Five years ago this month, the market reached its highest levels since the dot-com boom. At that point, I began monitoring a portfolio of blue-chips chosen on the basis of share-price performance. The idea was to see if momentum worked in a declining market, and also to see if a short strategy (selling losers) was as effective as a long one (buying winners). It's not - but buying winners unequivocally works. The long portfolio is up 28 per cent compared to an 18 per cent fall in the index.