Jump to content

clovelly

Members
  • Posts

    180
  • Joined

  • Last visited

Posts posted by clovelly

  1. personally, i would be more concerned if the curve fitting was in relation to the order entry logic, the buy and sell entry logic. the MM you can always tweak using the Risk factor. however, i don't rule out that there might be entry logic optimization in the master files. we can't tell because they won't decompile

    Yes, it is correct. But looking at their 32 library modules, each seems to cater for a different day of the month, with different parameters for the indis (stochastic, cci, MA, etc.). These indis are supposedly called by the main EA to determine the trade entries. I think the main EAs will reveal more curve-fitting codes designed for backtesting.

  2. Did anyone take a deeper look at price action in conjunction with S/R levels using renko bars?

    If we take e/j for example with a 10pip brick size, nearly each time when price reverses on support or resistance there are at least two same colored bricks in a row. That means you could pocket 10pips-spread over and over again.

    I just don't believe that what I'm seeing on historical data is what I would have gotten when running the chart live

    But I don't see the reason because S/R is based on fractals and there is a concrete price pattern to form a fractal

     

    edit:

    was too good to be true....S/R repaints badly...but maybe there are enough occasions where the pattern is for real

    The S/R is only drawn AFTER 2 bars have been passed. That's why you see 2 colored bricks in a row. To win 10pips, you need to be able to trade the past :) , which I doubt any broker will allow you to do.

  3. the easiest way would be to open an account at [link]www.4shared.com[/link] and upload your files there and then post a link here,

     

    If you do not know to open an account with 4shared.com, you can go to this http://stashbox.org/, click on the "Browse..." button to select the file you want to upload, then click on "Upload" button. Once the upload completes, just copy the link in the "HTML/Forum Codes" panel at the bottom of the screen and paste it here in your post.

  4. what values are these? dollars? pips?

    extern string Raschet_trala = "start / step in the currency, if 0 - Percentage;

    extern double start_trail_val = 50;

    extern double trail_val = 1; / / To use a trawl as TakeProfit set the minimum value> 0

    extern string proc_trail = "start / step as a percentage of the Depot;

    extern double start_trail = 10;

    extern double trail = 1; / / For use trawl as TakeProfit set the minimum value> 0

     

    There is some confusion in the code/comment: The comment says if start_trail_val = 0, then the trailing is based on percentage. But the EA coding itself is also confused:

     

    1. When calculating the trailing values, it uses a percentage of the account balance if start_trail_val > 0 (the start_trail_val is actually the % used in the calculation)

     

    2. However, when displayed on the EA panel, it will display ">Percent xx%/yy%" if start_trail_val = 0 :)

     

    The calculation code, not comment or panel display, will determine the real action by the EA. So, if you want "percentage", set start_trail_val > 0. If start_trail_val = 0, then the trailing values will be in dollars based on start_trail and trail

  5. my demo and one real accounts with newtec EA, big lots, Euro/usd- Euro/chf, from 20.05.2010:

     

     

    error - game over - tild

     

    live account and Demo : zero Dollars !!!!!!!

     

    very very happy-s*** EA!!!!!

    I am surprised why you went live with big lots when you were still not convinced about the EA (that was why you tested with the demo account)

  6. what's new in this 'new' version, nikkola? I had a quick look and found that the Stoploss is now externalised and the "Moving Average Sigma" indi is now calculated based on current TF (rather than M1 in the original), and some other cosmetic modifications. So, why 'new' ?
  7. The included .set files are only applicable for the Russian version of the EA, as some 'extern' names have been changed in the English version and hence no longer pick up the values defined in these set files.

     

    I found that these set files have been too optimised. As an example, in the settings for M30 EURUSD, the Buy/Sell stop price is 50 pips beyond the IB high/low, hence producing only 90 trades for nearly 1.5 years of trading on M30 ! That may be acceptable if it is consistent. However, using these set files, testing outside the test period in the reports (1/1/2009 - 1/5/2010) produces much worse results.

  8. Calls to DLL files always are a worry to me. Does anyone know of a DLL decompiler so we can see what is "under the bonnet". My concern is whether the EA has been optimised to fool any backtests as we have seen so many times before.

    I ran the backtest for a whole year without the DLL, and the result is EXACTLY the same as when it was run with the DLL. You may want to comment off the DLL calls to make it faster. Beware that the EA has been optimised for 2009, backtesting over the period before 2009 shows that it is a dud.

  9. Hi Clovelly,

    Just so I understand your comments about the video. You are saying that the EA in the video is not the same as the concept explained in his PDF report right? Maybe one should not associate the Dynamic Channel indi with the new EA but I was under the impression they are the same.. that the developer automated his depicted dynamic channel concept.

     

    Cheers

    Yes, that's what I meant :). The EA in the video has nothing to do with the Dynamic Channel. I suspect he "gave away" the indi (not a great one btw) as a lure to catch new buyers for the EA, which is now available.

×
×
  • Create New...