In essence, the single manager's strategy appears unchanged - both to him and to the client.


But the strategies of all the individual clients have changed, when they have allocated their money to him.


It's basically a misapplication of back-testing involving game theory. They think they are repeating, when in reality a great deal of learning on the fly and adaptation is taking place! They ARE doing something different.


Contrarian is figuring out what people have learned, and doing the opposite.


It's not because they are big that blows them out, being big is just a symptom of fluid discovery and concentration. It's because people have happened upon a way to pick what strategy other people are using.


It is worth noting that, if you bought at-the-market every day at 1:00, and sold at-the-market every day at 2:00 - and did this as reliably as the sun will shine every day for years and years - your strategy would be nearly immune to losses. Market particpants would at first be surprised, then compete to fade your orders non-redundantly, and eventually evolve to absorb and offset them perfectly.


So there is hardly anything "new" that works, whereas older things are safe by nature. If I reallocate some of my money from A to B, B has to allocate it to doing what A was doing anyway, or the whole dance falls out of whack. It's all evolved and interlocked.


elroy