Taming Martingale

Post Reply
AllanS
Posts: 7
Joined: Mon May 27, 2013 8:43 am

Taming Martingale

Post by AllanS »

No one in their right mind would risk a million dollars to make $1.

Very few have a spare million to risk, no matter the return.

Here's an idea that solves both problems.

Find a friendly broker, do what's needed to set up a secure, collective account, and convince 1000 people to contribute 1k each. We now have a million dollar war chest, and no one is losing sleep.

With a stop of 1R, the goal is to find 1:20 trades, with breakeven set at 3R.

Suppose we begin with R=10cents, stop 10 pips, and we're looking for the top or bottom of a 200 pip run using some clever breakout strategy. Things go badly. We martingale up to level 20, risking 25k. Finally, the trade goes our way. At 3R, we're +75k. At this point, we take enough off the table to cover accumulated costs and losses (maybe half the position), move the stop to breakeven, and let the rest run 200 pips to 20R. This will return around 120K, or $120 per person in the account.

The best case scenario occurs at level 23. By this stage, the war chest is half empty. But if this fully loaded position of about 400k finds a top or a bottom, we recover all the accumulated losses at 3R, and let the remaining 200k run 200 pips up to $4 million. ie. 4k to every member.

The worse case scenario occurs if we hit level 24 on the very first trade. Everyone loses 1k. There's really nothing left to do but find a dark corner and cry into your beer.

What's left to determine is this: Is it possible to devise a programmable 1R stop-loss strategy that has a very good chance of finding at least 3R within 23 attempts? Will this chance be good enough to justify risking 1k? Will enough of these trades run to 20R to justify investing 1k? Will the strategy attract 1000 participants?
AllanS
Posts: 7
Joined: Mon May 27, 2013 8:43 am

Re: Taming Martingale

Post by AllanS »

With standard Martingale, you end up with a huge position looking for a tiny return. I'm suggesting you let the winning position run to at least to 3R, with some running to 20R.

The 3R run cover costs, sets the trade to BE, and resets the Martingale to level 1. The rare 20R runs make the money.

I'd be very surprised if you can't devise a strategy that finds a 3R run within 20 attempts with very high probability. Once this is done, the rest will look after itself. I'm basically Martingaling an equity millipede.

Big players can afford this sort of strategy. Small players cannot, but the internet now allows small players to communicate and cooperate. "Divide and conquer" becomes "Unite and conquer." After all, banks (theoretically) are a lot of little people cooperating (but in fact are a lot of little people being used for the benefit of the owners of the bank.)

Every Martingale strategy will implode. Sooner or later, the market will throw up the losing sequence. But equally, sooner or later, an asteroid, a plague or WW3 will throw everything into chaos. Every system fails, sooner or later. The question is, what is the risk of ruin in this Martingale strategy relative to the time it takes to double your account? After withdrawing the initial stake, the risk effectively becomes zero.
Post Reply

Return to “Automated trading systems”