Micro martingale with small capital | Page 10 | Forex Factory

Options

Go to
Permalinks
Print Thread

Related Instruments

EUR/USD

Similar Threads

Martingale, Anti-martingale, and Compounding 40 replies

Martingale, Reverse Martingale, Modified Martingale, Maths 1 reply

Martingale Calculation EA (also reverse martingale) 3 replies

How to change this Martingale to Reverse Martingale ? 3 replies

Martingale vs. Non Martingale (Simplified RoR vs Profit) 1 reply

Attachments: Micro martingale with small capital
Exit Attachments

Micro martingale with small capital

  • #181
  • Dec 26, 2017 9:28pm Dec 26, 2017 9:28pm
Quoting LinkersX
I'd like to see your Martingale compiled on Linkers::X and back tested in 100% tick based environment with historical events included into the back-test. Just to see how your martingale strat performs in full market replay style back-test
Hi Linkers

To test with Data Quality 99.9% I did not do (have no means).

In this forum our friend (Mr. Giapel) has done that, please read. Only this EA will be suitable if the account type ECN and low spreads. If using data before ECN may not be suitable.

I only do Backtest with Live Account data history and testing forward Demo and Live Account.

Btw please try which one is best, if there are still shortcomings we are looking for a solution.
margiyono61 gmail com, Telegr @Margiyono
  • #182
  • Dec 26, 2017 10:26pm Dec 26, 2017 10:26pm
Hello! thanks for your ea.

I want to try it in a real account. I want to try with 500usd.
The ea is active but since several days the ea did not open any trades.

The ea really works? Please i will wait your response.

Thanks.
  • #183
  • Edited Dec 27, 2017 12:03am Dec 26, 2017 10:28pm | Edited Dec 27, 2017 12:03am
From 01/01/2016 to Feb-2017 (Wrong date), tick data, SL = 10 pips, spread = 1 pip.

Attached Image (click to enlarge)
Click to Enlarge Name: Mrg_Micro_Martingale.gif Size: 13 KB
  • #184
  • Dec 26, 2017 10:40pm Dec 26, 2017 10:40pm
Same settings and period, but spread = 1 pipette (doable on IC Markets). Drawdown is still huge (80%), and commissions would have eaten most profit.

Attached Image (click to enlarge)
Click to Enlarge Name: Mrg_Micro_Martingale_2.gif Size: 13 KB
  • #185
  • Dec 26, 2017 10:47pm Dec 26, 2017 10:47pm
Anything below 99.99% quality will show a holy grail system, so good data is paramount when testing any kind of scalping scheme, and still that is not enough (only live testing works). I can try to do some optimization runs if you want.

Attached Image (click to enlarge)
Click to Enlarge Name: Mrg_Micro_Martingale_3.gif Size: 12 KB
  • #186
  • Dec 26, 2017 10:56pm Dec 26, 2017 10:56pm
Quoting Abade
Same settings and period, but spread = 1 pipette (doable on IC Markets). Drawdown is still huge (80%), and commissions would have eaten most profit. {image}
Hi Adabe

Thank you for trying to do testing
Normally 2016.01.01 - 2017.02.28, with a default settting of approximately 10,700 transactions.
In tick data only 870s. Meaning to get Margin call at the beginning of the year?

- Spreads 1, ok
- What is the Stop Level?

Is it possible to do testing with the last period (about 2 months) to compare with BT Normal and Live Account results.

Okay good correction, can be searched date when there is a loss of prolonged there may be something that has not filtered conditions, to look for his solution. I'm sure it can be done.

The solution is now done Live Demo testing is done Backtest with the same period difference where and why.
IC Market can not with Pending_Order = true and Trailing_Active = true, IC sensitiv if direct error trade is disable until SL occurs.
margiyono61 gmail com, Telegr @Margiyono
  • #187
  • Dec 26, 2017 11:52pm Dec 26, 2017 11:52pm
Quoting elflacofabri
Hello! thanks for your ea. I want to try it in a real account. I want to try with 500usd. The ea is active but since several days the ea did not open any trades. The ea really works? Please i will wait your response. Thanks.
Hi Elflacofabri

Thanks.

To find out working or not, try Signal Level = 0. with Lotsize 0.01 (AutoLot=false) then restart MT4. (If there is no OP, there may be parameters that need to be adjusted)
After OP restore normal Signal_Level and restart MT4.
margiyono61 gmail com, Telegr @Margiyono
  • #188
  • Dec 26, 2017 11:57pm Dec 26, 2017 11:57pm
Quoting mrgmrg
{quote} Hi Adabe Thank you for trying to do testing Normally 2016.01.01 - 2017.02.28, with a default settting of approximately 10,700 transactions. In tick data only 870s. Meaning to get Margin call at the beginning of the year? - Spreads 1, ok - What is the Stop Level? Is it possible to do testing with the last period (about 2 months) to compare with BT Normal and Live Account results. Okay good correction, can be searched date when there is a loss of prolonged there may be something that has not filtered conditions, to look for his solution. I'm...
Attached Image (click to enlarge)
Click to Enlarge Name: Stops_Level.png Size: 110 KB


Stops levels = 0

It is a mean reversion strategy that uses low time-frame data, so it is very susceptible to strong momentum moves that don't revert back quick enough (ie:during news). The high relative DD (80%) means that it is VERY LIKELY to get a margin call on a real account. The high volume used means equally high commisions (on the low spread account), and those cannot be accounted for during the backtest, so the best bet is to do it live mainly because:

- The volume based commisions are a HUGE deal.
- The floating spread is a HUGE deal.
- As you said IC markets is sensitive to errors while trailing, but not all EA's are, the only way to know is to use it live.

I downloaded the most recent tick data, and will try to run an optimization for 2016 and run it on 2017 to see if the settings may be optimized realistically (we should aim to reduce DD in all cases). Here is the complete test from 2016/01/01 to 2017/07/26 using tick data on low spread (results are misguiding).

Attached Image (click to enlarge)
Click to Enlarge Name: Mrg_Micro_Martingale_4.gif Size: 12 KB


EDIT: I actually made a mistake on the dates i published early, the results don't go beyond march 2016, but these do!!
  • #189
  • Dec 27, 2017 12:36am Dec 27, 2017 12:36am
Quoting Abade
{quote} {image} Stops levels = 0 It is a mean reversion strategy that uses low time-frame data, so it is very susceptible to strong momentum moves that don't revert back quick enough (ie:during news). The high relative DD (80%) means that it is VERY LIKELY to get a margin call on a real account. The high volume used means equally high commisions (on the low spread account), and those cannot be accounted for during the backtest, so the best bet is to do it live mainly because: - The volume based commisions are a HUGE deal. - The floating spread is...
Hi Abade

Exactly
This EA is vulnerable to widening spreads and slippage in high movements.

When using Trailing_Active = TRUE and Pending_Order = TRUE, it is likely that there will be a "Modify error" error and result in "Trade is Disable". This transaction can not be in close until SL or TP unless MT4 is turned off. This is also one cause of long loss because the Chart TSB can not transact before the positions hit SL / TP.

With Trailing_Active = False and Pending_Order = False, it will not error but profit is smaller.

Commission is bigger than Profit -> This is true once it could happen due to Slippage on Close.

For OP there is Max_Slipage but for Close there is no Slipage limit.
margiyono61 gmail com, Telegr @Margiyono
  • #190
  • Dec 27, 2017 12:43am Dec 27, 2017 12:43am
Using Margin_Percent = 1% and StopLoss = 1 pip (the rest is default), i got this on Dukascopy tick data:

Attached Image (click to enlarge)
Click to Enlarge Name: Low_DD.png Size: 109 KB



Very nice DD and growth . The question is wheter these settings work live.
  • #191
  • Dec 27, 2017 12:52am Dec 27, 2017 12:52am
Quoting mrgmrg
{quote} Hi Abade Exactly This EA is vulnerable to widening spreads and slippage in high movements. When using Trailing_Active = TRUE and Pending_Order = TRUE, it is likely that there will be a "Modify error" error and result in "Trade is Disable". This transaction can not be in close until SL or TP unless MT4 is turned off. This is also one cause of long loss because the Chart TSB can not transact before the positions hit SL / TP. With Trailing_Active = False and Pending_Order = False, it will not error but profit is smaller. Commission is bigger...

Thanks for sharing. I believe one way to offset the risk from the modification error is to force a really tight SL with high TP using virtual stops (bid price), as you see from the tests a really tight stop reduces the DD significantly. I tried to upload the entire report but i got an error.
  • #192
  • Dec 27, 2017 12:57am Dec 27, 2017 12:57am
Quoting Abade
Using Margin_Percent = 1% and StopLoss = 1 pip (the rest is default), i got this on Dukascopy tick data: {image} Very nice DD and growth . The question is wheter these settings work live.
Good Job.

SL = 1 Pips means Margin_Percent can still be raised, this EA if loss or profit is still in the same candle and then immediately will Open again.

This EA for normal news cases (eg NFP, see Post # 68) I try live account still profit, but for big case like Brexit or EUR / CHF do not know yet.
margiyono61 gmail com, Telegr @Margiyono
  • #193
  • Dec 27, 2017 1:08am Dec 27, 2017 1:08am
Quoting mrgmrg
{quote} Good Job. SL = 1 Pips means Margin_Percent can still be raised, this EA if loss or profit is still in the same candle and then immediately will Open again. This EA for normal news cases (eg NFP, see Post # 68) I try live account still profit, but for big case like Brexit or EUR / CHF do not know yet.

Nice, thank you.

I'd rather keep the Margin_Percent setting as low as possible, and withdraw as soon as i am reaching the high volume trades as it usually means a lot more slippage.

And yup, i believe most red/big news are to be avoided as with any other mean reversion strategy. Will try some optimizations and let you know if i find something else.
  • #194
  • Dec 27, 2017 1:15am Dec 27, 2017 1:15am
Quoting Abade
{quote} Nice, thank you. I'd rather keep the Margin_Percent setting as low as possible, and withdraw as soon as i am reaching the high volume trades as it usually means a lot more slippage. And yup, i believe most red/big news are to be avoided as with any other mean reversion strategy. Will try some optimizations and let you know if i find something else.
Correctly.
Only if SL 1Pips for Pair GBPxxx may be too small especially GBPJPY.
For setting with Low SL, Martingale is not enabled (Default 9) except Marting_Min_Grid <SL.
margiyono61 gmail com, Telegr @Margiyono
  • #195
  • Dec 27, 2017 1:34am Dec 27, 2017 1:34am
Quoting mrgmrg
{quote} Correctly. Only if SL 1Pips for Pair GBPxxx may be too small especially GBPJPY. For setting with Low SL, Martingale is not enabled (Default 9) except Marting_Min_Grid <SL.

Yup, i would use a slightly bigger SL (2 pips maybe), considering the avg spread is 4 pipettes and i am using it as the max spread filter. These are the results for GBPUSD, tick data, 4 pipette spread, SL = 1 pip. Max DD is even better than the one on EURUSD:

Attached Image (click to enlarge)
Click to Enlarge Name: GBPUSD_Mrg.png Size: 141 KB


Do you think not using martingale would be a big disadvantage with your strategy?

PS: I note the consecutive losses counter is really big.
  • #196
  • Dec 27, 2017 1:50am Dec 27, 2017 1:50am
Quoting Abade
{quote} Yup, i would use a slightly bigger SL (2 pips maybe), considering the avg spread is 4 pipettes and i am using it as the max spread filter. These are the results for GBPUSD, tick data, 4 pipette spread, SL = 1 pip. Max DD is even better than the one on EURUSD: {image} Do you think not using martingale would be a big disadvantage with your strategy? PS: I note the consecutive losses counter is really big.
Great.

GBP is much better. But there are some brokers who apply high GBP Spreads.
For a small SL, try the attached EA (out of this topic). The result is much better.
Micro is part of Revo Power. What is done in Mico, can be applied in Revo.
Attached File(s)
File Type: ex4 Mrg_Revo_Power.ex4   86 KB | 383 downloads
margiyono61 gmail com, Telegr @Margiyono
  • #197
  • Dec 27, 2017 1:51am Dec 27, 2017 1:51am
Quoting mrgmrg
{quote} Great. GBP is much better. But there are some brokers who apply high GBP Spreads. For a small SL, try the attached EA (out of this topic). The result is much better. Micro is part of Revo Power. What is done in Mico, can be applied in Revo. {file}

Thank you, i am going to check it out
  • #198
  • Dec 27, 2017 1:54am Dec 27, 2017 1:54am
Quoting Abade
{quote} Yup, i would use a slightly bigger SL (2 pips maybe), considering the avg spread is 4 pipettes and i am using it as the max spread filter. These are the results for GBPUSD, tick data, 4 pipette spread, SL = 1 pip. Max DD is even better than the one on EURUSD: {image} Do you think not using martingale would be a big disadvantage with your strategy? PS: I note the consecutive losses counter is really big.
Qusetion : Do you think not using martingale would be a big disadvantage with your strategy?
Answer :
No, just info that the parameters are not working. Btw Marting is only a matter of means, just choose to use which one is best,
margiyono61 gmail com, Telegr @Margiyono
  • #199
  • Dec 27, 2017 1:56am Dec 27, 2017 1:56am
Hi mrgmrg
I'm a silent follower of your thread and I'm impressed by this promising EA.
Have you an idea of the selling price ?
  • #200
  • Dec 27, 2017 2:11am Dec 27, 2017 2:11am
Quoting mrgmrg
{quote} Hi Linkers To test with Data Quality 99.9% I did not do (have no means). In this forum our friend (Mr. Giapel) has done that, please read. Only this EA will be suitable if the account type ECN and low spreads. If using data before ECN may not be suitable. I only do Backtest with Live Account data history and testing forward Demo and Live Account. Btw please try which one is best, if there are still shortcomings we are looking for a solution.
If as you say, full 100% tick based back-test doesn't mean anything to you or your EA, how about you attach a micro REAL account with a trade explorer here
Linkers::X->theNextBigThing; in C++ HFT Event Driven Algorithmic Trading
2 traders viewing now