Usually performance is measured by some means of profit vs risk. Loss clustering is a very viable risk. Think of a strategy that has 100 winning trades and 50 losing trades, each 1% of account. Then think how the curve looks like if the losses are ...

Systematic trading

Kelly is far from the best ratio in trading. Experiment with a bit and you'll see why. You will get very volatile equity curves because kelly doesn't take ls clusterig (that you might also call DD) into account at all.

Systematic trading

Analyzing — So.. I started to analyze in-sample vs out-sample data. This is the first version of anything but there are already some interesting data here. What I have basically done is I sorted all strategies by in-sample fitness which is ...

Systematic trading

My recommendation: Look for higher timeframes. And read some Osler.

Systematic trading

Very nice curve, it's worth plotting two curves, one with logarithmic scale (my curves are logarithmic) and one with money management applied. With loaritchmic scale it's much easier to see how the system will work on different conditions although ...

Systematic trading

I have said this many times but out-sample is *the future* - it cannot be uses as any part of system. out-sample should be treated as future data, it cannot be used for any kind of validation, it's the same as testing strategy with real data in the ...

Systematic trading

No. Out-sample is the future. It is never used at any part of fitness or in any other way except for visual validation (that I do myself). edit: I also want to clarify that I have no interest in disclosing everything in my framework.

Systematic trading

Yes but only with in-sample data. Out-sample is the future, you can't use it on any part of any process if you want it to represent the future as it should.

Systematic trading

Yes it can handle multiple positions. I don't see why this would have anything to do with using out-sample as real out-sample.. Here is one run with about 8 hours of time from point zero (I started it after I switched back to original fitness algo). ...

Systematic trading

Here are two examples, I can repeat both my changing some aspects of fitness algorithm and running the program for few hours (the strategies that are combined to form these curves are the same on both examples). The curves will not be totally the ...

Systematic trading

Yes, at the final step I reset and repeat and see if the outsample is similar on all runs. I might post some pics later. I only do this to see if the outsample (the future) is similar to in-sample (the past/training period). It never is as good but ...

Systematic trading

If you do this, your out-sample becomes in-sample. By rejecting something based on out-sample you are using it ("the future") as part of your fitness calculation. If you do that, you just have two in-sample periods with different fitness algos. One ...

Systematic trading

The point of out-sample is that you don't have to believe anything, you can see how your system would have fared in the future. If you are finding systems yourself it's unfortunately very hard to use out-sample correctly. About curve fitting, there ...

Systematic trading

Is this in- or out-sample?

Systematic trading

Unless you already are using it, I would recommend adding an additional layer on top of your predictor, one that measures at what kind of market state the result is accurate. Market has some states that can be modeled and NN tends to work in one ...

Systematic trading

As I already said, I don't do that and yes, that would make out-sample in-sample.

Systematic trading

Simply said, I use out-sample for verification of method itself, not as part of the method. I don't think I can explain it any clearer than I have.

Systematic trading

Big problem with multi-instrument testing is that volatility tends to cluster across the instrument scope.

Systematic trading

Yes, that is correct if you do it as you state above. However I do it a bit differently, I use the out-sample to verify that my fitness algo provides me results that work both in-sample and out-sample. I use the out-sample to make sure that every ...

Systematic trading

No, that is incorrect. If you use out-sample in any way in your analysis, it becomes in-sample. It should be used as a verification tool of the whole system so that you will see how it behaves "in the future" that is out-sample. Real-time data works ...

Systematic trading