|
PureBytes Links
Trading Reference Links
|
Lincoln,
I too explored the K-ratio and gave up on it. It's concerned about how
straight the equity curve is while I want assurance about its propensity for
going up. I have toyed with modifying the Sharpe Ratio, which is the
average return divided by the standard deviation of the returns. Where the
standard deviation would normally look at all the deviations from the mean,
I have it consider only the drawdowns, excluding the equity increases. I
use this as an indicator when I add a system or commodity to a portfolio of
the whether the portfolio performance is improved by the addition.
My favoured way to calculate expected drawdown is to sample from the closed
trade history (or daily equity changes) thousands of times (Monte Carlo
style) and look at the 6.7th percentile of the drawdowns. This is the worst
drawdown I could expect in fifteen years.
A lot of people say to expect a drawdown of double the worst drawdown in
backtested results to allow for curve-fitting.
Paul
|