As investors experience the euphoria from one market high after another, the worrisome reality is that stocks could suffer a faster, more severe meltdown than during the 1987 crash. The proliferation of computer-driven algorithmic trading, also known as algo trading, has increased that risk, Barron's reports. An increasing amount of money is being entrusted to rules-based systems known as algorithms to pick stocks, place trades, mitigate risk, bet on volatility, and much more. Meanwhile, investors with long memories will recall that computer-driven program trading was a prime mover of the 1987 meltdown, and such automated strategies were a much smaller factor back then. (For more, see also: The Pros and Cons of Automated Trading Systems.)
Big and Fragile
Computer-driven quantitative trading strategies managed $933 billion of hedge fund assets as of the second quarter, according to data from Hedge Fund Research Inc. (HFR) cited by Barron's, up 87% from $499 billion in 2007. Additionally, rules-based, computer-driven index ETFs account for roughly another $3 trillion of investments, Barron's notes.
Meanwhile, there have been several events within recent memory involving sharp, unexpected drops in securities prices. The next big selloff could be made more severe by fast-acting computers that have an increasingly important role in the markets, Barron's warns. "The system is more fragile than people suspect," says Michael Shaoul, Ph.D., chairman and CEO of New York-based Marketfield Asset Management LLC, in comments to Barron's.
On Black Monday, Oct. 19, 1987, the Dow Jones Industrial Average (DJIA) fell by 508 points, or 22.6%. During a flash crash on May 6, 2010, the Dow dropped about 9% and the S&P 500 fell roughly 7% in just minutes of trading in mid-afternoon, before rebounding. A similar event occurred on Aug. 24, 2015, when the S&P 500 plummeted 5% in just minutes after the opening. The Dow dropped by 1,100 points in the first five minutes of trading that day, per CNBC, or by about 6.7%. (For more, see also: The Two Biggest Flash Crashes of 2015.)
In 1987, program trading set off a "poisonous feedback loop," as Barron's puts it, with computer-driven sell orders pushing down prices, which triggered yet more selling by these programs. An August 2007 selloff in quantitative funds that drove the S&P 500 down by 3.3%, known as the "Quant Quake," as well as the August 2015 edition of the flash crash, both appear to have been caused by similar feedback loops in computerized trading, Barron's observes.
Very smart people can design seriously flawed trading algorithms, or rules-driven quantitative funds. Long-Term Capital Management LP (LTCM) was a hedge fund built on quantitative strategies and it boasted two Nobel Laureates among its partners. The collapse of its high-risk, highly leveraged trading strategy in 1998 nearly brought down the broader market, until the Federal Reserve engineered a bailout, Barron's notes.
While the 1987 crash had program trading as a key cause, the vast majority of trades back then were executed through a slow process, glacial by today's standards, that often required multiple telephone calls and interactions between humans. Today, with the increased computerization of the markets, including the advent of high-frequency trading (HFT), trades often are processed within milliseconds. With incredibly rapid feedback loops among the algorithms, the selling pressure can build into a tidal wave within moments, wiping out fortunes in the process. Fasten your seatbelts.