I am again I fell for the so-called missing wheel, so that the performance of the breakout system represented quite interesting:
Very impressive, I think, but to question it critically. The time frame of the system is M5 and here again beat all the pitfalls of NT in the back to test simultaneously. Many trades look wonderful and run for several bars, that is longer than 5 minutes. Other trades again begun within the first 5 minutes after the trade was entered directly into the stop loss or profit target in the pure ....
... hmmm .... that should give us food for thought. This is to SameBarExits in the live trading can lead to bad consequences. Back in the test, I have only 4 values, Open, High, Low, Close Chart and also only in 5 minutes. How am I supposed to know which value of 5 minutes bars was the first to reach .... the high of the bar or the low of the bar ... exactly ... first you can not not ...
After I once again tried to come up with a multi-time frame strategy in more detail the possible error, were (of course) other errors or inconsistencies. The calculated rank did problems at once, the ATR-exits, and many others caused quite ugly code and quite ugly, not partly understandable, results.
have Ultimately I solved it in the simple manner and easy to count the SameBarExits. The output after the baking test looks for example like this:
have Ultimately I solved it in the simple manner and easy to count the SameBarExits. The output after the baking test looks for example like this:
DT_OpenRange_20 - CL 02-11 - Exits on the same bar: 143/253 (56.52%)
DT_OpenRange_20 - CL 02-11 - which winners ;: 140/143 (97.9%)
DT_OpenRange_20 - CL 02-11 - Of Losers : 3 / 143 (2.1%)
Results: It was even worse than feared. Almost 50% of trades were SameBarExits and almost all of them as late winner ... that means ... this should be checked manually, resulting in a horrendous and for me not worth time.
way out of this situation is for me to put the profit target is correspondingly higher, so that the exits on the same bar on acceptable 10% - 15% decline. Only then, can I trust the strategy in live trading.
Best regards
darth trader
0 comments:
Post a Comment