I keep hearing about traders acheieving only around 60-70% success rate with their trading. Why is this the case ? What are the factors that one should be looking out for to increase these odds?
Can they be improved upon?
It would be great to hear from traders who have overcome these kind of problems.
Yes, it can. My system runs between 93% to 97% accuracy to a specified target and under a defined MAE. Occasionally, the systems sees what I call the Out of Phase Performance Bracket, where it takes a rapid nose dive down to near 89% to 93%, before popping its nose back up, hammering the throttle forward and soaring into the high 90's again. Of course, it cycles back down to near the 93% and then levels off for a while. It has been following this cycle/patter for the past three (3) years now and I've decided to freeze the system and not over-tweak or over-optimize it to the point of breaking it.
I can live with the performance life cycle at this point and have actually developed a negative performance break-out theory that I am now testing, where I will use the failed trade patterns to artificially boost the overall trade performance. I'll do this by allowing the system to run, but turning "Off" the actual trade signal, when the system gets beyond its historical breaking point 97%. I'll then wait for the system to fail into its Out of Phase Performance Bracket (the high 80's) and then allow the trade signal come come alive again and take what should be a high probability trade sequence back up the ladder to the high 90's.
It should have an afterburner affect on the overall system accuracy. At least, that's my theory until the results dictate otherwise. Yes, it can be done, but not with inside the box thinking and blinders on, which is what you will find on most trading boards these days, unfortunately.
Example: You have so-called "experienced traders" who still don't know how to trade a simple Moving Average Crossing methodology, to its maximum degree of profitability. An MA Crossing methodology is not the most efficient way to trade, but its extremely simple, easy for the Newbie to get their head around and profitable when done properly - though not as profitable as a true Predictive Model for trading that actually works.
So, I would say, be careful when you ask for "expert" advice online and make sure that the people giving you advice, know what they are talking about AND can back it up with proof. I have on thread for Newbies on an MA Cross method that will be clearly demonstrated to work - to the degree that such a method can be profitable. It is a very good method for teaching some very important lessons to the Newbie, such as Magnitude and the importance of knowing that there are certain patterns in the market that love to repeat themselves on a very frequent basis. Those lessons can be carried over into more advanced forms of trading system development down the road and such lessons, once learned, can pay significant benefits in your research.
Learning about pattern recognition and the concept of magnitude early on in one's trading career, is priceless.
