I tried doing the math on this but got a little lost. I’d really like to hear other people’s thoughts about it.
Here’s the risk control strategy.
Say you like to risk a certain fixed percentage of your account per trade like 2% or whatever. Now suppose you halve the risk after every loss and then, if your current risk is below the original, you double the risk after every win until you get back up to the original risk size (where you stop). Then the total amount you could lose, even after nothing but losses, would be capped at twice the original risk. At least as a sum of percentage points.
Now compare this to a risk control system where you cut the percent risk by some fraction after a loss and then stay at that smaller risk size until you win back the percentage points lost during a drawdown. Then both strategies would take the same number of trades to scale in and out of a losing streak. (The first strategy pyramids down and up while the second strategy subtracts and adds, but it’s the same number of steps down and back up the ladder back to the original risk size.) Also both strategies should earn the same because both net zero percentage points as you move in and out of a losing streak. (I’m not really sure about that last part.) But obviously this second strategy has some positive risk of ruin.
Of course this ignores some pretty important stuff like compounding and reward/risk ratios. It also kinda confuses percentage points won and lost with amounts won and lost. But it does seem to make sense.
Zero risk of ruin?
IF this is a zero risk-of-ruin strategy, here are some more things about it:
Of course at some point you’d really be paper-trading, but if you mentally kept track of risk sizes, it would still be zero risk-of-ruin. Even for losing traders. (The winning traders would have their worst losing streaks interrupted. And the losing traders would at some point get to paper trading status and have to stay there.)
If you didn’t ‘paper trade’ the strategy past a minimal risk size, but kept the risk at that minimal percentage, then the winning traders would have some incredibly low risk-of-ruin. But the losing traders would still have a 100% risk-of-ruin only it would take them much longer to go broke because they were risking a minimal amount.
It makes sense philosophically because this is really the opposite of a martingale strategy which of course has a 100% risk of ruin even for winning players.
BTW it makes sense to me to call this a reverse martingale system but I just googled ‘reverse martingale’ and this is definitely not that.
More importantly than all that, do you think this is a practical/viable/good risk control strategy?
Here is one i will throw out there keep in mind money management strategys can only enhance a already flat profitable trade method.They not will make losing trades into winners.I have not tried this,but i cant see how it would hurt high percentage trades above 75 percent.
Trade with 2 1/2 percent of capital after each profitable trade raise it .40 percent after a losing trade drop it .20 percent never go higher than 5 percent of capital or lower than 1 percent.When you hit 5 percent revert back to the 2 1/2 percent of captal.This will enhance any streak of winning trades and you will be backing off any consecutive losing trades.The killer on this would be the chop(up and down) just like the market winner loser winner loser.That is why it is important that it is a high win percentage method so you take full of advantage streaks in a controlled fashion.The reason i thought about this was Oanda uses units so you can be quite specfic about your trade amounts.Now i just have to find that high pct method.
Some investors use some form of kelly criterion strategy.I dont know of any books Reaper that i have read Dr Thomas Carr,Toni Turner they spend just few pages on it.They do have entry and exit strategies percentage profit exit,price resistance or price support entry,average true range,parabolic Sar stop,price pattern entry.
With a reverse martingale strategy you increase the bet size after a win (or string of wins), and the risk per trade is always capped below by some minimum percentage. So that kind of strategy could never have a risk of ruin lower than the strategy of always sticking to that minimum percentage.
With this strategy you are decreasing the bet size after a loss and the risk per trade is not capped below but is capped above (by the original risk size). So this strategy would always have a risk of ruin lower than the strategy of always sticking to the original risk size.
And this strategy could even possibly have a zero risk of ruin. Because, disregarding compounding etc, the total amount lost, as a sum of percentage points of the original account, could never be more than twice the original risk. For the non-math geeks that’s because 1 + 1/2 + 1/4 + 1/8 + etc etc = 2.
Martingales are when you take on more risk when you lose. They are guaranteed to lose. So long as you’re playing the game against someone with alot more capital than you, which we are. It’s an especially heinous example of ‘gambler’s ruin’.
Reverse martingales are when you take on more risk when you win. They’re also bad.
What I was talking about was always reducing risk when you lose. So it’s completely different. Probably there’s something wrong with it like cutting into your ev too much. But it’s not a martingale or a reverse martingale. And you would definitely have a ridiculously low risk of ruin.
Technically speaking, if you’re using a fixed % risk (as opposed to a fixed $ risk or position size) you’re risk of ruin is zero because your losses get progressively smaller and smaller as your account declines, with your account balance approaching but never reaching zero. Of course, that assumes you don’t have a minimum trade size, or that you don’t have some other definition for ruin.
I would look at it inconjuction with your trading system based on how it operates. The frequency and lengths of winning and losing runs can influence the best choice of position sizing.