Is momentum investing in doubt?

Have investors wised up? This is the question posed by the recent underperformance of my momentum portfolio, which comprises the 20 shares that had risen most in the 12 months to June.

Granted, the underperformance is small – the result of losses on Halfords (HFD), Royal Mail (RMG) and Ferrexpo (FXPO) offsetting profits on Entain (ENT) and Reach (RCH) – and could easily be just statistical noise. And it’s exacerbated by the fact that the FTSE 350’s performance has been flattered by the oil majors: BP (BP.) and Royal Dutch Shell (RDSB) alone account for almost all the index’s small gain in the third quarter.

Nevertheless, the question must be asked. Common sense tells us that investors shouldn’t leave money on the table, so they should cotton on to profitable strategies and thereby bid away their profits. This happened to small stocks in the 1980s: after investors realized these had outperformed for decades, they piled into them with the result that they hugely underperformed in the 1990s. And John Cotter and Niall McGeever at University College Dublin have shown that a similar fate has befallen other once-profitable strategies. Might the same be under way with momentum?

Of course, there is massive evidence of its success. The fact that my portfolio has hugely outperformed in the past 10 years is only one data point here. It corroborates evidence from the US, international markets and non-equity assets, all of which shows that momentum investing works – a fact that is robust to different definitions of momentum.

Losing momentum?

But the sheer weight of this evidence is all the more reason to fear that momentum has stopped working. It’s understandable that investors might ignore a single isolated and unreplicated finding. But they cannot be so stupid as to ignore the large body of evidence that momentum works.

There is a reason to suspect that momentum will continue working and that this recent slip is just the sort of blip that can afflict the best investment strategies. It’s that momentum’s outperformance is due not simply to investor irrationality – to them underreacting to good news – but to the fact that they are riskier than other stocks and must outperform in order to compensate for this extra risk.

One of these risks is precisely the one I’m worried about. At any point in time you could reasonably think “momentum has done well but investors must have wised up to this by now” and therefore have reasonably avoided the strategy. It’s success, therefore, could be partly due to investors fearing that it had stopped working.

Another risk is that momentum can do very badly when the market falls a lot. When the market slumped last March, for example, momentum did even worse. It also underperformed during the worst of the financial crisis of 2008. Investors might reasonably avoid momentum stocks because of the danger they’ll do really badly in bad times.

And then there’s growth rate risk. Momentum stocks are often growth stocks or ones that have recovered from big difficulties. Both entail big risks: a lot can go wrong with expansion plans and recoveries can prove to be short-lived. Again, momentum should offer a risk premium on this account.

We should therefore expect momentum to do well simply because high risks should carry high rewards. Convinced?

You shouldn’t be. In the past 10 years, my momentum portfolio has beaten the FTSE 350 by 8.2 percentage points a year. That’s just too big a risk premium to be true.

What’s more, other risky strategies haven’t earned a risk premium at all. Value and high-beta stocks have underperformed the market in the past 10 years. What’s so special about momentum that it delivers a huge risk premium while other risky strategies don’t?

For me, these are genuine puzzles to which I don’t have a clear answer – and I don’t know anyone who does. My hunch is that I’d like stronger evidence from a less exceptional period than recovery from a pandemic before declaring momentum dead.

You might think there’s other evidence for the death of momentum. My negative momentum portfolio, which had done horribly for years, has soared in the past 12 months. For years, loser stocks used to keep losing. Recently, they have stopped doing so.

All of the profits from buying past losers, however, came between October last year and March, when the discovery of a vaccine raised hopes of an end to the pandemic, which triggered massive rises in the stocks hardest-hit by the pandemic, such as the airlines, Carnival (CCL) and Cineworld (CINE). Except for these few weeks, buying losers has lost you money.

What’s more, in the third quarter my negative momentum portfolio comprised not just big fallers but shares that just hadn’t risen much during the recovery from the pandemic because they hadn’t fallen much in the first place – such as Tesco (TSCO) and Reckitt Benckiser (RKT). In this sense, negative momentum stocks have been unusually defensive recently and have performed accordingly. That tells us little about the usual behaviour of past losers. 

On the defensive

As you’d expect, defensive stocks have lagged behind the market during the recovery of the past 12 months. As with momentum stocks, however, their longer-term performance has been impressive, comfortably beating the market over the past three, five and 10 years. And, as with momentum stocks, my portfolio merely corroborates other evidence from around the world showing that defensives do well.

There’s a plausible risk-based explanation for this. Fund managers fear that if they hold defensives they will underperform a rising market and so underperform their rivals, costing them business, a bonus and perhaps even their job. From their point of view, defensives are risky. And this risk pays a risk premium.

The mirror image of defensives are high-beta stocks. Whereas defensives have underperformed in the past 12 months, high-beta outperformed – thanks to the same autumn snapback that benefited past losers. They did, however, do badly in the third quarter thanks in part to losses on EasyJet (EZJ) and Asos (ASC).

But the longer-term picture is that high-beta shares underperform. Again, this fits with international evidence. Economists at AQR Capital Management have shown that high-beta assets (not just shares) underperform low-beta ones around the world. This, they say, is because of borrowing constraints. In theory, bullish investors should borrow to buy shares generally. In practice, though, they cannot do so to the extent they’d like. They therefore express their bullishness by buying high-beta stocks to get what is in effect a geared position on the market. This causes such shares to be overpriced on average and so underperform unless they enjoy the sort of exceptionally good news they got last autumn.

The performance of value stocks is stranger. They underperformed in the third quarter as a gain on Vectura was offset by losses on CMC Markets (CMCX), Polymetal (POLY) and Rio Tinto (RIO).  Nor did they enjoy the vaccine boost that high-beta and negative momentum stocks got. This is because my value portfolio is an odd mix. It has contained a few cyclicals such as Persimmon (PSN) and Tui (TUI), but also stocks that investors believe to have gone ex-growth such as tobacco stocks: the oil majors might be in both categories. The upshot is that there’s no obvious pattern to their performance. I’m not sure that value stocks should be seen as an asset class at all, but rather a mix of two different types of stock.

You might wonder what the point of all this is. Think of it as a way of mapping the market. Research has shown that some strategies have systematically beaten the market (momentum and defensives) and others have systematically underperformed. What we need to know is whether this is still the case, and under what conditions it is and is not true. There’s more to investing than using one’s individual judgment to pick individual stocks, and this exercise – I hope – demonstrates that.