A.I.M. Users Bulletin Board

[Adam]

Hi Toof, Yes I can see those problems with such filters as well. That's why I asked if anyone is using such methods successfully.

I'm skeptical about backtesting for two reasons:
1. past performance often does not reflect the future. In fact I read of studies on mutual funds that found that Morningstar high star funds don't do as well in the future, and that the star rating has no predictive value.

2. backtesting is like an unblinded medical study, in that we know the result and we can tweek, perhaps subconciously, the timing or settings to fit the chart.

As you say, such filters can reduce the trades and may require very volatile stocks to work, as well as reduce of the performance of AIM with other stocks.

My solution to this problem of deep-diving stocks, is to AIM ETFs. With ETFs there's no need for filters. Unless you're doing very specialized ETFs they're unlikely to make big drops and stay there. Of course the downside is the loss of volatility, which points out that everything has its price. Just pick your poison.