For example, will growing opposition to austerity in the euro area - embodied by Syriza in Greece and Podemos in Spain - lead to a re-eruption of the euro crisis? Will deflation in the region be merely temporary? Or might it become long-lasting either because it creates expectations of low inflation or because it leads to a vicious circle as governments impose more austerity in an attempt to stabilise debt-GDP ratios? How will the outcome of the UK general election affect the mix of monetary and fiscal policy? Will it take us nearer to leaving the EU? Will the fall in oil prices have mainly benign effects or might it instead trigger a financial crisis by reducing oil exporters' abilities to service their debts? Will the ECB finally undertake full-blown quantitative easing and if so will it do much good?
Questions such as these cannot be answered by appealing to knowable probabilities. Of course, we could attach a subjective probability to them - for example by saying there’s a 30 per cent chance of the UK leaving the EU - but such claims are arbitrary attempts to quantify the unquantifiable. Instead, what we have here are examples of Knightian uncertainty - dangers which cannot be quantified. Economic consultant John Llewellyn says: "Unknowable uncertainties outweigh the calculable risks."
Does this matter for investors? In one sense, no. Share prices are simple things: they can only go up or down. And the dangers of them doing so are reasonably well described by a statistical distribution in which moderate losses are less likely than a normal distribution while bigger losses are more likely. The things that will cause shares to move in 2015 might be unknowable and unquantifiable. But the likelihood of the moves is roughly quantifiable. In this sense, investors face risk - that is, quantifiable danger - rather than uncertainty.
However, uncertainty does matter in at least three other ways.
First, it can generate volatility. It’s another cliché that investors hate uncertainty. But again, the cliché is right. Ever since Daniel Ellsberg’s work in the late 50s, we’ve known that people prefer to bet upon known probabilities than unknown ones. Experiments by Tilburg University's Stefan Trautmann and Harvard’s Richard Zeckhauser show that this is true even when people are given the chance to learn about the extent of uncertainty they face. This uncertainty aversion means that assets which are thought to be exposed to uncertainty will be under-priced to embody not just a risk premium but an uncertainty premium. Such assets will therefore see big price rises if uncertainty is resolved and big price falls if it increases. This is a recipe for volatility.
Secondly, it means that volatility will vary over time. Uncertainty isn't merely something that is imposed onto financial markets from outside by pesky politicians. As Stanford University's Mordecai Kurz has shown, it is also endogenous: it arises from the process of trading shares.
This is because trading is a game of trying to anticipate others' beliefs; you make money by selling before others sell and buying before they buy. This means that prices can fall without any obvious trigger, simply if investors worry that others will worry; this is just what happened in mid-October and in mid-December when the FTSE 100 lost 400 points without much change in the investment environment.
The more uncertainty there is about the economy, the more uncertainty there is likely to be about others' beliefs. Such uncertainty will generate variations in volatility. There'll be periods of stability in which we don't worry that others will worry, interspersed with periods of volatility when we worry that they do. The transition from one state to the other can be sudden and unpredictable - as we see in the fact that there can be sharp spikes in the Vix index, a common measure of volatility.
Thirdly, all this has implications not just for equities but for government bonds too. Gilts, and especially index-linked ones, aren’t just risk-free assets but also uncertainty-free ones - at least if they are held to maturity. This means their prices will be high (and yields low) in uncertain times because investors will pay a high premium for protection against uncertainty. One reason why gilts did surprisingly well last year might be precisely because uncertainty increased.
This, though, is a two-edged sword. On the one hand, it means that uncertainty could keep real yields negative. On the other hand, though, it implies that if we do see a reduction in uncertainty we might see a sharp sell-off in bonds as their uncertainty premium falls. There has always been a big difference between bonds and bond funds - so much so that the two should be regarded as distinct asset classes. The difference is especially important now.
