Even the most cutting-edge quant funds rely on old-school human intuition

Image: Unsplash/Roman Mager
We may earn a commission from links on this page.

Financial markets are just as excited about artificial intelligence, machine learning, and big data as the technology industry. So-called “quants” have benefitted from this enthusiasm, with money from investors steadily flowing into their algorithm-driven funds.

Quant investing spans a vast range of strategies, but one of the most talked about methods is using artificial intelligence to sift through vast amounts of market data to uncover signals that humans can’t see. The idea is to use advanced mathematics and computing power, rather than traditional research and intuition, to gain an edge in the market.

Take PhaseCapital in Boston. Its chief investment officer has a PhD from Oxford with a focus on numerical optimization, artificial intelligence, and neural networks. Michael DePalma, its CEO, previously helped run quantitative investing strategies at AllianceBerstein.

And yet, for all its sophistication, the hedge fund says it still trusts an old-school indicator that investors have tracked for decades: the yield curve. When the spread between short- and long-term interest rates fell recently, the firm slashed its market exposure, DePalma told the Financial Times (paywall.)

DePalma later told Quartz that yield-curve analysis—such as the observation that short-term interest rates rising relative to long-term rates may signal the economy is sputtering—remains useful for helping assess risk. The slide-rule era tool is based on widely available bond price information, and economists have been documenting its predictive powers since the 1960s.

While there’s no guarantee that massive data sets crunched by algorithms will surface any sensible investing strategies, the yield curve is a tried and true measure. Yet the buzz around quants is so intense that investment managers may feel pressured to adopt some sort of algorithmic strategy, or else risk raising less money, DePalma said. Computer-driven hedge funds have doubled their assets under management, to $932 billion, following eight straight years of inflows since 2009, according to data-tracking firm HFR.

Some quant funds have justified investors’ excitement—Quantitative Investment Management’s $1.2 billion equity fund has gained 55% in just the first five months of this year, according to Bloomberg, while Renaissance Technologies’ equity fund rallied 14% over the same period. Overall industry returns for quant funds have been much more modest, at 1.4% so far this year. (The S&P 500 is up 8% over the same period.)

As a quant firm, PhaseCapital is obviously an advocate of advanced computing. But its reliance on the humble yield curve shows that letting algorithms loose on ever-larger data sets isn’t the only way to make money. As DePalma sees it, quant investing is as useful and potentially fruitful as other other investing strategies, some of which may have fallen out of favor recently. The hype around quant strategies is itself a risk—when any investment technique gets too popular, trades are crowded and investors stand to lose a lot of money (pdf). In that case, a dash of old-fashioned human intuition, relying on old-school tools, may be the wisest course of action.