It's an iron law: stockpicking is a great leveller over time

Reading the ever stimulating Jim Slater's new book on investment has set me thinking once more about the pros and cons of rule-based investing. Over the years Slater has investigated a number of different rule-based methods for buying and selling shares. He is perhaps best known now for his approach to the high-octane business of selecting growth shares, where he lays particular emphasis on relative strength and low price to earnings growth multiples as the key selection criteria.

Reading the ever stimulating Jim Slater's new book on investment has set me thinking once more about the pros and cons of rule-based investing. Over the years Slater has investigated a number of different rule-based methods for buying and selling shares. He is perhaps best known now for his approach to the high-octane business of selecting growth shares, where he lays particular emphasis on relative strength and low price to earnings growth multiples as the key selection criteria.

In the highly volatile and unbalanced markets of the past few years, this approach has had good and bad patches. For a while it worked exceptionally well, then there was a period where it was less successful, and so on, which seems to confirm the iron law of all markets, which is that stockpicking, however sophisticated the method, and however experienced the stockpicker, is a great leveller over time. As fashions and styles change so much in markets, and do so in ways that are hard to predict in advance, you need good fortune as well as hard work and skill to grind out those few percentage points of extra return that are all that distinguish the few great investors from the rest of the pack.

Where Slater has been a genuine pioneer, in my view, is in finding new ways to share his experience and enormous enthusiasm for the game of stockpicking with ordinary investors. Whether or not you agree with his methodology, there is no doubt that he has introduced a lot of investors to the kind of thought processes and the degree of analysis which are necessary if you are ever to aspire to become a successful manager of your own investments. His new book, which carries the eye-catching title How To Become A Millionaire, and is coauthored with Tom Stevenson, a former City Editor of this newspaper, essentially makes the case for starting young as an investor and harnessing the power of compound interest and the growth potential of shares and property to build your wealth over a period of years.

Among other things, if you have a mortgage endowment policy, the book advocates selling your policy and substituting an equity-based ISA instead (although the authors' assumption that investors should be able to earn a 15 per cent compound return from their equity investments in future seems over reliant on extrapolating the experience of the past 20 years). While personally still committed to the growth investing cause, Slater reveals he has now joined the ranks of those who acknowledge that for many investors, with limited time and resources, a tracker fund can be a sensible alternative.

When you think about it, of course, a tracker fund is merely an extended example of a rule-based investment system. In effect, the rule is "buy such shares as you need to have in order to replicate the performance of the main market index at all times". The particular merits of this rule are that it produces a relatively low-cost portfolio and one that guarantees to capture all the movements in fashion, style and external events that collectively constitute the concept of "market risk".

As market risk is the single most dominant factor in determining investors' returns, and one that cannot be diversified away, index funds inevitably produce portfolios that are inherently relatively efficient. The other great advantage they have, shared by all rule-based systems, but only if they are rigorously enforced, is that they ensure that investors do not succumb to their emotions by buying and selling at the wrong time -a proven way to lose money.

Are there other rule-base systems that do the same job as well or better? One which has demonstrated considerable staying power is the so-called O'Higgins method of picking portfolios of high-yielding large capitalisation shares. Depending on which variant of his theory you adopt, he advocated buying either one, five or 10 of the highest yielding stocks in the Dow Jones index at the start of each year, holding them for a year and then automatically replacing them using the same criteria.

In their new book Slater and Stevenson have taken a number of variants of the O'Higgins method and applied them to the UK market, to find out how they would have fared over the period 1984 to 1998.

Their method was to find the 30 largest companies by market capitalisation in the FTSE 100 index at the start of each year and then isolate the 10 stocks in that list with the highest dividend yields. Those 10 stocks constitute the core portfolio. Other variants include either single-stock or five-stock portfolios drawn from this initial list of 10. For example, the five-share selections are chosen by picking the shares with the smallest share price or lowest market value. All the figures are based on total returns, which assume that all dividends are reinvested rather than spent.

Over the entire 15-year period, during which the All-Share index grew by 17.1 per cent per annum, almost all the high yielding portfolios outperformed the index by between 0.9 and 3.8 per cent per annum. There are some anomalies - the fact that the second highest yielding share trounces the single highest yielding share in most years.

The question with all exercises of this kind is (a) whether they have any predictive power and (b) if the added returns compensate for the extra risk involved. Inevitably the figures are affected by the start date.

Take out the first three years in this sample (1984-86), for example, and the advantage enjoyed by the five and 10-stock portfolios is cut to 1 per cent a year, not a good margin. As Slater concedes, any one-stock portfolio is also unnecessarily risky, so they should be ruled out as well.I am prepared to wager that the five-stock portfolios owe their superior performance largely to the added specific risk of having only a partially diversified portfolio.

When the market as a whole falls, the smaller portfolios tend to fall further, which may well be a pointer to how they would perform in a future market downturn. Note how the periods of over and under performance relative to the All-Share index tend to run in two to three-year lumps.

What is not in doubt is that the 10-stock models are (a) emotion-free and (b) relatively low-risk strategies. You are unlikely to go far wrong with a portfolio of this sort, whatever the market conditions, provided you stick with it through the bad years as well as the good ones. But is it inherently better than an index fund, after adjusting for risk and costs? Maybe, but there is not that much in it, I would say.

davisbiz@aol.com