Someone who started investing in Wall Street in 1890, putting money in only during the month of September, would have lost three-quarters of the investment by 1994

INVESTMENTS

Can you make money in the stock market by following simple trading rules, such as "buy on the rumour, sell on the news", or "sell in May and go away"? There are many who believe that you can, although those who took the latter piece of advice this year would hardly have had much benefit from doing so.They would have missed out on the strong run that has taken the FTSE 100-share index to a new all-time high this month.

It's easy to dismiss such simple slogans as a waste of breath, but there may be more to them than you think. It is true that the academic world has taken great pleasure in recent years in puncturing many of the hoary old saws that passed for stock market wisdom in the days before we really began to understand how markets worked.

The bludgeon they have used, the doctrine of efficient markets, though much derided by market practitioners, remains one of the major contributions of the economics profession to the modern age. It and its variants have won a string of Nobel prizes for economists in the last 25 years.

The basic tenet of efficient market theory is not that markets have perfect information about what is going to happen in future. That is patently not the case. But what it does say is that, in a genuinely competitive market like the London Stock Market, you should not expect to be able to make above-average returns without taking above-average risk.

One reason is that there are sufficient smart and intelligent investors out there to ensure that any information that is publicly available will quickly be reflected in the buying and selling prices in the market. Or, to put it in plainer language, there are no free lunches in the stock market..

The evidence that supports the theory is generally powerful and convincing, but most academics now accept it is by no means universally applicable. The hunt for stock market "anomalies" - profitable buying opportunities that persist despite the market's general efficiency - remains a favourite pastime of academics. It also underpins the efforts of many professional investment managers. Those who class themselves as "value" investors, and spend their time looking for cheap or undervalued shares, are betting that not all shares are being priced efficiently by the market.

There's no doubt that anomalies do occur. Many ordinary investors have long been familiar with the so-called "January effect". This is the well- documented fact that share prices tend to perform better than average in that month. This, the academics readily concede, is not compatible with the view that the stock market is perfectly efficient.

Now we have some more evidence to the same effect. This month, a leading financial economist in the United States, Professor Jeremy Siegel of Wharton Business School, documented there is also a clear "September effect". His research, which involved combing through the movements of the Dow Jones Index as far back as 1890, showed conclusively that Septemberhas been a lousy month for investors, not just in America but in other big stock markets.

Professor Siegel calculates that someone who started investing in Wall Street in 1890 and put his money into the stock market only in September each year would, by 1994, have lost three quarters of every dollar invested. By contrast, someone who took his money out of the market each year in September but remained fully invested for the other 11 months of the year would by now have turned each $1 into more than $400. Such a strikingly divergent outcome is difficult to explain in either logical or behavioural terms, and is inconsistent with efficient markets theory.

Another man who thinks there is more to market anomalies than the academics would have you believe is David Schwartz, an American statistician who became interested in the stock market after selling his market research business in the late 1980s. With the energy of a self-confessed "boy from the Bronx", Mr Schwartz sat down in his study in the English countryside to make an exhaustive analysis of the daily movements in the UK stock market over the last 75 years.

Some years later, he has the results - an encyclopaedic quarry of trends and seasonal patterns in the behaviour of the UK stock market. His month- by-month analysis of how markets move through the year is now published as an annual, called the Schwartz Stock Market Handbook (perhaps inevitably, the boy from the Bronx also publishes this himself: call 01453 731 173 for details).

Mr Schwartz's labours make for fascinating reading. For example, here are his comments on some of the market's oldest and most trusted adages:

o"Sell in May and go away". The conventional wisdom here, says Mr Schwartz, is simply wrong. The market has actually risen seven years in a row in May since 1989. What actually gives May its bad reputation is that in a handful of well-documented years it fell very sharply.

o "Bull market bashes end with October crashes". It is true that years like 1987 have given October a bad reputation. But again the problem is that the month has experienced some bad falls. It is still profitable over the long run. What is more, the years of "big hits" are often well flagged in advance.

o "As January goes, so goes the year". This, says Mr Schwartz, is one saw that really has stood the test of time. When share prices rise in January, further rises often occur in the rest of the year - and the reverse holds when prices fall.

Mr Schwartz has many other examples of persistent trends. One interesting one is the trend for the stock markets to fall in the week before the Budget. This holds true even when the date of the Budget has been moved.

As a statistician, he knows as well as any academic the pitfalls of assuming that just because a price movement pattern has occurred in the past, it will recur. There is also the possibility that anomalies that did exist in the past and made money for investors tend to disappear as soon as they are documented and become generally known. To some extent, this has already happened with the January effect, which seems to be gradually moving backwards into December as more and more investors become aware of it.

Nevertheless, Mr Schwartz concludes that, whatever the academics may tell you, there is evidence patterns of predictable stock price movements do exist. None is universal, and nobody can expect to make money automatically by betting on them.

On the other hand, the knowledge prices have behaved in a certain way many times in the past should give investors confidence the odds are in their favour if they try to recapture the effect.

If you sold in May, page 26