Skip to main content

This Week's Big Dow Drop Was Ordinary, Not Extraordinary

Mark Hulbert says the Dow’s big declines may have been shocking, but it shouldn’t be surprising. It was overdue.

Monday’s big plunge, which took over a thousand points off the Dow, undeniably came as a big shock. And Tuesday's was almost as painful, shearing off another 879 points, or 3.15%, to close at 27,081.36 -- the index's worst two-day drop yet.

But these declines shouldn't have been so surprising.

On the contrary, drops as big as Monday’s, which measured 3.56% for the Dow Jones Industrial Average, are an inevitable feature of the stock market. Prior to Monday, in fact, the stock market had gone three times longer than average without a decline as big.

Why, then, are you complaining?

I base my comments on a groundbreaking 2003 study titled Institutional Investors and Stock Market Volatility, by Xavier Gabaix, now a finance professor at Harvard University, and three scientists at Boston University’s Center for Polymer Studies: H. Eugene Stanley; Parameswaran Gopikrishnan, and Vasiliki Plerou. The researchers came up with a formula that predicts the frequency of big daily market drops over long periods of time.

According to their formula, a drop of 3.56% or more will occur every 7.3 months, on average. As fate would have it, of course, prior to Monday it actually had been more than two years (since Feb. 8, 2018) since the Dow suffered a daily drop of at least this magnitude.

This suggests we were being unrealistic if we thought years like 2019, in which there was no drop as big as Monday’s, are the norm. On the contrary, last year led us to become spoiled rotten.

The researchers’ formula doesn’t mean that a 3.56% or bigger daily plunge will occur like clockwork every 7.3 months, I hasten to add. Instead, their formula predicts what will be the average frequency of such drops over long periods. So it’s possible that you will see no more days this year with drops as big as Monday’s, or a handful of them.

But over long periods of time, the formula turns out to be quite accurate. Consider the accompanying chart, which shows what the researchers’ formula predicts will be the number of daily drops of different magnitudes over any given 100-year period, versus the actual number experienced by the DJIA over the past 100 years.

Hulbert Chart 022520

While there is not an exact correspondence between the predicted frequency and the actual number, it is remarkably close. And the researchers report that there was an even closer correspondence when they measured big daily drops’ frequencies over longer periods than just the last 100 years and in many different foreign markets as well.

Why do big daily drops occur with such predictable frequencies? Gabaix says that it’s for two reasons: Stock markets are inevitably dominated by a relatively small number of large institutional investors, and there will be times when a number of them will more or less simultaneously want to get out of stocks.

From this perspective, it is almost irrelevant the specific reasons why the institutions all of a sudden get spooked in a specific situation.

In Monday’s case, of course, it was the coronavirus that was to blame. But the more salient point is that, given human nature and institutional dynamics, they will inevitably get spooked again at some point in the future -- and again and again.

We need to get used to that fact.