Seasonal investing has worked like a dream in the last 12 months. If you had followed the rule 'sell on May Day, buy on Halloween' you would have avoided a 2.5 per cent loss on the All-Share index last summer but bought just in time for a 26 per cent gain after inflation.
This continues a pattern. Since 1966 the All-Share index has delivered a total return after inflation of 7.9 per cent on average from Halloween to May Day, but has lost an average of 0.6 per cent from May Day to Halloween – even including dividends.
Nor is this pattern confined to the UK. Ben Jacobsen and Cherry Zhang at Massey University studied the entire history of stock markets in 114 countries and found that 87 of these had better returns from Halloween to May Day than from May Day to Halloween, with the difference being especially big in Western Europe. “The Halloween effect is a strong market anomaly,” they conclude.
But is it strong enough to trade on? Here, we run into the paradox of evidence. The stronger is the evidence that a strategy is profitable, the likelier it is that investors have wised up to it and therefore bid it away. And the evidence for seasonal evidence is indeed huge and should by now be well known: Jacobsen and Zhang’s evidence was first published in 2012. So perhaps investors have wised up.
And, indeed, seasonal investing has failed horribly quite recently. Had you bought on Halloween 2019 you’d have lost 17.6 per cent by the following May Day.
A single failure, however, does not invalidate a strategy unless it wipes you out completely simply because all strategies lose money sometimes. And, in fact, there’s little sign that investors have wised up to the seasonal anomaly and so eradicated it. The All-Share index has fallen in the last three May Day-Halloween periods. And the fact that the market did so well this April – historically the best month for shares – also suggests seasonality is still with us.
And there’s a good reason why investors haven’t wised up. Recall the best reason we have for why there are Halloween and May Day effects. It’s that, as York University’s Mark Kamstra and colleagues have shown, investors suffer from a form of winter blues. As the nights draw in we get anxious and depressed and so (at the margin) avoid equities which means that by Halloween they are cheap. By the same token, lighter evenings in the spring cheer us up, which causes us to push share prices up too far. In this way, stock market seasonality is a manifestation of ancient instincts: Halloween (and before that Samhain) marks a time of fear while May Day (or Beltane) is a time of hope and celebration. For investors to bid away these effects would require them to buy when they are nervous and sell when they are optimistic. That’s tough.
From this perspective it is not wholly correct to say that the success of buying last Halloween was due to the good luck of the discovery of a Covid vaccine being announced a few days later. The point is that depressed sentiment in the autumn causes investors to price in more bad prospects than good ones with the result that shares are more susceptible to getting nice surprises than nasty ones – although of course occasionally, as last spring, the nasty surprise is so nasty as to overwhelm this tendency.
We’ve reason to suspect therefore that seasonal investing still works. Not every time of course – in finance nothing is 100 per cent successful – but often enough to make money over the long term.
Many of you, I know, hate this claim.
And you’ve a good reason to. The fact that UK equities have underperformed cash from May Day to Halloween (since 1694 according to Zhang and Jacobsen’s data) is horribly counter-intuitive. It overturns our basic common sense that risky assets should on average outperform safe ones. It’s entirely understandable to resile from that.
But I fear there’s also a less reasonable motive for denying seasonal effects. We flatter ourselves that we make decisions rationally on the basis of relevant evidence. The idea that these decisions are instead swayed by the time of year contradicts this self-image. And a natural reaction to things that undermine our sense of ourselves is simple denial.
The time of year, however, is not the only reason to be wary of equities right now. The rise in Aim stocks suggests sentiment is too high; global share prices are high relative to the money stock; volatility is low; and the dividend yield on the All-Share index is below its long-term average. All these are bearish signs. Against all this, the best reason for sticking with equities is simply that they still have momentum on their side. (Of course there are bullish stories we can tell – but stories should be discounted very severely.)
Even if you attack zero weight to the evidence of seasonality – which you shouldn’t – there are therefore many reasons for caution.
