|
PureBytes Links
Trading Reference Links
|
Alex Matulich on 22/05/08 18:52, wrote:
Adam Hardy wrote:
Is anyone out there interested in 'optimal f' or 'secure f'
fractional money management techniques?
I wrote some software once to examine them in a Monte Carlo fashion.
I tend to follow Ralph Vince and optimal f up to the point where
Vince takes the 'biggest loss' from the trade history to which he
applies the formula.
Two problems with this approach:
1. There is no accounting for comfort level. Applying the biggest
loss in the history to the formula may well mean periods of 80%
drawdown.
2. The sequence of trades is assumed to be unchanged. The same
trades in a different order could result in different drawdowns.
The same trades selected at random over a longer period gives a
different result. Monte Carlo simulation gives you a feel for the
mean and standard deviation for max drawdown and expected equity
growth.
I don't know what Monte Carlo simulations are based on, programmatically, or
what they /can/ be based on, but are you able to define the kurtosis of the
distribution? I mean, a normal distribution would be less than ideal, I'd
definitely want something with fat tails.
Regards
Adam
|