|
Preface: I'm pretty new to trading and to this forum. Enjoying trading and this forum is excellent.
I have a few strategies that I backtested. Each one suggests trades on one or more stocks from a set, on a daily level (entries and exists are at market open and close only) with entries and exists defined and executed using software (but easy to manually override). This results in a handful trades per day per strategy. They seem to work in live trading -- at least for the last month or so. But I don't expect all of them to keep working forever.
My approach to things is pretty formulaic -- I do something if doing something with the same strategy (and strategy of choosing the strategy) has worked in the past. If trades were only noisy and independent, traditional statistics would do the trick. But clearly, markets are not simple noise, with various trades having unknown dependencies.
How do you figure out when a strategy stopped working? I was thinking to do a simple moving average, but when backtesting a rule like "stop trading when the moving average of trade results for this strategy is less than some number," this usually does worse than not stopping (at least for strategies that I selected because they have been profitable in the past). How do you think about exiting a STRATEGY rather than a trade? |
|
|
|
|
| Consider trend following each strategy separately, assigning weights to them and then dropping the lowest performers. TAA on the basket of strategies. |
|
|
