One of the best and simplest risk management devices is to follow the 10-month (or 200-day) moving average rule: you buy if prices are above their 10-month average and sell if they are below it. Mebane Faber at Cambria Investment Management has shown that this rule works well for the S&P 500, and I’ve shown that it also does well for the All-Share index, emerging markets and mining stocks.
And it’s not just equities where it works. It also does well for commodity prices.
To test this, I applied the rule to the S&P GSCI spot index: I chose this as it has a long history, but I suspect what follows will apply more or less to other indices which are tracked by commodity exchange traded funds (ETFs).
The spot index itself has more than doubled since 1989, implying a rise of 2.7 per cent per year. The standard deviation of these annual changes has been 22.9 percentage points, giving a Sharpe ratio of 0.12. If you had rigidly followed the 10-month rule at the end of each month, putting your money into cash when the rule told you to sell, you would have made 5 per cent per year before dealing costs. And because you’d have spent just over two-fifths of your time in a safe asset the volatility of these returns would have been lower, at 15.5 percentage points. That implies a Sharpe ratio of 0.32 – more than twice as good as the buy-and-hold strategy delivered.
This does not mean the rule works consistently. It fails whenever prices have a V-shaped pattern. If prices fall sharply enough to go below their 10-month average and then snap back, the 10-month rule gets us out of commodities when prices are low and causes us to miss the subsequent bounce.
Such failures are, however, outweighed by an occasional big benefit. The rule sometimes gets us out of long bear markets, which saves us fortunes. It told us to sell commodities in July 2014, for example, and not to return until April 2016. That saved us a 43 per cent loss.
This doesn’t mean the rule is perfect even in bear markets. In the 2008-09 crisis, for example, commodities fell more than 60 per cent from peak to trough, but our rule saved us only 24 per cent. This is because it missed the first leg of the bear market – when prices were falling but still above their 10-month average – thus giving us a 30 per cent loss. And it did not get us back into commodities until they had risen almost 40 per cent from their low point.
Now, I appreciate that many of you might not want to follow this rule. For one thing, frequent switches between cash and commodities incur dealing costs. For another, you need rigid discipline; if you hold on just once in the hope that a fall is just a dip when in fact it turns into a serious bear market you’ll suffer big losses. And, thirdly, the positive correlation between shares and commodities means you have exposure to commodities through equities and so might not need any more.
Nevertheless, there are two points here.
One is that commodities have just risen above their 10-month average. This suggests they are more likely than not to rise further.
The other is that this is yet more evidence for momentum: the 10-month rule works when price falls lead to further falls and fails when they lead to further rises. This corroborates research by economists at AQR Capital Management who have found that momentum effects exist not just in equities but also in commodities (and currencies for that matter).
Momentum, then, is ubiquitous. But why?
One possibility is that investors are Bayesian conservatives. They under-react to good news because they cleave too strongly to their prior belief that the asset isn’t very attractive. When they do this, assets that have enjoyed such news – be they equities, commodities or currencies – don’t immediately rise as much as they should. Instead, they drift up gradually latter thus giving us momentum.
Another possibility is that investors’ attention spans are necessarily limited: we cannot stay on top of everything. An asset that rises in price for any reason thus attracts attention and this in turn causes more buying which drives prices up even further.
The AQR researchers, though, have another theory. Momentum profits, they believe, might be a reward for taking on liquidity risk. When liquidity dries up, investors try to raise cash by closing their positions. Naturally, this is a bigger danger for the most popular positions. And these, by definition, will be those in assets that have risen most in price lately – that is, ones that have momentum.
This matters for equity investors. Momentum strategies did badly last year: my no-thought momentum portfolio (comprising the 20 biggest-rising large stocks over the previous 12 months) lost 22.4 per cent for example, although it has since stabilised. This could have been just bad luck – even the best strategies fail sometimes – or it could have been due to investors finally wising up to the power of momentum investing and, in doing so, bidding away its profits. My personal dial has oscillated between these two possibilities. If, however, momentum is so ubiquitous that it exists in most assets over long periods then it is likely to be a robust fact about equities too. And this makes me a little more confident that momentum investing might start to work again in stock markets.