My EA Gift to Traders | Page 41 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Screen shot indicator free gift 0 replies

DECEMA - Xmas gift! 6 replies

Vegas 1HR GIFT: Big opportunity coming! 5 replies

Attachments: My EA Gift to Traders
Exit Attachments

My EA Gift to Traders

  • #801
  • Mar 11, 2015 2:59pm Mar 11, 2015 2:59pm
Hi Dave,

I 'm following up on the volume filter. I was delayed in testing it out and just got to it now. What I'm finding is that the filter works correctly when used on its own. However, when I add another filter to get a combined backtest, I get no results all.

I'm using the latest version of the EA from page one, which as the volume filter included.

Kind regards,
drofwarc
  • #802
  • Mar 11, 2015 3:24pm Mar 11, 2015 3:24pm
Quoting drofwarc
Hi Dave, I 'm following up on the volume filter. I was delayed in testing it out and just got to it now. What I'm finding is that the filter works correctly when used on its own. However, when I add another filter to get a combined backtest, I get no results all. I'm using the latest version of the EA from page one, which as the volume filter included. Kind regards, drofwarc
It's probably because the back test produces no profitable results. The first thing to do is to extend the MaxBarsAgo on one of your criteria. I have found the Volume filter by itself is very restrictive in that it doesn't produce much. That is not to say that when used with other criteria it may help.

Enclosed is a .set file which produces results.
Attached File(s)
File Type: txt EUR_60_Dec2014_Feb2015_EMA_Volume.txt   13 KB | 217 downloads
  • #803
  • Mar 11, 2015 3:38pm Mar 11, 2015 3:38pm
Quoting alex8585
Dave , I got the verification from roboforex. see attached. What would be your suggestion on an ECN account to deposit my thought was to put 300 $ to take from the optimization results the 3 most good sets based on drawdown , profit , expected payoff , and run EA on three different charts for the same pair with the 3 different sets. Before i move to large amount i want to test it on small scale for couple months. Any ideas will be appreciated. Alex {image}
If you click on my profile page, you can see that the original RSI (running for around 4 weeks) is still strong along with 2 of the 3 components in the Traders Concept (which have been running continuously since the beginning of last week). That should give you a starting point.

It is really your business, but I would certainly run optimizations (last 3 months) on the following money management variables MaximumRiskPercentage, StopLoss, StopLossMultiplier, TakeProfitMultiplier, SetStopLossToBreakEvenPips, MinPipGainReset, DecrStopLossPercent plus whatever strategies and filters you desire to use. Even try a optimization of Time of Day.

The results for Trading Concept so far
Week 1 - Monday break even, Tuesday +100, Wednesday -100, Thursday -100, Friday +100, NETT 0
Week 2 - Monday +150, Tuesday + 150, Wednesday -150, NETT +150

Next week +- $200 daily, 4th week - run continuously.
  • #804
  • Mar 11, 2015 4:21pm Mar 11, 2015 4:21pm
I don't know whether many of you have noticed, but it you look on the hourly charts on the 7 majors, there is very prominent RSI and MACD divergence.

Another point, look at the 5/30 EMA and see how well it has served the current trend once again on the hourly charts.
  • #805
  • Mar 11, 2015 6:31pm Mar 11, 2015 6:31pm
Quoting Dave_C
{quote} It's probably because the back test produces no profitable results. The first thing to do is to extend the MaxBarsAgo on one of your criteria. I have found the Volume filter by itself is very restrictive in that it doesn't produce much. That is not to say that when used with other criteria it may help. Enclosed is a .set file which produces results. {file}
Hi Dave,

Using your set file, the volume filter worked. However, when I removed the EMA and straddle filters and used a candlestick filter instead, it stopped working correctly.

The issue seems to be related to the CSTimeframe setting.

I'm trying to backtest volume combined with a basic candlestick pattern (the reversal pattern I've used in the past) on the one-hour chart. I've set reqTimeframe to 60. When I set CSTimeframe to 0, the EA does not plot any trades. When I set CSTimeframe to 60, the EA does not plot any trades. When I set CSTimeframe to other timeframes (I tested 5, 15 and 240), it plots trades but not in the right places.

My setfile and EA with modded candlestick are attached.

Kind regards,
drofwarc
Attached File(s)
File Type: mq4 TradersGift_March 8_mod.mq4   102 KB | 237 downloads
File Type: txt EUR_60_Dec2014_Feb2015_Candlestick_Volume.txt   13 KB | 222 downloads
  • #806
  • Mar 11, 2015 6:59pm Mar 11, 2015 6:59pm
Hi Dave
I am running some optimisation test and was wondering once you have all the different sets of parameters how to you finally settle on which set to use
  • #807
  • Mar 11, 2015 7:13pm Mar 11, 2015 7:13pm
Quoting drofwarc
{quote} Hi Dave, Using your set file, the volume filter worked. However, when I removed the EMA and straddle filters and used a candlestick filter instead, it stopped working correctly. The issue seems to be related to the CSTimeframe setting. I'm trying to backtest volume combined with a basic candlestick pattern (the reversal pattern I've used in the past) on the one-hour chart. I've set reqTimeframe to 60. When I set CSTimeframe to 0, the EA does not plot any trades. When I set CSTimeframe to 60, the EA does not plot any trades. When I set CSTimeframe...
Insert this line at 1058 and recompile

PrevPrice = 0;
  • #808
  • Mar 11, 2015 7:27pm Mar 11, 2015 7:27pm
Quoting nobby911
Hi Dave I am running some optimisation test and was wondering once you have all the different sets of parameters how to you finally settle on which set to use
The set files included in the Trading Concept and Trading Stategies have boxes which are ticked. The values in Value column are the results of the optimization and are there by double clicking on the top line of the optimization results sorted by profit descending.
Attached Image(s) (click to enlarge)
Click to Enlarge Name: 2015-03-12 10_18_48-.jpg Size: 770 KB
Click to Enlarge Name: 2015-03-12 10_22_13-914382_ GlobalPrime-Live - EURUSD,H1.jpg Size: 196 KB
  • #809
  • Mar 11, 2015 8:06pm Mar 11, 2015 8:06pm
hi Dave

What i meant was once you have run optimisation do you always choose the most profitable set of parameters or do you also go by the drawdown % or some other criteria

I have been sorting by most profitable and then i scroll down until i find a set with the drawdow% less than 10%

I was just wondering what you criteria was for choosing an optimised set
  • #810
  • Mar 11, 2015 8:30pm Mar 11, 2015 8:30pm
Quoting nobby911
hi Dave What i meant was once you have run optimisation do you always choose the most profitable set of parameters or do you also go by the drawdown % or some other criteria I have been sorting by most profitable and then i scroll down until i find a set with the drawdow% less than 10% I was just wondering what you criteria was for choosing an optimised set
It's entirely up to you. Personally. if I see a drawdown, say higher than 25%, I look down to see if there is a standout low drawdown below but generally I go for max. profit.
  • #811
  • Mar 11, 2015 9:46pm Mar 11, 2015 9:46pm
When backtesting is the spread in pips or points ie for 2 pips on my five digit broker do i put in 20 for 2pips

Attached Image
  • #812
  • Mar 11, 2015 11:24pm Mar 11, 2015 11:24pm
Quoting nobby911
When backtesting is the spread in pips or points ie for 2 pips on my five digit broker do i put in 20 for 2pips {image}
All calculations are based on PointValue which is defined as .0001 (except for JPY and metals) so the value remains at 2 regardless of whether the broker uses 4 or 5 digits.

Here is the relevant code.

if(StringFind(Symbol(),"JPY") > (-1)) PointValue = .01;
else if(StringFind(Symbol(),"XAU") > (-1)) PointValue = .1;
else if(StringFind(Symbol(),"GOLD") > (-1)) PointValue = .1;
else if(StringFind(Symbol(),"XAG") > (-1)) PointValue = .01;
else if(StringFind(Symbol(),"SILVER") > (-1)) PointValue = .01;
else PointValue = .0001;
  • #813
  • Mar 11, 2015 11:40pm Mar 11, 2015 11:40pm
Thanks
The reason i ask is when i set the spread in tester to current for AUDUSD the Report said the spread was 16 so i took this as meaning i needed to add the spread as 20 roughly what the AUDUSD is with my broker
  • #814
  • Edited 2:48am Mar 12, 2015 1:33am | Edited 2:48am
Quoting nobby911
Thanks The reason i ask is when i set the spread in tester to current for AUDUSD the Report said the spread was 16 so i took this as meaning i needed to add the spread as 20 roughly what the AUDUSD is with my broker
That is an MT4 generated report. You would always specify 2 as defined in the Post #812. I will highlight that in the next help file update and that is what I nominate in MT4. I'm almost sure the drop down hasn't been amended for 5 decimal point brokers.
  • #815
  • Mar 12, 2015 2:30am Mar 12, 2015 2:30am
Dave ,

I was thinking exactly to adopt your way to run optimisation for the last 3 months on the open prices. Pick some sets and test them also on tick data . After that to take 3 of them i also think to put on different timeframes.

And every month to do the same in order to maintain updated to market conditions.

Can you please post set files that you currently use for the accounts on your profile , it will be much appreciated.

On Monday live account will start.

Alex
  • #816
  • Edited 2:57am Mar 12, 2015 2:36am | Edited 2:57am
Quoting alex8585
Dave , I was thinking exactly to adopt your way to run optimisation for the last 3 months on the open prices. Pick some sets and test them also on tick data . After that to take 3 of them i also think to put on different timeframes. And every month to do the same in order to maintain updated to market conditions. Can you please post set files that you currently use for the accounts on your profile , it will be much appreciated. On Monday live account will start. Alex
They are the same as at Post #1 in Trading Strategies or Trading Concept. Your methodology is spot on and I will update all the strategies at the end of Mar. There may be some benefit in gaining good entries by using the new volume fllter. As you know RSI hourly is my favourite. Hope you capitalised on my highlight on divergence earlier today!
  • #817
  • Mar 12, 2015 2:59am Mar 12, 2015 2:59am
Quoting Dave_C
{quote} They are the same as at Post #1 in Trading Strategies or Trading Concept.
I can see that all the accounts are running on 60 min timeframe ?

Also i forgot to ask if you gonna add support/resistance from a higher timeframe ? I had also the thought , could we check ATR values are increased or decreased for an X bars ago.

Lets say that we have 10 -15 candles with low ATR value and then we get a candle that is out of that range , so we could jump in that direction some kind of breakout but based on momentum.

Just my suggestion bring here for discussion.

What you people believe on this.
  • #818
  • Mar 12, 2015 3:10am Mar 12, 2015 3:10am
Quoting alex8585
{quote} I can see that all the accounts are running on 60 min timeframe ? Also i forgot to ask if you gonna add support/resistance from a higher timeframe ? I had also the thought , could we check ATR values are increased or decreased for an X bars ago. Lets say that we have 10 -15 candles with low ATR value and then we get a candle that is out of that range , so we could jump in that direction some kind of breakout but based on momentum. Just my suggestion bring here for discussion. What you people believe on this.
I'm not at my computer now but from memory SR defaults to 60 mins. Can't remember whether it is configurable. ATRMaxBarsAgo can be added.
  • #819
  • Mar 12, 2015 3:18am Mar 12, 2015 3:18am
Quoting Dave_C
{quote} I'm not at my computer now but from memory SR defaults to 60 mins. Can't remember whether it is configurable. ATRMaxBarsAgo can be added.
AtrMaxBarsAgo will be very valuable . Yes it configurable but for daily SR it needs to be written 1440 which is too large as number and i am afraid it will create some issues.

But as soon as i get home will check i come back to you.
  • #820
  • Mar 12, 2015 11:24am Mar 12, 2015 11:24am
Hi Dave

From what i have found out it appears that you enter the spread in points as in 20 for 2 pips etc

. Strategy Tester as of MT4 build 509 now enables you to fix the spread by setting the value in points in the small “Spread:” field on the right just above Optimization. 25 points as an example means 2.5 pips.

also if the data used is exported to MT4 using Tickstory (dukascopy data) the spread is set at the point of export and to change it you need to use the tickstory tool MT4 data editor.
I wondered why changing the spread in MT4 wasn't making any difference to my back testing.
5 traders viewing now