Sassan Zaker at Bank Julius Baer says that just as there is - in theory at least - a trade-off between risk and return, so there can also be a trade-off between alpha and uncertainty. A high alpha might be just a reward for exposing yourself to uncertainty, in the sense of unquantifiable dangers.
In one sense, this must be true. Alpha is the return we get controlling for risk. This means that, to measure alpha, we must not only know which risks are relevant, but also be able to measure them. But we often can't do this. We know that market risk - a share's covariance with the market - is important and measurable. But what other risks should we control for, and how do we measure them: tail risk, liquidity risk, cyclical risk, correlation risk and so on? In the absence of definitive answers to these questions, alpha is uncertain.
There's a more important sense in which alpha is the counterpart of uncertainty. Many strategies which (sometimes) generate alpha do so by taking on uncertainty. For example:
■ Investing in private equity, venture capital and micro-cap stocks means owning companies with little cash flow and unknowable prospects, often dependent upon new technology. Such companies aren't so much risky, in the sense that their pay-offs have a known probability distribution, but rather genuinely uncertain.
■ Behavioural finance strategies, such as buying momentum or defensives, expose us to the uncertainty of whether, or when, investors will wise up to the cognitive biases that cause such shares to deliver good returns and so eliminate their alpha. This is an especially nasty problem because recent returns can't help us. For example, defensives have done well recently. Does this confirm that they are great investments? Or have they risen because investors have finally cottoned onto this fact, and so bought them and in doing so have bid up their prices to levels from which future returns will be poor?
■ Investors in resource stocks take on uncertainty. There's the uncertainty of whether exploration will yield the hoped-for discoveries of reserves; the uncertainty about the possibility of a catastrophic operational failure such as BP suffered with Deepwater Horizon; and the uncertainties of whether investing in countries with cultures of corruption and poor corporate governance will lead to large losses such as Bumi has seen.
■ Any strategy using leverage exposes you to the uncertainty (I call it so because it cannot be quantified) of whether your debts can be refinanced. As banks discovered in 2007-08, this question cannot be taken for granted. Any strategy of borrowing short to lend long - banks' traditional business model - is an alpha-uncertainty strategy.
In many common ways, therefore, chasing alpha means taking on uncertainty. Which poses the question: are investors rewarded, on average, for doing so? This question is not easily answerable. Uncertainty, by definition, is not quantifiable. We cannot therefore know what its price should be and so cannot say for sure whether stocks that appear to have offered alpha were really underpriced.
What we do know, though, is that the cliché that investors hate uncertainty is true. They hate it more than risk. This means you'd expect them to normally avoid uncertain shares and strategies, which should mean that investors brave enough to take it on should get good returns. If you're rewarded for taking on risk, you should be even better rewarded for taking on uncertainty.
However, three things argue against this. One is that investors might not realise they are taking on uncertainty. They might instead be blinded by overconfidence into thinking they are onto a sure thing. This was the fate of banks in 2007-08. This warns us that uncertainty can be dangerously overpriced.
Secondly, stock-specific uncertainty, like stock-specific risk, is diversifiable because you can spread it by holding other shares. You couldn't, for example, have quantified the danger that Bumi would suffer 'financial irregularities'. But you could have diversified that uncertainty. Economic theory - and common sense - tells us that diversifiable risk shouldn't generate extra returns. The same should be true for diversifiable uncertainty.
Thirdly, says Mr Zaker, investors' aversion to uncertainty will vary with their general sentiment. Uncertainty strategies will be 'cheap' when sentiment is depressed but less so when it is not. This implies we should expect alpha to disappear if sentiment worsens.
Herein, though, lies some hope for stock-pickers. Investor sentiment right now is quite depressed; one gauge of this is that Aim stocks, which are sensitive to sentiment, are at a three-year low relative to the All-Share. This suggests that investors are shunning uncertainty more than usual, which could mean that alpha-hunting strategies are underpriced and are about to pay off well.
This, though, is not my principal point. The point is merely that what is true so often in economics is true also for alpha - there's no such thing as a free lunch.
