“Quant” Hedge Funds

IN February 1996, a chess-playing computer named “Deep Blue” defeated Grandmaster Garry Kasparov in what was the first chess victory of a computer against a reigning world champion. This was not only a significant moment in the history of competitive chess, it signaled that computers, with the right algorithm, could surpass humans.
In financial markets, hedge funds began to imagine a day where machines, not people, would be responsible for stock trading. If a computer could defeat the world’s best chess player with the right algorithm, couldn’t there soon be a day when machines would also compete with humans as stock traders?
The answer is a resounding yes. Today, quantitative hedge funds are the source of 27% of all U.S. stock trades, up from 14% in 2013. These “quant” hedge funds line up to hire the best mathematicians and computer scientists among a growing number of graduates pursuing careers in quant investing.
They compete to create the best algorithms, which use statistical models to select the most attractive trades. Using high-powered computers, stock traders are able to analyze trillions of gigabytes of data.
The long term prospects of this approach to stock trading are seemingly great. These algorithm-based models can perform tasks and analyze data at a rate impossible for humans. It, therefore, makes sense that quants now have 29% of stock-trading volume — about as much as individual investors — and have more than doubled their share of hedge funds assets to $932 billion of investments.
But does the influx in quant hedge funds indicate that, just as Deep Blue defeated Grandmaster Kasparov, computers have now surpassed humans’ abilities at stock trading?
The answer is a conditional no. Since 2009, quant hedge funds have been performing poorly, losing money in four of the last five years and making many wonder whether or not the growth of quant investing has permanently harmed returns. Throughout recent years, traders were far better off using more traditional approaches than quant-fund investing.
And yet, the aftermath of the financial crisis was a particularly turbulent period for markets that often abandoned the historical patterns quant algorithms rely on. According to quant-hedge funds, it is unfortunate but not unsurprising that quant models would fail in these recent years, when less predictable politicians and central bankers steered the economy.
Even if quants have become better at predicting the direction of the market when it is acting freely, they have not developed a strong ability to predict the behavior of policymakers. Humans, it seems, are still the best at predicting the movements of an economy directly guided by other humans.
If this is true, then quant investing strategies, and the algorithms they rely on, may come to succeed once again as markets descend into more stable and familiar terrain. The next Grandmaster may yet be a stock trading computer.
