It's clear that share prices are seasonal. What's not so clear is why. Mark Kamstra at York University in Toronto believes it is because investors' moods are seasonal in two related ways.
One is that in the autumn we become anxious and nervous whereas we become more hopeful in the spring. This means that shares fall in the autumn, to levels from which they subsequently rise. And they get another boost in March and April as we become more optimistic. These mood swings are embodied in our ancient pagan festivals. Halloween, or Samhain, is a time of fear whereas May Day is a celebration of hope and fertility.
Secondly, we are prone to the winter blues; as the nights draw in we get a bit depressed. And depression reduces our willingness to make sacrifices for future benefits. This reduces demand for shares, which offer such benefits.
These two theories have a testable implication. If they're right, we'd expect shares in the southern hemisphere to have a different pattern to those in the north, because their autumn is our spring, and vice versa.
My table summarises the evidence here; it's based on price changes in local currency since 1970, which is when MSCI's data begin.
| Equity returns since 1970 | |||
| Nov-April | May-Oct | All periods | |
| Australia | 6.5 | 1.1 | 3.8 |
| UK | 10.2 | -0.8 | 4.6 |
| UK-Australia | 3.7 | -1.9 | 0.8 |
| Local currency terms. Source: MSCI | |||
This shows that Australian shares also do better between November and April than between May and October, which seems inconsistent with Professor Kamstra's theory.
However, this is simply because almost all national stock markets are correlated with each other. If we look at differences in returns, we see the pattern we expect. Between Halloween and 30 April, the UK outperforms Australia by 3.7 percentage points. But between May Day and Halloween, it underperforms by 1.9 percentage points.
Granted, this difference is flattered by outliers; the UK massively outperformed Australia in our winters of 1975 and 1977. But if we look merely at the number of times the UK has outperformed Australia, there's also a seasonal pattern. The UK has beaten Australia in 24 of 44 November-April periods, but only in 18 of 45 May-October ones.
The obvious implication here is that now might be a good time to buy - especially as the sell-off in the last few weeks (which has been greater in the UK than Australia) is consistent with this pattern.
But there's another implication. It's that investing is not a wholly rational process. Like our pagan ancestors, we are prone to seasonal mood swings - and, perhaps, with less rational justification.
