Join our community of smart investors
OPINION

When momentum works

When momentum works
March 30, 2022
When momentum works

Momentum investing works. We’ve known this about US equities for 20 years, but it is also true in international markets, commodities and foreign exchange. And it’s true in the UK, too. My model momentum portfolio has risen by an average of 9.7 per cent a year in the past 15 years, compared with a gain of only 1.7 per cent a year in the FTSE 350.

But, of course, no strategy works all the time. My momentum portfolio has suffered periods of underperformance, such as in 2012, 2019 and in recent months.

This poses the question: can we identify – or better still predict – the circumstances in which momentum works or doesn’t? The answer is: only partially.

To get this result, I took annual changes in my momentum portfolio relative to the FTSE 350 and looked for correlations between it and other economic variables (of course, my portfolio is only one of countless ways of measuring momentum's performance, but I suspect that what's true of it is true of other such measures).

The strongest such correlation is with the US dollar. Momentum tends to underperform when the dollar rises against sterling or the euro. This happened in 2012, 2014 and in the past few months. Conversely, a weak dollar such as in 2006, 2010-11, 2017 and 2020 sees momentum do well.

This is perhaps no accident. It’s consistent with the theory that momentum sometimes does badly in bad times. On such occasions, investors seek out familiar assets such as dollars. And so momentum does badly when the dollar does well and vice versa.

Consistent with this is another correlation. Momentum tends to do badly when gilts do well and well when gilts do badly. This too suggests than momentum is unusually risky: it does well in periods when investors want safer assets.

Sadly, however, these correlations are no help for anybody wanting to try to time the use of momentum investing, because we cannot reliably predict when gilts or the US dollar will do well or badly.

There is, however, one element of predictability here. It’s that momentum tends to do well after the yield on the All-Share index has been high. A high yield in 2008-09 and in the spring of 2020 led to momentum beating the market in the following 12 months. When equities generally are cheap, momentum tends to be especially cheap.

This also fits the theory that momentum is unusually risky. Equities are cheap when investors are avoiding risk. When they do so, they avoid momentum even more. And when they rediscover their appetite for risk momentum does even better than equities generally: the massive outperformance by momentum during the recovery from the Covid-induced slump in early 2020 fits this pattern.

With the yield on equities now below its long-term average, this is reason to be wary of momentum strategies.

What is striking, though, is just how few strong correlations there are between the pattern in my chart and other things. Interest rates, for example, have no ability to predict momentum’s ups and downs: this is one context in which the yield curve does not work. And while there are positive correlations between momentum and commodity prices, these are not especially high.

In fact, even if we combine those few indicators that are well-correlated with momentum, we don’t have great explanatory power. The US dollar, gilt returns, dividend yield and commodity prices all put together explain only around two-fifths of the variation in momentum’s annual relative returns. This leaves a lot unexplained.

This is a good thing. The fact that momentum’s performance is only lightly related to identifiable risk factors suggests that the strategy does well in most market conditions. My chart above shows that there have been only a few periods when momentum significantly underperformed the market, and these have been short-lived.

This, in turn, suggests that a good part of momentum’s great long-term performance is due not to risk factors but to investors making systematic errors of judgment. For example, they are too quick to book profits by selling winners, and they underreact to good news about shares they are neglecting. In this sense, momentum is an attractive strategy.

Does this mean you should blindly buy rising stocks to the neglect of everything else?

Yes, such a policy would have served you well in the past: only eight funds in Trustnet’s database of all-companies unit trusts have beaten my no-thought momentum portfolio over the past 10 years. You might not want to do this, however, simply because it feels so counterintuitive.

What you might consider instead, though, is factoring momentum into stockpicking. Think very carefully before selling good-performing stocks; don’t be put off a share because it has had a good run recently; and don’t hold on to losing stocks in the hope they’ll get even. Even if you are not a pure momentum investor, at least keep it on your side.