Goldfish Don't Have Defensive Moats
To argue that many investors and all markets are like goldfish, continually forgetting the recent past as soon as they swim around the corner of their cute seaweed festooned castle, is to do our piscine friends a disservice: goldfish have a memory span of up to three months. Investors sometimes struggle to be consistent for three hours consecutively.
It often seems like people have a switch in their heads marked “Risk”. While this is in the “OFF” position they’ll happily fill their boots with any old rubbish stock as long as enough other people are buying it. Yet as soon as life’s rich uncertainty rears its ugly head the switch gets set to “ON” and they head for cover behind the seaweed in the hope that their tiny castle can protect them. Memo to markets: risk is always with us, no matter how bad your memory – and castles in goldfish bowls don't have defensive moats.
To invest on the basis that something is always about to go wrong is the sign of an intelligent investor. To figure out that the time to really invest is after everyone else has got scared about this and headed for the hills is the sign of an intelligent and wealthy one. The evidence of markets suggests most people aren’t that smart, yet we’ve not exactly suffered from a lack of learning opportunities about these types of situation in recent years.
We had the Asian financial crisis in 1997; the dotcom crash of 2000; then the tragedy of 9/11 and finally the subprime crash of 2007. So we've had plenty of practice in dealing with situations of heightened risk; but it doesn’t seem to make the slightest difference to market behaviour. As soon as everything gets a bit scary everyone heads for the high ground of whatever the latest safe haven is supposed to be.
This wouldn’t be irrational behaviour if people spent most of their time surveying events from the high ground anyway. Being partially invested in supposedly lower-risk stuff like government bonds or gold or tulip bulbs or conch shells is a perfectly rational approach to the true uncertainty of the world. But mostly people don’t do this and operate in a risk on/off binary world that they judge from the performance of markets and the signals given off by price movements.
This signalling seems to a genuine problem. Instead of looking at stock fundamentals and a margin of safety people look at the price (see Quality Signalling for Quality Stocks). It’s a common way for people to judge the quality of something that they don’t know a lot about and, in fairness, a lot of the time a stock’s price does signal something useful about its quality. Unfortunately there are also times when it doesn’t and these are usually in periods of increased uncertainty and forced selling due to margin calls and liquidity problems.
This lack of fundamental analysis means that many investors are actually not as expert as they think they are and this opens them up to the problem of risk perception. A classic example of this is the study by Paul Slovic who looked at 30 different risky situations and asked a range of different groups, including some risk analysis experts, to rank the risks. The non-expert groups ranked nuclear power as the number one risk in the world, yet the experts put it at number 20. Despite Japan’s attempts to prove them wrong the truth is that you’re much more exposed to radiation danger through having X-rays than you are from a nearby nuclear power station: but that’s not the way people perceive it. Risk is subjective not objective: and that's true even when experts measure it.
The problem is not really about risk, but about the perception of risk. If this feeds through into stock market behaviour we would expect to see that the degree of risk aversion exhibited by participants is related to their perception of risk in the world. This is far from a new idea – Werner De Bondt and Richard Thaler long ago demonstrated that portfolios of “loser” stocks – firms whose prices had been impacted by negative announcements – outperformed portfolios of winners over three year periods.
In particular there seems to be something about the fear of extreme events that causes the risk switch to be flipped to "ON". Paul Slovic and Elke Weber have developed a risk model that takes into account perceptions of risk. This psychometric risk model measures risks on two dimensions: dread and uncertainty. Uncertainty is the degree to which we’re familiar, or not, with the risk. Dread is the degree to which we feel we’re under the control of potentially catastrophic circumstances. Risks which are both dreaded and unfamiliar tend to provoke extreme reactions.
The risk on/off behaviour of stockmarket participants is conditioned by uncertainty and dread. Uncertainty is endemic in markets anyway, but when this is reinforced by the apparent possibility of catastrophe, as in the failure of the world’s financial systems and the ruining of personal finances, then dread takes over. When that happens then we stop thinking and start reacting.
Affect and Risk Aversion
We’ve known for a long time that we have two mental systems for handling risk – a cognitive, analytical one and an affective, reactionary one (see Risk, Stone Age Economics and the Affect Heuristic). In conditions of stress and fear the latter takes over. As Slovic and Weber explain:
“When decision makers are asked to make a single choice based on the description of possible outcomes of risky choice options and their probabilities, rare events tend to be overweighted as predicted by prospect theory (Kahneman and Tversky, 1979), at least partly because the affective, association-based processing of described extreme and aversive events dominates the analytic processing that would and should discount the affective reaction in proportion to the (low) likelihood of the extreme events occurrence.”
So what we’re seeing as investors move in and out of stocks in a co-ordinated and irrational way is, oddly, an entirely predictable (although wildly unprofitable) feature of risk perception. Many investors, both professional and private, probably handle the normal uncertainty of markets with reasonable aplomb. What sends them into a tizzy of risk on/off behaviour is the introduction of dread risk, the perception of globally and personally catastrophic events over which we have no control. Stuff like the Euro imploding or the US having its credit rating reduced to something lower than a vegan's chlorestrol level.
Of course, as we’ve seen before in Black Swans, Tsunamis and Cardiac Arrests, the possibility of a calamitous, one-off event is always with us; it’s just that we’re not normally aware of it because of disaster myopia. But when we're brought face to face with the consequences we don’t behave entirely sensibly. As Gerd Gigerenzer noted in Out of the Frying Pan into the Fire, in analysis of the reaction to the 9/11 attacks:
“Dread risks can cause direct damage and, in addition, indirect damage mediated though the minds of citizens. I analyze the behavioral reactions of Americans to the terrorist attacks on September 11, 2001, and provide evidence for the dread hypothesis … An estimated 1,500 Americans died on the road in the attempt to avoid the fate of the passengers who were killed in the four fatal flights.”
Were these people behaving irrationally?
Only to the extent that they didn’t understand the risks they were taking. Of course there was the possibility of another terrorist attack, and there still is, but travellers were probably safer flying in the US after 9/11 than before. Similarly for investors. Dread risk is always with us but we shouldn’t shy away from this at the point the rest of the world recognises this. The very awareness of the possibility of global catastrophe lessens the likelihood.
But the risk is never, ever zero. Which is why investors should get a friend to superglue their risk switch so that it’s permanently set to Risk := “ON”. After all, the defensive value of a moat in a goldfish bowl is precisely zero.
Related articles: Quality Signalling for Quality Stocks, Ambiguity Aversion: Investing Under Conditions of Uncertainty, Risk, Stone Age Economics and the Affect Heuristic