Announcement

Collapse
No announcement yet.

Partner 728x90

Collapse

hft and backtest

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    hft and backtest

    Hello everyone, I have attached a picture with the problem. During the backtest when there are discharges hft the entrance of the trade is too early, or on the first candle from which the hft. is there any way to avoid this? I tried pace_of_tape paris, but also eliminates other inputs that are valid.
    for example in this case the algorithm was to make a single input, not treoltre to the problem delĺingresso anticipated.
    Does anyone have any idea how to fix this problem?
    thanks for the answers
    Max
    Attached Files

    #2
    Hi Max, I'm not sure I follow your request - do you look for alternative ideas to incorporate further filters in your strategy?

    Comment


      #3
      hi bertrand and thank's for the answer.
      yeah,it is exact. a filter that removes these discharges hft that does not allow you to have reliable results.

      Comment


        #4
        What would HFT discharges then be? The spike you saw on your chart?

        Comment


          #5
          too much data in a limited time, less than one second.
          in the graph set in renko 10 are created over two candles in a second. for me this is considered hft

          Comment


            #6
            Looks to be classic news related perhaps to me? 1600 time...

            You could not predict those events but be sensitive when they would be most likely seen for your market based on observation and then choose to stand aside or use less size.

            Comment


              #7
              I was interested in the backtest .. I know that in reality it is not possible, but maybe we could act in the backtest ..

              Comment


                #8
                Sure, for backtesting an example here would be to include a time filter in your conditions, to avoid for example triggering it during classic news times or lunch hours.

                Comment


                  #9
                  sorry I did not mean this, but for example if the bar now has the same instant as the previous one, you should not run inputs ..

                  Comment


                    #10
                    You could also compare the timestamps of the bars to ensure they would not be equal or a min range spaced apart. The access would be given by Time[x]

                    Comment


                      #11
                      yeah good idea.
                      thank you for the assistance

                      Comment

                      Latest Posts

                      Collapse

                      Topics Statistics Last Post
                      Started by argusthome, 03-08-2026, 10:06 AM
                      0 responses
                      88 views
                      0 likes
                      Last Post argusthome  
                      Started by NabilKhattabi, 03-06-2026, 11:18 AM
                      0 responses
                      48 views
                      0 likes
                      Last Post NabilKhattabi  
                      Started by Deep42, 03-06-2026, 12:28 AM
                      0 responses
                      30 views
                      0 likes
                      Last Post Deep42
                      by Deep42
                       
                      Started by TheRealMorford, 03-05-2026, 06:15 PM
                      0 responses
                      33 views
                      0 likes
                      Last Post TheRealMorford  
                      Started by Mindset, 02-28-2026, 06:16 AM
                      0 responses
                      68 views
                      0 likes
                      Last Post Mindset
                      by Mindset
                       
                      Working...
                      X