Martingaling - Looks so good until the inevitable | Page 8 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Looks profitable, Need good Coder help 6 replies

Martingaling for a Living 81 replies

PathFinder Trading Platform. Looks good. 1 reply

USD/KRW looks like a good short 0 replies

HEDGE looks good 43 replies

Attachments: Martingaling - Looks so good until the inevitable
Exit Attachments

Martingaling - Looks so good until the inevitable

  • #141
  • Aug 17, 2014 5:51am Aug 17, 2014 5:51am
Quoting Proximus
{quote} I just want to prove my point, but i dont think you get it with your tiny brain...But then i just looked at your previous 50 or so posts and see that you are quite a troll man. You are ForexFactory's chosen resident troll, i mean atleast 8000 of your 10k posts are just trollings. {image} Look in the mirror Magix and smile
The only point you've made is that you live in a world of fantasy...

You take trading wisdom from Hollywood productions and base market data on backtest reports.

Wake up son...

This is real life...not a dream...not a movie...

Out of my 10k posts, there are also 3500 charts....your point is invalidated.

Post more pictures, more movie quotes and more backtest data...you will still never be able to carry on a valid conversation until you start working with the market from a realistic standpoint. Stop being a moron. Start reading what is in front of you. Erase everything you think you know, because it's all based on bullshit...

It's only after you do this, you may be able to gain any sort of understanding.
Money Can't Buy Happiness. Poverty Can't Buy SHIT! You Choose!
  • #142
  • Aug 17, 2014 6:08am Aug 17, 2014 6:08am
Well I like the way we are proving what happens when you average down .. or "post average"...

You're value decreases significantly every time you simply "refuse to lose"

I think I might take a profit here...

Nite...

*and here's to a very special little man ... Just adorable
H.D
  • #143
  • Aug 17, 2014 6:11am Aug 17, 2014 6:11am
Quoting pipster1
Here is the concept. You are confident that the probability you of losing 3 in a row is extremely remote.......
Pipster,

If either (1) the trades are independent (i.e. no serial correlation exists), or (2) one or more trades in the group of three has positive expectancy, then martingale is merely adding unnecessary risk, as we can simply (in case (1)) take consistent risk on each trade, and (in case(2)) take only the individual trades that deliver positive expectancy. That is the same reasoning used in point #3 in post #126.

The only occasion you'd want to use martingale across the group of 3 trades would be if the individual trades delivered negative expectancy, but the group as a whole somehow delivered positive expectancy (and ideally with probability improving following each failed trade in the group). A possible example might be if you were trading breakouts, and your testing proved that a profitable breakout always occurs within 3 attempts of price breaking above/below a certain point, out of a consolidation. Hence if the first two attempts were fakeouts, the third trade would (on balance) turn the group as a whole into overall positive expectancy. But even then the martingale is unnecessary, as we could simply ignore the first two breakout attempts from each consolidation and trade only the third. This would lower risk exponentially, but with the drawback being that we would miss out on income on the occasions where the first or second attempts did result in profitable breakouts.

Hence this becomes yet another illustration of weighing up desired return against a risk model that is personally acceptable (i.e. the trader is willing to embrace the additional risk imposed by the martingale. As always, to increase return linearly by using progressive staking, we must increase risk of ruin exponentially).

I've posted the attached XLS many times before. It displays the probability of incurring X consecutive losses in Y trades, given a system with a win rate of Z%.
For example, the table shows that a system with a 90% win rate has a 39% probability of encountering 3 consecutive losses within any 500 trade sequence (which is possibly higher than many people would expect). Because the table is simply based on binomial probability formulae, it does not take any kind of serial correlation into account, i.e. any 'clustering' or 'anti-clustering' of wins and losses would skew the table values accordingly.

David
Attached File(s)
File Type: xls Probability of X consecutive losing trades.xls   26 KB | 358 downloads
  • #144
  • Aug 17, 2014 6:22am Aug 17, 2014 6:22am
[quote=hanover;7679685]{quote} I've posted the attached XLS many times before.

Thanks Hanover. I don't Martingale other than occasionally tier a trade that is moderately against me where PA has skewed, but not hit my stop, but then my goal is to only break even.
  • #145
  • Aug 17, 2014 6:53am Aug 17, 2014 6:53am
Quoting VaBikePacker
{quote} It depends on maximum margin % usage. Keep that under 6%, run some calcs and I think you'll be shocked at what your account can survive..Now, the practicality of putting your account through it is a whole other matter certainly... Martingaling a $100 account at .10/pip, 50 lev...hmm...my initial reaction is it shouldn't take long to fry $100 if you're stacking trades closely and have a bit of a bad initial entry..Your free margin drops by $15-$20 on each entry depending on the mood of your broker at that moment it seems... ...sitting the...
No, you're right Virginia, 50:1 leverage is too low to martingale in. That's why I haven't coded the subroutine into any EA's that I have made. We here in the states are at an amazing disadvantage to the rest of the planet in that regard. If we get off to a bad start, we are done in a day, unless we use a very large grid, and even then we can't keep up because our profit ratio will be lower when the market returns to the mean.

It sucks, but it is still fun to watch others do it.
If you are good at something, never do it for free--Joker
  • #146
  • Aug 17, 2014 7:28am Aug 17, 2014 7:28am
Quoting Proximus
{quote} B.U.L.L.S.H.I.T i smelled bullshit there, but just for fun i modified my ea to exactly that: Change the trade direction when the previous closed trade was a loss And double the lot size, while reseting it after a winner Starting off with 0.01 lot, a 500$ account just as many of you would,run through 2010-2014, through 5 pip to 100 pip STOPS (TP=SL) in 2.5 pip increments, and this is the result: {image} {file} worst outcome graph death in 44 trades {image} Also there was a jackpot which made a 400% gain,however it was 1 of 39 optimization...
Umm, the guy died of cancer 10 years ago I think..but he was very successful pit trader..he had 155 consecutive wining months in his trading carrier..I don't know what was his MM in that sense that maybe he martingaled only when he was sure that he will win, I don't know, but the point was that even the pros use it to some extent...it's not a failure as some are trying to make it here.. It is dangerous and it has it own merit but you have to use it as that, and not just blindly following the manual.
A loaded gun is dangerous too, and it's up to you to use it wisely and not just go on a shooting spree in a local shopping mall.
  • #147
  • Aug 17, 2014 7:38am Aug 17, 2014 7:38am
Every thread in which this magix comes becomes a mess lol (close observation).

Magix my humble request to you is to be civilized a bit , and respect other's view too.
  • #148
  • Aug 17, 2014 7:47am Aug 17, 2014 7:47am
Quoting Magix
I already know of 1 member I don't want to compete against...and he is using our house bot! lol LINK Of course...as things go...it really is a matter of I will pay $100 for somebody to prove me right or wrong....
//----
Quoting Magix
Hayseed...you gotta couple grand kicking...you down with your ideas? (I'd love to see what you got under the hood!)
//----


hey magix...... ...... actually i did deposit 2k in that account on the 14th, but it was to test an ea.... just a coincidence ..... really had nothing to do with prox or this thread, i was just clownin around makin it appear as it did.....

but i'll rather compete against your link above, the counterrip 100 guy.... and give'im the 100% head start..... ......
//-----

both the martingale and hedge subjects seem to bring out the worst in some people..... funny.... both of those subjects interest me beyond belief..... so much so, there is a separate computer for each subject in my office.....

my live ea experience and years of studying, suggest both martingale and hedging/'opposing trades' have some positives to offer...... if, and only if, their negatives can be controlled......h
//-----
Attached Image (click to enlarge)
Click to Enlarge Name: rendezvous with rama.png Size: 373 KB
to trade and code, keep both simple... no call to impress....h
  • #149
  • Aug 17, 2014 8:03am Aug 17, 2014 8:03am
Quoting hayseed
but i'll rather compete against your link above, the counterrip 100 guy.... and give'im the 100% head start..... ...... //----- both the martingale and hedge subjects seem to bring out the worst in some people..... funny.... both of those subjects interest me beyond belief..... so much so, there is a separate computer for each subject in my office..... my live ea experience and years of studying, suggest both martingale and hedging/'opposing trades' have some positives to offer...... if, and only if, their negatives can be controlled......h...

Feel free to compete against him...he is a member of another forum, testing our current Voodoo Flagship.

The Counter Rip Logic, was a Trigger on Grid logic that I coded for an M1 trader. It has since been added to an MT platform to trade in multiple time frames and is actually a pretty fantastic earner.

The primary account he is running, has actually shot past the 100% marker, LINK.

Risk is always on the table and I can even go with you on the hedging angle...All honesty, I've not been successful with the hedging as applied to these, but under the MT platform, a currency can be long on a fast frame and short on a longer...so, in reality, much the same...

People that need to chime in with zero relevance always kind of piss me off...posting for the sake of posting can be done in trade journals, the members lounge or they can keep to their facebook/twitter accounts...it is the feed of misinformation that really gets me heated.

I don't care how people trade, what they trade, or why they take trades...

The only thing that really concerns me, is if they have made these decisions based on their own relevant findings, vs having a bias based in bullshit.

How many in the forum are afraid of just running an EA? And really, why??

Something automated that can take a trigger based on their own trade logic, but yet, they refuse to open themselves to these.

Past the Martingales, cost averaging, grids...just a simple trigger...

If they can't overcome a fear of using their own logic as is automated, what chance do they stand when it comes time to just push the trigger when they do get a live call...

Much respect for ya, Hayseed...

Money Can't Buy Happiness. Poverty Can't Buy SHIT! You Choose!
  • #150
  • Aug 17, 2014 10:58am Aug 17, 2014 10:58am
Can we go back to martingales?

I assume everybody agrees a simple martingale is bad business.
Maybe some of you are trading with refined martingales and are they profitable? All flavours I've tried are bad, commercial or not, and the biggest problem of them all is that you get small profits for days, weeks or months but you cannot be reasonably sure that the system is positive.

My current approach is, as I said in a previous post in this thread some pages away, to run many many martingales in order to increase the speed profit and reach the account target before an 'incident' appears.

Has anyone tried this before? With any success? I'm on my 6th month testing now and will have relevant stats in a couple months but if someone has already done it...

t.
  • #151
  • Aug 17, 2014 11:27am Aug 17, 2014 11:27am
Quoting Magix
{quote} The answer to this, is yes.
I've seen the data, but does not satisfy me (even if it shows impressive results!).

My gut feeling is that 6 or 10 martingales are better than one, but still too slow to reach the target. I need to be able to stop in 3-5 weeks, no more, in order to have repeatability and enough stats, I don't like to have a system that with all probability will eventually blew the account (my opinion only) running for months.

100-200 martingales do the job for me, more would be better but we reach soon the limit of MT4 or the broker (max open orders).

t.
  • #152
  • Aug 17, 2014 11:30am Aug 17, 2014 11:30am
Quoting Rap Skallion
Interested in what you guys come up with in terms of EAs. As for me I have had my fill of programming gw basic, Qbasic, Fortran, C,C++, Mt4, they now just give me a headache. Too much like what I did before I retired.
Okay, I'm sorry about that. For me I will be messing around with Magix's trigger on grid mechanism. I will test this week, next week my horse goes in the next round of the contest that Magix is holding over at his place.
If you are good at something, never do it for free--Joker
  • #153
  • Aug 17, 2014 11:42am Aug 17, 2014 11:42am
Quoting tacticat
{quote} I've seen the data, but does not satisfy me (even if it shows impressive results!). My gut feeling is that 6 or 10 martingales are better than one, but still too slow to reach the target. I need to be able to stop in 3-5 weeks, no more, in order to have repeatability and enough stats, I don't like to have a system that with all probability will eventually blew the account (my opinion only) running for months. 100-200 martingales do the job for me, more would be better but we reach soon the limit of MT4 or the broker (max open orders). t....

6-10 martins per trade pair??

Guessing you are working on the nano levels, correct?
(not saying this as a bad thing, just wondering)

Prior to my work here and at the other forum, I was doing some testing with a Group at WWI. I used to run 4-6 EAs per pair on a single account and do agree that it can have a tendency to balance the account. Tests I was working on were a 10k start, just to ensure that they would make it through the week, and would separate results on a spread sheet.

From there, I shifted my focus to just single traders per account...lower the DD, or lower the account size start and increase the potential for gain.

There is definitely more than one way to skin a cat.

If you check through my post history and attached files, maybe you'll find something that you can add to your heap o traders to augment up to the number of EAs running...

After this, sorry...I got nothing else to offer on the multiples subject.
Money Can't Buy Happiness. Poverty Can't Buy SHIT! You Choose!
  • #154
  • Aug 17, 2014 12:04pm Aug 17, 2014 12:04pm
Quoting Magix
{quote} 6-10 martins per trade pair??
My setup:
- One single EA (on single chart). I don't like to have a lot of charts and EAs running.
- Operating in 32 pairs with that EA (28 currency+2 gold +2 silver). The more the better, I've tried also with indexes, up to 42 instruments at the same time, but difficult to equalize.
- EA opens a BUY and a SELL martingale in every pair at start and opens more BUY and SELL martingales whenever a new possible range is detected. At any time there are at least 64 martinagales, typically 150, more when market is in clear trend.

Quote
Guessing you are working on the nano levels, correct?
Sorry, I'm fairly new to trading jargon. I suppose you refer to the volume of operations? Martingales start always with 0.01 and increase volume in the series operation arithmetically. Grid is also widening (and dynamic ranges are used), so it's difficult to open the 6th level of any martingale (0.1 lots approx). With leverage 1:500 and 3K euros system can usually cope for the first weeks and target (double) or stop (account blown) is usually reached in 3-5 weeks.

The life cycle of a system like this is quite stable, system starts and during the 1st 4 or 5 days the DD gets -500 more or less, the second week we see DD up to -2000, the 3rd week usually the equity starts oscillating with higher maximums and higher minimums till TP (doubling the 3K and reaching 6K) is reached. 30% of times though the account blows.

With that setup a system can be started every 5 days or so and is not correlated, so we have more stats to validate that the ratio of TP is over 60% or 70%.

t.
  • #155
  • Aug 17, 2014 12:14pm Aug 17, 2014 12:14pm
Quoting Magix
{quote} ....it is the feed of misinformation that really gets me heated...... How many in the forum are afraid of just running an EA? And really, why?? Something automated that can take a trigger based on their own trade logic, but yet, they refuse to open themselves to these. Past the Martingales, cost averaging, grids...just a simple trigger... If they can't overcome a fear of using their own logic as is automated, what chance do they stand when it comes time to just push the trigger when they do get a live...
Man, can I relate to this one. From my start in Forex like many I heard of and saw many EA's. I wouldn't go near them with a DEMO let alone a live account! LOL

What was it Magix? Maybe just 4 or 5 months ago I was laughing in my thread with another member about EA's? We were having fun making up names for E A. The best one we came up with was Empty Account...LOL That's how "Dead Set Against Them" I .... WAS. But not anymore.

I don't know crap about coding or how to build or modify one (yet) but I sure do understand the "logic" of trading. Being a bit of a computer geek in other areas and just in general, a pretty logical sort, I ..... "Opened Myself Up" to the pure logic of it all. I'll never go back.

Thank you for turning the light on for me with your, "getting my feet wet" EA brother. I'm hooked ... and that's official!

I've learned in my short experience so far that, just like with manual trading, some personal discretion is always a good thing too.

Much respect for both you and Hayseed.
Ghost Rider - WWTBMD?
  • #156
  • Aug 17, 2014 12:51pm Aug 17, 2014 12:51pm
Quoting tacticat
{quote} My setup: - One single EA (on single chart). I don't like to have a lot of charts and EAs running. - Operating in 32 pairs with that EA (28 currency+2 gold +2 silver). The more the better, I've tried also with indexes, up to 42 instruments at the same time, but difficult to equalize. - EA opens a BUY and a SELL martingale in every pair at start and opens more BUY and SELL martingales whenever a new possible range is detected. At any time there are at least 64 martinagales, typically 150, more when market is in clear trend. {quote} Sorry,...

In working with these, you are already starting to see the pattern of cycles. Some days they seem too good to be true, others, they are hemorrhaging money to the point of death.

The market works in trend and chop cycles, some of which I have discussed here, some at the other forum. Figuring these out can allow you to sit out the shit and only play the hops...it can be frustrating to watch, but in the end run, a day of no loss, beats the hell out of potential account loss.

Hedging and dynamics can definitely help to equalize the load and being unfamiliar with the tool you are playing with, am really at the end of any suggestions.

Personally, I look for steady gains without a large draw...finding this can allow a larger sized lot to generate a larger sized profit, without the same tax. Beit a single pair or a balance of others.

Our flagship, limits concurrency...built in by it's author, this allows us to only play so many occurrences of any single currency, so that if it goes into runaway, it isn't robbing the bank, while we are trying to make a deposit. It too has the ability to trade all available currencies but there are some that I have found staying out of, to be of benefit, including GOLD...Won't touch that with anything I play with and rarely find a solid excuse to trade it manually. The risk on that, especially on a grid/martingale is more than I am willing to play with.

Nice to see other approaches to market, even if under the same ideas. Definitely gives a little more food for thought.

Money Can't Buy Happiness. Poverty Can't Buy SHIT! You Choose!
  • #157
  • Aug 17, 2014 1:09pm Aug 17, 2014 1:09pm
Quoting Magix
{quote}Some days they seem too good to be true, others, they are hemorrhaging money to the point of death.
And that's the reason I want to run many instrument and in both directions and in different levels. I need to 'stabilize' the combined outcome.

Quoting Magix
...it can be frustrating to watch, but in the end run, a day of no loss, beats the hell out of potential account loss.
With a short-period cycle of a month it's usually the opposite! When I see the equity going down I see the 'springs' loading and ready to release. Big DDs anticipate a short cycle, sometime to stop but sometimes to TP .

Quote
Personally, I look for steady gains without a large draw...finding this can allow a larger sized lot to generate a larger sized profit, without the same tax.
Agree. But that's difficult, difficult to win with limited risk. In order to keep my peace of mind I tend to see a multi-martingale cycle as a single operation. You risk 3K, you win you get 3K you lose you lose 3K, easy. Limited risk, in fact, for me, putting the whole account at risk is the safest thing one can do with automated trading.
One can cheat himself thinking that he can set up filters to limit the risk... till the ultimate catastrophe appears (software bug, platform freezing, broker stealing,...)


Quote
Our flagship, limits concurrency...built in by it's author, this allows us to only play so many occurrences of any single currency,
Same for me, but only used when the account (a real one) has less than 3K to start the cycle.

An example of cycle. It started some 12 days ago (half-life), with 3K. We see that equity (red) dropped till 2K (free-margin 1K) and that's the best case scenario... After 12 days we're in 5K, almost in target. Right now 104 martingales active, max exposure on GBP pairs as one can expect.

t.
Attached Image (click to enlarge)
Click to Enlarge Name: screen-capture-533.png Size: 53 KB
  • #158
  • Aug 17, 2014 1:26pm Aug 17, 2014 1:26pm
Quoting tacticat
{quote} And that's the reason I want to run many instrument and in both directions and in different levels. I need to 'stabilize' the combined outcome. {quote} With a short-period cycle of a month it's usually the opposite! When I see the equity going down I see the 'springs' loading and ready to release. Big DDs anticipate a short cycle, sometime to stop but sometimes to TP . {quote} Agree. But that's difficult, difficult to win with limited risk. In order to keep my peace of mind I tend to see a multi-martingale cycle as a single operation....
Looks very nice, tac.

When sergien had his TE up, he was loaded on a couple of different tfs, though not sure of his settings, it was taking pant shitting amounts of trades for the account size...this, definitely has it's risks and rewards, but his took off like a rocket.

The hedging logic, seems to be what balances you and adds equity, and thus would explain the "Springs Loading" effect. You have loaded and profited in one direction and now on the major retrace, you are looking for the motherload.

This will be subject to change once we are back in a regular trade cycle.

Summer session is coming to a close and after next NFP, we should see normalization...if you can really call it that...with extra volume, the pace can increase and the session trades become a lot more relevant.

Would be interesting to see how you fair in the regular months...I'd be willing to bet that with a more normalized cycle, you will probably see shorter profits, but less draw down on the equity...

Very kewl, though...

Something of your own creation or is this one of the commercial traders?
Money Can't Buy Happiness. Poverty Can't Buy SHIT! You Choose!
  • #159
  • Aug 17, 2014 1:44pm Aug 17, 2014 1:44pm
[quote=Magix;7680035]{quote} Looks very nice, tac.
[quote]
But it's only martingales... and there is not yet enough data to determine if it's worth to run with real money. I've run only 2 times with real money just to verify the differences with demo in 2 brokers. The good news is there are no relevant differences, even with big news and platform freezing or spreads widening.
Quote
I'd be willing to bet that with a more normalized cycle, you will probably see shorter profits, but less draw down on the equity...
Not sure. I'm starting independent cycles every 4 days since March (that's easy because it's only one EA ). There are times that a couple cycles fail consecutively but there seems to be not clear rule. The average win ratio from the beginning is more than 60%.

Although the system shuts down when it reaches TP/SL (never in all tests after 6 weeks!) I sometimes let a system run for more time, with no changes (0.01 volume always, no MM).
This is one of these systems, started 26th March: in June and July it take a good hit, and on that time cycles were failing 50% of time.
Attached Image (click to enlarge)
Click to Enlarge Name: screen-capture-534.png Size: 49 KB


t.
  • #160
  • Aug 17, 2014 1:48pm Aug 17, 2014 1:48pm
Quoting Magix
{quote}Something of your own creation or is this one of the commercial traders?
My code. I'm not a trader, have no trading experience. I have friends in the trading business and I'd like to verify the myths out there: martingales don't work, sure?

But to be honest, I tried a similar setup with a commercial martingale, and the results are quite similar once you force the martingale to forget about the entry conditions, enter both senses and reenter several times when new ranges appear.

t.
0 traders viewing now