Past performance is not indication of future results.

We cannot have it both ways. On the one end we talk about reversion to the mean. Then on the other end we talk of performance chasers.

People shouldn't chase performance of Bill Miller. So people shouldn't buy LMVTX. Then, people buy LMVTX and "we told you so". Then, but wait, don't lose the faith in Bill Miller because he will come back.

We say don't look at performance or just the start rating. We cite reversion to the mean. Then we say Glinsman is good - just look at his past record. Why? Shouldn't Glinsman have been sold just like Miller should have been sold? Why are we to keep faith in fund managers when they falter, but not sell their fund when they have done too well? Why is it that we are not supposed to buy funds like CGMFX but we never say sell LMVTX or the fund in question?

What about funds like SCMLX and SCMVX? Incredible returns that's why we buy. But wait, let's be selective. Let US decide which managers we like. For THOSE managers let's look at the M* rating and recommend that we buy. For others we don't like, let's not look at the star rating. Then when funds we recommend do badly "keep the faith", but funds we don't like are "risky", and we tell them that. And if the funds we don't like do badly, then "we told you so".

A manager does stop working. Persistence of performance is ONLY persistent until the fund gets noticed. Then CRASH! BANG! Happenned to Miller. Happenned to Schneider. Happenned to Muhlenkamp. Happenned to Nygren. Happenned to NFJ. No wait, did not happen to NFJ. THAT is why Oppenheimer is out and NFJ is in. It is really that simple. No need to do extra analysis.

Nobody can predict the future. No one can say OCC is better/worse than NFJ. Changing manager indicates to shareholders that management is trying to do something constructive besides simply "keeping the faith". If the OCC manager has a stellar record, so does NFJ. And guess who's held up better? OCC or NFJ?