Q: I run a quant based fund that uses AI techniques as a forecasting tool which in turn filters through a classic Modern Portfolio Theory model in a market neutral environment. The model forecast 1 week into the future and the portfolio weights are rebalanced every week.

Initially we always felt that we wouldn’t use a stop loss as the rebalancing each week would in effect be our stop loss. We are now however rethinking this philosophy and wanted to know if you had done much work on stop losses for portfolios and more specifically for portfolios run in a market neutral environment.

A: You haven’t told me whether you run a hedge fund or a mutual fund because there is a big difference. The latter has to be invested. If you are the latter then you use some of the ideas that were covered in my Interview with A Portfolio Manager – a back issue of Market Mastery.

Now I’m assuming by market neutral you basically mean that you have a short position for each long position? If that’s the case, then you probably don’t need stops. However, I don’t know the sort of things you find yourself dealing with. If I had a better idea of those, then I’d probably have a lot of suggestions for you.

For example, one of my supertraders ran a fund in London that was full of risk-adjusted positions. He really didn’t understand “Tharp Think” too well and I really didn’t understand his thinking. However, once I understood what he was doing and got him around to “Tharp Think” he made some very important changes to his portfolio and last year had his best year ever.