some more discussions on subject here: http://www.elitetrader.com/vb/showthread.php?s=&threadid=64970&perpage=6&pagenumber=1
I write computer software to test such ideas. I find stopping losses gives results that I do not expect. For example I might observe that a moving average crossover system is unprofitable. I then test if the inverted moving average system is profitable. Since I buy when prices are decreasing and sell when prices are increasing I choose to manage risk with a stop loss rule. Occasionally the stop loss shows a trade that I do not expect. When I model the effect of slippage and commission on the trading system performance may be further reduced. The best performing systems that I recall testing are simple trend following systems. The most profitable method in many cases is a buy and hold forever system. Drawdowns can be great when trading a buy and hold method. I recall testing systems that reduce both volatility and profitability compared to buy and hold. It may seem strange to want to earn less profit than buy and hold, but the draw downs might be reduced enough that I can follow the system. I do not believe I can follow a buy and hold method.
Even though the trader made the wrong decision...at the time he thought it WAS the right choice. It's like speeding around a slow driver then getting caught. If you knew you were going to get caught..you wouldn't do it, same as a trader making a bad trade. They all seemed like good ideas at the time. oh and btw, there are three positions: Long, short, doing nothing cm69
If you are consistently losing money it is because you are consistently inconsistent. Even if you fade your trading signals you will remain consistently inconsistent.
THE best way is to fade another person that consistently ignores stops & pulls a plug too soon on winning trades.
The law of numbers states that even an idiot will get 2-3 out of 10 correct by default. There is however a man amongst us who has consistently got 10 from 10 incorrect for the last 6 years and has still 2 years to run. You may fade his calls with impunity.
It doesn't work because being profitable doesn't only depend on getting the direction right ( long/short). If you don't know when to get out of your trade , you can still lose even though your prediction of the direction of the market was correct. In fact, I would say that knowing when to exit is a lot harder than knowing when to enter a trade !! If you constantly exit "too" early, the ratio of your average winning/average losing trade becomes bad. If you exit too late, aiming for big gains, you will give away most of your profits, and many winning trades will end as losing trades... all money managament. Many newbies don't realize this, only focusing on their entries...
It will be a good thing to create a software that send to the market the opposite orders a person sends, then take 10 (or 50) newbies (who are thinking are trying to win actually) and let the sotware run placing the opposite order they thing they are placing. (then see what happens). May be its a crazy idea. May be not.
Don't think this would work because the majority of people who lose in trading do so because they let their losers run and cut the winners short. Maybe this program would give a higher % of winning trades from the start but people will either be greedy and not take the profit or be fearful of losing it and would cut the trade short with very little profit. The other problem with having a vhigh % win rate is you know you are not wrong very often and if you don't stick to strong risk managment the minority of the time when you are wrong will eat your account up.