Join our community of smart investors

Momentum wins again

Our no-thought momentum portfolio has continued to beat the market – which is puzzling
October 13, 2017

Yet again, momentum investing has done well. My no-thought momentum portfolio (which comprised the 20 best-performing stocks with a market capitalisation of over £500m in the 12 months to June 2017) rose by 8.3 per cent in the third quarter, thanks to big gains in stocks as otherwise diverse as Fevertree, Ferrexpo and IQE.

This extends great long-run performance. In the past five years, momentum’s 150 per cent return has beaten all but three funds in Trustnet’s database of all companies’ funds. And in the past 10 years it has beaten all but eight such funds.

This, of course, is not an idiosyncratic finding of mine. It corroborates research from the US and around the world that momentum investing pays off well – and not just in equities.

Benchmark portfolio performance    
 In Q3Last 12MLast 3YLast 5YLast 10Y
Momentum8.332.582.6150.1196.3
Negative momentum-1.717.6-11.2-21.1n/a
Value-0.62.535.472.0-34.2
High beta2.717.30.520.6-16.9
Low risk-1.34.930.546.765.2
Mega caps0.37.48.226.110.6
FTSE 3501.17.613.733.921.5
Price performance only: excludes dividends and dealing costs 

Why? Part of the answer is that investors tend to underreact to news. If they see a company announce some good news they stick too much to their prejudice that the stock is a mediocre one and so don’t buy it strongly enough, with the result that it’s price doesn’t rise sufficiently immediately after the news. Another part of the story might be that investors have limited attention. Lots of stocks simply aren’t on their radar, and it is only when their attention has been grabbed by good performance that they are attracted to the stock.

In themselves, explanations such as these are insufficient. They run into the question: why don’t well-informed investors buy momentum stocks quickly, thereby bidding up their prices to levels from which subsequent returns are only average? Why hasn’t the momentum effect been bid away?

It can’t be because investors don’t know it: research on momentum is now abundant. A more likely explanation is that momentum is risky. The risk here isn’t particularly market risk: on average momentum’s beta has been close to one. Instead, momentum carries an element of cyclical risk; momentum tends to do badly in downturns, except if sterling falls. This risk paid off in the third quarter when the yield curve steepened slightly, indicating that the market believes the chance of recession has receded slightly.

But there’s another risk – regret risk. If you pick stocks conventionally and underperform, you can console yourself with the thought that you followed best practice and due diligence and the clichés of backside-covering management-speak. If you simply buy past winners and fail, you have no such consolation. Instead, you’ll regret your unconventional behaviour. In this sense, momentum’s losses would hurt you more – and quite likely cost you your job if you’re a professional fund manager. The fear of behaving like a maverick means that momentum is risky, and so requires a risk premium.

Frankly, though, I’m not sure how convincing this is. In the past 10 years, momentum has outperformed the FTSE 350 by 9.5 percentage points a year. This is a massive risk premium. Can this really be only a fair reward for cyclical and regret risk? I’m honestly not sure. The returns from momentum are so huge as to puzzle me.

The converse of momentum doing well is that negative momentum – the worst performers in the previous 12 months – does badly. We saw this in the third quarter, too, with the portfolio falling 1.7 per cent. Except for a rally earlier this year – perhaps in part because some stocks that fell immediately after the EU referendum bounced back – this too continues a long-term pattern.

It’s easier to explain this than it is the high returns from positive momentum. The difficulties and risks involved in short-selling mean that the smart money cannot profit much from the overpricing of stocks that have experienced bad news. They do not, therefore, force down such share prices sufficiently quickly.

Value investing also had a lacklustre quarter, losing us 0.6 per cent. Here, though, we must distinguish three types of value stock. Some shares have high dividends to compensate investors for their cyclical risk. These, however, did well for our value portfolio in the quarter: Vedanta, Galliford Try and Persimmon posted gains. Others have high yields because their price has slumped. These hit our portfolio hard. Carillion and Pearson both fell: they were also members of our negative momentum portfolio.

Yet others have good yields because they are dull stocks, which investors perceive to lack growth opportunities. These too slipped slightly in the quarter: Vodafone, SSE and Centrica.

This last fact is linked to another – that my defensive portfolio also fell slightly. This has wide significance. Most UK equity income funds also fell in the quarter – most famously, Woodford’s equity income fund. They did so because they have a bias towards large defensive stocks, many of which had a bad quarter.

That poor quarter, however, merely slightly reverses what has been good long-run performance. My low-risk portfolio has beaten the market over the past five and 10 years. As with momentum, this is no idiosyncrasy but rather is corroboration of a trend that has been found around the world over the long term. This, though, poses the question: has the longstanding defensive anomaly finally been bid away as investors have at last wised up to it?

I can’t be sure of course, but I suspect not, for three reasons. One is the good performance of momentum. This raises the question: why should investors have wised up to defensives' good long-run performance but not to momentum’s?

Secondly, we have a powerful risk-based reason to believe that defensives should outperform on average. It’s that they carry benchmark risk. If the market does very well, defensives will probably underperform, which would cause a fund manager owning them to lose his bonus and perhaps his job. For this reason, many managers are underweight in defensives, causing them to be underpriced on average.

To see the third reason, look at the performance of my high-beta portfolio. It has been the mirror image of defensives’ performance: it’s had a good quarter and 12 months after years of underperformance.

Now, there’s a reason for that underperformance, pointed out by economists at AQR Capital Management. Many investors, they say, cannot borrow to buy stocks if they are bullish of the general market. Faced with this constraint, they instead take a geared position on the market by buying high-beta stocks – those they expect to outperform a rising market. This leaves such stocks overpriced on average with the result that they subsequently do badly. The converse of this, of course, is that low-beta stocks are underpriced on average and so do well.

And here’s the thing. Twelve months ago, UK investors were not especially bullish; they were fretting about the effects of Brexit, among other things. They weren’t therefore as exposed to high-beta stocks as usual, nor as underweight in defensives. The result was that high-beta stocks have since done well while defensives have underperformed.

This might well have been just temporary. If so, we should see the normal pattern resume soon, with defensives doing well and high-beta less so. If I’m right – and it’s only an if – then we should see the poor performance of some equity income funds also reversed.

Finally, a reminder of the purpose of this exercise. The point is not to show how clever I am in spotting ways of beating the market: these portfolios are merely simple tests of the ideas established by researchers years ago and don’t contain any original ideas. Nor even is it to recommend particular strategies: momentum in particular carries risks you might not want to take. Instead, these portfolios are intended to be ways of testing hypotheses in real time: is the CAPM right? (No.) Does value investing pay? (Yes, much of the time but not in recessions.) Does momentum work? (Yes, but we’re not entirely sure why.) Personally, I’m more interested in testing hypotheses than I am in making money or boosting my ego.