Reverse Martingale - two winners in a row

[Zwaen]

Careful. You are using reason and logic. And what you calculate from what you just said is that the stated strategy will increase volatility without increasing return. Ouch.

I liked your previous post, very good 'summary' of the goals of a trader or investor. I personally devoted the last 2 years to reducing sd only.
For fun I put the strategy above in my test-lab and indeed, var up. Had not expected this at first and had to think about it, but it probably is due expanded ranges due to streaks, both up and down. Amazing to see!

 

[Snuskpelle]

So there are two different use cases here:
A. Independent trades
B. As part of scaling up a single trade off the same signal/alpha (e.g. a presumed trend).

For A) there's absolutely no point, it just comes off as a gambling exercise (see first thread response by FreeGoldRush). A better (if for some reasons problematic, see e.g. leveraged ETFs) alternative is to scale by account size, i.e. compounding.

For B) this scheme is likely too aggressive, as your highest quality alpha is likely at the start. Consider scaling the trade linearly (add the same amount at each step) instead, or even less at each step. Either way, no point overthinking this because how you structure a trade does not magically add alpha. Spend your time looking for alpha, not playing around with various trade structures.

 

[SimpleMeLike]

What do you do if you lose several trades in a row?

LMFAO HAHAHAHAHAHHAHAH this is FUNNNNYYYYYYY.

Good question.

 

[PBS]

LMFAO HAHAHAHAHAHHAHAH this is FUNNNNYYYYYYY.

Good question.

I lose the last winner's profit (say 3R) and then lose 1R, 1R, 1R, etc. Why is it funny?
I am not trading like that yet. I am just giving it a thought.

 

[GaryBtrader]

I am considering a trading system in which you play to get two winners in a row, where the second winner is risking the profits of the first one. After that, or after a loss, the risk reverts to the standard risk. It is modified reverse Martingale position sizing.

For example - standard risk is $100 and the sequence follows:
- trade 1: risk $100, profit: $300
- trade 2: risk $300, win: $600

- trade 3: risk $100, profit: $200
- trade 4: risk $200, loss: -$200
- trade 5: risk $100: profit: $250, etc.


Will that work in the long term?

This will work long term as long as your trading strategy has positive expectancy, i.e. your average return per trade is positive and your percentage winners is enough to keep you afloat. Like anything, if you have a losing strategy you're going to go bust eventually.

It would actually be pretty straightforward to simulate this even in Excel. But you need to model the strategy first which means you need a large number of sample trades to determine average return, standard deviation, etc.

 

[PBS]

I ran a few simulations and indeed, it does not make sense - increasing drawdowns, not increasing returns. However, I will implement one idea: paying out the profit, if the account balance is above starting line, after hitting two winners in a row - this allows me to stay extremely focused, take only quality trades and avoid the feeling of giving back paper profit.