He began from the premise that if professional investors are better than retail investors at interpreting publicly available information, then they should be as likely to beat the market by investing in stocks based a long way away from them as they are when they invest in more local stocks. After all, the public information released by, say, a California-based company is equally available to a California-based investor as a Connecticut-based one. The two are on a level playing field, so skill at interpreting information should lead to better performance by either.
But there is no such outperformance. Mr Sulaeman studied fund managers' trades between 1999 and 2010 and found that when they bought and sold shares in companies based outside their home state they did no better than the market.
However, when they traded shares in companies based in their home state, there was outperformance. It's local knowledge that generates good returns, not general investment skill.
You might think this is because fund managers can exploit scuttlebutt information about local companies. They hear rumours about how well or badly things are going in the head office of a company based down the road, and can invest profitably on this.
But this doesn't explain their success. If they use this sort of information, their outperformance should be persistent. But it's not. Mr Sulaeman found that the local stocks that fund managers bought outperformed their sells only in the calendar quarter in which the trades happened, and no longer.
Instead, he found that the outperformance was concentrated around the time of earnings announcements, and especially around the time of bad earnings news. This suggests that fund managers' success isn't because of their intellectual ability but simply because they have local contracts who warn them of impending bad news. Mr Sulaeman says: "Institutions on average are not skilled and their superior intraquarter performance is more likely to reflect opportunistic access to short-term local information."
The old cynics might therefore have a point: what matters isn't what you know, but who you know. This is entirely consistent with one version of the efficient market hypothesis. If all publicly available information is already embedded into share prices, we can only make good risk-adjusted returns by exploiting private information.
Is the same true of the UK? It is almost impossible to say, simply because so many fund managers and companies are based in London, and so we don't have enough regional variation to replicate Mr Sulaeman's work. However, his work does suggest that fund managers who incur the huge costs of London office space might have a good rationale for doing so.
