Thursday, January 10, 2008

Measuring Trading Performance of My Expert Advisor System

At the end of day 35, my expert advisor experimental trading fund stood at $1050.79. So far, it only made a gain of 1.8% from an initial starting fund of $1031.85. To provide a better picture of its trading performance over the period of 35 days, I performed a modified Sharpe-Ratio type calculation, known as the standard score.

Standard Score
"Each standard score has a corresponding percentage rank. This measure allows you to determine, assuming everything stays the same, what percentage of time periods (e.g. weeks) will be winners versus losers. It is based on absolute trade performance, not percentage gain. This is very important if you're trying to make a living by trading (will there be a paycheck next week?)"

Based on the calculation, my EA standard score is 0.007, and its corresponding percentage rank is 0.5. The result simply implies that I only have a 50% chance to make money on a trading day using my EA system. Obviously, this is not good enough. And if stays like this, my EA fund is definitely going nowhere. Hence, improvements are needed.

Possible Improvements
Future work in the following aspects is to be considered for improving the score, and to guarantee my future paycheck:
1) Increase the average winning size /book profits more often (implemented) and devise new trailing strategy
2) Increase the number of winners/ filter out underperformed markets
3) Reduce the average losing size / devise new exit strategy for losing trades

I will also be using a new trade history table to review my daily performance from now onwards.

No comments: