It's tempting to attribute their performance to investors' mistakes, or cognitive biases. Risky stocks seem more glamorous, which causes investors to neglect dull but reliable shares. The hunt for the 'next Microsoft' or 'next Google' leads them to scramble for risky 'growth' stocks such as Facebook and so overlook 'get rich slow' strategies such as buying defensives. And overconfident investors overestimate their ability to spot good stocks and so prefer speculative investments to more reliable ones.
One thing, however, suggests these explanations aren't the whole story. It's that defensives' good performance is so widespread and long-standing. Nardin Baker of Guggenheim Investments and Robert Haugen of Haugen Financial Systems have found that the least volatile 10 per cent of shares subsequently outperformed the most volatile between 1990 and 2011 in every one of the 33 major stock markets they studied. And researchers have found that defensives have done better than they should in the US since at least the 1930s.
These facts pose a problem for the cognitive bias theory. Investors should at least some time and somewhere have wised up to defensives' good performance and so bid up their prices to levels from which subsequent performance has been poor. The fact that this hasn't happened (yet!) suggests that the defensive premium is rooted in something more permanent than investors' irrationality.
But what? A new paper by David Blitz and Pim van Vliet of Robeco Asset Management and Eric Falkenstein at Walleye Trading suggests some possibilities.
One, they say, is leverage constraints. The capital asset pricing model (CAPM) predicts that everyone should hold the same bundle of equities, with the only difference between investors being that cautious ones hold lots of cash and few shares, while adventurous ones should borrow to hold more shares. Many investors, though, cannot borrow freely. If they want higher risk and return, therefore, they will buy higher-beta shares. This will cause these to be overpriced and so deliver sub-par returns, the converse of which is that defensives will be underpriced and so offer good returns.
This explanation is consistent with the fact that defensives have done well in the last three years while the overall market has risen. Tougher financial conditions in 2010 and a 'reach for return' at a time of low expected returns increased investors' appetite for high-beta shares in 2009-10, leaving lower-beta defensives especially underpriced then.
A second reason why defensives are so often underpriced is that investors don't care only about volatility and return, as the CAPM assumes. Some also care about their relative returns. For them, defensives are risky because they would be expected - normally - to underperform a rising market.
This provides an opportunity for retail investors. It's professional investors who have reason to worry about underperforming a rising market, as this could get them sacked. Those of us who don't need to fret about this can therefore buy underpriced stocks without worrying about the risk they carry.
There's a third reason for defensives to outperform, say Messrs Blitz, van Vliet and Falkenstein. It's that some fund managers actually want upside risk. Funds that enjoy big returns attract inflows from naïve retail investors and earn their managers bonuses and flattering publicity. Riskier stocks are more likely to offer this than defensive ones: "We lost 10 per cent while our rivals lost 20" isn't as good a marketing campaign as "We made 40 per cent while the market rose 20". If you want top-decile performance, then, you'll often avoid defensives. So these will be underpriced.
Now, none of this guarantees that defensives will always outperform; even the best strategies will fail sometimes, even over quite lengthy periods, simply because shares are volatile. What it does mean, though, is that the tendency for defensives to do well is due not merely to corrigible irrationality, but to reasonable behaviour that results from the very structure of financial markets.
