stevehopwoodforex.com
https://www.stevehopwoodforex.com/phpBB3/
Print view

Martingale debunked by Gary
https://www.stevehopwoodforex.com/phpBB3/viewtopic.php?t=547		 Page 9 of 11
Author:  themaxx [ Wed Jun 12, 2013 3:45 pm ]
Post subject:  Re: Martingale debunked by Gary
garyfritz wrote: So with 0.02-size bets, you make about as much profit as the Martingale makes with as much as 0.64-size bets. WITHOUT the gut-wrenching losses and ridiculous risks.

AND, since the 0.02-size non-Mart bets have a fixed and known risk size, you can actually scale them up much larger. If you determined the Martingale's 0.64-lot bet was your maximum tolerable risk, imagine what would happen to your profits if you traded a NON-Martingale strategy with, say, 0.10 or 0.20 or 0.50 lots per bet...
Gary could you please clarify what position sizing method the "scaled non-Martingale" and the "Non-Martingale" uses? I don't understand what you mean by "Scale = 2.0 means you double your position size". Is it just 1% of equity and 2% of equity (or 2.5% of equity if scale = 2.5)? Thanks!

Pete
Author:  garyfritz [ Wed Jun 12, 2013 4:58 pm ]
Post subject:  Re: Martingale debunked by Gary
Actually the simulator spreadsheet might be a bit confusing on that point. It uses **FIXED** position sizes. That's why it can go so negative if the simulated system starts losing. If you were betting a percentage of your account, you couldn't go negative.

So, in that spreadsheet, you specify the win size & loss size in cells B6 & B7. The "position size" is a multiplier of that specified win/loss size.

Let's assume you set Win$ to +200 and Loss$ to -100, and the Non-M scale value (cell B9) to 3.0.

* The "Non-Mart" system wins 200 when it wins, and loses 100 when it loses. That's a basic position size of 1.

* The "Scaled non-Mart" system does the same, but it multiplies the position size times the scale value in B9. So it wins 3*200 = 600 when it wins, and loses 300 when it loses. Its position sizes are 3x larger than the "non-Mart" system (because B9 is 3.0).

* The "Martingale" system starts out at the basic Win$/Loss$ values, multiplied times the Martingale progression in cells B13-U13. So with the default progression of 1,2,4,8, etc, you start out winning/losing +200/-100. If you lose, then on the next trade you double your lot size (because C13 = 2), and you win or lose +400/-200. If you lose again, on the next trade your lot size is 4x larger than the base lot size (because D13 = 4), and so on. When you win, you reset the Martingale level to 1 and return to cell B13 in the progression.

Does that clarify it?

I should probably modify it to use percentage risk instead of fixed position sizes...
Author:  themaxx [ Thu Jun 13, 2013 10:04 am ]
Post subject:  Re: Martingale debunked by Gary
garyfritz wrote: Does that clarify it?

I should probably modify it to use percentage risk instead of fixed position sizes...
Yes, perfectly. I always knew Martingale was a steaming pile of £$%^& - this spreadsheet clarifies it visually.
Author:  garyfritz [ Mon Aug 26, 2013 11:54 pm ]
Post subject:  Re: Martingale debunked by Gary
FYI: Some observant types might notice that a series of posts from the last few weeks have suddenly disappeared. They ran afoul of Steve's intolerance for things that might endanger his members.

A member posted claims that it was safe and very profitable to trade Martingales if you did it properly. On several occasions I questioned this, and showed math examples that indicated his claims could not be true -- unless I totally misunderstood his approach. I asked him for clarification. After my latest post proving (I think) that his claims could not be true, he posted "Enough has been said. It is up to you to figure it out. I cannot say any more due to a confidential agreement with original creator of this system."

Steve blew a fuse (his words :)) and basically said "Either you prove your claims, and show people HOW they can trade it safely, or I will remove the posts to prevent unsuspecting members from blowing up their accounts based on your unsubstantiated claims." The member declined, and asked Steve to remove the offending posts.

There is a slim possibility that this member actually does have a safe and working Martingale approach. But if he will not explain it, Steve will not permit undocumented boasting that might lead others to ruin.

More likely, in my opinion, is that this member has not sufficiently studied his approach, and doesn't fully understand his risks. He merely THINKS it's safe because he has not (YET) spun to the wrong chamber in his game of Russian roulette. He may make incredible returns for a good while, and I wish him luck -- but most likely he will eventually fall victim to the Martingale's explosive self-destruction like so many unwary traders before him.

Our firm belief is unchanged: Martingales, the way most people trade them, are guaranteed to blow your account sooner or later. If you trade them with defined risk, as in my updated spreadsheet, then they are not much more likely to blow your account than a standard fixed-risk method -- but they don't provide any benefit either.

If anyone wants to claim otherwise, they had better back it up with clear explanations that can be verified by others.
Author:  dudest [ Tue Aug 27, 2013 11:13 am ]
Post subject:  Re: Martingale debunked by Gary
@Gary:
LIKE.jpg
Author:  nanningbob [ Fri Nov 29, 2013 12:20 pm ]
Post subject:  Re: Martingale debunked by Gary
garyfritz wrote:FYI: Some observant types might notice that a series of posts from the last few weeks have suddenly disappeared. They ran afoul of Steve's intolerance for things that might endanger his members.

A member posted claims that it was safe and very profitable to trade Martingales if you did it properly. On several occasions I questioned this, and showed math examples that indicated his claims could not be true -- unless I totally misunderstood his approach. I asked him for clarification. After my latest post proving (I think) that his claims could not be true, he posted "Enough has been said. It is up to you to figure it out. I cannot say any more due to a confidential agreement with original creator of this system."

Steve blew a fuse (his words :)) and basically said "Either you prove your claims, and show people HOW they can trade it safely, or I will remove the posts to prevent unsuspecting members from blowing up their accounts based on your unsubstantiated claims." The member declined, and asked Steve to remove the offending posts.

There is a slim possibility that this member actually does have a safe and working Martingale approach. But if he will not explain it, Steve will not permit undocumented boasting that might lead others to ruin.

More likely, in my opinion, is that this member has not sufficiently studied his approach, and doesn't fully understand his risks. He merely THINKS it's safe because he has not (YET) spun to the wrong chamber in his game of Russian roulette. He may make incredible returns for a good while, and I wish him luck -- but most likely he will eventually fall victim to the Martingale's explosive self-destruction like so many unwary traders before him.

Our firm belief is unchanged: Martingales, the way most people trade them, are guaranteed to blow your account sooner or later. If you trade them with defined risk, as in my updated spreadsheet, then they are not much more likely to blow your account than a standard fixed-risk method -- but they don't provide any benefit either.

If anyone wants to claim otherwise, they had better back it up with clear explanations that can be verified by others.
Hi Gary, sorry I have not kept up with these posts. I could have helped you out. I have studied martingale systems and methods thoroughly over several years. I think I have a very good understanding of how they work and why they fail. In reality and if I explain this right, you will understand why all martingale systems fail if left to them selves.

1 Biggest problem #1. The one problem I could never solve and never did was the taking of a trade at a peak of a move. If you programmed the EA to trade with the trend it would take a trade at a peak of a move and then it would go against you. The same problem if you counter traded the trend. No matter which way you programmed the EA to work, no matter what indies you used, you always ended up with a trade at the peak of a move and then a reversal. I labeled this trade at the time the death trade and that name has kind of stuck at FF. Now you have the problem of: that is the trade you have to get out of because it is the one that will kill you. So the question became when do you bail on that trade or series of trades because it aint going back.
2. This lead to the second biggest problem. The range of the martingale. For example if you trade 1,2,4,8,16,32,64. You are now 7 levels deep. If your reentry is 100 pips you have a range of 700 pips in which to get the re-tracement. If the re-tracement during that 700 pip run is never the amount or 100 pips from your re-entry point to even the trade you are cooked. Remember you could go to 699 and not get to the next re-entry level and it retrace to 501 (198 pip re-tracement) and not hit your clear out. So on some points you need more than 100 pip re-tracement for the martingale to work.
3. So you say no problem- you go 200 pips and your range is doubled to 1400. But now you need a reentry to return 200 pips to breakeven the move and face a potential trade of 398 pip move retracement that would never make it work. Well, you may not get a 200 pip move on one of your reentries so your goose is cooked. Same thing with 300,400 or more reentry levels.
4. #2 and #3 are insolvable. Your range will either be too small and price will over run your range or your range will be too big and never retrace to a solution. Both of these issues cannot be resolved, there will always be a price action move that will beat either scenario.

How to solve the problem? In my studies I came to the conclusion that you could recover 3-4 levels deep and still come out ahead even if you took the occasional loss at at level 4 or 5 and you didnt double your lots. If you only kept 3 level maximum going and close a level every time you reentered a level so the max loss you would ever take would be 3 levels. Recovery would work mathematically if you stayed out of the death trade. I never went past 5. I just ate the loss. Because these were so few and far between eating the potential death trade kept me from experiencing it and enough recovery trades made breakeven that my winners would overcome the bailout trade. However ...............

The real issue becomes how good of a trader are you? I seldom use recovery any more because I have learned not to fight a trend and be on the wrong side of it. If the trend reverses just close out and you can still multi level trade but only with the trend on your side. That is a discipline that is the hardest for a trader to do. Gee. if I just hang in there a little while longer .............. I will do that if the trend is still on my side, I will hang in there, however, if it is not it is time to eat the trade and move on.

I have never seen a Martingale system that has been able to solve problems 1, 2, or 3. Multi level trading or using part of you total lots to enter a trade is used by professional traders all the time but it is to add to winning positions or phasing into a trend direction not to bail themselves out of a bad trade going in the wrong direction.

Just my thoughts from someone who really tried to solve issues 1, 2, and 3.
Author:  giailang [ Fri Nov 29, 2013 1:32 pm ]
Post subject:  Re: Martingale debunked by Gary
Thanks NB. Totally agree of fitting the series within the right trend, and splitting volume to smaller pieces for that series.

Re the member who has faith in Martingale: If he can watch realtime PL chart of his EA, it can prove your opinion better.
I myself blew one account by grid-scheme EA (same lot size spreads over the grid, much safer than the Martingale you are mentioning of), with perfect EQ/BAL chart and acceptable DD, but real FD killed it after half year of real trade. Martingale is much more harmful than flat grid because of exponential allotment.

To prove the statement, he can take 10point3 EA (or DLM) and do backtest, then forward test. If he's lucky, $100 may become $2000 in a week, then badluck arrives, only one day is more than enough to toast the newly gained 2k account. The said EA has good signal detector for some market conditions, but not all. It can burn your account because of using Martingale scheme at wrong time, wrong trend.
Author:  nanningbob [ Sun Dec 01, 2013 2:34 am ]
Post subject:  Re: Martingale debunked by Gary
giailang wrote:Thanks NB. Totally agree of fitting the series within the right trend, and splitting volume to smaller pieces for that series.

Re the member who has faith in Martingale: If he can watch realtime PL chart of his EA, it can prove your opinion better.
I myself blew one account by grid-scheme EA (same lot size spreads over the grid, much safer than the Martingale you are mentioning of), with perfect EQ/BAL chart and acceptable DD, but real FD killed it after half year of real trade. Martingale is much more harmful than flat grid because of exponential allotment.

To prove the statement, he can take 10point3 EA (or DLM) and do backtest, then forward test. If he's lucky, $100 may become $2000 in a week, then badluck arrives, only one day is more than enough to toast the newly gained 2k account. The said EA has good signal detector for some market conditions, but not all. It can burn your account because of using Martingale scheme at wrong time, wrong trend.
10.3 I remember that EA. Was going to make me a bundle. I thought I had the problems 1,2,3 licked with that one and then it ran into the Oct. 2008. The thing is it almost caught the right direction of the trend and I would have made a killing ............ instead ............. well live and learn
Author:  giailang [ Sun Dec 01, 2013 6:27 am ]
Post subject:  Re: Martingale debunked by Gary
nanningbob wrote:10.3 I remember that EA. Was going to make me a bundle. I thought I had the problems 1,2,3 licked with that one and then it ran into the Oct. 2008. The thing is it almost caught the right direction of the trend and I would have made a killing ............ instead ............. well live and learn
Yep, our successful is made of broken bricks of failure. At least, 10.3 has good signal detectors for the normal market conditions (e.g. applicable when S/D is not distorted by currency intervention or not collapsed by extreme news).

Thanks to your idea, I made a set of detectors since 10.3, and have clearer view about money management. Also thanks to SH with his Beast series that helps me to correct my view of "recovery scheme" using the Fibonacci Allotment, especially when SH warned me when I tried to revive the beast. Since then, I use realtime PL plotting to see what's really happening to the EA.

From my point of view, I'd like to find the market conditions that tend to falsify the signal detectors I have. It means that even your 10.3's detectors are still good if we use it appropriately.
Author:  Thweis [ Thu Dec 19, 2013 10:54 am ]
Post subject:  Re: Martingale debunked by Gary
We also employ grid based EAs in our trading, however they all have a build-in equity protection and by combining several in a risk managed portfolio one can build a method with a favorable risk/reward ratio. In the end it doesn't matter if you have frequent small losses or rare large losses. Hower a system like ours which employs the latter must be traded automatically as it will be mentally demanding to trade it manually. So by cutting your losses short with a grid you remove the negative aspects associated with it.

Steve I am new to this forum coming from donna's page and I would love to hear your thoughts about our way to employ a grid.
http://www.thweis.com/index.php/our-mehod

v/r
Martin

Image
All times are UTC Page 9 of 11

Powered by phpBB® Forum Software © phpBB Limited