Join our community of smart investors

Momentum's failure

Momentum investing has done terribly in the past three months. There's a message in this fact for all stockpickers
Momentum's failure

Momentum investing has stopped working. Having massively outperformed the FTSE 350 in the 10 years to August – rising by more than 250 per cent against the index’s less than 50 per cent gain – my no-thought momentum portfolio has lost more than 20 per cent since then, more than twice as much as the index has fallen.

In part, this is because almost all stockpicking strategies have had a hard time. Because a few big stocks such as AstraZeneca and Royal Dutch Shell have beaten the market in the past three months most shares have underperformed market indices, which are weighted by market capitalisation. Because of this, most stockpickers have underperformed the market. According to Trustnet, only 27 of 253 unit trusts in the all companies sector have outperformed L&G’s index fund since August.

This, though, doesn’t explain why momentum has done so badly as to underperform every one of those 253 funds.

Nor can the blame be placed upon one stock or sector doing badly, Momentum’s failure has been evenly spread across many stocks. Premier Oil, Burford Capital, Softcat and Oxford Biomedica have all done badly recently.

So, what can explain momentum’s terrible performance?

One possibility is simple bad luck. The standard deviation of returns on momentum relative to the FTSE 350 since 2004 is such that the underperformance we’ve seen in the past three months has a 1 per cent chance of occurring. That seems very unlikely. But there have been 56 non-overlapping three-month periods since I began the portfolio, so there’s an almost 50:50 chance of such a loss occurring once in this time.

There is, however, a much more troubling possibility. Perhaps investors wised up to the merits of momentum investing in the summer. If so, they would have bid up the prices of momentum shares too much, with the result that they became overpriced and so subsequently fell.

There’s a precedent for this. In the 1980s, for example, investors realised that small stocks had beaten the market for decades. So they piled into them, with the result that they underperformed the market for the following 10 years. This fits a pattern. John Cotter and Niall McGeever at University College Dublin have shown that several stock market anomalies have weakened or disappeared in recent years.

If this is the case, it should trouble all stockpickers. It means that any hitherto good strategy can suddenly fail because investors have wised up. It also means that we cannot trust fund managers’ track record. Good past performance might be because the manager has followed a strategy to which investors have now cottoned on to, causing his holdings to become overpriced.

Although this explanation seems plausible, it suffers from a problem: why should investors only have wised up now? We’ve known for years that momentum works. The first solid evidence for it was published in 2002 by Narasimhan Jegadeesh and Sheridan Titman and lots of corroboration has followed. Why should it have taken so long for investors to cotton on?

The answer might be that they’ve been scared to do so because momentum investing is in fact risky. It’s just that this risk has materialised recently.

Victoria Dobrynskaya at Moscow’s National Research University has pointed out a curious feature of momentum strategies – that they tend to underperform the market a lot when the market itself does badly. They have high downside betas. The historic success of momentum investing, she believes, is at least partly a compensation for taking on the risk of losing a lot when the market falls.

The fact that momentum has done badly recently when the market itself has fallen is entirely consistent with this. And there have been many other times when momentum has underperformed a falling market, such as in the first three months of this year.

 

 

If this is the case, then momentum should outperform again when the market stabilises.

Or will it? One difficulty I have with this story is that there is one obvious period when momentum did not underperform a bear market: the crash of 2008-09. If momentum did not do especially badly then – when pretty much all other risks materialised – can we be confident that it really is especially risky? And if momentum holds up in truly terrible times, why should it pay out such a huge risk premium in normal times?

What we have here, then, is a genuine puzzle. We just don’t know whether the failure of momentum investing is permanent or temporary.

In truth, though, this is not just a story about momentum investing. The same problem, at some time or other, afflicts all stockpicking strategies. All suffer periods of underperformance: value investing, and Warren Buffett, did badly during the tech bubble, for example. Only with hindsight can we tell for sure whether the underperformance is temporary or not. But of course investors need foresight rather than hindsight. It is devilishly hard in real time to distinguish between the investor who has the courage to stick with a proven strategy and the stubborn fool who carries on in the face of contrary evidence.