“Never short a dull market” is an old saying on Wall Street. That’s good advice for the next couple of months, when stock-trading volume typically is lighter than average. Stock traders fret that these summer doldrums will sink share prices, but history suggests there’s nothing to worry about.
That doesn’t mean the US stock market will perform well this summer. It just means that light trading volume is not a reason, in and of itself, to worry.
To start, let’s review the volume data. The chart below plots each month’s average trading volume when expressed as a ratio to its trailing 12-month moving average. On average since the early 1970s, this volume ratio begins to decline in the last half of June, drop further in July, and August has the lowest ratio of any month.
(Note that this ratio is the proper trading-volume metric to focus on, since volume itself has skyrocketed over the past 50 years. Simply averaging each month’s raw volume total would have been skewed by this secular trend.)
To test whether a dull market should be sold short, I calculated the correlation between a given month’s volume ratio and the S&P 500’s SPX,
total return in that same month. Over the entire period since 1970 there was no statistically significant pattern.
It is interesting to note, however, that in some decades the correlation was positive while in others it was negative. For the decades of the 1970s and 1980s, for example, there was a strongly positive contemporaneous correlation between trading volume and the S&P 500’s return. In those decades, it might have paid to sell a dull market short.
Since the early 1990s, in contrast, there has been a weakly negative correlation between trading volume and the S&P 500. In these decades the stock market tended to do better when volume was lower than when it was higher. In such an environment, of course, you most definitely would not want to sell a dull market short.
A topic for another discussion would be why this correlation has changed. Perhaps high-frequency trading has skewed the volume data? That’s an intriguing hypothesis, since computers don’t take summer holidays.
I next measured whether a given month’s trading volume had any value in predicting the S&P 500’s return over the subsequent quarter, six months, and 12 months. I came up empty, both overall and in each of the separate decades.
The point of this column is different: the overall exchange trading volume is not itself a helpful market-timing indicator, and light trading volume this summer is not something to worry about.
Mark Hulbert is a regular contributor to MarketWatch. His Hulbert Ratings tracks investment newsletters that pay a flat fee to be audited. He can be reached at firstname.lastname@example.org