysts confidently asserted it had something to do with the Senegalese money supply, but others pointed to revised monthly figures showing a poor tuna haul off the Peruvian coast.” We can guess at what’s made things move, but these guesses are often no more than that—and the opinions of two experts can diverge wildly (and still both be wrong.) Analysts are often similarly bad at predicting which stocks will fly, and which will flop—in one 35-year study, researchers found that the top 10% of stocks recommended by analysts had less growth than the bottom 10% they advised against.

That hasn’t stopped us from trying, however, or from holding onto a sense that since markets are moved by people, we therefore must be able to understand them, because we are them. It’s much easier to tell a story that attributes the rise or fall of numbers to investors’ fear or greed than to acknowledge that, for many, algorithm trading, and the volatility it may produce, is far more inscrutable. We haven’t surrendered to the robots—not yet, anyway—but you can understand, and maybe even forgive, a little bit of Hal 9000-inspired dread.

Levine, for his part, thinks there’s something to be said for present-day market moves being a mixture of the two: “My general view is that market moves are emergent properties of many humans acting on their individual beliefs and desires, and that computers are tools that allow the markets to do what they were going to do anyway, but faster.” Those gut feelings about investor fears or confidence may be no more wrong than they were before, but we’re simply seeing their effects on a grander and speedier scale.

A separate question, of course, is whether all this market volatility can or should be ascribed to algorithm training as it has been. Despite recent relative placidity, markets have always been volatile. The last three months’ ups-and-downs may have been choppy, but they’re by no means historic. It may also be true that we’re simply talking more about the most minute market moves. “It’s in our face more,” Nancy Tengler of Tengler Wealth Management told Bloomberg (paywall). “We have too much focus on the day-to-day or minute-by-minute or second-by-second movements.”

Blame it on the robots if you must—it’s an easy out for those struggling to understand what’s happened already, and what’s going to happen next. The trouble is, it’s not very useful for everyone else. This may be an occasion where panicky analysts may have to face the fear, and acknowledge they didn’t know all that much before, either.