Money management & Martingale

[maildigger 10]

Has anyone ever thought about a Martingale (=if you lose you double your position the next trade) money management?

 

In the long run it seems a bad idea due to the possibility of a losing streak and you can't keep adding contracts.

 

What if you combined it with a strict daily money management? Let's assume the total maximum loss you will take is X and your daily goal is 2X or more.

[brownsfan019 17]

It might work if you impose other rules to it, but then the question becomes - are you even performing a martingale strategy?

 

It's something I've thought about before as well and in the end, it was a slippery slope (for me).

 

There's only one way to find out - do some backtesting and/or live testing and see how it goes. Keep detailed records. It could be an easy analysis if you do it on paper, while trading the way you currently do. See where you are at the end of the day.

[torero 10]

Ask the traders who pyramiding up or down while in winning positions or losing positions and read Market Wizards how these guys become rich. It was pyramiding while their positions were winning, not losing. Evidence is proven everyday that Martingale don't work. I think the Dot Bomb Bubble years were an extreme major example but no one can last forever fighting the market when proven wrong (even the Bank of England couldn't do it vs Market and Soros).

[notouch 10]

Nick Leeson traded using Martingale money management and look what happened to him. I strongly recommend watching the movie Rogue Trader if you're not aware of all the facts.

[waveslider 10]

There is a program called "market system analyzer" that can test all of these different money management techniques. I think they have a free one month trial.

 

My opinion is that you add on one contract per trade after reaching $XX account value.

 

Doubling position size almost always seems like a bad idea, so does pyramiding as you win if not done properly.

 

In pyramiding your entry in the same position as you win you may end up with an inverted pyramid as you raise your average entry price. Subsequent entries should always be smaller in size.

 

The trick to money management is to take a lot of trades, and manage them according to your individual approach. I see there being two approaches:

 

You are either a high probability trader, or a trader who has larger average wins than average losers (high profit factor).

 

A high probability trader wants to keep his maximum losers as low as possible. As long as the losers are not too big and the probability is high enough, the account will remain near its profit peak.

 

High probability trading normally means your losers are larger than your winners. Your winners are small, but consistent.

 

A high profit factor trader can have a low probability, but still be profitable as the size of the winners take care of losses. A high profitability trader has smaller losses and larger winners.

 

The high profit factor trader may have more extended draw-down periods and more time in the market (exposure).

 

So decide what your style is, and add size little by little as your account size grows.

 

WS

[rolange 10]

Has anyone ever thought about a Martingale (=if you lose you double your position the next trade) money management?

 

In the long run it seems a bad idea due to the possibility of a losing streak and you can't keep adding contracts.

 

What if you combined it with a strict daily money management? Let's assume the total maximum loss you will take is X and your daily goal is 2X or more.

 

I have considered it and the allure has attracted many others too. Interesting reading on another forum (Forex)and it seems to possibly work if extremely tight criteria are observed.

1. Trading micro lots on a basket of FX

2. Removing profits from accounts on a regular basis

3. Looking at historical volatility and acknowledging that days where there can be a move of 300+ pips without a retrace.

 

Far better to take small hits and ride the winners in my opinion.

[Dogpile 10]

check out the 'Kelly Criterion' -- a related topic.

 

Kelly criterion - Wikipedia, the free encyclopedia

 

Kelly Criterion % = W - [(1 - W) / R] = What % of capital to bet each time where;

W= Win % (Winning Trades / Total Trades)

R = Win/Loss Ratio (Expected Dollars Won When You Win vs Expected Dollars Lost When You Lose)

 

The optimum bet for the greatest growth of bankroll is making the full bet suggested by the Kelly criterion, but this produces a volatile result.

[maildigger 10]

Can give an practical example of that formula ?

[Dogpile 10]

So if Win 53% and win $ vs loss $ is 1.0 (equal):

 

Kelly Criterion suggests you bet 6% of your bankroll on each trade.

 

So if Expected Win % is 65% and win $ vs loss $ is 0.9 (equal):

 

Kelly Criterion suggests you bet 26.1% of your bankroll on each trade.

 

You can simulate this in excel by having it generate random numbers and excel can project the next 10,000 or 100,000 trades -- simply by taking your percentage win rate. You can then graph the results. The variance is absolutely sick. I suggest you do this long before you ever consider this as a money management technique because you wouldn't be able to stomach it.

 

The Kelly forumula is an interesting one to observe though. It is a nice mathematical way to show the interraction of % vs dollars won.

 

Personally, I think the key to trading is to get your % very high first with very reasonable 'stops'. This takes a long time. But once you do this, you can simply leverage your trades and grow capital without big variance.

 

Variance is a subject that is just hard for people to really comprehend. If your win rate is low and you seek big wins... you can just go on some awful streaks as the variance associated with low % plays out over time.

[maildigger 10]

How would this be used with trading futures ?

 

60% wins, 40% losses, R= 1.25 (incl commissions)

 

So:

 

0.60 - ( 1 - 0.60 ) / 1.25=

0.20 / 1.25 = 0.16

 

So risk 16% each time ... How many ES contracts should I trade if I use a 100K account?

[Dogpile 10]

<<So risk 16% each time ... How many ES contracts should I trade if I use a 100K account?>>

 

if this is a serious question --- dude, 16% per trade?

 

60% and 1.25 ratio is fine if you do lots of trades. that isn't my style but I know some traders with numbers like that and they do great.

 

Personally, I would err on the side of being very conservative with contract size if you are getting 40% losers.

 

If your style is to do lots of trades per day, smaller size trades can add up and you will not get the big dollar drawdowns associated with trading bigger size after having a bad run. a 'bad run' is never that far away if you are getting 40% losers. moreover, many 5-lot ES trades in a day adds up quickly if you trade a lot and can really get that 60%/1.25 target consistently.

[maildigger 10]

I used the Kelly formula and this is my outcome. I would be very hesitating to risk 16% of my account in a trade :doh:. Normally I would risk at max 1% in a trade but I'm checking out if there are superior money management strategies.

[JKonstenius 10]

I have tried many systems that don't work. What I like about martingale is that it works on paper. Have looked at many markets for trading this strategy. Seems as though markets that let you bid and ask are the best such as stocks. Have not been able to find any automated software for this method. I first got the idea from a guy I met trading wheat using the you can't lose trading commodities book. He doubled down into massive quantities of futures before he took a small profit. One intersting idea was to do this with a spread so your range would be known in advance such as crack or crush.

 

Jeremy

[momentom 10]

The thing that kills martingale is you are risking big to make small. Also, the distribution of the losses in a live sample WILL put you out of business when the row of losses is long enough - and it will be eventually.

[Sparrow 10]

Some more info on trading and Kelly

 

some more info

more detailed

 

Edit:

 

Oh i was wrong in the original post, using the formula can lead to successful trading. Larry williams who used it for winning the futures worldchampionship switched to another strategie though.

[JKonstenius 10]

Was searching the net for more info and found this interesting product.

 

http://www.stocksystem.com/

 

 

Aparently they use a linear averaging down system vs the non linear escalation of martingale. I plan on using thier idea in a Roth using blue chips.

[Clym 10]

This post reminded me of a martingale system that i bought back around 1990 when I was first interested in trading.

 

I did a little searching and it looks like the author has died, the software is no longer for sale, but the documentation is here.

 

the algorithm explained

http://www.petereliason.com/sts/book_ch11.html

main page

http://www.petereliason.com/sts/title.html

[Trvlwanderer 10]

For day and swing trades, martingale may certainly blow you out quickly. In fact, I would surmise that fading out of a losing position will show a greater bottom line. I've done both and averaging down shows consistent deeper losses in my journal while anti-martingale is consistently better.

 

I have, over the last year, begun a very strong practice of averaging up into winning trades and the results have been staggering. Especially considering retracements and insuring the trend continues. This is also backed up with a quick check of the journal. Some trades have shown a loss from the last buy, but profits have already been locked at that point. A small dip from the last buy is minute in contrast to the averaging in previously.

 

I still chuckle when I see martingale in action at casinos with roulette. The odds do not change each turn! With a stock, the same is likely true. Cut your losers, they may continue to lose, and in most cases do.

 

My $.02

[jperl 11]

For day and swing trades, martingale may certainly blow you out quickly. In fact, I would surmise that fading out of a losing position will show a greater bottom line. I've done both and averaging down shows consistent deeper losses in my journal while anti-martingale is consistently better.

 

I still chuckle when I see martingale in action at casinos with roulette. The odds do not change each turn! With a stock, the same is likely true. Cut your losers, they may continue to lose, and in most cases do.

 

My $.02

 

Trvlwanderer,

You need to make a distinction here between averaging down and scaling in. I would agree that arbitrary averaging down is a losing proposition, but scaling-in is not. The distinction has to do with the concept of risk tolerance used in managing a trade. If for example you have a strict risk tolerance say 2% of your account balance, then scaling in within the boundaries of your risk tolerance becomes a viable technique to limit loses. See the thread [thread=2189]"Trading with Market Statisitics VI: Scaling in and Risk Tolerance"[/thread] for a more complete discussion.

[Trvlwanderer 10]

Trvlwanderer,

You need to make a distinction here between averaging down and scaling in. I would agree that arbitrary averaging down is a losing proposition, but scaling-in is not. The distinction has to do with the concept of risk tolerance used in managing a trade. If for example you have a strict risk tolerance say 2% of your account balance, then scaling in within the boundaries of your risk tolerance becomes a viable technique to limit loses. See the thread [thread=2189]"Trading with Market Statisitics VI: Scaling in and Risk Tolerance"[/thread] for a more complete discussion.

 

I may have been misunderstood. I will scale in if the trade is still playing out and I need a better fill....but scaling in to lower my entry on a trade that is getting away from me is not a good idea. At least that is what my experience over the years has proven.

 

Now investing....different story. scaling in over time while my margin of safety is still viable is very effective for solid stocks.

 

Maybe I misunderstood the OP.

 

Trvl.

[metalhead 10]

This post reminded me of a martingale system that i bought back around 1990 when I was first interested in trading.

 

I did a little searching and it looks like the author has died, the software is no longer for sale, but the documentation is here.

 

the algorithm explained

http://www.petereliason.com/sts/book_ch11.html

main page

http://www.petereliason.com/sts/title.html

 

that domain seems to have expired. Any more info about this anywhere else?

[Dr.Lo 10]

Van Tharp wrote "THE" book on position sizing, in which he does an in-depth analysis (lots of tables, charts and statistics) of Martingale, anti-Martingale, Kelly criterion, optimal f, and many other types of position sizing algorithms (93 different ones, total). It's called "Van Tharp's DEFINITIVE GUIDE TO POSITION SIZING". It also covers essential information about mental errors, biases, fallacies, and costly mistakes to avoid. It's big (380 pages), and it's not cheap (around $150-200 US). He also has a two dvd set that covers the essential "meat" of the book, but of course leaves out much of the detail. I love the way his mind works (the contextual perspective that is relatively unattached to any specific content, which he got from his training in neuro-linguistic programming, and working with thousands of traders with vastly different beliefs and methodologies, plus he's a Ph.D. shrink), even though I might not personally agree with all of his beliefs, but in this specific area, I haven't yet seen anyone in or near his class. Just my two cents, and I could be wrong - it's happened before, and it probably will happen again - LOL! (trader humor, we all have losing trades, don't we?)

What's the bottom line? Just like practically all of the great traders profiled in the Market Wizards series say, "You've got to find what works for you". In my opinion, this applies equally to method, technique, time-frame, markets, psychology and position sizing. Position sizing in particular must be approached with an understanding of personal, individual risk tolerance and risk aversion, and account size and our own objectives are also vital considerations and unique elements of a successful total trading plan. For people who want a "just do this" answer, look elsewhere; for those who want a broad understanding within which to find themselves and meet their goals, you might just end up loving Van Tharp as much as I do.

[waveslider 10]

Excellent post, thanks Dr. Lo

[Dr.Lo 10]

Thank you, Waveslider, for your kind acknowledgment!

I appreciate knowing that 'someone is out there', if you know what I mean.

After your response, I remembered that Dr. Tharp offers a FREE position sizing trading simulator for three levels of difficulty (for the higher levels of difficulty, he charges around $200, but one is under NO obligation to purchase the full game) which can be found at:

http://www.iitm.com/products/Trading-game.htm

In my opinion, when you can consistently kick ass at the first three levels, you've gotten the basic idea pretty well. The key word there is "consistently", NOT "kick ass" (Van has commented on some of the winners of trading championships having fantastic luck, and the statistical probability of using those same position sizing algorithms with success on large samples being practically nonexistent. Remember, in large sample size, streaks of multiple winners and losers WILL OCCUR, and one's system will collapse if bet size is not small enough to survive those streaks).

I should probably state for the record, in the spirit of full disclosure, that while I have personally purchased many of Dr. Tharp's products and services, I have NO financial ties with him or his excellent organization, the International Institute of Trading Mastery (I.I.T.M.).

Playing the free game enough times should clearly reveal the biases and limitations of many types of inadequately thought-through position sizing algorithms. Martingale systems, in particular, can ONLY work under the boundary conditions of 1. your goal is a (typically) $1 win per series of bets 2. you have an UNLIMITED bankroll and 3. your counter-party will allow an unlimited number of bet size doubling. While it may have some purely theoretical validity, it has consistently failed miserably in the real world, in both gambling and trading.

This quote sums it up best, to me:

"In theory, theory and practice are the same.

In practice, they aren't" Yogi Berra.

LOL!!!

All the best!

[AgeKay 10]

I still chuckle when I see martingale in action at casinos with roulette. The odds do not change each turn!

 

I actually tried that once in a casino. I was betting on 'red' and it took 16 'black's in a row to blow me out. It actually kept going until there were 21 'black's in a row. I just thought to my self: What the... Anyway, I learned the lesson to never double up again.

 

It's the same when it comes to trading. You might backtest your strategy 10 years but then comes on day you have never seen before and you lose everything. So I would avoid it if you want to do this long term.

 

I have thought a lot about position sizing and also done a lot of analysis on it and came to the conclusion that a big variance in your position size will have a negative effect on your results (due to the unpredictability of your trading results). You don't need varying position size to trade profitably. Sure after you are comfortable with your trading strategy then you might want to tweak it a bit.

 

Regarding Tharp, as far as I know, he is still not trading professionally, he makes his money by selling his "trading education" and coaching (he does coach big professional traders, but he does not teach them how to trade, he helps them with their psychology)...