Join our community of smart investors

In praise of defensives

Defensive stocks do better than they should – and we've good reason to expect them to continue to do well
June 2, 2020

UK equities have risen since the autumn – if you look only at defensive stocks.

My portfolio of low-risk stocks is now almost 10 per cent higher than it was in early September and is up by a similar amount over the past 12 months, even ignoring dividends. This is not a quirk of how I formed the portfolio: it comprises the 20 shares with a market capitalisation over £500m with the lowest betas over the last five years. Low-risk sectors such as tobacco, food retailing and utilities are also now higher than they were in the autumn.

This is not what conventional theory predicts. My defensive portfolio has had a beta of 0.9 in the last 12 months. That means it should have fallen only slightly less than the market. But in fact in the last 12 months it has risen almost 10 per cent while the FTSE 350 has fallen almost 15 per cent. Low-risk stocks have therefore delivered a positive alpha, at least if we control only for market risk.

Why? In retrospect, it seems that low-risk stocks were unduly cheap a year ago as a result of their underperformance in 2018-19.

But this is only part of the story. Defensive stocks have done well for years. In the last 10 years, my defensive portfolio has risen 69 per cent while the FTSE 350 has risen only 25 per cent.

This is part of a long-term global pattern. Back in 1972 Fischer Black, Myron Scholes and Michael Jensen showed that in the US low-beta stocks had done better than they should since 1931. More recently, separate research by Nardin Baker and Robert Haugen and by economists at Robeco Quantitative Investments has found that low-risk stocks outperform around the world. And economists at AQR Capital Management have shown that this is true of assets such as commodities and bonds as well as equities. They have also show that it is buying low-risk stocks that has been the key to Warren Buffett’s great success.

The good performance of low-risk assets is therefore a robust fact – one that is resilient to different periods, different countries, different markets and different definitions of low risk. It is perhaps the strongest deviation there is from efficient market theory: only the momentum effect comes close in terms of weight of evidence.

So, why do defensives do so well?

One theory, proposed by AQR economists, attributes it to borrowing constraints. In theory, bullish investors should borrow to buy shares generally. In practice, though, many cannot borrow either because banks won’t lend to them or because their mandates forbid it. They therefore express their bullishness by buying high-beta stocks – ones they think will be a geared play on a rising market. This causes high-beta shares to be overpriced and low-beta ones to be underpriced.

This theory has a testable prediction. It implies that low-beta stocks should do especially well when borrowing constraints are most binding. One of these circumstances is when prices have fallen a lot, because that is when collateral is low. Sure enough, low-beta stocks have indeed beaten the market since March (consistent with this theory, they also delivered a high alpha in 2009).

Recently, however, Josef Zechner at Vienna University has proposed another theory. He shows that investors are happy to pay high prices for shares with positive coskewness – that is, ones that do especially well when the market does exceptionally well. Because defensives do badly in such circumstances, investors ignore them, which causes them to be underpriced and so to deliver good average returns. In effect, investors get a premium for taking on negative coskewness.

Exactly why they have this preference is unclear. It could be for the same reason that they pay too much for small speculative stocks; they like the small chance of large, quick gains and need compensation for stocks that offer slower steadier profits instead.

This explanation, though, runs into a problem: why haven’t smarter investors by now wised up to this error and so piled into defensives, thereby eliminating their under-pricing.

It could be because, for fund managers, 'low-risk' investments are misnamed. They carry the danger of underperforming in bull markets – as indeed they did between 2015 and early 2019. Because fund managers are judged on relative performance, holding defensives exposes them to the risk of losing bonuses and even their jobs, and it means risking being left off financial advisers’ best buy lists and thus being ignored by investors.

If this explanation is right, it is doubly good news for private investors. If 'low-risk' shares do well because they are in fact risky for many investors, then we have a strong reason to suspect they’ll continue to do well on average as they should carry a risk premium. And it means that if we hold defensives, we’ll be compensated for taking on a risk that shouldn’t trouble us. We’ll get something for nothing. There is such a thing as a free lunch.

Of course, this doesn’t mean they’ll outperform month in, month out. Even the best strategies underperform sometimes – and defensives might well do so if the market rises further. What it does mean, though, is that if you must bet against the efficient markets hypothesis by being an active stockpicker then investing in defensive stocks is the best-evidenced means of doing so.