Scaling down your position size when you lose - don't do it! I have modelled it extensively and you end up flatlining. The issue is, if you risk 2% of your capital - and you lose, you don't now want to consider your capital at 98%, because 2% of that will be around 1.9. Lets say you have win/lose ratio of one out of 3, and by winning we mean 2*risk. That is a breakeven ratio. But not if you have scaled down your position size. Your win never brings you back to square one. You have to back your strategy in this game. You have to know how much drawdown you can take before reaching hte dreaded point of no return, and you have to have the confidence that your system will not let you get there. If you don't have this - go back to the drawing board.
no - if you hit 50% you are stuffed, period. I don't know what the point of no return is, but 50% has to be well past it - at this point you have to figure that something is seriously wrong with your trading strategy. What i am saying is, if you have 100k, and you apply the 2% rule, and you lose, then you have 98k left. Don't now apply the 2% rule to the 98k, because if you lose again, you will have 96.xK left. Now if you win on try 3, (assume your target is 2*risk) then you will only recover 3.x%, not 4% - which puts you at less than starting, and you will now be in a vicious circle. Remember that your system is a function of risk vs reward as well as the ratio of wins to losses. If you downsize your position on losses, then you introduce a 3rd factor - the sequence of events. And the first two give us enough headaches to worry about
Been there, done that, clawed my way out of it. Learnt many leassons in the school fee days and am more strategic in choices. Still working on the emotional aspect but getting there and making better choices based on better experience and choosing better stocks and positions.
OK, but then your risk reward ratio must be like 500 to 1 or otherwise you will be betting on a sequence of events that will give you x number of positive trades. The first scenario is unlikely and the second re-iterates my point - if you downscale your position size, then you introduce the sequence of events as a criteria for success
Scaling capital has the following effect: let initial capital be "C" then a positive trade will give C*1.02 and a genative trade will give C*0.98 A sequence of trades ... C * 1.02^p * 0.98^n because multiplication is commutative. with p number of positive trades and n number of negative trades. And because multiplication is commutative the order with which the trades happen is not important. You make money as long as p > n. so any strategy that ensures p>n, will make money in the long run. You cannot run out of capital - I ignore small issues like trading costs. Not scaling will take C to zero with any sequence (any) that where n-p>50. OK, given that would be a pritty crappy strategy. PS: if you fix your growth expectation (say 15%) then that allows you to compute the p/(p+n) ratio - some say this is 0.6, that is 60% successfull trades.
Actually you need a p>n by 2.02% in order to break even, is that your point? So that if your strategy gives a win ration of less than 52% you are stuffed. OK. This is interesting since 52% is about the win-ratio of the bank in roulette. ;-)
So the tradeoff between your suggestion and the traditional approach is that in your approach you should avoid any sequence that has n-p>50 and in the traditional scaled capital approach you should avoid any strategy with p/(n+p)<0.52. makes sense now.