As long as you assume 50/50 minus trading cost your expectancy is negative. No amount of money management (i.e. martingale) will change that. Therefore mathematically to last in the long run you need an infinite account. With a fixed size account the best you can do is choose your desired probability of ruin and from that compute you return. Note that that will not change your long term expectancy as ruin is inevitable. If I remember correctly from the top of my head (it's been a long time but you can easily do the calculation by making a few assumptions) for a probability of one ruin every 50 years your return is less than CD by a factor 10-15. In any case it's a loosing proposition (i.e you can't make enough before ruin happens). Now if you have a stationary return process and you know its parameters (mean/std) then that's a different story... Ninna
It can be done with very little provided you can trade small lot sizes. If you have a 100 share minimum, or 1 contract minimum, you can't consider it with less than oh... 20k+. And that's ehh. If you can trade 1 share, or odd lots, and not get killed by execution fees, you can martingale with even a $1k account; the problem is your profits won't be enough to be "meaningful," i.e. you'll make $5/day. Let's take BAC, closing price 13.97. Let's say you buy 100 shares at 14.00, profit target of 14.25. If it goes to 0, you lose $1400. $14.00*100=$1400 You buy another 100 shares at 13.50=$1350 At $13.00, you buy another 100 shares.=$1300 At $12.50, you buy 200 shares.=$2500 $12.00, 200 shares.=$2400 $11.50, 200 shares.=$2300 $11, 300 shares.=$3300 $10.50, 300 shares.=$3150 $10, 300 shares.=$3000 $9.50, 300 shares=$2850 $9.00, 400 shares=$3600 $8.50, 400 shares=$3400 $8.00, 400 shares= $3200 $7.50, 400 shares=$3000 $7.00, 600 shares=$4200 $6.50, 600 shares=$3900 $6.00, 600 shares=$3600 $5.50, 800 shares=$4400 $5.00, 800 shares =$4000 $4.50, 800 shares=$3600 $4.00, 1200 shares=$4800 $3.50, 1600 shares=$5600 $3.00, 2000 shares=$6000 $2.50, 3000 shares=$7500 $2.00, 4000 shares=$8000 $1.50, 8000 shares =$12000 $1.00, 12000 shares = $12000 $.75, 20000 shares =$15000 $.50, 40,000 shares =$20,000 --------------------------- Pretty much impossible to lose money structured like this, however your original target is $25. Your total capital required is $151,350 You own 99,700 shares in a worst case scenario, at an average cost of $.65/share. It's basically impossible to lose money if you choose the right company. If you can trade 10 share lots, you only need $15,000. The problem is, your risk/reward becomes SO inverted - 6000:1ish, that one blowup (e.g. 2008 financial crisis can kill you). However, if you take profits on all the bumps on the way down and choose the right companies and are diversified, you can make money... espcially doing this with commodities like OIL, corn, etc.
SPY is 10 times worse then the futures because you will run out of margin much faster. Dude, listen to me. You have to care about the % return because you have to measure it against the alternative which is putting the money in a fixed income asset. If one can make 5% in a bond that's 50k a year guaranteed every year. Why would you risk losing the entire million dollars when you can make 50k a year. Then you state 50k a year would be cool. LOL. Good, then put the money into a long term bond. That bond will far outperform the martingale about 98 times out of a 100. You do see that right? Most of the time you won't get the chance to martingale. That's the problem! In a normal up trending market you will be long 100 spyders on a million dollar account. LOL.
This kinda goes to the idea of "never take a loss". There is a rational reason for this which is that when one enters a trade one has a hypothesis that something will happen -- if it doesn't then then one might hypothesize that there is no particular edge in selling. So, you want to hold all your losses until they become winners. The problem is, as Maverick said, to do this you will need to use very low leverage. In the end your equity curve will look just like the index. What is missing from the original thesis is that there was an implicit hypothesis that to gain the EDGE that you would exit. I probably wouldn't try a method like this but if you were to try it.. probably doing it on the way up would work best. This would be essentially a trend following strategy..
This is an illusion. Diversified negative expectancy doesn't give positive expectancy (except in some rare and strange cases but not here). It just makes it harder for you to realize it and gives you a false sense of confort. Ninna
That's not right...you need to double your shares every 50 cents not keep buying the same amount. You'd reach millions of shares very fast
How is doing it on the way UP the best idea? Do you mean with a short position? Price can go up indefinitely. I'd never do that. Forget using BAC. Companies can go to 0. SPY can't.
Edit By my math if you bought BAC every 50 cents double down you'd be buying over 250 million shares your last trade
He meant add to your position when it goes with you instead of against you. Anyway just like martingale it doesn't turn negative expectancy into positive one. Ninna
I think the difficulty becomes choosing a profit target. Sell too early and miss out on potential MASSIVE profit. I wish I still had my shares of QLD from back in the day, but I sold them a while back.