Momentum > Seasonality: September Stock Returns

By in
Momentum > Seasonality: September Stock Returns

Today we want to take a deeper dive into what history can tell us about potential S&P 500 price performance over the balance of 2021 given the index’s already strong YTD return. We’ve recently covered the basics of seasonal equity returns with long run data back to 1928, and our conclusion was that September can be pretty dicey but Q4 usually makes up for that and then some.

To sharpen up the analysis, we will focus on returns from 1980 – present. This is a long enough timeframe to give us a reasonable dataset and also reflects what we’d call the “modern” US equity market. Many of you know that the original explanation behind stock market seasonality was the annual flow of capital between American farmers and the banking/financial system. This, of course, is no longer a relevant issue so the last 4 decades of return data should be free from its effects.

Four points to today’s analysis:

#1: Let’s start with the S&P’s average performance for September and its surrounding months:

  • Average price return for January through August (1980 – 2020): +6.3 pct
  • September (1980 – 2020): -0.6 pct
  • October through December (1980 – 2020): +4.3 pct

And here is the volatility around those readings:

  • The standard deviation of returns for January through August: 11.4 points
  • September: 4.4 points
  • October through December: 8.8 points

#2: With that in mind, let’s now see whether a positive or negative S&P performance in the 8 months leading up to September have any bearing on that month’s return, and what that means for the balance of the year:

The S&P has had a positive YTD return from January through August during 29 years since 1980, or 71 pct of the time; the average return was +11.8 pct. During those years:

  • The average return in September: +0.21 pct
  • October through December: +3.1 pct (excluding 1987, +4.0 pct)
  • Average total annual return: +18.3 pct

The S&P has been negative from January through August during 12 years since 1980, or 29 pct of the time; the average return was -7.0 pct. During those years:

  • The average return in September: -2.6 pct
  • October through December: +7.3 pct
  • Average total annual return: +0.6 pct

Takeaway: a positive performance for the S&P from January through August does, on average, correlate with positive (albeit modest) returns in September. It also leads to a decent Q4 (+3 pct) to top off the year. Conversely, negative returns in the months up until September are followed, on average, by a bad September. While the S&P typically rebounds in Q4, total returns during those years are near flat.

#3: How does this data look when we have a really strong January – August? Two points:

2021 has delivered a historically remarkable performance thus far. The standard deviation of returns for the S&P from January through August is 11.4 points since 1980. With an average S&P performance of 6.3 pct over the same period, that means a return of +17.7 pct is one standard deviation from the average or statistically significant. Since this year’s 20.4 pct return YTD save September is well over that threshold, we used that one standard deviation return over the average as our cutoff to mark positive outlier years. Here are those years and how they performed:

  • Six years (ex 2021) returned greater than 17.7 pct from January through August since 1980, or 15 pct of the time: 1986 (+19.7 pct), 1987 (+36.2 pct), 1989 (+26.6 pct), 1991 (+19.8 pct), 1995 (+22.3 pct) and 1997 (+21.4 pct).

    The average return here: +24.3 pct.
  • During the same years, the average return in September was -0.7 pct and it was also down 0.3 pct in Q4.

    Excluding 1987, it was down 0.4 pct in September and up 4.3 pct in Q4.
  • The average total return during these 6 years was +26.1 pct. Excluding 1987, they were up 30.1 pct.

Takeaway: when the S&P has a very strong stretch from January to August it does typically see a down September but still advances in Q4. The one exception was 1987, where the S&P declined by -23.2 pct during the last three months of the year. Although that year was an outlier, we understand why already cautious investors could use it as one more case study to justify de-risking their portfolios.

#4: In contrast to our last point, here is how the S&P performs in September and Q4 after exceptionally poor returns during the prior months. One standard deviation to the downside of the average S&P return from January through August is -5.1 pct since 1980. Here are the years that returned worse than that threshold and how they performed:

  • Six years returned worse than 5.1 pct from January through August since 1980, or 15 pct of the time: 1981 (-9.6 pct), 1990 (-8.7 pct), 2001 (-14.1 pct), 2002 (-20.2 pct), 2008 (-12.6 pct) and 2010 (-5.9 pct).

    The average return here: -11.9 pct
  • During the same years, the average return in September was -5.0 pct, but it was up 3.2 pct in Q4. Excluding the outlier of 2008, it was up 8.4 pct in Q4.
  • The average total return during these 6 years was -10.6 pct.

Takeaway: especially bad momentum for the S&P heading into September leads to a particularly negative month. Although the index typically has a solid rebound in Q4, it’s not enough to salvage the year with negative double digit total annual returns on average.

Bottom line: every September is different, and the key variable is whether YTD returns have been strong or weak. Since 2021 has been exceptionally good, history says this September should be fairly stable. Yes, that may feel like a mini-bear market after the steady gains we’ve seen YTD and yes, you can still take some risk off the table here without much fear of missing out on the next move higher. In the end, however, the work we’ve shown you here today is a good reminder that momentum is a durable investment factor. Stronger, in fact, than seasonality…