|
PureBytes Links
Trading Reference Links
|
John Sweeney S&C wrote:
>
> Oops! I shouldn't have implied anything mysterious. It just seems to me to
> be an odd idea: that one constant fraction of capital should optimize
> portfolio growth. I'd like to find the original thinking behind that so I
> could examine it closely.
>
> John
>
Again, there is no "one constant fraction " to "optimize" growth. But
the Optimal F concept (3 Ralph Vince books are the original, thorough
works on this) does give an optimal fraction of the portfoilio to bet
for optimal growth. But that fraction can change depending on changes
in your system's performance. I am sure that Stocks and Commods has
published articles about Optimal F.
The idea isn't that odd but it does require some assumptions. It is
based on system trading. If you have done proper testing, you can break
your system's performance into a series of numbers (averages). Then
assuming that your system is robust and will continue to perform
similarly in the future, you can use those numbers (ie % winners,
biggest loser, avg trade, # trades) to create an optimal percentage to
risk on each trade. Again, though, it can change if your performance
changes (primarily if you achieve a larger than expected biggest loser)
so it is not constant.
This is a very simplistic explanation of Optimal F, take a look at those
books for more.
To go back to fixed fraction. It is just a variation of Optimal F.
Depending on the profitability of a system optimal f can be quite high
(30-40%), which can be quite scary in real life. So to make drawdowns
more manageable, you use a lower percentage. Vince points out that ANY
deviation from optimal f destroys that overall profitability, but that
is the price we pay to sleep at night.
Eric
|