No announcement yet.

Improvements in Drawdown Statistics

  • Filter
  • Time
  • Show
Clear All
new posts

  • Improvements in Drawdown Statistics

    Hi guys,

    I crushed some numbers on the FMAM Services Forum thread about his real drawdown. Myfxbook and Simpletrader statistics show less than 6% Peak Drawdown.

    When I did the math, the real drawdown was over 40%.

    Do the ST statistics copy the myfxbook data? If no,
    How do SimpleTrader statistics differ from myfxbook?

    We give you access to best FX Signals submitted by Professional Traders delivered inside Free Live Trading Room and through Automated MT4 Trade Copier.

  • #2
    They calculate the dd using the close price of the day when looking at old trades in the past

    Sent from my Nexus 6 using Tapatalk


    • #3
      Hey, exactly as Okda says.

      In an ideal world, it would be good to try and take each trade and simulate it, but in reality its simply to intensive, and you'd need an insane amount of tick data.

      So we take each trade, and look at price at the end of each day and simulate what the DD would be at that time.


      • #4
        Thank you both. I am assuming that this calculation is only on New data ( once the ST-Sender is attached to the account). How is the historical data handled?

        Instead of using tick data, can we use the OHLC of the daily to cut down the needed resources to give a more accurate drawdown calculation?

        My example is a cut and paste from the FMAM discussion.

        Account Balance: 15,500 USD
        Sell EURUSD Lotsize 1.0
        Opened Feb 1, 2016 13:10
        Closed Feb 29, 2016 14:26
        Opened 1.08842
        Closed 1.87440
        Peak drawdown was on Feb 11,2016 when EURUSD reached the High of Day at 1.1376
        Actual pip drawdown of 492 PIPS


        In order to properly calculate any system's RR value, we need accurate results.


        • #5
          This is only for old data as far as I know, for the new data it is updated I think with every update sent from the sender ea every 30 seconds, am I right?

          Sent from my Nexus 6 using Tapatalk


          • #6
            I didn't realise just what a sham the stats are. So you might as well not trust them, unless you go through yourself trade by trade,

            Hedge is there a methology you go through that shows these difference easily or is it a trade by trade situ.

            Hearing this makes me wonder if what I thought signed upto last month is actually legit.

            Sent from my SM-N910F using Tapatalk


            • #7
              Using daily OHLC isn't really feasible either, because we can't assume that a trade was held throughout an entire day, through the high or low. To make that assumption is worse than using mid night close (As most sites do). As trades may be open for minutes at a time, the only real way to accurately work out DD historically would be to use historical minute data. Which would be hard because you also have brokers who use different symbols names - so again matching them may not be that possible.

              To say the stats are a sham is a bit of an over reaction - We have tried to keep everything in line with Myfxbook because this is the industry standard that is used.

              I'm sorry if that's not good enough for some people, but we can only do what we can do with historical data. In terms of the data moving forward - we do indeed store the maximum equity DD an account goes into and pass that data to our server when it heartbeats (which is on a 30 second interval).


              • #8
                I think the analysis that HedgeBitcoin has provided is incredibly insightful, and shows the limitation of drawdown statistics as presented here and on myfxbook. While I wouldn't say the stats presented here a 'sham', they have clear limitations and at the very least these limitations should be clearly described (i.e. 'the drawdown metrics presented here represent end-of-day drawdown only, and not the drawdown that was experienced across the lifetime of each open trade').
                Last edited by philtre71; 08-22-2016, 06:40 PM.


                • #9
                  Can it be done in a way, that your server handle one signal per day and calculate it's true dd using minute chart?
                  A single signal per day is not that much and not all signal has several thousands of trades and most of them doesn't go back more than a year so most of the signals won't need much work to be done
                  And this will put you on the top of the market as the only site that can't be cheated to hide dd of past trades
                  Also in the signal page there could be an indicator that show software the past trades were already checked using the minute chart or still in queue

                  This could be a manual script run by one of your admins on each signal in the market place and when all signals are done, will be ran on new signals only and also one by one

                  Sent from my Nexus 6 using Tapatalk