I would like to put a simple question, something that concerns me but it's probably quite simple and common on the day to day of sistematic trading, I suspect.
I've been designing a bunch of strategies. Most of them are real rubish from the very beginning, others don't support a quick backtest though looking good on its preliminary results, others come together with extremes drawdawn though beeing profitable most of the time (too risky for me). BUT I have a couple of them that give me good results, and I mean with good results: Good ratios: Maximum Drawdawn (Ok), Profit factor (Ok), return/DD (Ok), Walf Fordward Eff. (Ok). Great! Holy Grail! Eureka!
My concern is that those promising results are only presents for a single index. On other markets like FX, Commodities, even other indexes they simply and clearly fail no matter the parameters I use. So, what do you do with those strategies? Go ahead on that single market or just discard them because of lack of robustness?
I have the feeling to discard them despite all the sweat, blood and tears invested. I don't trust them if they just pass the exam in one index. But on that single market they are outstanding! Right now I'm passing and continue with more ideas, trying to cold down my mind and come back to those strategies later...
I'll appreciate your wise comments
The following user says Thank You to Enric66 for this post:
You have the opinion that a strategy must work in multiple markets to be considered "robust." Maybe you should ask yourself why that has to be true. Why does a strategy that works on the ES also have to work on the Euro currency? Or, does an ES strategy have to work on the TF or NQ - and if so, how good does it have to be?
Examine your mindset about this requirement you have - that a strategy should work on multiple markets to be robust. I'm not saying it is right or wrong. And as long as you are satisfied your requirements are based on sound logic, then definitely throw away strategies that don't meet your requirements.
I hope this helps!
If you have any questions please send me a Private Message or use the futures.io "Ask Me Anything" thread
The following 4 users say Thank You to kevinkdog for this post:
Sorry for stepping in with my opinion in Kevin's AMA, but if a strategy fails to perform similarly on a highly correlated instrument, then it's a sure sign of curve fitting and you shouldn't trade it.
Due to time constraints, please do not PM me if your question can be resolved or answered on the forum.
Need help? 1) Stop changing things. No new indicators, charts, or methods. Be consistent with what is in front of you first. 2) Start a journal and post to it daily with the trades you made to show your strengths and weaknesses. 3) Set goals for yourself to reach daily. Make them about how you trade, not how much money you make. 4) Accept responsibility for your actions. Stop looking elsewhere to explain away poor performance. 5) Where to start as a trader? Watch this webinar and read this thread for hundreds of questions and answers. 6) Help using the forum? Watch this video to learn general tips on using the site.
If you want to support our community, become an Elite Member.
The following 4 users say Thank You to Big Mike for this post:
well maybe true in many cases but I disagree on that if it is inteded as a matter of fact. ES and YM are highly correlated markets(just to make an example that fit) but that doesn't mean that a strategy that works well on ES works the same way on YM, yes it should, in many cases it does, but it's not an if-then equation tu run or not a strategy, in my opinion.
My opinion instead is that there are very few strategies that are worth trading their own and I still didn't find any. But there may be a lot of average strategies that put into a portfolio could give you a quite stable performance. So the goal should be building a portfolio and make it grow, not working to find the best and robust strategy that will make you money forever(I know we all agree on it, but worth repeating)
The following user says Thank You to montanajtt for this post:
If you make the excercice to curve fitting a graph wouldn't you indirectly curve fit its highly correlated as well?
I didn't expect high performance on all markets for a single strategy. But I'm shocked at seeing a very good result on a single one and miserable results on another ones. I would have liked to see good results on one and at worst flat results on the others. Why do I have this paradigm? Well probably I've read too many times about this concept of robustness. Anyway if it weren't true which would be the sense of using CCI out of Commodities, or MAs everywhere, etc?
I hope my comments don't exceed your patience. But if I may, let me put my question in other way: For a mechanical trader all the rules to define a strategy are absolutely accurate, no exceptions. Once the strategy is 'manufactured' there's a mechanical process to test it and this last process should have its accurate rules. Here's the question. On these rules do you consider the results in just one market or do you expect some minimum for other markets as well?