Am I wasting my time backtesting my strat?

Discussion in 'Strategy Building' started by DROBBY, May 6, 2020.

  1. DROBBY

    DROBBY

    Hey no problem man, this ain't some secret sauce 100 000 lines of code algo I spent 10 year developing. Its just the simplest trend following strat. I find it work best on the 4h personally, test it out and make sure you get the system and that it fits your style.
     
    #31     May 6, 2020
  2. DROBBY

    DROBBY

    Nah, I don't do fundamentals. But I know a guy who only trade the most obscure shit coins using fundamentals only. Hes making bank but hes some kind of savant lol. Everyone else doing that since 2018 is getting fragged. But he just glide on by making bank on most of his trade.

    So yeah its def possible, but you have to do your own research on this one, I really don't know much.
     
    #32     May 6, 2020
  3. easymon1

    easymon1

    no thank you please.
     
    #33     May 6, 2020
  4. SunTrader

    SunTrader

    LIB, no not liberal but look inside bar testing.

    Otherwise all kinds of assumptions (mostly bad) are made about where and at what price a fill happened historically.

    Even then I personally only trust forward testing - i.e. money on the line in actual trades.
     
    #34     May 6, 2020
  5. Sekiyo

    Sekiyo

    Backtesting is to gather evidence it doesn’t work.
    Not to gather evidence that it does work.
     
    #35     May 6, 2020
  6. Hi Drobby,

    If you are getting a 50% win rate with 2.5-to-1 RR, which is at the low end of the performance numbers you mentioned, then this corresponds to an approximate overall profit factor (gross gains from the winning trades divided by gross losses from the losing trades) of 2.5.

    My understanding is that consistently profitable traders who can achieve profit factors over 2.0 are extraordinarily rare. Some famous traders/authors have said that if one gets a PF over 2.0 in backtesting, then it's likely there's something amiss with the tests.

    BTC is an unusually volatile market and maybe your numbers are right, but best to double-check and probably also to test additional time periods.

    Good luck to you.
     
    #36     May 6, 2020
  7. DROBBY

    DROBBY

    Yeah we will see. At the end of the day that's just numbers on an excel spread sheet. Crypto is unique in the sense that it trends all the time. I was shocked to see quarter long chop c on some forex and indices chart when I first looked at them few weeks ago. That might explain why I could get such an amazing projected ROI with a simple trend following strat.

    Also bear in mind that as far as indicies/bond/forex is concerned my strat is gonna be on high TF (8h or 12h, not sure yet, test will bring clarity). So it is much easier to get a 2.5 to 1 R/R. If I was day trading and making 10 trades a day then it would be another story. Some very successful day traders make about the same amount of trade in a good day than I will do in a month. If they risk 1% of capital on each trade with a strike rate of 50% and 2 to 1 R/R, then theyd be making such return it would be insane.

    Assuming 10 trades a day, 200 trades a month, with these stats, thats 165% ROI a month. Safe to say we might not be comparing apples to apples.
     
    #37     May 6, 2020
  8. MichalTr

    MichalTr

    11 was just example. No special meaning. To check if you know anything about probabilities. So the good news is - you know something.

    "I just know I can potentially make a very good ROI based on a calculated risk." - and here is a bad news - nope, you can't know that unless you test your strategy in action. Like I said before (twice) - backtests mean nothing. Sorry.

    Of course I wish you that. That would be nice achievement. And as crypto markets have less pro participants, maybe that would be even easy. But I thought that you want to use it in "normal" markets.

    "As you probably know the order in which you multiply stuff does not affect the end result." - that one was funny. Yea, I know math a bit ;)
     
    #38     May 6, 2020
  9. MichalTr

    MichalTr

    Ok, here comes the magic part... You should also make excel spreadsheet with calculation of your future millions... Saw it too many times.
     
    #39     May 6, 2020
  10. DROBBY

    DROBBY


    Ho yeah thats for sure, time will tell.
     
    #40     May 6, 2020