Is September really the worst month for investors?

This article was originally published in The Telegraph.  

SEPTEMBER and October are Marmite months. For some they evoke Keats’s ‘mists and mellow fruitfulness’; for others they are simply a reminder that another summer has been and gone. Investors need to hold both thoughts at the same time. That’s because while most of the stock market’s returns are secured during the winter months the onset of autumn has historically been a dangerous time. The market’s big ‘events’ have tended to occur in these two months. Volatility is higher at this time of year. 

I was not surprised, therefore, that the first full week of September saw the late August rally fizzle out. Returning from their holidays, investors have found plenty to fret about as they get back into their routines. The timing of the peak in interest rates remains uncertain as the oil price creeps up again; China casts a lengthening shadow, including a $200bn two-day hit to the value of Apple; earnings are still falling, and valuations leave little room for error.  

The conventional wisdom is that the autumn months are a time to be cautious. But what’s the actual evidence? Fortunately, it’s easy to put the seasonality myths to the test. The historic data on the world’s most important index, the S&P 500, are freely available on the web (hat tip to the excellent moneychimp.com monthly returns tool). So, the many stock market adages about the year’s good, bad and ugly months can be measured. 

Using monthly data all the way back to 1950, it seems that investor anxiety about September may be justified. The month has fewer positive returns than any other, and more negatives. It is the only month of the year in which there have been fewer rising months than falling ones since 1950. The average return over the 73 years is also the worst of any month at -0.8%. 

While there are three other months that have recorded a negative average return, February, June and August are only just in the red. The tendency for the market to rise over time means that the September slide really is an outlier. There are seven months of the year where the average rise is the same or greater than the average September decline. The past three years have all experienced significant market falls in September of between 4% and 9%.  

Why should September be such a dud for investors? There are plenty of theories, most of them unconvincing. One says that investors come back from their holidays in August with a renewed determination to sort out their portfolios. These have often drifted during the quieter trading days during the summer. Out go the underperformers in a kind of new school year clear out. Another more plausible scenario has big mutual funds engaging in ‘window dressing’ ahead of their year-end, typically at the end of September. 

What about October? This is perhaps the month that investors are most scared of, given its association with the crashes of 1929 and 1987.  But the reality is different. On average, shares have risen by 0.8% in the tenth month, kicking off an eight-month run in which only February’s 0.14% average fall blots the copybook. The other seven months from October to May have on average seen the market rise since 1950. October also has one of the best hit rates over that period, rising on 45 occasions and falling on just 28. 

The myth about October being a bad month seems to relate more to its volatility than its average performance. Some of the most terrifying days in stock market history have occurred in October, including the 1987 crash when the Dow fell by 23% in one day. Three of the other biggest falls happened in October 1929. There have been more daily swings of more than 1%, up and down, in October than in any other month. It’s a rollercoaster. 

Again, it is hard to point the finger at a good reason for this. It has been suggested that the Presidential elections in November play a part, with investors positioning themselves in the month ahead of the vote. But these occur only every fourth year, so that seems to be something of a stretch. The 1987 crash was clearly exacerbated by the Great Storm and the fact that no-one in London could get to their desks, but that was a one-off.  

So, are September and October good times to invest? The averages would suggest it’s a favourable time to put money to work in a phased way. The old adage about Selling in May and Going Away has persisted because of the measurable tendency of markets to perform worse over the summer on average. Buying in at a yearly low point makes sense. 

Incidentally, the same seasonal effect can be identified in many markets around the world. September is the worst month of the year in the US, Germany, Britain and Japan, and the tendency for markets to outperform during the winter is consistent across all of these markets. 

The cost of implementing a strategy based on seasonal variations, however robust the averages are over time, means all this is more noteworthy from a theoretical than a practical standpoint. It’s also fair to say that no-one invests over a 70-year time horizon and there will be long periods when the markets could move against you.  

As with other seasonal adages like the January Effect, the underperformance of September and the volatility of October may be more interesting than actionable.

Important information - investors should note that the views expressed may no longer be current and may have already been acted upon. Overseas investments will be affected by movements in currency exchange rates. Reference to specific securities should not be construed as a recommendation to buy or sell these securities and is included for the purposes of illustration only. This information is not a personal recommendation for any particular investment. If you are unsure about the suitability of an investment you should speak to one of Fidelity’s advisers or an authorised financial adviser of your choice.