How Algo Trading Is Worsening Stock Market Routs

Computerized, self-driving vehicles are being hailed as an important advance in highway safety. In the stock market, by contrast, computerized trading algorithms are being accused of reckless high-speed investing that is exacerbating recent price plunges and bursts of volatility.

"Rules-based strategies will take emotion out of it so when selling happens, more selling can occur and when buying happens, more buying can occur," is what Todd Rosenbluth, director of ETF and mutual fund research at CFRA, told Barron's. 

At the open on Friday, the S&P 500 Index (SPX) was up 0.81% from Thursday's close. As of 10:15 AM New York time, this gain had increased marginally, to 0.85%.

As alluded to by Rosenbluth, algorithmic trading, also called program trading, can create self-reinforcing trends that unfold at lightning speed, far beyond the ability of human investors and traders to keep up and undertake evasive action. In fact, some programs are designed precisely to follow trends, and the recent correction in stock prices was intensified as these algorithms suddenly shifted from buying to selling.

"It's just not normal, this action [was] like a rubber-band broke," according to Keith Lerner, chief market strategist at SunTrust Private Wealth Management, in remarks to Barron's. Momentum investing, in which investors chased the hottest stocks, was a factor in the market's rapid rise. Now the opposite effect may be in play, with selling pressures accelerating.

"The truth is that the market is just as irrational and divorced from fundamentals on the way up as on the way down. It is the nature of the markets more so today than ever, as a result of the computerized high-frequency trading strategies of the Wall Street wise guys. What we've watched this week is herd behavior on steroids," is how Washington Post business and economics columnist Steven Pearlstein puts it.

Indeed, the problem is heightened by the fact that many of the investment firms that create trading programs and algorithms follow similar, if not identical, decision rules. The key episode in the 1987 stock market crash, the Black Monday drop of 22.6% in the Dow Jones Industrial Average (DJIA), was the result of a "poisonous feedback loop" among these programs, as an earlier Barron's article described it.

In the ensuing years, trading algorithms have become increasingly more pervasive, controlling an increasing percentage of total transactions, and thus adding to the dangers. (For more, see also: Could Algo Trading Cause a Bigger Crash Than 1987?)

A recent study finds that nearly $8.8 trillion of financial assets worldwide were controlled by trading algorithms as of 2016, and that this figure is projected to grow at average annual rate of 8.7%, reaching $18.2 trillion by 2025, according to Business Wire.

Meanwhile, data from JPMorgan Chase & Co. found that quantitative and passive investing strategies account for about 60% of all equity assets, a figure that roughly doubled in ten years, per Bloomberg. By contrast, the same data indicated that only about 10% of equity trading volume now comes from human individual investors and discretionary investment managers. Passive strategies include index funds, whether structured as traditional mutual funds or as ETFs.

Defenders of automated trading systems cite several key points in their favor: consistency, discipline, elimination of emotion, and analytical rigor. They assert that such programs bring more logic into the trading and investing process. Also, defenders of high speed trading see economic logic in market prices adjusting quickly to changing fundamentals, or changing perceptions of fundamentals.

On another front, in response to the experience of 1987, major trading venues such as the New York Stock Exchange (NYSE) have implemented so-called circuit breakers or trading curbs that temporarily halt trading in the midst of a steep selloff. The idea is to stem a panic, and let market participants step back and take a breath.  (For more, see also: The Pros And Cons Of Automated Trading Systems.)