MGH System (Martingale Grid & Hedging)

Sincerely hope，Intentioned people who be able to systematically sort out the detailed rules for this strategy, like Doc document, or flow chart made by Ata.

So that users like me with poor English can learn and research.

many thanks

Stage 4 was pretty crappy.

1/1/2004 - 5/12/2005
4,525 pips total with 1 series with a serious DD of 12,562 pips. Yuck!
Went to x8

May be a suggestion, set a limit on losses? How many positions where the system is winning and how many where it is not? If we have more figures where MGH developed winning positions than where it didn’t, hence set a limit in pips to losses may seem like a good strategy?

Perhaps a few interesting ideas can be found in the SnowBall system: https://sites.google.com/site/prof7bit/snowball

Hi Spyderman.
I see in your trade explorer that you do very well in this strategy
I test 50 pips grid also with good results at the time, a little slow but without any stress.

regards

I don't use Martingaling per se (I don't double up) and can't hedge (I'm in the U.S. -- which blows, by the way), but I can offer a couple of tips nevertheless with respect to grid sizing, since I use directional grids for certain types of trades.

A hedging grid system is non-directional, meaning that you set up your hedging grid in relation to current price without regard to the direction the pair is moving. In comparison, Martingaling is a counter-trend system that trades in a direction opposite the trend and waits to take advantage of a retrace. The two systems, I think, are distinct, since a gridded hedging system does not necessarily require Martingaling to be successful.

At any rate, two things are critical with Martingaling: (1) grid size; and (2) point of entry.

With respect to grid size: too big a grid and you never hit your levels; too small a grid and the doubling of lot size essential to the success of the system becomes quickly overwhelming, since the first position is 1x lots, the second 2x lots, the third 4x lots, the fourth, 8x lots, and so on and so forth. Martingale systems that double the lot sizes as each level in the grid is hit without limits as to the doubling or drawdown generally fail in short order.

Set a sensible limit as to the total size of the net position that will result from all the levels being hit without retrace in relation to your equity. For example, if you have $10,000 equity, limiting the doubling to .01 lot, .02 lots, .04 lots, .08 lots (for a total of .15 lots) would probably be reasonable, since the total exposure would be less than 2 x equity. Additionally, using stop losses will provide additional equity protection should the anticipated retracement not occur.

Determining an appropriate grid size, in my mind, is pair specific. For example, a pair like EUR/CHF experiences far less intraday movement than a pair like GBP/NZD. While you might arguably space your grid 5 pips apart for a pair like EUR/CHF (Daily ATR is currently 18 pips) (point in fact, that would probably not be worth it due to spread), that would make little sense with GBP/NZD, whose current Daily ATR is currently about 120 pips. In short, the grid should be sized to the pair, taking into account an ATR of some kind, whether it be Daily, Weekly, or some other time frame.

As far as point of entry is concerned, I would suggest that if you are going to use a Martingale system at all, you do not do it without regard to obvious S/R. For example, if you follow USD/JPY at all, you'll realize that 103.00 is an area that has provided resistance in the past. Price is driving up currently toward that level. Purely as an example, you would set up the following grid to take advantage of a retrace that the pair is likely to experience at the 103.00 level:

Sell 1x lots at 102.40/set target price for position to 102.30
Sell 2x lots at 102.60/move target price for net position to 102.50
Sell 4x lots at 102.80/move target price for net position to 102.70
Sell 8 x lots at 103.00/move target price for net position to 102.90

You can also use Pivot Points or Fib retracements/extension levels for this purpose, looking for retracements at particular levels such as the S3/R3 or the .764/.236. (Me personally, in the above example, I would probably just fashion a single entry order to short USD/JPY at the resistance level ... .)

Ultimately, you can see why an EA might be useful for this system. You either have to keep your eyes on the charts to adjust the target price as your grid levels are hit or have an EA that will adjust the target prices for you as the grid levels are hit.

Lastly, and again while I cannot use hedging here in the United States, pay attention to swap/rollover fees. They are frequently not the same on the buy side as they are on the sell, and some of them can be quite onerous (GBP/NZD swap: 3.76 on the sell side, a whopping -8.32 on the buy).

In my opinion the "beauty" of this strategy is:

1. that it combines the advantages of Hedging, Grid and Martingale, and
2. the final profit of a basket comes when the market starts trending. In other words: it is not a usual Martingale which "bets" against the trend, but its Grid/Martingale component "bets" against the retracements of a trend, and profits nicely when the market moves again in trend direction.

Diversification (i.e. running it on multiple pairs), position sizing and setting reasonable SL/TP levels are vital prerequisites for this strategy, which should be seriously considered if anyone is planning to trade it on a live account. The real challenge is finding the "right settings"...

So much for USD/JPY resistance at 103.00 ... . Blew right through that on weak Japanese trade data.

But if you'd taken that Martindale trade example, you'd only need the pair to retrace 40 pips to hit your target price ... .

Just a quick update on the pairs I'm trading...for what its worth.

Has any tried to run this with a broker like Oanda?

Oanda's set-up is kind of different in the respect that you don't buy lots, but rather single units of the currency. I imagine that this is to allow for micro account traders to trade more proportionally to their equity size. With brokers like FXCM (not to knock them; most brokers are set up the same way), you're confined to 1K increments. Naturally, if you have huge equity, this is not a problem, but for smaller account traders (I would say $10,000 or less), Martingaling even with 1K lots could get you in trouble fairly quickly, depending upon how your grid is set up.

Now, setting up with 1 unit sizes, you could double 1, 2, 4, 8, 16, 32, 64, 128, 256, 512, 1024 before trouble would begin to appear on the horizon. However, I guess the trouble there would be if you consistently hit your target price around the 3 or 4th doubling and very rarely hit it at higher doubling levels. You'd frequently make next to nothing ... .

forex warrier ea google serch is based on this concept

Hi guys. How is everything?

I've been working on trying to mitigate the high drawdown as possible. So far I have not succeeded, but I'm working on it and in new ways to create the grids. I hope soon to give you good news and ideas to work on it.

Best wishes to all.

Hi ZebraSquirl!

Yes, Oanda is a good broker to try this system starting with 1 unit. I think there is some other MT4 brokers with nano accounts starting with 100 units (0.001 lots).

The only problem with Oanda is that hedging is not allowed due the NFA rules, even in Oanda Eaurope division.

The only solution for that is to open 2 subaccounts, one for Buy's and another for Sell's.

Kfx.

Look forward to seeing what you come up with.

I use ForexBrokerInc. for micro/nano accounts. They offer 1,000 unit lots and you can trade .01 lots.

Hi spyderman.

1.000 units (0.01 lots) is micro, not nano/cent account. I think nano/cent account would be 100 units (0.001) ..... [Not 0.0001,Sorry my mistake in previosus post, fixed now]

example in USD micro account ........... 1 pip x 0.01 lots = 10 cents.

example in USD nano/cent account ... 1 pip x 0.001 lots = 1 cent.

As mentioned before, keeping the TP reasonably small (e.g. 30-60 pips) and unchanged - i.e. even after initiation of H2, H4, H8 hedges - would give a chance to get out of baskets sooner and therefore limiting the risk of large DD (which is obviously increased while waiting for larger profits after multiple hedging).

Also, setting a reasonable and tolerable fix SL is a must.

This kind of points out why forbidding hedging is stupid. The work around for that is to buy in one account and sell in another account. It is stupid that I have to set up two accounts to do this.

Unfortunately, no good, similar workaround for FIFO ... .

Hey spyderman, you have a PM ...

Anyway, later I'll post an entry criterion for the system, which I'm working.

Sounds good. Do you have some example of the worst trade (SL reached) and how many pips you lost vs. how many pips you win?

But in a "Hedge x8", you can have more than 500 pips DD ... If the TP is 60 we will need 8/9 win trades to cover 1 loss. Surely I'm wrong and maybe I have not understood your point. It would be great you post some chart.