Looking at our readers' portfolios, it seems that many investors have 'risk barbell' equity holdings: big weightings in both defensive stocks and in speculative ones. This can be a very intelligent approach because for a holder of defensive stocks, speculative stocks can actually be a way of reducing overall risk.
You might be surprised by this. After all, speculative stocks are tremendously volatile. But this alone does not ensure that they add to the total risk of one's wealth. Consider the other risks that speculative stocks carry:
■ Sentiment risk. If investors become more optimistic, they are - pretty much by definition - more likely to buy speculative Aim stocks than dull defensives. They might even sell the latter in order to buy the former. This implies there can be a negative correlation between defensives and speculative stocks.
■ Behavioural risk. History shows that, whereas defensives have generally done better than they should, glamorous lottery-type shares have often done worse. But there is always the danger that anomalies such as these will disappear as investors wise up. This suggests there might be a negative correlation between defensives and speculatives - because if investors have cottoned on, defensives will do relatively poorly and riskier stocks relatively well in future - or a zero one, because the two anomalies are separate and so the chances of them disappearing are separate.
■ Idiosyncratic risk. Speculative stocks carry lots of this. The questions of whether an oil explorer will find big reserves, or whether a new technology will succeed are largely independent of what's happening in the rest of the market. As Alex Coad of the Max Planck Institute concluded from a survey of corporate growth, "firm growth appears to be an idiosyncratic and fundamentally random process".
These three risks imply that correlations between speculatives and defensives should be low or negative. And history suggests this is often so. Looking at the last five years of monthly returns, correlations between Circle Oil (a popular Aim stock) and National Grid or GlaxoSmithKline have been 0.08 and 0.02 respectively. That between KSK Power and Glaxo has been 0.08. That between Asos and Glaxo has been 0.03. And so on.
Such low correlations suggest a fascinating possibility - that if we have a portfolio of defensives, we might be able to reduce risk by replacing a defensive stock with a speculative one.
This is because defensives tend to be correlated with each other, partly because even the most defensive shares carry some market risk, partly because they are sensitive to sentiment risk - doing relatively well as sentiment deteriorates and relatively badly as it improves and partly simply because they are sometimes in the same sector. Replacing a defensive stock with a speculative one might therefore reduce the correlations between our holdings sufficiently to reduce the riskiness of our portfolio, even though the speculative stock is more volatile in itself than the defensive.
Let's do some numbers. Imagine you have a portfolio of 10 defensives, in which the monthly variance of the average stock is 30 - that of ABF or SSE in the last five years - and the average correlation between them is 0.5. The variance of this portfolio is 16.5 - a monthly standard deviation of 4.1. Now, let's replace one stock with a share that has a variance of 280 and zero correlations with defensives; this is roughly the properties of Circle Oil in the last five years. The variance of your portfolio falls to 16.3. Selling a defensive to buy a speculative stock thus reduces risk. This is because Circle introduces some zero correlations into your portfolio, and this diversification benefit just outweighs the contribution made by its extra volatility.
Of course, there are caveats to this. For really volatile stocks - and many are more volatile than Circle - their volatility contribution exceeds their correlation benefit. And if you hold less defensive stocks, you might not gain so much diversification benefit from switching towards speculative stocks because the latter might be more correlated with your existing holdings by virtue of greater exposure to sentiment risk.
Nevertheless, the general point holds. Because the risk of speculative stocks is largely idiosyncratic, their marginal contribution to the riskiness of a defensive portfolio can be small or even negative.
But what of the returns on a 'risk barbell' portfolio? Economic theory says that a switch to speculative stocks from defensives should raise expected returns because speculative stocks expose us to more sentiment risk and thus raises the beta of our portfolio - its sensitivity to market risk.
If this is the case, then a mixture of speculatives and defensives might have a better Sharpe ratio than defensives alone - similar (or lower) risk and better returns.
However, the advantage of the higher beta is slight; the beta on many Aim stocks (and in fact Aim itself) with respect to the All-Share is only around 1.0, against around 0.5 for many defensives. And history shows that this not generated the returns it should have.
But you can be forgiven for thinking that history won't repeat itself in the next few months. If investors' sentiment improves - say, because the US's fiscal cliff is resolved satisfactorily, or because the global economy picks up or simply because it's the right time of year - then speculative stocks should do well. And if they do, those of you with a light sprinkling of them among defensive stocks might well be holding much more mean-variance efficient portfolios than you might think.
MORE FROM CHRIS DILLOW...
Chris blogs at http://stumblingandmumbling.typepad.com