A Very Quick One Today

A series of commitments this morning and early afternoon have made it impossible for me to write a normal length piece today.  Instead, I’ll simply offer a couple of quick data points about seasonality – or the lack thereof – that I find interesting.

First, as has been widely reported, the S&P 500 (SPX) just finished its best September in 15 years.  Last month’s 3.53% advance was above even last September’s 2.02% rise and fell short only to 2010’s 8.76% jump in recent history.  We noted earlier this month that while September is statistically the worst month of the year, whether we go up or down in any given September is essentially a coin flip, even while considering the four year down streak that was broken last year.

There is indeed truth to the “September Scaries”, but data show that a decline in September is far from pre-ordained.  Yes, September is statistically the worst month of the year for the S&P 500 (SPX).  It is only one of four that show net declines over the past 25 years, but it is by far the worst.  (According to Bloomberg data, January has averaged -0.19% this century; February -0.46%, June -0.25%, and September 1.51%).  Yet over that period, the likelihood of a September decline is only slightly greater than that of a coin flip: 12 of the past 25 Septembers have seen positive returns for SPX. 

Looking ahead, we see that October is a month that has averaged solidly positive returns of 1.31% over the prior 25 years, but we have a fair dispersion of results as well.  The odds are a bit better for an up month, 10 of 25, but that is hardly a resounding consensus.  When we narrow the data set to Octobers that follow positive Septembers, we see that five of them resulted in lower SPX returns for that month.  If you’re looking for a definitive pattern, I’ve yet to find one.

One pattern that is intriguing is the recent preponderance of down days to start the trading month.  Of the nine months prior to this one, six have begun with SPX trading lower, for an average decline of 0.41%.  To be fair, SPX was sharply lower this morning when I collected the data – it’s now modestly positive – but a decline would fit in this pattern.  It’s probably spurious, though it could be a bit of a hangover from month- or quarter-end position markups.  I know some of you doubt that it occurs, but I assure you from my experience dealing with institutional investors that it does occur.  How else might we explain yesterday’s late elevator ride higher.

Source: Interactive Brokers