Increasing Size During Drawdown
- Thread starter oldtime
- Start date
there are a lot of ways to trade and an infinite number of ways to do it badly,don't trade til you have a setup,if that setup gets you stopped out 40 times ,stop using it,skim thru this thread and read all of the op's entries,this works in all liquid markets,it will tell you several times a day a week a month where the market is going,then you just have to sit and wait for opportunies to enter, making money is easy, not giving it back is hard http://www.elitetrader.com/vb/showthread.php?s=&threadid=127338&perpage=6&pagenumber=2 i would suggest that you hand chart the rth's of your instrument for a bout six months, you will learn its personality,its players,become familair with the repetition and confident in your trades,will spoil you to the point that you wont take the bad setups..only change in the last 25+ years is that we now have multiday charts like the,c one in aapl above ,you can now compare the setups in multiple markets, when you get multiple instrument's lining up all together,its 90% of the time gravy,if you have the patience you could trade these setups about once a nonth and catch some great moves
ah, I'm almost never flat, my mother was protestant and we were taught to work all the time
I'm not a hunter, no skill that way, don't know how to be quiet in the woods
I'm a trapper
we just plod along all day checking our traps and resetting them making all kinds of noise along the way
Trading is like flipping a quarter just because you got the last 40 trades wrong it has no bearing on the outcome of your next trade. When your in a draw down a profitable system will correct itself right. So your trying to double up your position obviously to make your money back sooner which in return can turn a small drawdown into a large drawdown. Why not double up your position when you get 40 trades right at least then your possibly going to lose money gained instead of digging a deeper hole.
I think if you have an edge you should always worry about protecting your 100%. The deeper you dig into your 100% the more time it takes you to get out of the hole. The old addage if you lose 50% it takes a 100% gain to get back to where you were. I think how you handle your draw downs defines you as a trader. Everybody can make a great trade not everybody can manage a bad draw down.
Maybe though a compromise would be to have a max risk of 2% of your account no matter what.So if you risk 1% per trade you can martingale to max risk 2%. Or whatever is comfortable without digging to much of a hole.
Just speculating!
I think if you have an edge you should always worry about protecting your 100%. The deeper you dig into your 100% the more time it takes you to get out of the hole. The old addage if you lose 50% it takes a 100% gain to get back to where you were. I think how you handle your draw downs defines you as a trader. Everybody can make a great trade not everybody can manage a bad draw down.
Maybe though a compromise would be to have a max risk of 2% of your account no matter what.So if you risk 1% per trade you can martingale to max risk 2%. Or whatever is comfortable without digging to much of a hole.
Just speculating!
I should clarify my statement also. The word 'Martingale' gets thrown around as a title for any type of scaling system. That is horribly inaccurate.
Martingale is a very specific betting system in which bet size is doubled with each loss. The idea relies on an unlimited bankroll. It really has nothing to do with scaling or cost averaging. It is, by definition, a revenge trading system and a surefire way to blow up an account. Here's why...
Nobody truly has an unlimited bankroll. The beginning bet size must then be small enough to allow for > 25 bets. This would give you a strong statistical confidence that a blow-up event wouldn't happen in your lifetime (six sigma event)
If you construct the betting system in that manner, you'll be making roughly 0.1% annually. This is actually very intuitive, because you are trading a system that now essentially carries almost no statistical risk. One would expect to realize an even lower return than short term treasuries if the bet actually carries less risk than the treasuries.
Let's assume that this is just unacceptable, and you at least want to be getting 3% annual. Your bet size must increase roughly 30X. Now you are in 4.8 sigma territory. There is now a slight blowup risk sometime during your lifetime.
Now let's assume you want something more like 10% annual. This would require 100X initial bet size, taking you into roughly 4 sigma territory and blowup frequency jumps to about every 30 years on average. So you now have a pretty good chance of blowing up.
20-30% annual target brings you into a realm where a blowup during a normal lifespan is very probable. A 50% target makes blowup a virtual certainty. All of this is very intuitive.
In the end, a martingale system is a losing system. A scaling system is much different, and is a method of averaging the cost basis of the investment in recognition that you cannot pick absolute bottoms with regularity. Scaling systems frequently include no bet size increase, or at least very limited increase. Losses are not exponential as in the case of the martingale.
In reality, scaling is simply a method of giving up some potential return for the ability to trade at greater frequency and slightly higher probability of profit.
Martingale is a very specific betting system in which bet size is doubled with each loss. The idea relies on an unlimited bankroll. It really has nothing to do with scaling or cost averaging. It is, by definition, a revenge trading system and a surefire way to blow up an account. Here's why...
Nobody truly has an unlimited bankroll. The beginning bet size must then be small enough to allow for > 25 bets. This would give you a strong statistical confidence that a blow-up event wouldn't happen in your lifetime (six sigma event)
If you construct the betting system in that manner, you'll be making roughly 0.1% annually. This is actually very intuitive, because you are trading a system that now essentially carries almost no statistical risk. One would expect to realize an even lower return than short term treasuries if the bet actually carries less risk than the treasuries.
Let's assume that this is just unacceptable, and you at least want to be getting 3% annual. Your bet size must increase roughly 30X. Now you are in 4.8 sigma territory. There is now a slight blowup risk sometime during your lifetime.
Now let's assume you want something more like 10% annual. This would require 100X initial bet size, taking you into roughly 4 sigma territory and blowup frequency jumps to about every 30 years on average. So you now have a pretty good chance of blowing up.
20-30% annual target brings you into a realm where a blowup during a normal lifespan is very probable. A 50% target makes blowup a virtual certainty. All of this is very intuitive.
In the end, a martingale system is a losing system. A scaling system is much different, and is a method of averaging the cost basis of the investment in recognition that you cannot pick absolute bottoms with regularity. Scaling systems frequently include no bet size increase, or at least very limited increase. Losses are not exponential as in the case of the martingale.
In reality, scaling is simply a method of giving up some potential return for the ability to trade at greater frequency and slightly higher probability of profit.
well, martingal is the only system I know of that has no risk of ruin if you have infinite money
thanks for the good info
I really just started the thread a long time ago to see if anyone was working with size
I was surprised how many replies I got that were just knee jerk and not really well thought out
I've been experimenting with size again recently in a paper account, so I just bumped to see if anybody had any ideas
I've seen some impressive tests on DCA, but like you say, that is not the same as martingale
thanks for the good info
I really just started the thread a long time ago to see if anyone was working with size
I was surprised how many replies I got that were just knee jerk and not really well thought out
I've been experimenting with size again recently in a paper account, so I just bumped to see if anybody had any ideas
I've seen some impressive tests on DCA, but like you say, that is not the same as martingale
No system has any risk of ruin if you have infinite money...
