Join our community of smart investors

On uncertainty

Uncertainty matters for investors at least as much as risk. But we are not always rewarded for taking it on
March 5, 2019

We all know that as investors we face lots of risk. But we face uncertainties, too. And the distinction matters.

It was first made back in 1921 by the University of Chicago economist Frank Knight. Risk, he said, is something we can quantify, as when we say there’s a one-in-six risk of a die rolling a one. Uncertainty, however, cannot be quantified.

An obvious example of the latter is political uncertainty. We cannot quantify, for example, the likelihood of a no-deal Brexit. Of course, some try to do so when they attribute (say) a 20 per cent chance to it. But this isn’t useful. We’ll never see many parallel paths of history from today, one in five of which lead to a no-deal. So we’ll never know whether such statements are true or not. They merely tell us about the speaker’s state of mind. And some guy’s state of mind is an unreliable guide to reality.

Political uncertainty, however, is only one of our problems. Another is: will the future resemble the past? Since 1900, UK equities have given an average real total return of 5.5 per cent per year with a standard deviation of 20.9 percentage points. This seems to suggest that equity investing is a matter of risk alone – that there is (say) an almost one-in-four chance of a loss of 10 per cent or more in any year.

Such an inference is wrong. It holds only if the future resembles the past. But will it? We’ve reasons to fear not. In the 20th century share prices were depressed for long periods by fears of revolutions, inflation, war and crisis. As these fears receded prices rose. But those were one off relief rallies that won’t recur. Also, economic growth and hence dividend growth was strong in the latter half of the 20th century, but it mightn’t remain so in our era of secular stagnation. And investors might well have wised up to the good value of equities, thereby driving prices up to levels from which subsequent returns will be lower.

These are all dangers we cannot quantify. This means we face what Richard Bookstaber in his book The End of Theory has called 'radical uncertainty'.

To see another source of uncertainty think of the market as being an ecosystem containing different species. This raises the question: which species dominates? One type is the trend follower or momentum investor. Another is the value investor. The former sells falling stocks; the latter buys them. If the latter are sufficiently numerous, prices will be stable. If the former are, they’ll be volatile. We cannot say confidently in advance which it will be: volatility is prone to sudden spikes and falls. We therefore face what Stanford University’s Mordecai Kurz has called endogenous uncertainty – uncertainty that arises not from the macroeconomy but from the structure of the stock market itself.

All these are sources of uncertainty for the aggregate market. Individual stocks face more uncertainties. An obvious one is that companies with short histories or which operate with new technologies or in unfamiliar markets are especially difficult to value. Yet another is that apparently sound companies can suddenly get into trouble because of fraud, mismanagement or an ill-judged takeover.

But there’s a perhaps bigger uncertainty: technical change. Even large companies can be killed by being on the wrong side of changing tastes or technology – what Joseph Schumpeter called creative destruction. Only one in eight of the US’s largest 500 companies in 1955 are still in the largest 500 today. And only two of the UK’s largest employers in 1907 are independent stock market-listed companies today: Prudential and WH Smith. History warns us therefore that the prospects for even the biggest companies are uncertain.

Herein lies a big difference between risk and uncertainty. Some risks diminish over time as losses are followed by gains: if you bet heads on the toss of a coin you might lose on one or two tosses but you’ll win roughly half the time over enough tosses (assuming you’re not wiped out by short-term losses!). Uncertainty, however, increases with time. We have a fairly good idea of the strengths, weaknesses, opportunities and threats facing a particular company over the next 12 months. But if we’re honest, we haven’t a clue about them over the next 30 years.

All this raises the question: are investors rewarded for taking on uncertainty? In theory they should be. Back in 1961 Daniel Ellsberg showed that people hate uncertainty more than they hate risk: they were happier to bet on drawing a red or black ball from an urn containing 50 of each (a risky bet) than they were from an urn containing unknown proportions (an uncertain bet). This suggests that investors avoid assets which are prone to lots of uncertainty, which means these should be cheap and so offer good returns for those brave enough to take on such uncertainty.

Evidence here, however, is mixed. On the one hand equity valuations are low when political uncertainty is high. But on the other, stocks that carry more uncertainty don’t seem to do better than those that don’t. Aim stocks, for example, have under-performed mainline ones since the index was launched in 1995. And growth stocks have generally underperformed value stocks over the longer run.

It’s possible, therefore, that there’s an uncertainty premium in the aggregate market but not in many individual stocks. Which is perhaps yet another argument in favour of tracker funds.