Newly published research by the University of Exeter’s Centre for Finance and Investment analysed the performance of UK equity performance in 2,175 segregated UK pension funds between 1983 and 1997. Author Professor Tonks found weak evidence of persistence of outperformance from one year to the next, based on averages of the top quintile (some 400 funds). On the strength of it, he suggests inter alia the FSA was wrong to hold that past performance is of no practical purpose in selecting active managers.

Apart from the fact the returns on what may be only part of a multi-asset pension fund portfolio are not as accurate as daily valued unit trusts, the findings even at face value do not support any practical strategy. Statistical evidence derived from large populations does not translate easily into a world where investors need to select a much smaller group of managers, or maybe only one manager, not 400. Can the expected outperformance justify paying higher management charges? Can managers be switched, as active selection implies, without incurring very high frictional costs? Not every year, surely, Professor? There may be ways for investors with a healthy scepticism about performance persistence to play the active management game, at least with some of their money, but this is the daftest I’ve seen yet.

TAGS: Active management, performance