According to some intriguing research from the London Business School’s stock market number crunchers, a favorite nugget of investment wisdom may have to be abandoned. The late Sir James Goldsmith said that “when you see a bandwagon, it’s too late.”

He meant that by the time an investment story is in the news, the smart money is already riding the wave. Only fools join the party at this late stage.

Now the LBS’s Elroy Dimson, Paul Marsh, and Mike Staunton have turned this conventional wisdom on its head with an impressively exhaustive study of share price movements between 1900 and 2007.

It shows that simply buying last year’s winners can be a reliable way of beating the market.

Their study, published in ABN Amro’s Global Investment Returns Yearbook, confirms that momentum works.

Quite obviously, at big market turning points like those in 1975, 2000, and 2003, it can also be a recipe for losing money. But on average, over the very long haul, it pays to go with the flow.

“Momentum, or the tendency for stock returns to trend in the same direction, is a major puzzle,” the LBS three said.

“In well-functioning markets, it should not be possible to make money from the naive strategy of simply buying winners and selling losers. Yet there is extensive evidence that momentum profits have been large and pervasive.”

The numbers certainly back up the claim. In one of LBS’s studies, which analyzed all fully-listed stocks between 1955 and 2007, the shares which had outperformed the market most in the previous 12 months went on to generate an annualized return of 18.3% while the market’s worst laggards rose by 6.8% on average. Over that period the market as a whole rose by 13.5% a year.

Arguably, those figures, impressive as they are, are conservative. That’s because the portfolios created were weighted by company size (like FTSE’s indices). Using an equally weighted portfolio in which smaller companies have the same impact as bigger ones resulted in a 25.6% rise for last year’s winners and a 12.2% rise for the losers.