I don't know too. Perhaps it sounds better to quantify the future when presenting to yourself/wife/boss/capital allocators/investors or whoever that maybe ... "Using monte carlo simulations, there is a 90% chance that our portfolio returns might fall between -10% to 20% for the next 12 months .... assuming no material changes to market environment etc etc." This beats saying .... "I honestly have no clue what's going to happen next year, for sure things are going to change but in the long run we sure are going to make some serious money."
Jokes aside, apart from monte carlo understating the severity of drawdowns, with a perfect smooth equity curve, random scrambling of trades or equity curve would be unable to detect flawed assumptions in the original model (e.g breakouts based on lunar cycles), or soon to change market environments (low vix in 2007). All this does nothing to address the quality of inputs in the trading model to give the 'confidence' of going forward. The number of people using a particular framework should have little bearing on your work but most important what your conclusions from your research lead you too developing your own risk framework.
The following 3 users say Thank You to Ming80 for this post:
Learn C# first or manually find a decent system first
As most of you will know I have a model which I have deceloped manually in excel, but since it only has N= 55 trades, I am not too confident in its long term potential (and it has large DD ~ 90%).
So now my next decision is whether I should
1) Continue to manually develop this model in excel and after 10 more trades try to wing it live
2) Learn the basics of C# and convert this model into NInjascript and take it from there ( I already have two more trading ideas in the "R&D" department of my factory that I may put into the prototype phase).
These ideas could be accomplished in a one month timeframe.
Just wondering what some more experienced members like @kevinkdog @treydog999 would advise me to do.
Mathematically win rate does not matter, only expectancy (average trade value). Psychologically win rate does matter, especially if you are trading only 1 or 2 systems. This is because trading something with a 30% win rate while profitable with positive expectancy you may not be able to psychologically hold out a drawdown for 6 months + even though that is within probable bounds. However if you have a basket of 20 systems, then the psychological factor of that single 30% win rate system is buffered by the rest. even if you had 5 of the systems with that kind of win rate, your draw downs would not be as magnified since they would be able to help counter act each other.
The following 4 users say Thank You to treydog999 for this post:
No, if you are using a low level language. But if you are using something like ninjascript or easylanguage there are going to be limitations due to functions that are lacking or made for ease of use. But if you are programming from the ground up, you can pretty much make anything.
The following user says Thank You to treydog999 for this post:
Just like any craftsmen, you have to learn and develop your skills. So yes, you will not be able to make extremely complex programs at the start. But as you learn and grow more advanced things will be within your grasp.