Time passes and the caravan of luminaries in the world of finance and economics moves on with it. How widely, I wonder, does the name of Paul Samuelson still resonate among those under the age of 35? This wise and feisty man, once described with feeling by a contemporary as "a human mainframe", has good claim, along with his great rival, Milton Friedman, to have been the most influential economist of the post-war period.

In a career that has left virtually no corner of economics untouched, he was the man who first took the philosophies of Keynes and propagated them far and wide across the world. He was the first American economist to win the Nobel Prize for economics and wrote what proved to be the standard textbook on economics for a whole generation of students.

Samuelson's work has covered so many different subjects - his collected works run to six volumes and more than 400 academic papers - that it is easy to overlook his contribution in specific fields.

As it happens, one of the areas where he has had the greatest influence is in investment. As an admirer of Keynes, who was a noted speculator as well as a professional economist, he was one of the first economists to try and understand why markets should behave in the way that they do. A paper he wrote in the 1960s, dryly entitled "Proof That Properly Anticipated Prices Fluctuate Randomly", was the basis of what has since come to be known as the concept of efficient markets.

This is the notion that competition is such in large and well capitalised stock markets that it is difficult for anyone, even professional investors, to make consistent excess returns from them, excess returns in this context meaning higher returns than the market averages, after allowing for the different degrees of risk adopted. He later summed up his philosophy as "no easy pickings, no sure thing gains".

When Samuelson first propounded this potentially radical idea, it was greeted with astonishment, bordering on outrage, on Wall Street, which was hardly a surprise when you begin to think through the implications. Never one to want to avoid making waves if the opportunity arose, Samuelson himself concluded in a 1974 article that the best thing most professional fund managers could do was to pack their bags and go home. But, as a pragmatist, who like Keynes has always liked to dabble in the markets himself, he also had some serious advice for ordinary investors.

This was that the best thing they could do, once they had decided to invest in equities in the first place, was to put the bulk of their money into a low-cost index-tracking fund. He has long been a fan of the Vanguard Group, a fund management company which sells exactly such index-tracking funds to millions of American investors and has compiled a consistent above-average performance.

The line of reasoning, as those who have read this column will know, is an argument which I believe holds as broadly true today in general as it did when Samuelson first propounded it. It does not mean, as Jason Hathorn, a director of the specialist fund management group Buchanan Partners has kindly pointed out to me, that there are a number of ways in which investors cannot still hope to add value by adding a slice of active management on top of their basic indexed portfolio.

One of the arguments for investing in a professionally managed European or emerging markets fund, for example, is that there are good grounds for believing that other markets are not yet as efficient as the UK and United States and therefore that it will still be possible to find profitable pricing anomalies there.

Like Barclays Global Investors, whom I wrote about recently, Buchanan Partners believes that it is possible to use quantitative stock selection methods to exploit specific areas of inefficiency in the UK market. Their track record to date suggests that they may well be right, although how long these particular anomalies will persist is something which is by definition unknowable.

Samuelson would not disagree. In two recent articles for the Journal of Portfolio Management, he looked back on the 30 or so years of research that has been carried out since he first started developing the efficient markets hypothesis. His conclusion is that, while many individual anomalies have been shown to exist, the "jury of history" has failed to find "systematic inefficiency that exercisers of judgement could use to achieve excess risk-corrected returns". In other words, nobody has yet succeeded in disproving the basic thrust of the efficient markets concept.

Note that Samuelson's argument now, as before, is not that it is impossible for some professional fund managers to achieve excess returns. He explicitly allows the possibility that a handful of exceptionally talented individuals, such as Warren Buffett, can outperform the market on a consistent basis, though most will not, partly because they tend to trade too much and fritter away some of their gains in high dealing costs.

His case has always been that such exceptional investing talents will normally use their talent to make money for themselves, rather than rent themselves out to others; and that it is difficult and/or expensive for ordinary investors to take advantage of such talents even if they are able to identify them.

One thing that tends to happen, for example, is that most of the excess returns that the good managers achieve are simply absorbed by the higher fees that the managers with the best records are able to charge. This is one reason why the unit trusts with the best performance records tend also to be the ones with the highest initial charges and often the highest annual charges as well. As an ordinary investor, you pay for what you get, but what you get, quite often, once you tot up the costs, is not worth paying for, compared with what you get from indexing.