That's a good statement! I can look at the same chart for weeks at a time and continue to find new things despite having traded for decades. Bur I'll add, many struggling posters shoot themselves in the foot. They become defensive and argumentative when trying to seek information which shuts down the flow of information to them.
I don't care if Virtusa doesn't want to post anything, he is too cocky. I was not expecting much from this post anyway, since there are about 40 replies and no one cared to check the data. Mr broken record even said that there are only 50 trades and doesn't want to bet his house on it, when there are clearly over a hundred trades. Others argued about the input dataset, not a single one checked if the input criteria is correct, which is not, the algo does not enter on 4 consecutive candles, that was to check if anyone that replied actually checked it. By chance, and purely by chance we got that it is better to use a trend forecaster in order to exit and that should improve the algo. That is the only bit that we can salvage from this thread, the rest is garbage. Call me defensive again, but the level is too low
You are not considering circumstances, people from different cultures express themselves differently.
Hello Drawdown Addict, Well my friend. Are you an algo developer WannaBe, or are you an an Algo Developer Professional? If you knew all the answers to your questions, why did you even create to thread? Three ET members, are you challenging you on obtaining more trades, so what does that this means for your end goal? How will you get better, if you never get challenged? Get out of your feelings, and back test 20 years or more if you can, then post back. LOL, I can make 1000 algos look good using only 6 months of data or 50 trades. That is EASY. Why not use 1 month of back testing data? How about 4 weeks? It should be overwhelming common sense to the serious algo trader to use as much back test data as possible for a Robustful proven Edge. Then you can make a better bet, right?
I did tell you already that you should go ahead with your 20 years of data, that no one is stopping you, I don't see the point of continuing the argument. You are so sure that the more data you use the better is the algo, ok, so go ahead and use it. I don't see why you want to argue about it.
Hello Drawdown Addict, It is not about what I should do. My specific question to YOU is simply, Why are YOU using 6months of back testing vs 20 years of back testing data? What is your answer?
I did answer that already, 10 posts ago. I am just replying out of politeness but I am not going to repeat myself. Read the whole thread slowly and you will see why I use a small data set. It is not a big deal anyway.