The Statistical Odds of a Seasonal Equity Correction
As most readers know, equity markets are arriving at a seasonally unfavourable period. Traditionally equity markets “take a breather” in the third quarter before blasting off again for the ubiquitous year-end rally. Additionally, the August to October period has notoriously coincided with major market downturns (1981, 1987, 1990, 2000, 2007) and sharp corrections (1998, 2011, 2015).
Some may argue that, in this new era of central bank hegemony, past market behaviour no longer serves as a guide, and that traditional market indicators are less relevant. We, however, believe that the central bank stranglehold on the markets is temporary, as market forces can only be suppressed for so long. While quantitative easing and zero interest rate policy (ZIRP) have clearly suppressed the importance of traditional valuation metrics and other market indicators, we believe that it is a mistake to turn a blind eye to the past. Those who forget history are doomed to repeat its mistakes.
With this caveat, we review in this Commentary the seasonal performance of the S&P 500 since World War II. The first chart below shows the monthly breakdown of the 70 monthly data points since 1948. The bars show our calculations of the mean monthly returns. The calculations confirm empirical evidence of a strong year-end rally (November and December), as well as strong equity performance in the Spring (March and April). Conversely, August and September have traditionally been the weakest period, with negative mean monthly returns. The numbers just above the date x-axis show the percent of occurrences that the month returned a positive return over the past 70 years. The big winner, unsurprisingly, is December, during which 3 out of every 4 Decembers have seen positive S&P 500 returns. On the other end of the scale, less than 1 September out of 2 has managed to post positive equity returns.
We took are seasonal analysis of S&P 500 returns a step further by introducing a precondition. We asked, given the index performance over the first half of the year (January to June), how has the performance of more the volatile months of July through October varied?
We divide the 70 observed first half S&P 500 performances into three groups:
- Positive first half (S&P 500 up over +2%)
- Negative first half (S&P 500 down over -2%)
- Flat first half (S&P 500 finishing June between -2% and +2%)
Positive First Halves
Since 1948, the S&P 500 has posted 41 positive first halves of the year. The subsequent monthly break-down is shown in the next chart. The mean monthly returns are higher for July, August, and September compared to all 70 observations, while October has been a tad bit weaker. The percent of winning months is higher across the board when the S&P 500 starts the year strong. In sum, a strong start to the year helps attenuate the damage in the turbulent third quarter.
Negative First Halves
Since 1948, the S&P 500 has posted 19 negative first halves of the year. The next chart of July to October returns starkly contrasts with the above performances following positive starts to the year. The percent of positive July, August, and September returns plummets along with the mean monthly returns for these months. At the same time, we see that October has often marked a bottom, as returns for this month bounce back. When markets correct in August / September, October has seen an unusually high percent of positive monthly returns (68.4%, that is, over 2 Octobers out of 3 post positive returns).
Unchanged First Halves
The final case is when the S&P 500 does essentially nothing in the first half of the year. We counted 9 cases of January to June S&P 500 performances between +/- 2%. The final chart of July to October returns shows that boring first halves of the year getting exciting – for bears – in the August to October period. However July has tended to post strong gains a high percentage of the time after a flat first half.
With the S&P 500 up +8.24% in the first half of 2017, the most likely scenario is the Positive First Half (second chart in our Commentary). Statistically, we can still expect some slowdown in the rally in August / September, but portfolios should reinvest aggressively in S&P 500 companies by October. For those interested in the seasonal trends on a sector-by-sector basis, the charts can be consulted under the tab “Seasonality Charts” here.
Like to give your opinion? Go to Comments