In this interview he describes the method 11:40 http://video.google.com/videoplay?d...O4OYrALk2rAt&q=charlie+rose+soros&vt=lf&emb=1 If he's got a profit on the year he goes for it(trades bigger) and risks his profit trying to hit a big year. If he is not doing as well or losing in the year he cuts back. It looks to me that is a bit irrational to use this method based on the fallacy to play with the 'markets money', now if he is doing this because its less stressful(because losing the principal would be too painful or because he's got a low drawdown tolerance ) or the clients complain less then he is right of course
I confess I didn't watch the video. I don't like his political views. However, reading your explaination of his method, it does make sense to me. If his system/strategy is working he goes in heavier, when it doesn't seem to be working he lightens up. What's wrong with that? It's basically playing the equity curve!
this means simply to always press your winners and cut size when you're doing poorly..also you want to be bigger when you're right and 'going for the jugular' when your are really right.
Suppose January my account equity is $ 10,000 and by September my equity is $ 100,000. I might not trade the same position size if account equity increases. I might use a risk budget that is a percentage of equity, say 1 % of equity. So if I experience a winning streak then my positions are bigger, and if I experience a losing streak then my positions are smaller.
I'm just saying its ilogical to use this method(have nothing against anti-martingale). the 'principal' he refers to has last year's(and previous years) profits so in theory he should have no problem losing that on a GBP type trade. but he does, on the day 365 of the year he wants to risk all the year profit, few hours later it becames 'too painful' to lose it and it needs to be protected. its seems to be mainly a psycological method(rooted on the market's money falacy)rather than a mathematically sound one