The Longer You Play, The More You Lose

By: Andrew Smithers | Thu, May 20, 2004
Print Email
The Sunday Times, 16th May, 2004

Unlike most articles about investment, which tell people how to make money, this one will try to persuade you not to lose it. Shares, bonds and property are all overpriced and even more recondite things like gold seem to be singularly lacking in appeal.

Cash is the thing to hold and we are lucky in the UK, compared with other countries, that money on deposit gives a decent return. This is a rather negative and disappointing view, so I had better start by apologising. I am sorry that I haven't a more upbeat story to tell you, but please blame the markets rather than me.

The problem with investing in shares is that the rewards fluctuate hugely. They don't just do this from one year to another, but over decades. As reasonable people would suspect, long periods of high returns can only be obtained if shares get thoroughly overvalued and they are inevitably followed by long periods of poor returns.

We are in the early years of one of these poor periods. This is illustrated in Chart 1 which shows the returns from equity investment in the US. For the 30 years or so up to the end of the millennium, returns were wonderful and even after the subsequent falls they have still be extremely high for longer term investors. Starting with any year you like from 1973 to 1993, the return on the shares has been well above its long-term average and has typically been around 10% in real terms, which is more than 3% above the 100 year average. The position is not so extreme for the UK, where shares are less expensive, but this is only a moderate consolation as the UK market seems to follow the US rather than have a life of its own.

Unfortunately, what is good news about the past is bad news for the future. The common sense view that good times can't last for ever is strongly supported by the evidence. This shows that stock market returns do not follow a random pattern, like tossing a coin, but that good times are followed by bad ones and vice versa.

If you play roulette, the chances of red coming up on the next spin of the wheel are not influenced in the least by the number of times that red has come up recently. Stock markets are different; the past is a guide to the future. But it's not much of a short-term guide. We know today that the chances of poor returns over the next few years are very high, but we don't know very much about the chances over the next year.

This again is very like playing roulette. You know, or you should know, that the longer you play the more certain you are to lose money. Although the odds against you on each spin of the wheel are small, over time this small disadvantage turns into a near certainty of loss. Investing in overvalued stock markets, like those we have today, is very similar. People play roulette despite the odds being against them, and they sometimes win. But this doesn't make it a sensible thing to do. Holding shares today may make money over the next 12 months, but the chances are in favour of investors losing. Over the next five years the chances of losing are even higher.

Because markets give poor returns after periods of good ones, it follows that they can be valued. There are two correct but different ways in which this can be done and they are illustrated in Chart 2 for the US stock market and Chart 3 for the UK one. As you can see the two correct ways track each other closely, which must of course be true of any valid measure of value. (There are of course a lot of other incorrect ways for valuing the stock market. These tend to be very popular with stockbrokers, who like to pretend that shares are always worth buying.)

The two different ways of valuing stock markets are based on two different approaches. One, which uses the so called q ratio, depends on the fact that the fundamental value of companies in aggregate is the same as the value of their assets, after deducting their debts. (This is not the same as their book value, as allowance must be made for inflation.) The other method of valuing shares depends on their earning power. It is the based on the PE multiple, but allows for the fact that in the short-term earnings can fluctuate dramatically.

Investors should be aware that, because earnings can go up and down so sharply, this means you cannot use this year's PE multiple to value the market. A great example of this is 1932, when in PE terms the market was probably at its most expensive ever, because profits were so depressed, but it was also about the best year ever to buy shares.

Using either the q ratio or the cyclically adjusted PE, and looking at either the US or the UK stock market, the most optimistic conclusion that is possible is that shares are only around 45% overvalued.

If shares offer poor returns, bonds are obviously one alternative. Sadly the prospects here aren't very good either. Government bonds yield around 4.5% in the US and 5% in the UK. As the Bank of England is aiming at an inflation rate of around 2% p.a. this suggests that the real return, after allowing for the impact of rising prices, will be around 3%. This is a little on the low side, particularly when the budget deficit is so high and the economy appears to have very little spare capacity. There is a high chance that the Bank of England will feel it necessary to push up short-term interest rates and bond prices are more likely to fall than to rise when this happens.

Current bond yields are not much better than the return that can be obtained on cash, and keeping money on deposit in today's conditions has the added advantage that your investment can't go down in price. As equity markets usually overshoot on the downside, there is a strong chance they will move from being overvalued today to being significantly undervalued in a few years' time. Having cash to invest then will be a great advantage.

Stock markets usually overshoot. They not only get overvalued, they get undervalued as well. If Wall Street fell by a third it would be fairly valued, but on past experience it could easily become undervalued and fall to half its current level.

As the prospects for shares and bonds are bad, some investors might be tempted by property and it certainly looks as if house prices are headed up again this year. But house property in the UK looks just as dangerous.

Chart 4 below shows the value of the UK housing stock has probably never been as high as it is today. The last time things were as out of line as they are today was in 1973, when we had the infamous secondary banking crisis.

Investors should remember that when markets fall economic conditions will be worse than they are today. Economies go up and down and shares are more likely to be overvalued when things look good. But this will have almost no impact on the value of shares, which is why I haven't referred to the state of the economy at all in explaining that shares are expensive. The frequently heard view that shares are good value because the economy is in good shape is nonsense. It doesn't necessarily follow that good times go with overvalued shares and bad times with cheap ones, but it is more often true than the other way around.

World stock markets have been weak recently. No-one can of course be sure why this has happened. However, it seems that investors are getting worried that the era of exceptionally low cost of borrowing of US dollars is coming to an end. While this worry is fully justified, it doesn't mean that stock markets will continue to decline in the short-term. Indeed, the recent decline may now fuel speculation that interest rates will not go up in a hurry. Equally, fashion may decide that now is the time to exit. What we know is that markets are overvalued; what we don't know is whether the next spin of the roulette wheel will turn up red, black or even green, the bad 'once in thirty-seven' chance, when nearly all the punters lose.


Andrew Smithers

Author: Andrew Smithers

Andrew Smithers
Smithers & Co.

Smithers & Co. Ltd. provides advice on international asset allocation to about 100 clients based mainly in Boston, London, New York and Tokyo. Our work is based on the fundamental belief that no one's judgement is better than their information. We believe that our clients' decisions will be helped if we can provide them with important information that is not otherwise available to them. We therefore concentrate on research which aims either to tackle issues in greater detail and thoroughness than is otherwise available or to tackle issues of importance which seem to have been generally overlooked. Examples of the former include our work on stock market valuation, the profit distortions arising from the use of employee stock options and the underlying secular problems of Japan's economy. Examples of research into areas which have otherwise been largely overlooked include our work on the Japanese life insurance industry.

Our approach to research is also different. The standard approach bases market projections on economic forecasts of major economic aggregates, such as GDP and inflation. Stock market, bond and currency forecasts are then derived from the way these estimates differ from the consensus. We consider this approach to be flawed in two ways. It places excessive reliance on the ability of any particular analyst to produce forecasts which are consistently better than average. It also ignores the evidence that stock markets tend to lead economies, rather than the other way around. In contrast, we put greater emphasis on "information arbitrage", in which we include identifying factors which have been overlooked, drawing on data and academic research which have not yet been exploited and pointing to inconsistencies in the implicit forecasts of different markets.

Andrew Smithers, founder of Smithers & Co., is also columnist for London's Evening Standard and the Tokyo Nikkei Kinnyu Shimbon's Market Eye, and is regularly quoted in the New York Times, Barron's, Forbes, The Economist, The Independent, and the Financial Times.

Copyright © 2002-2007 Andrew Smithers

All Images, XHTML Renderings, and Source Code Copyright ©