This column has been written by John Rekenthaler, vice president of research for Morningstar.
In the early autumn of 1987, the U.S. stock market slumped. The Dow Jones Industrial Average had doubled over the previous two years, reaching its all-time high of 2,722 in August. The euphoria quickly subsided. By mid-October, profit-taking had knocked 20% off the Dow's peak value, and the financial media's tone had turned ominous. As with January of this year, October 1987 was one of those times when the market felt as if it was teetering, on the edge.
On that date, it lost its balance--down 22.6% on Black Monday, Oct. 19.
Black Monday demonstrated that the stock market was vulnerable to unintended consequences. Although worries about market structure had always existed, the concern was muted until Black Monday. That day's huge drop, seemingly unsupported by economic news (it's not as if a giant investment bank were shut down over the previous weekend, as in the 2008 crisis), suggested that the foundation was broken. The losses, it was said, "cascaded" upon each other.
Portfolio insurance was fingered as the culprit. The newly popular investment technique sold into market downturns, to protect portfolios against further declines. These sales were mostly done with futures contracts. The stock market would slide; the portfolio insurers would sell index futures; the price of the stock-market futures would fall because of the sell orders; investors would treat the declining futures pricing as an early warning signal and then sell stocks; the stock market would slide. Rinse and repeat.
The explanation sounded credible. There certainly was some truth to it. Portfolio insurance had exploded in popularity, growing from almost nothing two years prior to becoming a major chunk of futures trades. What's more, orders tended to come in a bunch. We can track that activity, as well as the effect of those trades on the futures markets. What we don't know is to what extent the change in futures prices affected the stock market.
In its post-crash study, the U.S. Securities and Exchange Commission hedged its conclusion on portfolio insurance, stating that it was not the primary cause of the woes but was a "significant factor" in making a bad day worse.
Since Black Monday, the SEC has been quite aware of unintended consequences, thanks much. As have other market observers. The financial media has delighted in telling ghost stories about derivatives.
The new danger?
During the past several years, the scrutiny has switched from futures/options to indexing.
The scrutiny started with indexed exchange-traded funds, in their case not because they were indexed, but rather because of ETFs' trading mechanisms. Periodically, those mechanisms have misfired, thereby temporarily cratering the price of the victim ETF and potentially spooking stock-market investors. In this instance, the link between the offending security and the market behaviour is even more tenuous than that between index futures and stocks in 1987, as 2010's "flash crash," initially blamed on ETFs' problems (and oh they had problems), was later found to have had many sources.
The attention has now moved to the practice of indexing itself. Many active managers have advanced the theory that growth of market-capitalization-weighted index investing has caused a "bubble" in stocks that belong to popular indexes and led to neglect among those that do not. When this bubble pops, the argument goes, disappointed index investors will leave those index funds, thereby forcing the funds' managers to sell the popular stocks, which will further depress their prices, which will lead to more redemptions, and so forth. Like Black Monday, but in slow motion.
Most recently, the claim has moved to strategic-beta investing. Research Affiliates' Rob Arnott, the most visible proponent of that approach, has co-authored an article entitled, "How Can 'Smart Beta' Go Horribly Wrong?" Taking the index-bubble argument one step further, Arnott wonders if the boom in strategic-beta ETFs that seek to exploit such factors as value strategies, momentum investing and low-volatility stocks, among others, has gone too far. Has the academic discovery of these "anomalies" sowed their own destruction, as new monies have boosted the price of securities to the point where they must eventually kersplat?
Arnott writes: "We foresee the reasonable probability of a smart-beta crash as a consequence of the soaring popularity of factor-tilt strategies." Irony, thy name is Arnott! I am less convinced than Arnott is by the data that he presents, but that is a topic for another time. Suffice it to say that strategic-beta indexing has been added to the list of items that may cause unintended stock-market consequences.
Wrapping up
1) It's hard to separate the effect of unintended consequences--because of a flaw in market structure and/or the bubble-making behaviour of stock investors--from the carnage that normally occurs during bear markets.
2) However, there is at least some validity to the concern--and possibly a great deal.
3) There is more reason to worry now than in the past. Besides exchange-traded derivatives, there is now a huge marketplace in custom derivatives, plus the existence of ETFs, plus the surge in indexing, plus the invention of high-frequency trading. To go further than that would be to tell a campfire tale. We don't know if any of those items will prove destabilizing. But we also don't know if they will not.
None of this is intended as a warning against U.S. stocks. That is where most of my own assets have been invested, are invested, and almost surely will be invested for the foreseeable future. This note is instead meant as a word to the wise. Many U.S. stock-market characteristics have changed in recent years. Don't be surprised if a few glitches occur because of that.
