Forex Factory (https://www.forexfactory.com/)
-   Trading Systems (https://www.forexfactory.com/forum/71-trading-systems)
-   -   MGH System (Martingale Grid & Hedging) (https://www.forexfactory.com/thread/497448-mgh-system-martingale-grid-hedging)

FxMasterGuru Aug 10, 2014 3:28pm | Post# 401

One more "brainstorming" idea about TP:

ORIGINAL:
First Batch: 10 or 30 pips.
Hedgex2: 100 pips
Hedgex4: 250 pips
Hedgex8: 1000 pips
Hedgex16: ? pips

ALTERNATIVE I:
First Batch: 10, 30 or 60 pips (as per Modified Strategy)
Hedgex2: 30 vs. 60 pips (Original vs. Modified)
Hedgex4: 30 vs. 60 pips (Original vs. Modified)
Hedgex8: 30 vs. 60 pips (Original vs. Modified)
Hedgex16: 30 vs. 60 pips (Original vs. Modified)

  1. In this case we would use Hedge trades simply to get the maximum we would get from First Batches.


ALTERNATIVE II:
First Batch: 10, 30 or 60 pips (as per Modified Strategy)
Hedgex2: 120 pips (Original vs. Modified)
Hedgex4: 240 pips (Original vs. Modified)
Hedgex8: 500 pips (Original vs. Modified)
Hedgex16: 1000 pips (Original vs. Modified)
Hedgex32: 2000 pips (Original vs. Modified)

  1. This alternative multiples the maximum profit we would get from First Batches by the number of the most recently used Hedge multiplier.


corliz Aug 10, 2014 4:14pm | Post# 402

Hi

Why complicate stuff.

A martingail system will always profit IF YOUR POCKETS ARE DEEP ENOUGH? So the death trade comes when you catch a trend.

A grid system will buy (enter) trades with a trend.

Why not combine the 2 systems?

While martingail goes into massive drawdown the grid system will pile up the pips. It will hedge you against drawdown.

The martingail will double the lots with every entry but the grid will not double its lots. Maybe the distance of the grid must be smaller than the martingail.

You will always be heavily towards the martingail. Can this be made in ea?

spyderman Aug 10, 2014 4:24pm | Post# 403

One more "brainstorming" idea about TP: ORIGINAL: First Batch: 10 or 30 pips. Hedgex2: 100 pips Hedgex4: 250 pips Hedgex8: 1000 pips Hedgex16: ? pips ALTERNATIVE I: First Batch: 10, 30 or 60 pips (as per Modified Strategy) Hedgex2: 30 vs. 60 pips (Original vs. Modified) Hedgex4: 30 vs. 60 pips (Original vs. Modified) Hedgex8: 30 vs. 60 pips (Original vs. Modified) Hedgex16: 30 vs. 60 pips (Original vs. Modified) In this case we would use Hedge trades simply to get the maximum we would get from First Batches. ALTERNATIVE II: First Batch: 10, 30 or...
Ultimately that's a decision for each trader to make. Its nor as if everyone has to take the same
My greater concern is lot sizing. Typically I base that on desired risk factored with stop loss size. Obviously that's not applicable here so have to try and determine in advance our risk to help determine sizing

hanover Aug 10, 2014 8:06pm | Post# 404

My idea is to have a Max. DD of -5.000 in a grids serie, including open and closed positions. At this point we will close all trades. This represents a 5%, 10%, 15%, 20% of your account ... Depending on how aggressive trader you are. Another idea is close all trades in some hedging level, for example close all trades at x32. Any suggestions would be greatly appreciated.
If I may attempt to read between the lines, you're attempting to devise a system that is capable of surviving, despite its having a martingale component.

Price movement is pretty much random, hence you must assume that if you trade for long enough, you'll encounter every conceivable price pattern; or put another way, all price patterns are equally likely to occur, albeit in an unknown sequence. Grid systems are especially vulnerable, in that they don't perform analysis specifically aimed at identifying and trading only situations that offer potential bias. If the price sample being traded is ostensibly random, then, it also means that all systems, no matter how simple or complex, have an equal chance of success or failure.

If you are serious about risk management, you must attempt to calculate the probability of account loss, and then somehow find a MM model that reduces this risk as much as possible. I expect that this will prove to be your biggest challenge.

We all know that grids and martingales are dangerous but please abstain from making comments debating whether these techniques are good or bad.........
No comment.
........ and focus on the system.
Well, I did my best, given the latitude that you allowed me.

bailingzhou Aug 10, 2014 11:54pm | Post# 405

Hi Why complicate stuff. A martingail system will always profit IF YOUR POCKETS ARE DEEP ENOUGH? So the death trade comes when you catch a trend. A grid system will buy (enter) trades with a trend. Why not combine the 2 systems? While martingail goes into massive drawdown the grid system will pile up the pips. It will hedge you against drawdown. The martingail will double the lots with every entry but the grid will not double its lots. Maybe the distance of the grid must be smaller than the martingail. You will always be heavily towards the martingail....
yes, it's good idea. I have thought it before.
While it bring smaller risk (use gird to hedge), also bring smaller profit.
And the problem is the SL/TP of your gird. There are many details confused me.
Welcome your wise opinion~

corliz Aug 11, 2014 4:58am | Post# 406

{quote} yes, it's good idea. I have thought it before. While it bring smaller risk (use gird to hedge), also bring smaller profit. And the problem is the SL/TP of your gird. There are many details confused me. Welcome your wise opinion~
Hi

I do not understand "smaller profits"? You get double profits when you hit a long trend. It is like having 2 separate EA's. In normal conditions both will profit. In trend both will profit again, while one hedges the other's draw down.

Only problem with martingale is if you hit a long trend.

Grid will make profit because it follows the trend. Grid can even enter at pull backs

Martingale will profit as soon as trend turns around.

Grid only starts when 2nd trade of Martingale enters. Close all (Martingale and grid) when martingale profits. See what the grid gives.

Somebody write a EA and lets see what happens?

sugarman Aug 11, 2014 7:24am | Post# 407

Hi Kfx

This is a very well thought out system and probably the best I have seen with regards to grid trading, however like all grid and martingale systems you will I think eventually blow your account.
Great work anyway.

gridmaster Aug 11, 2014 9:18am | Post# 408

is this correct as i understand modified version of mgh.
buy/ sell 1 lot on grid level 1 .............tp=10/30/60
grid level 2 buy / sell with 2 lot.................tp=10/30/60
grid level 3 buy / sell with 4 lot.....tp=10/30/60
as so on. this correct '''''
in other word buy stop / sell stop order having multiplying lot on having constant grid level having fix take profit.

is this correct.......

Paracelsus Aug 11, 2014 9:44am | Post# 409

I agree to concentrate our efforts more on risk and money management than technical issues, as it seems that small modificiations (incl. entry direction at round numbers and First Batch targets) do not really matter.

Each Martingale strategy has a "Death Pattern" or "Death Trail", i.e. a series of specific market moves which lead to the death of the account. Similarly to an imaginary "Death Labyrinth", which has thousands of paths to safe exits, except for one, which leads to a death trap. I call this one path the "Death Trail" which we should avoid by reducing its statistical chance of getting on it.

What can reduce the statistical chance of getting on the "Death Trail":

1. Reducing the lots (with a given margin).
2. Reducing Target Profit. The purpose of it would be getting an early chance to start a "new sequence", i.e. a new First Batch again. Each time we get back to "First Batch" we reset the distance to the "death exit" and statistical chance of getting on the the "Death Trail". In other words: at Hedgex16 the statistical chance of travelling through the "Death Trail" is significantly higher than at the start of a "First Batch" or even at "Hedgex2".

In addition to the reducing the statistical risk of getting on a "Death Trail" it is equally important to collect reasonable profits, so finding the balance between these two would be the ultimate goal.

Ata-Turkoglu Aug 11, 2014 10:38am | Post# 410

I agree to concentrate our efforts more on risk and money management than technical issues, as it seems that small modificiations (incl. entry direction at round numbers and First Batch targets) do not really matter. Each Martingale strategy has a "Death Pattern" or "Death Trail", i.e. a series of specific market moves which lead to the death of the account. Similarly to an imaginary "Death Labyrinth", which has thousands of paths to safe exits, except for one, which leads to a death trap. I call this one path the "Death Trail" which we should avoid...
Hi Paracelsus..

Thak you for sharing info about the "Death Trail" possibility.. regarding to that I would also like to share my conclusion about MGH after studying it since the thread started (and I am still studying it):

Let's say we are trading the 5M Timeframe AND we believe the 1H trend is LONG:

Then, if we luckily have started buying inside an Elliott-Wave 2 retracement (or at least a place near fibo 50, 61.8) of a long trend on 1H chart, our SIMPLE strategy will work so nicely.

Even if we have to open a HedgeX2 short trail for a while, since we assume the big trend is upwards, sooner or later price will go up, leave 3 HedgeX2 shorts open, but will close the SimpleX1 long trail that we had to leave open earlier and it will end up with a successful HedgeX4 long trail with a nice profit over+250 pips.

---> but I have to repeat: This situation could happen only if we could have started inside a 2nd Elliott Wave of a long trend of 1H chart (or fibo levels 50,61.8).
So.. this means we end up with the exact same question all Elliott Wave fanatics ask themselves: How will we be sure we are inside an Elliott Wave 2 of the overall trend?

In a nutshell: If you want to trade with MGH strategy, try to detect a 2nd EW and start there. Starting anywhere else is pure luck or at least it will easily lead you to x8 or higher hedge situations I think.

P.S: Replace all my "long" words with "short", and all "short" words with "long" and you will get the downtrend version of the same idea.

Happy Trades

spyderman Aug 11, 2014 10:59am | Post# 411

is this correct as i understand modified version of mgh. buy/ sell 1 lot on grid level 1 .............tp=10/30/60 grid level 2 buy / sell with 2 lot.................tp=10/30/60 grid level 3 buy / sell with 4 lot.....tp=10/30/60 as so on. this correct ''''' in other word buy stop / sell stop order having multiplying lot on having constant grid level having fix take profit. is this correct.......

No I think what FX meant by the 10/30/60 is that would be the variance in take profit on x1 based on whether you had 1, 2 or 3 orders open. The TP for all 3 trades is 1 level away from the 1st entry. So if you have just entry 1 and it hits TP then 10 pips on a 10 pip grid. If you have 2 open, the 1st is 10 pip TP and the 2nd is 20pips...so 30 pips total. If you have 3 orders open the 1st is 10 pips, 2nd 20 pips and the 3rd is 30 pips...so 60 pips total. 10/30/60

spyderman Aug 11, 2014 11:21am | Post# 412

{quote}Price movement is pretty much random, hence you must assume that if you trade for long enough, you'll encounter every conceivable price pattern; or put another way, all price patterns are equally likely to occur, albeit in an unknown sequence.

David,

Much respect and thanks for commenting. I was wondering when you would stumble by, as I've followed your thoughts on many martingalesque threads over the years. As always your take is always well thought out and logical.

I tend to agree with your thoughts in general and on this specifically, especially about the quoted nature of price movement being pretty random, especially in the short term... What is price doing next...up? down? flat? No knows.

However I would hesitate to conclude that, because price movements in general are random, that that would necessarily mean that a trader will encounter every conceivable price pattern or that all patterns are equally likely to occur.

The reason I say that is because, while price movement may be random, there are non-random forces in play in the market that will, at the least mitigate, if not prevent certain "patterns" from manifesting themselves.

For example, in a totally random data set, it would seem evident that one possible outcome of market movement would be a price action that continues up and up with no retracements whatsoever. However, that isn't feasible in the markets we trade...especially forex.

Nation currencies are involved, and those nations interests are tied to the valuation of their currencies. Nations can and do take steps to keep their currencies from inflating or deflating past a certain point. They're not always entirely successful, but the effect of what they attempt isn't random and does affect price patterns.

Investors, fund managers, national banks, etc. take profits and the like based on price movement and that changes order flow, so again those actions effect price patterns in a non-random fashion.

It would seem possible then, and perhaps that is just wishful thinking, to devise a trading system with elements of martingale and grid trading, that is flexible enough to take advantage of these non-random price movements. At least enough to siphon off some profits from the system.

Paracelsus Aug 11, 2014 12:42pm | Post# 413

1 Attachment(s)
Well, it seems we have a group of enthusiasts and experts getting together here. All inputs are appreciated. Ata-Turkoglu is right about getting an "edge" with the "First Batch" entry, however, waiting for the "right moment" might rob us of precious other opportunities to make profit.

Getting back to the "Death Trail": with any Maringale, its statistical chance/occurance should be calculated. On purely mathematical basis it would be practically impossible, however, a 10-year backtest would give a "good idea".

For example: If the strategy gets onto the "Death Trail" once every year (statistically) losing 50% of the account value, but makes 15% per month, then - statistically - it is winner strategy (11 x 15% - 1 x 50% = 110%). Lets say it encounters the "Death Trail" twice a year: 10 x 15% - 2 x 100% = 50% per year. Still not bad. Of course, the two 50% losses may come one after an other, so I would experiment to find the right balance between profit and "Death Trail" risk until the statistical chance of getting on the "Death Trail" would be less than ONE occasion per year. Again, there are no guarantees, one has to play the statistical chances as in any gambling.

One might argue that it is not a "trading strategy" in its real sense. I agree. But I could not care less from my luxury yacht cruising the Caribbean Islands...

Click to Enlarge

Name: Mine is bigger.jpg
Size: 78 KB
P.S. If my entries resemble of FxMasterGuru's, it is not a coincidence... Something has happened to his login, so I am "replacing him" for now...

Ata-Turkoglu Aug 11, 2014 1:08pm | Post# 414

...One might argue, that it is not a "trading strategy" in its real sense. I agree. But I could not care less my luxury yacht cruising the Caribbean Islands... {image}
That was a good one

Happy Trades

hanover Aug 11, 2014 6:21pm | Post# 415

David, Much respect and thanks for commenting. I was wondering when you would stumble by, as I've followed your thoughts on many martingalesque threads over the years..........
Spyderman,

Great points.

You’re correct, of course. The probability of price moving (say) 50 pips in an hour is always going to be greater than its moving 5,000 pips. Or you could just as easily have argued that there are an infinite number of possible patterns, and since nobody lives long enough to trade for an infinite amount of time, it’s impossible for them to encounter every conceivable pattern.

The only thing I would query in your post is that IMO central bank intervention is perhaps one of the biggest causes of unexpected spikes and outlier-type patterns.

I still stand by what I said as a broadbrush principle, even if it’s not absolutely correct, i.e. that it’s nigh impossible to anticipate all of the permutations that the market will throw at you, and thus build a system that is absolutely safe from an incalculable number of one-off possibilities. Or put more simply, how can you expect the unexpected?

IMO this is an absolute must read for anybody who’s developing a system, martingale or otherwise. These guys are much more experienced and knowledgeable traders than I’ll EVER be (soultrader was something of a legend when he used to post here at FF). Here’s the conclusion that they reached: One thing is for sure. no matter how good your system appears to be, no matter how much money it has already made and no matter how much confidence and testing you put into it - sooner or later a market condition will surface which you haven't seen before. we never know what is around the corner, you could run the gauntlet and have 10 years or more of great results. but sooner or later the market will throw off a wild one.

I guess that summarizes the message that I was trying to convey.

I want to respect the OP’s wishes, and not disrupt the flow of the thread, so I’ll leave it at that. And in any case, there’s no point in repeating material that I’ve posted many times in other martingale threads.

I wish everybody great success with this system.

Respect,
David

spyderman Aug 11, 2014 6:40pm | Post# 416

{quote} Spyderman, Great points. You’re correct, of course. The probability of price moving (say) 50 pips in an hour is always going to be greater than its moving 5,000 pips. Or you could just as easily have argued that there are an infinite number of possible patterns, and since nobody lives long enough to trade for an infinite amount of time, it’s impossible for them to encounter every conceivable pattern. The only thing I would query in your post is that IMO central bank intervention is perhaps one of the biggest causes of unexpected...
Thanks again David. I agree whole-heartedly about the unexpected showing up and I'll take a read on your link for sure. Thanks again for stopping by.

spyderman Aug 11, 2014 6:44pm | Post# 417

Seems like Kfx has gone on sabbatical. Hope all is well. I pm'd him, so we'll see if he's still around.

Paracelsus Aug 12, 2014 4:14am | Post# 418

{quote} “One thing is for sure. no matter how good your system appears to be, no matter how much money it has already made and no matter how much confidence and testing you put into it - sooner or later a market condition will surface which you haven't seen before. we never know what is around the corner, you could run the gauntlet and have 10 years or more of great results. but sooner or later the market will throw off a wild one.” I guess that summarizes the message that I was trying to convey.
Excellent observations and very true quotation, David. Thanks. I totally agree about its validity for "trading systems".

As I have already mentioned I consider this Martingale/Hedge strategy not a "trading system" in its classical sense. Why? For this strategy does not really matter which direction the market goes, it should just go "somewhere" without encountering a few specific price movement patterns which lead to the "Death Trade". (The theoretically possible pathways to the "Death Trade" I call "Death Trails".) In this sense, it is more like a gambling strategy applied to the Forex markets, so it may be wiser to approach it with a "GAME THEORY" mindset.

There are infinite price patterns (pathways) from "First Batch/First Trade" to one of the final outcomes: Winning or Losing. Among these there are infinite "Winner Trails", however, there are only "finite", i.e. limited number of "Death Trails". Just as a very simplified Roulette example: If you played "red/black" Martingale and you bet on red, the "Death Trail" of sequences would be: all black. In this case there is only one "Death Trail". This strategy's potential outcomes are much more complex than in roulette, not to mention other variables (TP, SL, MM) which makes it a more promising "game" than roulette, but also more challenging.

It seems that this thread is losing stream, especially that the inventor, KFx has not commented on any brainstorming ideas. Anyway, even if this thread dies off, future Martingale enthusiasts should find some "pearls" in this thread, so I encourage them to read it.

spyderman Aug 12, 2014 7:21am | Post# 419

1 Attachment(s)
Just a quick update on my EurUsd 50 pip grid.
You might notice a small modification...

Click to Enlarge

Name: EURUSD!H1.png
Size: 62 KB

kaczucha11 Aug 12, 2014 9:37am | Post# 420

1 Attachment(s)
Today's trades +60 pips

Click to Enlarge

Name: GBPJPYM5.png
Size: 57 KB

regards


© Forex Factory