Well that largely depends on who you ask. But typically it refers to a rally in stock prices in the month of December, or sometimes specifically the trading days between Christmas and the second trading day of January. The reasons given for why it occurs include traders closing loss making trades for tax purposes (and then reinvesting) or just good all Christmas euphoria, but in truth nobody knows.
With the lack of real evidence out there I've analysed the last 21 years of movements in the stock market to see is there is any actual evidence of a Santa Rally.
Does the Santa Rally exist?
The Santa rally is often more talked about in the US. If you look at the stock market returns for December for the last 21 years versus the rest of the year, it certainly looks promising. The table shows the price movement in the S&P 500 for January to the end of November for each of the last 21 years, versus the price movement in December. Click to enlarge
- The first thing to notice is that in 15 out of the last 21 years there has been a Santa rally
- Secondly, what has happened in the previous 11 months has little bearing on whether a Santa rally will occur or not.
What about in the UK stock market & can you predict a Santa Rally?
I thought I'd take the research even further and look at whether the Santa Rally phenomena exists in the UK. Also can you predict when a rally occurs? Does what happens in the period before December make a difference at all?
The table below summarises the results for the FTSE 100 over the last 21 years:
To sum up:
- The stock market has risen in 19 of the last 21 years in December!
- With an average return in December being of 1.91%
- Year to date performance or the preceding 6 month market movements give no indication as to whether a santa rally will occur or not.
- Neither do market movements in the months just before December make a difference. The only years when a Santa Rally didn't occur were 2002 and 2014, yet there are no tell-tale signs as to why there wasn't a rally.
Yet 19 out of 21 is statistically very significant. In theory in any given month you can either make money or lose money. So the binary nature of the outcome would lead you to expect the distribution to be nearer 11 good vs 10 bad years (or 10 good vs 11 bad) rather than 19-2. Of course that is slightly simplistic but just look at the columns for September, October and November - the spread of good and bad years is much more in line with expectations.
So there is no specific indicator in the preceding months that you will or won't get a Santa Rally in December, but the odds are that one will occur.
The parallels with 2011
In my recent commentaries I've drawn parallels between what occurred in the Autumn of 2011 and Autumn 2015. Indeed it was how I had an inkling that a rally in October this year was a distinct possibility, despite October typically being the worst month for stock markets. So what occurred in the final months of 2011? The chart below shows how both the US stock market (in red) and the FTSE 100 performed. After a sell-off in late November the market headed higher into the year end.
What about the years without a Santa rally?
While 2002 disappointed and the market slowly ground down into the year end, 2014 was slightly different. The December 2014 figure of -1.36% actually masks a 6% rally in the final 2 weeks of 2014. So you could argue there was a Santa rally after all!
Can you predict the scale of a rally?
In theory a rally is simply when you make some money. So making 0.1% would be a rally as there is no definition of what constitutes a surging market. But can you get an indication of the size of any potential rally? Well the answer is apparently so. The table below is a replica of the table above but coloured coded. Each month has been graded from green to red, with a the strongest positive stock market moves in dark green while the biggest falls are in dark red. So you can pick out unusually strong or weak months.
What becomes apparent is that the strongest Santa rallies (and indeed the only fall) are preceded by strong market swings, both positive and negative. Essentially providing a springboard to a stronger December. A fairly flat Autumn usually leads to a weaker rally. So perhaps if the last weeks of November can swing the -0.49% 'to-date' figure for November further negative it will set up a strong December rally. But like all axioms nothing is certain.
Summary
- The Santa rally does exist and based on recent history is statistically likely to occur in 2015 - although far from guaranteed.
- Yet the exact timing of any rally is the difficult to pin down but usually occurs at some point after the last two weeks of November
- A market fall in the next few weeks will increase the chances of Santa rally this year.
