Robustness testing SQ X question
4件の返信
Hey guys, I wanted to ask a question regarding the robustness testing with SQ X.
– I’ve chosen MC retest methods 5 tests with 100 simulations (Except the parameters change test which I do independently afterwards).
– I’ve chosen the following filters, the 2 default ones and the ebook recommended one:

All my strategies (Which nicely passed 15 years of real tick data according to my main filters previously used) failed this one and I wanted to make sure I’m doing this right.
That’s a random example of one of these strategies MC:

As you can see it miserably fails the 95%/original Ret/DD test with 1.84 to 10.69.
My questions are:
1) 5 tests with 100 simulations means 100 simulations, in which every of this 5 tests parameters is being randomised so 100 per strategy, or is it 100 simulations per test, so 500 per strategy on each run?
2) Is it the correct way to look on the MC tests? running those 5 tests together, or should I run one by one and eliminate strategies by the above rule?
3) Do you strictly discard strategies that don’t live up to the 50% ret/DD at 95%? even if the ret/dd is positive at 100% and you like all other features of the strategy? (low stagnation, good PF, high win % etc)
Thanks.. I’m having trouble with SQ robustness, I feel like I’m discarding beautiful strategies sometimes, but I don’t want to hold on to any of them just out of hope 🙂
0
What are your cross check settings? Defaults are not matching ebook ones. Double check it
0
What are your cross check settings? Defaults are not matching ebook ones. Double check it
Those are the settings:

Yes they are different, but then I’m not sure how to filter results :/
イリヤ
0
my opinion is this – you cant make more than 1 MC test at a time, you need to do them one by one
利益を上げるアルゴトレーダーになりたいですか?私たちは2014年序盤からStrateQuantソフトウェアを使い始めました。現在では、あらゆるタイプの市場に対応するEA構築に関する膨大なノウハウを持っています。私たちはこのノウハウ、アプリ、ツール、そしてすべての最終的な戦略をリアル・トレーダーの皆様と共有しています。参加したい方は、以下にご記入ください フォーム.
0
4件の返信を表示中 - 1 - 4件目 (全4件中)