An average long run value of the equity premium of up to 8% is arithmetically unsustainable: within a few decades, profits and dividends would absorb the whole of national income.
In the 20th century, returns on equities substantially exceeded returns on safer assets. The size of the difference depends on the time period, the country and the method of calculation. But estimates of the average long-run value of the equity premium are mostly in the range of 3 per cent to 8 per cent.
Equity investment is riskier than cash or bonds. But not that much riskier. If the average return were even 3 per cent higher, medium to long-term investment in equities would be almost certain to yield more: the risk would become negligible. In countries such as Britain and the US with large equity markets, 5 per cent or more of national income is needed to compensate people for worrying about the value of their stocks.
So why is the historic equity premium figure so high and will it be as high in the future? Estimates of the prospective equity premium determine the cost of capital to business. Such estimates are central to every long-term economic decision – how to fund pensions, whether to build nuclear power stations, what to do about climate change.
Some explanations of historic levels of the equity premium imply that it will continue to be high in future. Perhaps there are many people who really need to be richly compensated to invest in equities. Even a small probability of large loss causes them great distress, or puts their job at risk. They invest a lot of emotional energy in turning nervously every morning to the share price pages of the Financial Times. They twitch over their BlackBerries in the airport lounge. They pay fees and commissions to financial advisers in generally unsuccessful pursuit of better returns.
If these are indeed the explanations, then the moral is to discipline yourself to avoid these traps. Accept that there are more important things to worry about than a falling stock market, recognise that daily share price movements are meaningless noise, restrict yourself to reviewing your portfolio only once a year. Stick with a few good stocks and take a cynical view of the claims of financial advisers. You will then gain most of the additional return from equity investment for little of the added cost.
But other explanations of the high equity premium offer less reassurance: they imply that the future cannot be like the past. If you examine figures at the higher end of estimates of the historic premium, you realise that they are arithmetically unsustainable: within a few decades, profits and dividends would absorb the whole of national income.
Perhaps the past century was just a very good time for equity investors. It was the age of inflation, in which real assets did better than nominal assets: the age of the large public corporation, in which big, professionally managed companies came into being and a regulatory framework was put in place that enabled small savers to trust business people with their money. Shareholders benefited from changes in the economic environment that will not happen again.
All historic analysis of investment performance suffers from survivor bias. Returns on the successful investments of the past are generally higher than you can expect on similar investments in the future. Investment managers advertise their best funds, not their dogs; unsuccessful hedge funds close. The equity markets of western Europe and the US are the bourses that remain open. People who put their money a century ago in Russian bonds, Chinese equities and Argentinian tramways lost most of it and there are today no analysts of these markets to tell the sorry tale.
Since none of these explanations is wholly convincing, the most plausible account of the equity premium paradox is that there is a bit of truth in all of them. That would imply that a future premium would be much lower than a historic premium, but still surprisingly high. But with real returns on indexed bonds below 2 per cent around the world and much lower in the UK, an equity premium at the low end of estimates of the historic range would offer a prospective real return on stocks below 5 per cent. That is a lot less than most people expect and many are counting on.