Thinking clearly about risk is difficult: we are influenced by hopes, by regrets, and by an inability to evaluate new risks.
One of the first shares I ever bought was in a small Scottish shipyard called Robb Caledon. The shipyard was, for practical purposes, bankrupt. But a bill to nationalise the shipbuilding industry was going through Parliament. It prescribed a formula for the price. This was based on the market value of the shares in what, at least for Robb Caledon, had been happier times.
If the bill passed, the government would pay around £1 a share. If the bill failed, the shares were worthless. Since the Labour government had a parliamentary majority and a manifesto commitment to nationalisation, the shares seemed to me a good buy at 40 pence.
When I rang a stockbroker, he advised strongly against. He pointed out that shipbuilding nationalisation was a foolish policy, one of many which were leading the country to ruin. He reminded me that the shares had already risen from 20 pence a few weeks before. And he pointed out that I could lose my whole investment.
All these points were true and irrelevant. They are characteristic of the mistakes we make in thinking about risks. We find probability difficult, and confuse our hopes and desires with our expectations. We let our current judgment be swayed by what we might have done in the past. And we assess risks individually, not as part of a portfolio. The most you can lose if you invest £1000 in Robb Caledon is the same as your loss if you invest £5000 in Shell and its price falls by 20%.
More recently, I sent a colleague on an assignment in Kazahkstan. He came to see me to say that Kazahkstan was more dangerous than London and he thought the company should buy life insurance to protect his wife and children. Since I was not very sure where Kazahkstan was, I was not in a position to dispute the premise, but could challenge the argument. I understood that his wife and children needed protection, but didn’t they need that protection just as much when he was walking around the streets of London (where insurance could be bought cheaply) as when he was on the steppes, or whatever they have in Kazahkstan? I argued in vain.
There is a risk you will be killed in a road accident. Almost one man in a hundred (many fewer women) dies that way. How much would you pay for extra safety features that would half that risk, such as airbags and crumple zone protection? £1,000: perhaps as much as £2,000? Now consider how much I would have to pay you to cross a minefield in which there is a 1 in 100 chance of your being killed. I bet the answer is a lot more than £1,000. Experiments show that people need to receive much more to accept a new risk than they will pay to reduce an existing risk.
All of this behaviour financial economists describe as irrational. This is a bad phrase. I cannot show that you are wrong to make these decision – to buy extra insurance when you are engaged in risky activities, or to assess new and old risks differently. But I can show that people who do behave in these ways will lose money to people who behave “rationally”.
Now maybe those who pay out do not mind. After all, insurance companies, lottery organisers and derivatives traders all make profits. Holders of insurance policies, lottery ticket holders and derivatives buyers on average lose. But they go on insuring, entering the lottery, and buying and selling complex financial instruments, so presumably they get something out of it. But if the object of business is to make money, then the business management of risk ought to follow the principles of rational behaviour.
Rational behaviour is often counter to our deep psychological instincts. We often think about unlikely events more often than is justified by the probability that they will actually occur, particularly if the outcomes are drawn to our attention. That is why we are more interested in the jackpot on the National Lottery than the – often equally likely – chance of dying of a heart attack before the draw takes place. And why we ask for large sums to insure an unnecessary risk even if the probability of it materialising is extremely low.
We overvalue certainty. Asked to choose between a holiday in Torremolinos and a fifty per cent chance of one in Barbados, many people choose Spain: but reduce the probabilities to ten per cent and five per cent and Barbados becomes more popular. We are influenced by regret, so that our choices are influenced by what might have been. That is why my stockbroker thought I would be reluctant to purchase shares I could once have had more cheaply: which of us has not entertained similar feelings? And we are readier to gamble with money we have just won than with money that is already in our pocket. The house money effect: if someone gives you chips, you rarely go straight to the bank to cash them in.
But most of all, we tend to think about risks one by one rather than by reference to the totality of the risks we face in our lives or our businesses. Many investors were put off Robb Caledon shares by the prospect of a total loss. I would have been daft to invest the whole of my life savings in them: yet if I have a diversified portfolio, I can take large risks on individual components of it. So long as we think about risk in “irrational” ways, people who trade risks will make money at the expense of people who bear risks.