Announcement

Collapse
No announcement yet.

Partner 728x90

Collapse

holy grail system?

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

    #31
    Originally posted by shooly76 View Post
    what about median renko bars? these are candlestick renko bars. I have included a screenshot of chart using median renko. I ran a backtest using median renko..and total net profit was over $55,000 more including 1 tick for slippage.

    (median renko) the entry price looks correct to me..this is just the generic chart from backtest results..the chart I use for live viewing has my custom indicators on it.

    also I was viewing a chart today in sim mode (default renko)..and system traded exactly where and when I wanted (I use heiken ashi2 renko on my chart). now I will switch to heiken ashi2 median renko.
    • The shipping Renko bars are a total fiction and useless for backtesting. The reasons have been discussed multiple times in these fora. Try a forum search on "Renko" threads which have a response from me. There should be some interesting stuff in those, as at today, 32 threads. Here is one as a starter: http://www.ninjatrader.com/support/f...ad.php?t=68224
    • Any Renko BarType that shows the wicks (and hence possible price action) can be backtested provided one takes appropriate action to compensate for the offset of the open of bars, from the close of the previous bar. That typically is then a slippage of "half the brickSize + 1" tick. Anything less is also a fiction, because it will make almost all entries have positive slippage with respect to entries. While this positive slippage will be true in some cases, in others it will not.
    • Of course, one can always use a MTF setup and use Signals off the (overlapping) Renko bars, and entry on the secondary time frame (best would be 1-range or 1-tick).
    • Heiken-Ashi based on any Renko represents 2 levels of massage of the primary price data. Not bloody likely to be of any use.

    A word of advice. In Backtest/Optimization, one should always be a cynical, skeptical, conspiracy-theory nut: any overly positive results must be viewed with extreme prejudice, and the (often) mistaken assumptions that may have been overlooked should be found and routed out.

    I can tell you some very interesting war stories about our "pie in the sky" dreams when we first tested already successful strategies against various Renko bars.
    Last edited by koganam; 10-30-2014, 08:19 AM.

    Comment


      #32
      Im starting to understand now.. where do I get BetterRenko bars?

      Comment


        #33
        Originally posted by shooly76 View Post
        Im starting to understand now.. where do I get BetterRenko bars?
        My source is here:

        https://www.bigmiketrading.com/downl...load.html?view

        However, elite membership is required to download bar types and indicators.

        Comment


          #34
          thanks..

          I have just started the month of october in market replay on the current automated strategy.

          I am using median renko bars for market replay..just to see results based on my innaccuate backtesting/monte carlo..they were backtests afterall..market replay should give me very close results to actual live performance..correct?

          that being said..are the median renko bars ok for market replay?

          Comment


            #35
            market replay- TF 12-14 10-01-14

            very nice first day..135 ticks

            I will report more after I finish october testing..I did change some settings and parameters..most importantly.. PT and SL levels..

            Comment


              #36
              All bars should be ok for market replay. The issue with custom Renko bars is the fake open which is usually the bottom of the body in an up bar or the top of the body in a down bar. If that part of a bar is used in the strategy I would think the results will be inaccurate.
              eDanny
              NinjaTrader Ecosystem Vendor - Integrity Traders

              Comment


                #37
                total net profit was(-$4770) for october market replay. in one day it lost $2830..very flat/choppy day. so clearly I have a ton of work to do. but I can see that the days it takes 100+ trades..it does badly. the days w fewer trades, the system performed better.

                there are a few things to improve:

                avoid super-flat market days. I will first cut the trading day in half..and maybe increase renko ticks..maybe try some different entry triggers and adjust SL and PT levels.

                Comment


                  #38
                  Originally posted by shooly76 View Post
                  thanks..

                  I have just started the month of october in market replay on the current automated strategy.

                  I am using median renko bars for market replay..just to see results based on my innaccuate backtesting/monte carlo..they were backtests afterall..market replay should give me very close results to actual live performance..correct?

                  that being said..are the median renko bars ok for market replay?
                  I think that for a median renko chart market replay will give you results that are closer to reality than a backtest.

                  However, you cannot simulate slippage with the market replay.

                  Comment


                    #39
                    I did include 1 tick for slippage in market replay

                    Comment


                      #40
                      Originally posted by shooly76 View Post
                      I did include 1 tick for slippage in market replay
                      Means nothing. Market Replay is a replay of the market, so what you specify for slippage means nothing; the fill engine does not take it into account. Unless of course, you are actually counting the slippage when you calculate where to place your orders. Orders will be filled per where you place them.

                      Comment


                        #41
                        The shipping Renko bars are a total fiction and useless for backtesting. The reasons have been discussed multiple times in these fora. Try a forum search on "Renko" threads which have a response from me. There should be some interesting stuff in those, as at today, 32 threads. Here is one as a starter: http://www.ninjatrader.com/support/f...ad.php?t=68224
                        Any Renko BarType that shows the wicks (and hence possible price action) can be backtested provided one takes appropriate action to compensate for the offset of the open of bars, from the close of the previous bar. That typically is then a slippage of "half the brickSize + 1" tick. Anything less is also a fiction, because it will make almost all entries have positive slippage with respect to entries. While this positive slippage will be true in some cases, in others it will not.
                        This is useful to know.. I'm using a traditional Renko system (with no wicks and the strange "missing" brick at reversal) and using Stop orders to enter and exit at the reversals instead of waiting for the reversal brick to fully form e.g.

                        if (Close[0] > Open[0]) EnterShortStop(DefaultQuantity, Open[0], "Short-Stop");
                        if (Close[0] < Open[0]) ExitShortStop(Open[0], "SL-reversal", "");

                        Wanted to clarify a few things -
                        1) Since I'm not using Market orders at the Open, I get around the whole issue of the missing brick and the phantom open, correct?
                        2) During backtest simulations (minute-level data, not tick-level), Ninjatrader actually looks at the minute level prices to ensure fill at the Stop levels are possible, and doesn't blindly fill at the phantom Open tick, correct?
                        3) In live auto-trading, how do i write code in my strategy to monitor the prices "intra-bar" and place a Simulated SS Stop, so the Stop is submitted only when the Stop level is reached intra-bar? Note that because the Renko doesn't draw the missing reversal bar, we have to monitor the price crossing the reversal point and insert a Stop order intra-bar.
                        - I don't want to leave a naked Stop order at all bars because i'm operating at very small Renko brick sizes of 2-3 ticks.
                        - Note that I also don't want to de-select "Calculate on bar close" because it will evaluate the entire system tick-by-tick, and i want the Renko bricks to abstract the noise during trend bars and not get stopped out.

                        Comment


                          #42
                          Hello kbellare,

                          Thank you for your post.

                          1. I do not fully understand you meaning behind the Market orders at Open, are you referring to historical fills or real-time fills?

                          2. In Live testing (that is on real-time data), the Renko bars are moving tick-by-tick as they are based on tick data. I recommend you review the details at the following link for Renko bars: http://ninjatrader.com/support/helpG.../bar_types.htm

                          3. In this case, you could set up the system using CalculateOnBarClose = False to monitor price action, but use FirstTickOFbar to calculate conditions for order submission. This idea on separating calculate on bar close and on each tick is detailed at the following link: http://ninjatrader.com/support/forum...ad.php?t=19387

                          Comment


                            #43
                            Did you ever work out this strategy?

                            Dolfan

                            Comment

                            Latest Posts

                            Collapse

                            Topics Statistics Last Post
                            Started by Prasanna, Today, 02:53 AM
                            1 response
                            14 views
                            0 likes
                            Last Post NinjaTrader_ShawnB  
                            Started by KevinWalker, Yesterday, 07:48 PM
                            1 response
                            10 views
                            0 likes
                            Last Post NinjaTrader_ShawnB  
                            Started by Leeroy_Jenkins, Today, 03:00 AM
                            0 responses
                            11 views
                            0 likes
                            Last Post Leeroy_Jenkins  
                            Started by PeterInBrisbane, Today, 01:07 AM
                            0 responses
                            16 views
                            0 likes
                            Last Post PeterInBrisbane  
                            Started by WaveAtElliot, Yesterday, 05:11 PM
                            1 response
                            17 views
                            1 like
                            Last Post hdge4u
                            by hdge4u
                             
                            Working...
                            X