|
PureBytes Links
Trading Reference Links
|
: I'm trying to understand how to think of MDD.
It is my opinion that the Max DrawDown statistic in the TS system report is
almost useless. It is the best "worst" case scenario. A more realistic
approach is to run a Monte Carlo Simulation / a Probability of Ruin
calculation on a system, to find out what is drawdown you may likely face in
the future, with any particular system. Bill Brower has a pretty good MCS
(he may still be on this list - we haven't heard from him in a while - and I
forgot the exact spelling of his website. It may be
www.insideedgesystems.com . Bill, correct me if I'm wrong.). I highly
recommend you get that.
With the information from a MCS, let's say that I now know that I have a 4%
chance (or less) of seeing an X amount of drawdown. Then for every (X+margin
requirement) dollars I have, I'd trade 1 contract [even then, this approach
is aggressive.]. The idea is that you allocate enough money to survive the
drawdown and yet still be able to trade. Since the margin requirement
changes with the market action, you have a limited way to account for the
variablity of the markets. Lately, for me, the formula is now (X+Margin +
small fudge factor). :-)
|