Autop Dynamic Grid EA - Enhanced version of Auto-Profit 3.0 | Page 12 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Dynamic Fibonacci Grid 1,227 replies

Dynamic Fibonacci Grid - interactive trading 281 replies

Dynamic TakeProfit & Dynamic StopLoss EA/AI/Algorithm 0 replies

Martingale EA Dynamic Grid, Multipler 1.6 0 replies

Grid, Burst Grid, Grid Risk and Grid Profit 1 reply

Attachments: Autop Dynamic Grid EA - Enhanced version of Auto-Profit 3.0
Exit Attachments

Autop Dynamic Grid EA - Enhanced version of Auto-Profit 3.0

  • #221
  • Jul 21, 2016 3:39am Jul 21, 2016 3:39am
By using "Open Price Method" the backtest has managed to get through the "OrderModify error 130".

Although the results with this method are not too accurate, they are accurate enough to mark when a "Martingale DEATH SEQUENCE" happens.
  • #222
  • Jul 21, 2016 4:23am Jul 21, 2016 4:23am
Quoting kornrogers
{quote} Same here EDO... Another suggestion, i don't know if this is possible, as we know this EA in ranging markets is 100% secure, in a large trend that's when the problem occur. As the distinctive feature of this EA is the fact that it expands the grid when there is a trend, and you already inserted the hedge feature, we could also increase the lot size in the hedge size by simple increment and the retracement side with the normal multiplier. So for a 2 multiplier it would work like this. Retracement side - 0,01;0,02; 0,04; 0,08; 0,16; 0,32;...
it is very hard to decide whether it is a trend or not.

There are so many false break out in this market, especially on GBP pairs.

You think it is a trend and it turns out a false break out, a even larger DD will happen.
No one wants to be defeated~
  • #223
  • Jul 21, 2016 5:23am Jul 21, 2016 5:23am
Quoting Barlam
{quote} Hi all, whether You may needed or not. As promised attached You find my results for EURUSD M1 BT from 01.01.2016 until 18.07.2016 in 4.3 and 5.1 with the "same" settings as suggested by Edo.
Thank you for the settings Barlam.
For those who wonder, this settings may be the most profitable ones, but does not care about any filter and keeps on entering the market, waiting for a retrace at any point. Hence a big move (such as a news hike) can lead to extreme DD.

However I'm still testing similar settings with success, surviving the 5000 trades threshold easy.

Quoting bobby2
Hi Edorenta, first, thanks for all your hard work and sharing it here. It is appreciated. Subscribed to thread. Started testing on a demo starting today. I'm running V5.2 on 3 pairs, AUDUSD, USDCAD, and USDJPY, starting with a Bal of US$1000. The last time I tested one of the grid EAs the DDs were so bad I never bothered with it much. I'll let this run 24/7 for a few days and see what happens. Thanks again.
Hey Bobby, DD is ultimately inherent to grid strategies, now this strategy is about how to limitate it, maximizing our Calmar ratio (Profit over max DD).

Quoting MasterGuru
@Edorenta Thank you for this very promising EA
Thanks for joining the journey MasterGuru. Just as Zilla you're a man of wise words on this forum and I appreciate your feedback and your backtests. I'll look into my code to optimize its speed, however I'm now close to 2000 lines of code and loads of TP/Step/Hedge calculation functions and calls that make tick testing quite tough.

Quoting jonirrenicus
{quote} it is very hard to decide whether it is a trend or not. There are so many false break out in this market, especially on GBP pairs. You think it is a trend and it turns out a false break out, a even larger DD will happen.
I totally agree with you. If we were able to identify trends that well we wouldn't be working on such strategies hence a scale-hedge would come along with fewer profit if the market retraces and hits our retrace-TP.
You'll always miss 100% of the shots you don't take.
  • #224
  • Jul 21, 2016 5:27am Jul 21, 2016 5:27am
With any Martingale, the "death sequence" is practically unavoidable on an average "retail trader" account.

On the other hand, if the EA can produce average 500% (non-reinvested) profit with average 3 total crashes per year, it translates to 200% profit per year, assuming that the profits are withdrawn on a regular basis.

In summary: a good Martingale setting is not necessarly the one without crashes, but one with net yearly profits including the unavoidable crashes.
  • #225
  • Jul 21, 2016 5:41am Jul 21, 2016 5:41am
Quoting jonirrenicus
{quote} it is very hard to decide whether it is a trend or not. There are so many false break out in this market, especially on GBP pairs. You think it is a trend and it turns out a false break out, a even larger DD will happen.
I totally agree with you. If we were able to identify trends that well we wouldn't be working on such strategies hence a scale-hedge would come along with fewer profit if the market retraces and hits our retrace-TP.


I understand you guys, and it was just a suggestion. I know that this would lower the profit, but my idea was to change it to be more safe, with less drawdawn when heavy trends occur.

Beside this, I don't think you totally understand me, cause we would not need to know if a trend is on or not, cause the hedge(trend) positions would be open exactly on the same step as the retracement.

I believe we could have the both ways, the one that masterguru is talking about, with bigger profit, and some crashes, and the one with less crashes and less drawdawn.
  • #226
  • Jul 21, 2016 6:10am Jul 21, 2016 6:10am
This is the result From 13/7/2016 in my real account, I will post the details after few hours/days. Still in great condition and small drawdown
still using 4.3 with default setting，just changed the martingale multipler to 1.1
Attached Image (click to enlarge)
Click to Enlarge Name: image.jpeg Size: 157 KB
  • #227
  • Jul 21, 2016 6:41am Jul 21, 2016 6:41am
Hello

First I want to apologize for my bad English, this is not my native language. [IMG]file:///C:\Users\Jean\AppData\Local\Temp\msohtmlclip1\01\clip_image001.gif[/IMG]
Thanks Edorenta for your share.
I am trying to make a long backtest Autop 5.2 from the 01/08/2008 to today.
But there is a problem.
Please look at capture from the journal.
Thanks for help
Attached Image (click to enlarge)
Click to Enlarge Name: Capture journal MT4 Autop 5.2.JPG Size: 50 KB
  • #228
  • Jul 21, 2016 7:53am Jul 21, 2016 7:53am
On 7/21/2016. at the start of the Asian session, I changed my demo acct. to 4 pairs, usdcad, eurchf, audjpy, and audusd. v5.2, settings from post 212, TF = M1. Began with an actual Balance of $992.11 after losing $8 on 2 pairs during the previous 2nd half of London and full NY session. The results are better. There were a few big hits, but as long as the account survives and recovers that's OK. Trade Explorer synched from 7/21 is here http://www.forexfactory.com/#acct.81-tab.report
  • #229
  • Jul 21, 2016 8:09am Jul 21, 2016 8:09am
Quoting sagera
This is the result From 13/7/2016 in my real account, I will post the details after few hours/days. Still in great condition and small drawdown still using 4.3 with default setting，just changed the martingale multipler to 1.1 {image}
Hi,
thank you for the details. Would you let me know what was your maximal DD ?
  • #230
  • Jul 21, 2016 8:21am Jul 21, 2016 8:21am
Hi Edo,

I am following the FT with version 4.3. Today during news of ECB the Ea opened extra 2sell n one buy trade with the already existing grid in which all trades had tp level set. Those weird trades didn't had any tp placed. Was that normal? I checked journal but there was no error. FYI, I have disabled add filter.
  • #231
  • Edited 8:51am Jul 21, 2016 8:28am | Edited 8:51am
As the below backtest shows: SHORT - and even medium - TERM results can be very deceiving for stability (i.e. without the crash scenario).

The EA was working perfectly (with an experimental parameter set) from January 1st, 2012 for almost 22 months straight, until the "death sequence" happened on October 18th, 2013.

Attached Image (click to enlarge)
Click to Enlarge Name: Autop Dynamic Grid EA 21 months.png Size: 18 KB
  • #232
  • Jul 21, 2016 8:39am Jul 21, 2016 8:39am
this perfectly demonstrates why when trading martingale every trade could be your last. only trade these systems if you are willing to withdraw profits regularly


Quoting MasterGuru
As the below backtest shows: SHORT - and even medium - TERM results can be very deceiving for stability (i.e. without the crash scenario). The EA was working perfectly (with an experimental parameter set) from January 1st, 2012 for almost 22 months straight, until the "death sequence" on October 18th, 2013. {image}
  • #233
  • Jul 21, 2016 8:40am Jul 21, 2016 8:40am
Quoting MasterGuru
As the below backtest shows: SHORT - and even medium - TERM results can be very deceiving for stability (i.e. without the crash scenario). The EA was working perfectly (with an experimental parameter set) from January 1st, 2012 for almost 22 months straight, until the "death sequence" on October 18th, 2013. {image}

At least 4/5 times of profit..that goes with your idea of withdraw whenever we double the account at least. The problem is if the death sequence occurs in the first place
  • #234
  • Jul 21, 2016 8:53am Jul 21, 2016 8:53am
Quoting kornrogers
{quote} At least 4/5 times of profit..that goes with your idea of withdraw whenever we double the account at least. The problem is if the death sequence occurs in the first place
Exactly! Which also means that anticipating a certain number of "crash events" one should have at least 2-3x "reserve deposit" available to be able to start over.
  • #235
  • Jul 21, 2016 9:04am Jul 21, 2016 9:04am
Quoting edorenta
{quote} thank you for the settings barlam. For those who wonder, this settings may be the most profitable ones, but does not care about any filter and keeps on entering the market, waiting for a retrace at any point. Hence a big move (such as a news hike) can lead to extreme dd. However i'm still testing similar settings with success, surviving the 5000 trades threshold easy. {quote} hey bobby, dd is ultimately inherent to grid strategies, now this strategy is about how to limitate it, maximizing our calmar ratio (profit over max dd). {quote} thanks...
nice work

are you going to add the time filter??
  • #236
  • Jul 21, 2016 9:22am Jul 21, 2016 9:22am
Quoting MasterGuru
As the below backtest shows: SHORT - and even medium - TERM results can be very deceiving for stability (i.e. without the crash scenario). The EA was working perfectly (with an experimental parameter set) from January 1st, 2012 for almost 22 months straight, until the "death sequence" happened on October 18th, 2013. {image}
Could you show me your experimental sets? I'm curious about the steps and multiplier you used. As well as the trend filter / hedge settings.
(Btw an equity stop at 30% of the account might have saved you from this death sequence, the question now is to know whether it would have been hit several times across your test or just along this sequence.)
You'll always miss 100% of the shots you don't take.
  • #237
  • Edited 9:48am Jul 21, 2016 9:37am | Edited 9:48am
Quoting Edorenta
{quote} Could you show me your experimental sets? I'm curious about the steps and multiplier you used. As well as the trend filter / hedge settings. (Btw an equity stop at 30% of the account might have saved you from this death sequence, the question now is to know whether it would have been hit several times across your test or just along this sequence.)
Well, I have run multiple experimental sets over the past 24 hours and it was one of the two attached below.

Currently I am testing settings with $10k deposit and 20% Equity Stop by which I can simulate multiple small ($2k-$5k) account crashes without actually crashing the test account. According to the preliminary results the first 20% DD happened after 15 months with an approx. 8% average monthly profit until that point, i.e. only 2.5 months of profit were wiped out, although the recovery time might be longer due to Dynamic MM. Not bad Calmar Ratio if it remains so...
Attached File(s)
File Type: txt Autop v5.2 Masterguru 1.txt   6 KB | 548 downloads
File Type: txt Autop v5.2 Masterguru 2.txt   6 KB | 507 downloads
  • #238
  • Jul 21, 2016 9:56am Jul 21, 2016 9:56am
Quoting MasterGuru
{quote} Well, I have run multiple experimental sets over the past 24 hours and it was one of the two attached below. Currently I am testing settings with $10k deposit and 20% Equity Stop by which I can simulate multiple small ($2k-$5k) account crashes without actually crashing the test account. According to the preliminary results the first 20% DD happened after 15 months with an approx. 8% average monthly profit until that point, i.e. only 2.5 months of profit were wiped out, although the recovery time might be longer due to Dynamic MM. Not bad...
Thank you for your time, looking forward to your preliminary results.

Quoting Neo1
{quote} nice work are you going to add the time filter??
What time filter are you talking about?
You'll always miss 100% of the shots you don't take.
  • #239
  • Edited 10:40am Jul 21, 2016 10:17am | Edited 10:40am
Quoting MasterGuru
{quote} Well, I have run multiple experimental sets over the past 24 hours and it was one of the two attached below. Currently I am testing settings with $10k deposit and 20% Equity Stop by which I can simulate multiple small ($2k-$5k) account crashes without actually crashing the test account. According to the preliminary results the first 20% DD happened after 15 months with an approx. 8% average monthly profit until that point, i.e. only 2.5 months of profit were wiped out, although the recovery time might be longer due to Dynamic MM. Not bad...
If you use martingale multiplier, you will blow your account many times (e.g.: 2016/4), you can see Barlam's post #200, start from $500 from 2016/1 and the result is great but if you start in 2016/4, your account blew up very soon. We can use martingale with some luck but we may meet the death sequence when you are in the very beginning... So we must study the death sequence of this EA. First we can see and regconize the sequence then we may code.

But MasterGuru, may you provide the 2013 trade record here? Then we can see the trade start from which day and use different indicators to see the sequence. Regards!

P.S.: Below are EURUSD 2013/10 H1 and H4 chart, you may observe the huge price action at 16/10 to 17/10. This action is very sudden from ranging to upward trend.
Attached Image(s) (click to enlarge)
Click to Enlarge Name: Screen Shot 2016-07-21 at 11.13.46 PM.png Size: 41 KB
Click to Enlarge Name: Screen Shot 2016-07-21 at 11.14.18 PM.png Size: 40 KB
  • #240
  • Jul 21, 2016 10:44am Jul 21, 2016 10:44am
Seems like Troya's account is doing pretty well so far : http://www.forexfactory.com/troya7634

Troya would you share your settings? Thx
You'll always miss 100% of the shots you don't take.
1 trader viewing now