brooksrimes
Junior member
- Messages
- 16
- Likes
- 0
After many years of hard work, I have been unable to find/create a system to trade.
In some ways, deep down, I feel that trading may be too complex to systematize, so I look at discretionary trading methods. But even discretionary trading has rules. In order to "prove" that a discretionary trading method "works", you either have to do one of 3 things: trade it in real time, trade it in a simulated mode or back test it. #1 and #2 take a lot of time and then you only know how it works over a relatively short period of time.
#3 allows a test spanning years but can be difficult and in some cases near impossible to code. This also gets into all the complex and time consuming issues of optimization, avoiding curve fitting, walk-forward testing, etc. I often find that a system will work well for a given period of time but becomes unprofitable or has unacceptable low profits over a larger test. Occasionally, I have also come across systems that generate decent profits but have unacceptably large drawdowns.
I feel like I'm stuck in a rut and don't know how to get out. I have reached a certain level of competancy in coding and testing systems, but don't know how to get to the next level.
I certainly don't expect anyone to give me the results of your hard work, but assuming you are a successful system trader : ), I would appreciate any nudges into the right direction.
Brooks
In some ways, deep down, I feel that trading may be too complex to systematize, so I look at discretionary trading methods. But even discretionary trading has rules. In order to "prove" that a discretionary trading method "works", you either have to do one of 3 things: trade it in real time, trade it in a simulated mode or back test it. #1 and #2 take a lot of time and then you only know how it works over a relatively short period of time.
#3 allows a test spanning years but can be difficult and in some cases near impossible to code. This also gets into all the complex and time consuming issues of optimization, avoiding curve fitting, walk-forward testing, etc. I often find that a system will work well for a given period of time but becomes unprofitable or has unacceptable low profits over a larger test. Occasionally, I have also come across systems that generate decent profits but have unacceptably large drawdowns.
I feel like I'm stuck in a rut and don't know how to get out. I have reached a certain level of competancy in coding and testing systems, but don't know how to get to the next level.
I certainly don't expect anyone to give me the results of your hard work, but assuming you are a successful system trader : ), I would appreciate any nudges into the right direction.
Brooks