This piece is built upon Eric’s discussion about the complex nature of asset price determination at his own blog at philosophyofmoney.net earlier this week.
Here is a chart of the South Korean and US stock markets so far this year, alongside some news items from the Korean peninsula:
As Eric mentions on his blog, it would have been reasonable to anticipate more volatility in Korean assets than we have seen. The language and displays of military strength on both sides is a source of concern. However based on a chart of the Korean stock market, you would barely notice anything significant had occurred at all, let alone the suggestion that we could be facing a major global conflict.
This resilience (so far) of markets is part of a broader trend. The realised volatility of most major assets was low in the first quarter of the year in spite of the hoo-ha around rising geo-political risks.
That markets have been so quiet, while the world is scarier, raises interesting questions about what causes asset prices to move. Intuitively we have a sense that “uncertainty” and volatility should go hand in hand. There is also a tendency to feel that it is possible for “uncertainty” to increase and decrease regularly, and that it is possible for the environment to be ‘unusually uncertain.’
This is a little odd. From a philosophical standpoint, what we don’t know now was presumably part of what we didn’t know before – having more information can’t technically make us less certain about what will happen.
For example it must be the case that a Trump election victory was part of the probability distribution three years ago, or else it couldn’t have happened. But now we are certain that Donald Trump is the US President, political uncertainty is deemed to have increased.
What we are actually saying is that we believe that the likelihood of extreme negative events has become more certain.
There are two ways that this can happen:
1) New information reveals that we were flawed in our original beliefs. Human beings typically overweight “known unknowns” and underweight “unknown unknowns,” that is, we effectively ignore risks that we don’t know about.
So, when a black swan is discovered it means that from the individual’s point of view there is greater uncertainty. You previously would have re-mortgaged your house to bet on a swan being white, now you are not so sure.
2) Time horizon matters. Probability distributions capture the range of outcomes over a particular period of time. Investors have not priced in the chances of the earth being swallowed by the sun because, although it seems certain to happen, it is very unlikely to take place within the time horizons of the average investors. It seems likely that there will be many market participants out there who treat global warming or the high levels of Japanese debt in the same way.
This seems to be a somewhat silly analogy, but consider how European investors have treated the various elections across the continent. They are often telegraphed well in advance (elections come in regular cycles) but the “uncertainty” that goes with them only tends to manifest itself in markets in the months beforehand.
As a result the simple passage of time can be significant.
These two issues mean there is deep complexity in how we should think about uncertainty. We should spend more time thinking about unknown unknowns, but how much? The time horizon of the average investor matters but is not static: the cohort of market participants is constantly changing and even individuals will worry more about the short term at some points than at others.
So what about Korea? A third observation on uncertainty
Why then haven’t markets started to factor in a greater likelihood of a deeply negative outcome?
The most obvious answer would be that actually the chances of conflict haven’t increased. North Korean sabre-rattling is not a new phenomenon (though our sense is that the South Korean equity market and currency tended to react more to these phases in the past) while the rules of mutually assured destruction may still hold as a force which prevents conflict.
You might also argue that other forces have been more important drivers of markets and that we cannot know the counterfactual. Perhaps equity markets would be even higher had tensions not increased in Korea. This is a real possibility; the South Korean equity market has actually de-rated as it has gone up, due to the strength of profits delivery.
Valid (and untestable) as these arguments are, there could be something else going on. Uncertainty only matters for investors insofar as it is uncertainty “within the rules of the game.” Nuclear war could arguably fall outside those rules: what does it matter the weight of bonds versus equity in your portfolio in an apocalyptic event?
Again this may seem extreme, but it is interesting to note that there was no impact on US stocks in and around the Cuban missile crisis, possibly for the same reasons. In 2008 there was a correlated fall across assets as investors went to cash, but if a true systemic collapse means that you cannot get cash out of an ATM, all bets are off. Ditto for a communist revolution. How can investors factor in such risks? You do not change your strategy at monopoly because there is a (significant) probability that your opponent will tip over the board in frustration.
We have written a couple of pieces on how volatility and uncertainty do not go hand in hand and how a period of low volatility doesn’t mean we are “due” a crash. But the fact that delivered volatility has been low of late is enough to reveal just how deeply complex the price determination mechanism is and how wary we should be of any simple model which says that “prices simply respond to news.”