MGH System (Martingale Grid & Hedging) | Page 42 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Simply "grid" system, no hedge, no martingale 9 replies

ECN && STP && Scalping && Hedging - Broker Review - Oct 2009 24 replies

Martingale, Reverse Martingale, Modified Martingale, Maths 1 reply

Simple grid manual system without martingale 8 replies

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

Attachments: MGH System (Martingale Grid & Hedging)
Exit Attachments

MGH System (Martingale Grid & Hedging)

  • #821
  • Edited 5:24am Nov 8, 2014 5:10am | Edited 5:24am
Quoting SwingMan
{quote} My analysis show just the opposite ... Decreasing BatchX / Trade2 lot size would definitely decrease drowdown. In order to increase the profit we can increase the base lot size.
Sorry, I misunderstood. The "variable" was not clear to me, i.e. it referred to decreasing BatchX/Trade2 lot sizes.

Well, it sounds like an explorable idea, which can be "stretched" to the extreme even by leaving "BatchX/Trade2" entries completely out of the equation... How about that?

It would definitely reduce DD even for larger baskets, which could safely include Batch5 and even Batch6. Accommodating Batch5 and even Batch6 entries safely (i.e. with tolerable DD) would significantly (exponentially) reduce the chance of hitting a "Death Trail".

I would not be surprised that if our final conclusion will be that even a Batch10 is needed to avoid the "Death Trail" by 99.99% certainty.
  • #822
  • Nov 8, 2014 5:23am Nov 8, 2014 5:23am
Quoting Paracelsus
...Well, it sounds like an explorable idea, which can be "streched" to the extreme even by leaving "Trade2" entries completely out of the equation...
Correct, I also thought about this.
I have developed a system with "very safe" entrys. The little problem is that sometimes just after the entry occurs small reverses, therefore am I come to this thread and believe that just your simplification on two-steps grid may be helpful to avoid early exits.
  • #823
  • Edited 5:52am Nov 8, 2014 5:26am | Edited 5:52am
I will give you my thoughts and experiences on MGH EAs thus far, and specifically related to an EA I purchased from a failed MQL5.com signal provider earlier this year. I have tried using this EA on EURUSD as was successfully shown by the signal provider until about May, when the EU peaked at 1.4 and then came crashing down suddenly and blowing up the signal account.

I was unable to duplicate the phenomenal returns of the SP, nor was I able to repeat my 120% gain of December 2013 using a different GM EA (with no hedging). Definitely, a hedging aspect added on a GMEA increases profitability and also risk - but margin requirement is obviously reduced as both sides of a trade are taken (sorry to US investors that can't hedge).

I see two ways of successful employment of MGH EAs:

  1. You can improve "account survivability" by filtering entries using indicators or time, but this will reduce the EA's frequency of trading (and thus profits). This approach may be mandatory on pairs with large price fluctuations, including EURUSD. Eg. filtering EA to operate only during Asian session.
  2. Or keeping the EA simply MGH with no entry filters. This is only possible on a tight-ranging pair such as EURCHF.

In either method, pair-specific analysis to determine the best settings is mandatory: every pair behaves differently with variances in volatility, pip cost, spreads, etc.

​For me I am sticking to a simple approach, namely focusing on: a) grid size, and b) martingale factor. Also, the EA I currently use opens two first trades - in a sense this lowers the "logarithmic / exponential" effect of the martingale (albeit marginally), since the grid now goes: Level 1 (2 x 0.1 = 0.2), L2 (0.11), L3 (0.12), L4 (0.13), etc. (for example). Therefore I am sticking to EURCHF-only for now.

I am hoping to analyse and determine other pairs suitable for a simple MGH strategy without filters, since I pursue a HFT-strategy.

Folks should always remember: account survivability is the 'name of the game' - profits are secondary.

"In the chart, the height of a bar is equivalent to how often a certain price appears in the price curve. The red bars are based on the EUR price in Swiss Francs (EUR/CHF) since October 2011; the green bars are the EUR price in US Dollars (EUR/USD) in the same time period. You can generate such a chart with the Distribution script. We can see that the red EUR/CHF price distribution is quite narrow and ends abruptly on the left, as if cut with a knife. In contrast, the green EUR/USD price distribution is much wider with a broad peak in the middle. It resembles a bell curve (more or less), which indicates a random price distribution. The EUR/CHF distribution however is no bell curve. It reveals a market inefficiency that can generate profit in automated trading. In this case, the inefficiency is the result of the CHF price ceiling that was established by the Swiss National Bank in September 2011. It prevents the EUR from ever dropping below 1.20 CHF. This market inefficiency is used to your advantage in Zorro's Z5 strategy." (http://www.zorro-trader.com/manual/en/strategy.htm)

Attached Image(s) (click to enlarge)
Click to Enlarge Name: Clipboard01.jpg Size: 137 KB
Click to Enlarge Name: mql.png Size: 22 KB
  • #824
  • Nov 8, 2014 5:58am Nov 8, 2014 5:58am
Quoting bluepanther
I will give you my thoughts and experiences on MGH EAs thus far, and specifically related to an EA I purchased from a failed MQL5.com signal provider earlier this year. I have tried using this EA on EURUSD as was successfully shown by the signal provider until about May, when the EU peaked at 1.4 and then came crashing down suddenly and blowing up the signal account. I was unable to duplicate the phenomenal returns of the SP, nor was I able to repeat my 120% gain of January 2014 using a different GM EA (with no hedging). Definitely, a hedging aspect...
I don't want you to be sad, but if a grid EA can work safety only with indicators and/or choosing the most ranging pairs - it is very bad.

I think and it is my 4 years experience (especially with grid-hedge strategies) that using indicators or any other "signals" to try to make a grid EA safer, it is just a waste of time. It just delays the MC.

Using multi currency on one account: it is a good risk management, because of diversification. And to use many pairs (up to 28) no mandatory to know their "behavior", because of the causes mentioned above. It is enough to know that a pair rather trending or rather ranging, and you can choose a startegy that most suited to it (grid stategy, just you cab use "grid against trend"; "grid or pyramid with trend" or "range break out grid", like the well known Sure Fire Hedging Strategy).

But it is not enough in itself, of course!

Try to determine the direction of a pair or just the probability of it: it is just an unncessary energy-killer activity.
  • #825
  • Nov 8, 2014 6:50am Nov 8, 2014 6:50am
Quoting bluepanther
I will give you my thoughts and experiences on MGH EAs thus far, and specifically related to an EA I purchased from a failed MQL5.com signal provider earlier this year. I have tried using this EA on EURUSD as was successfully shown by the signal provider until about May, when the EU peaked at 1.4 and then came crashing down suddenly and blowing up the signal account. I was unable to duplicate the phenomenal returns of the SP, nor was I able to repeat my 120% gain of December 2013 using a different GM EA (with no hedging). Definitely, a hedging aspect...
Thanks for the idea, Bluepanther! I guess, the next step would be to fine tune MGH on EUR/CHF instead of on EUR/USD and probably by using SMALLER BatchX/Trade2 lot sizes as per SwingMan's recommendation.

I have agree with Moncsicsi, i.e. "guessing" the direction BETTER can be misleading, as it may just mask potential "Death Trails", i.e. we should clearly see the chance of hitting one.

Further ideas are welcome! Please join the discussion.
  • #826
  • Nov 8, 2014 7:05am Nov 8, 2014 7:05am
I share 3 EAs, I found them somewhere on the NET.
If somebody feel spirit, test is on visual test mode, with the attached settings, so you can learn its strategy.

I know well that this thread is for MGH strategy, and I don't want to make an OFF TOPIC, but I think to watch, learn something else is sometimes necessary to make further improvement on the current strategy.

So my intention with sharing other EAs in this thread is to "wake up new ideas" to make the MGH strategy much better.
Attached File(s)
File Type: rar Experts.rar   48 KB | 1,268 downloads
  • #827
  • Nov 8, 2014 7:30am Nov 8, 2014 7:30am
Quoting moncsicsi78
I share 3 EAs, I found them somewhere on the NET. If somebody feel spirit, test is on visual test mode, with the attached settings, so you can learn its strategy. I know well that this thread is for MGH strategy, and I don't want to make an OFF TOPIC, but I think to watch, learn something else is sometimes necessary to make further improvement on the current strategy. So my intention with sharing other EAs in this thread is to "wake up new ideas" to make the MGH strategy much better. {file}
It is much appreciated your intention
  • #828
  • Edited 1:19pm Nov 8, 2014 12:35pm | Edited 1:19pm
Quoting moncsicsi78
{quote} I don't want you to be sad, but if a grid EA can work safety only with indicators and/or choosing the most ranging pairs - it is very bad. I think and it is my 4 years experience (especially with grid-hedge strategies) that using indicators or any other "signals" to try to make a grid EA safer, it is just a waste of time. It just delays the MC. Using multi currency on one account: it is a good risk management, because of diversification. And to use many pairs (up to 28) no mandatory to know their "behavior", because of the causes mentioned...
@Moncsicsi: I definitely agree with you on diversification, but any MGH will require considerable margin on one pair alone - on 28 pairs you would be looking at starting your account with a minimum of $100,000 (and perhaps I am underestimating this amount - that's less than $4000 per pair). Of course, depends on grid size, martingale factor, starting lots (0.01), and pairs' volatility. I quite agree that filters for entry is detrimental to the performance (and profits) of an EA, but what about time filter (eg. Asian session only)? This way risk is controlled better because volatility is lowered (exaggerated movements are less likely), and a smaller grid can be utilised to preserve profits on the lower activity.

Keep in mind: diversification mitigates pair-specific risk but concurrently increases inherent risk, the same way that hedging mitigates trend risk but increases inherent risk.

It is important to analyse and review current pair behaviour (especially range), since some pairs exhibit specific behaviour due to extraneous factors (eg. interest rates/carry trading, commodity pairs, being a "major pair", market sessions, central bank interventions, etc.).

Yes, the whole concept of hedging is to "catch" the market regardless of direction - attempts to predict are both contrary and futile.

  1. 'Time spent in reconnaissance is seldom wasted.' John Marsden
  2. 'If you know the enemy and know yourself you need not fear the results of a hundred battles.' Sun Tzu

​...and thank you for the ex gratia EAs!

Quoting Paracelsus
{quote} Thanks for the idea, Bluepanther! I guess, the next step would be to fine tune MGH on EUR/CHF instead of on EUR/USD ...
Yes, ultimately I am hoping to increase volume on EURCHF to lower spreads with my broker, so it is more efficient for MGH EAs to trade on!
On a serious note, EURCHF exhibits its tight range due to central banking intervention - a market inefficiency as observed by Zorro that we can take advantage of!
  • #829
  • Nov 8, 2014 12:55pm Nov 8, 2014 12:55pm
Quoting bluepanther
{quote} @Moncsicsi: I definitely agree with you on diversification, but any MGH will require considerable margin on one pair alone - on 28 pairs you would be looking at starting your account with a minimum of $100,000 (and perhaps I am underestimating this amount - that's less than $4000 per pair). Of course, depends on grid size, martingale factor, starting lots (0.01), and general market volatility. I quite agree that filters for entry is detrimental to the performance (and profits) of an EA, but what about time filter (eg. Asian session only)?...
Bluepanther: if I use 28 pairs for any type of grid trading, -maybe it is unbelievable, but there is no need to deposit a huge amount! The reason is simple: there will always be more pair, which are in profit while the other will be in loss! The pairs which are in profit, increase the free margin.
And that's why I use more pairs...
One example: if we go long trade on the 28 pairs at the same time, there will always be more or less pairs, which will be in profit! (And this is the same situation if we start trading with short orders.) Because the probability of going down every pair at the same time is relatively smaller! (And vice versa.)
The pairs which is going in to profit, sooner or later will be closed - the number of the pairs in trading will decrease.

But as I wrote before: this not a hole startegy, it is just a portion or view of the hole strategy.

And I also wrote that we can combine different type of grid strategies at the same time. For example the more trending JPY pairs we should use "trending grid" strategies, or "range break out grid". At the same time on the ranging pairs we should use "grid against trend" strategies.
And yea, I think closing all pairs together in a common profit is better then closing each pairs alone, because the "good pairs" (which are in profit) will be able to collect more and more profit while the "bad" pairs (which are in loss) try to go profit. When we close everything, we can see that some pairs were in loss in itself (maybe they can cause a MC if they were just alone on the account...) , but we can close in spite of this because there are other pairs, which profits are enough to close everything in net profit.
  • #830
  • Edited 1:18pm Nov 8, 2014 1:06pm | Edited 1:18pm
Quoting moncsicsi78
{quote} Bluepanther: if I use 28 pairs for any type of grid trading, -maybe it is unbelievable, but there is no need to deposit a huge amount! The reason is simple: there will always be more pair, which are in profit while the other will be in loss! The pairs which are in profit, increase the free margin. And that's why I use more pairs... One example: if we go long trade on the 28 pairs at the same time, there will always be more or less pairs, which will be in profit! (And this is the same situation if we start trading with short orders.) Because...
Understood. I am speaking from the standpoint of the EA working on pairs in separate grids - you speak of a basket-trading strategy. Definitely a good idea!

In that case pair-specific analysis is less important (especially if the EA does all that for you automatically), as perhaps too the considerable margin as pairs are managed conjunctively.
  • #831
  • Nov 9, 2014 5:52pm Nov 9, 2014 5:52pm
Here is a new EA version 3.1
For the calculation of lots for each batch and trades 1 or 2, was used a special algorithm in order to reduce the drawdown (see "MGH BatchLots").
You can test different grid values, take profits equal to 0 or 5, lot sizes, etc., and report what comes out...
I hope there are not too many errors this time...
Attached Image (click to enlarge)
Click to Enlarge Name: EURUSDM15-v3.1.png Size: 18 KB
Attached Image
Attached File(s)
File Type: ex4 MGH BatchLots.ex4   10 KB | 769 downloads
File Type: mq4 MGH BatchLots.mq4   6 KB | 1,045 downloads
File Type: ex4 MGH EA_v3.1.ex4   66 KB | 986 downloads
  • #832
  • Nov 10, 2014 8:48am Nov 10, 2014 8:48am
Hallo SwingMan,
thanks a Lot for the Last EA!!!
Looks Great for me.
Edwin
  • #833
  • Nov 10, 2014 10:53am Nov 10, 2014 10:53am
still the old system, v2.3
test, GJ, today...
the 3 set of trades produced a total of 15 positions, and +38$ profit /clear, after already from the 27$ commission deducted/
would the new system make considerably different results, or there is no significant difference between the 2 versions.
Attached Image (click to enlarge)
Click to Enlarge Name: MGH GJ.png Size: 27 KB
there is always, always another trade!!
  • #834
  • Edited 1:06pm Nov 10, 2014 12:14pm | Edited 1:06pm
Based on reported backtest results by Kfx (as mentioned before) it seems that with the current settings the "Death Trail" could not be avoided even with 50% Global SL, so based on multiple suggestions in this blog I would support diversification between multiple (20+) pairs to reduce the risk of a "Death Trail".

In addition, reducing BatchX/Trade2 positions (as per SwingMan) would give a chance to further reduce the chance of a "Death Trail".

Even with a slight technical glitch after the weekly VPS restart the forward test is still looking good (even "too good"), i.e. +9% net profit over the past 2 weeks. It also means that there is still room to accept less profit for less risk.

Attached Image (click to enlarge)
Click to Enlarge Name: MGH 1110_1.jpg Size: 886 KB
  • #835
  • Edited 4:41pm Nov 10, 2014 4:16pm | Edited 4:41pm
Here is a new EA version 4.1

- Calculation of different lot sizes for each batch and trade 1 or 2, as in version 3.1
- Trend direction is calculated with EMA Slope on M15 for the period 32, which roughly corresponds to the H1.
- Variable grid intervals calculated with the ATR. For this reason I tested/recomand the M15 time frame.
(For me it is always difficult to take or to optimize fixed grid spaces, so the ATR)

If you test test this version, you'll probably find that it is more stable than 3.1.
Attached Image(s) (click to enlarge)
Click to Enlarge Name: EURUSDM15-test41.png Size: 32 KB
Click to Enlarge Name: EURUSDM15-report41.png Size: 50 KB
Attached File(s)
File Type: ex4 MGH EA_v4.1.ex4   76 KB | 1,133 downloads
File Type: tpl MGH EA_v4.1.tpl   1 KB | 992 downloads
  • #836
  • Nov 10, 2014 5:01pm Nov 10, 2014 5:01pm
Here the test results with "Basket_TakeProfit=5"
Attached Image (click to enlarge)
Click to Enlarge Name: EURUSDM15-report41-$5.png Size: 50 KB
  • #837
  • Nov 11, 2014 1:12am Nov 11, 2014 1:12am
Awesome Swingman.
Unfortunately no 90 % modeling quality ????
  • #838
  • Nov 11, 2014 2:19am Nov 11, 2014 2:19am
Hi,SwingMan
Found 4051 errors, what am I doing wrong?

Attached Image (click to enlarge)
Click to Enlarge Name: 4051eror.PNG Size: 46 KB
  • #839
  • Nov 11, 2014 2:51am Nov 11, 2014 2:51am
Quoting yczzd
Hi,SwingMan Found 4051 errors, what am I doing wrong? {image}
No idea what your platform makes.
In any case I see that "Order Grid = 0", and so the put of an TP=0 caused the error message.
Attached Image (click to enlarge)
Click to Enlarge Name: USDCHFM15.png Size: 21 KB
  • #840
  • Nov 11, 2014 3:15am Nov 11, 2014 3:15am
Quoting SwingMan
Here is a new EA version 4.1 - Calculation of different lot sizes for each batch and trade 1 or 2, as in version 3.1 - Trend direction is calculated with EMA Slope on M15 for the period 32, which roughly corresponds to the H1. - Variable grid intervals calculated with the ATR. For this reason I tested/recomand the M15 time frame. (For me it is always difficult to take or to optimize fixed grid spaces, so the ATR) If you test test this version, you'll probably find that it is more stable than 3.1. {image} {image} {file} {file}
@SwingMan

You mentioned: "Variable grid intervals calculated with the ATR"

Could you please describe how grid interval calculation is done based on ATR? Thanks!
0 traders viewing now