US stocks, as measured by the popular Dow Jones Industrial Average (“the Dow”), closed down 21 points (-0.12%) on February 28, in the process snapping the index’s 12-day winning streak which had been the longest for the Dow in more than 30 years back to January 1987. Perhaps even more noteworthy than the length of the win streak is the fact that on each of the 12 days of the streak, the index closed at an all-time high, tying the index’s record for most consecutive days closing at an all-time high, which was also achieved back in January 1987. In addition to the Dow, other popular market barometers such as the S&P 500 Index (US large cap stocks) and the Russell 2000 Index (US small cap stocks) also hovered near their all-time highs throughout much of the last month. In fact, on February 27, the last day of the Dow’s win streak, all three indexes closed at an all-time high.
In addition to the win streak and the achievement of all-time highs, many market watchers and pundits have noted how “quiet” the markets have been of late. By this they mean that the volatility of the market has been unusually low—that is, the magnitude of the market’s daily moves both up and down have been relatively benign.
One relatively simple and intuitive way of measuring the volatility of the market is to track the number of days during a specific time period in which the markets moved up or down by more than one percent in any given day. Exhibit 1 below plots the number of daily moves greater than one percent in either direction during the first two months of the year for this year and the prior three years as well.
As you can see in the chart, through the end of February neither the S&P 500 nor the Dow experienced a daily move of one percent or more in either direction. This comes in sharp contrast to the experience of last year where the S&P closed either one percent higher or lower 23 times, while the Dow did so on 24 occasions. Suffice it to say that the experience thus far this year makes 2014 and 2015 appear like white-knuckle roller-coaster rides.
A more statistically-sophisticated way of measuring the volatility of the market is to calculate what’s known as the standard deviation of returns. Exhibit 2 plots the standard deviation of daily returns for the S&P 500 Index on a monthly basis from January 2014 though February 2017. By this measure too, it’s apparent volatility has been noticeably absent of late.
With markets at all-time highs and volatility unusually low, it’s reasonable for investors to question whether either of these facts about the current market environment contain any reliable or actionable information about the future near-term prospects for the stock market. One way to answer this type of question is to “review the tape.” That is, to examine the historical record of the market’s performance.
First, we’ll address whether or not the market achieving a new all-time high contains useful information for investors going forward. For evidence, we can look at the performance of the S&P 500 Index for the better part of the past century and compare the returns on the index in the year following months where the market closed at a new monthly all-time high to those months where the market did not close at a new all-time monthly high. Exhibit 3 shows that from the beginning of 1926 through the end of 2016, the proportion of annual returns that have been positive after a new monthly high is roughly similar to the proportion of annual returns that have been positive after any index level. Looking at this data, it is clear that new index highs have historically not been useful predictors of the future direction of the market.
Next, let’s examine whether recent stock market volatility levels contain any statistically reliable information about future stock returns. A simple, visual way to assess whether recent stock market volatility and future returns are related is to graphically plot historical observations of one month’s realized volatility (as measured by the standard deviation of monthly returns) and the next month’s return which we’ve done in Exhibit 4.
To use non-technical verbiage, you can “see” that the relationship between the two appears “flat.” That is, recently volatility—neither high nor low—has not statistically indicated whether future returns will be either high or low. This relationship can also be confirmed through a more technical regression analysis which further confirms that there has been no statistically reliable relationship between recent volatility and future returns.
Another relatively intuitive way to see if recent market volatility and future returns are related is to look at average returns across different market environments in terms of recent volatility. In Exhibit 5, we take monthly returns for the US stock market (represented here by the Fama/French US Total Market Index, a measure of the performance of the broad US market) and break them up into three categories based on the previous month’s standard deviation of daily returns. Consistent with the graphical representation displayed above, this analysis further confirms a lack of a statistical relationship between recent volatility and future returns.
As the table demonstrates, average returns in months when the previous month experienced low volatility (25th percentile or below) were slightly lower than when then previous month experienced high volatility (75th percentile or above) yet were slightly higher than the average of all months included in the analysis. Because stock returns have been noisy, however, these differences in average returns have not been reliably different from zero (i.e., in stat-speak, the differences were not statistically significant). In other words, at a glance there does not seem to be an economically meaningful difference in average US stock market returns based on the volatility of the market in the prior month.
What can we take away from all of this? As it directly relates to the two questions posed at the beginning of this commentary, we’ve learned that: (a) new all-time highs in equity markets have historically not been useful predictors of the future direction of the market and that (b) neither has the recent volatility of stock returns. Put simply, we believe that all investors would be well served to remember that while positive realized returns are never guaranteed, stock investments have positive expected future returns regardless of current index levels or recent volatility.
There is considerable academic evidence that an investment strategy attempting to forecast short-term price movements is unlikely to be successful. Forecasting short-term stock market performance based on whether the market is at or near an all-time high or based on recent volatility is no different.
A more effective strategy in our perspective is commit to and diligently adhere to a strategy of letting markets work for you. Over long periods of time, stocks have historically been terrific wealth builders, having generated long-term returns well in excess of those of long-term Treasury bonds, short-term Treasury-bills and inflation, as displayed in Exhibit 6.
Ultimately, we believe that developing an asset allocation to match up with your unique risk tolerance and investment objectives, and staying disciplined and rebalancing in all market environments, remains an effective way to pursue your long-term investment goals.
- Kathmere Capital Management Investment Committee