|
PureBytes Links
Trading Reference Links
|
>Bob - I don't see why it's not appropriate to "fade" the optimal f
>value by a certain percentage ....say -20% rounded down..... so if
>the recommended leverage is "10" then use "8".
That is one approach. The percentage seems a bit arbitrary to me.
>OR a better idea is to use calculus here and determine the point at
>which the optimal f begins to decrease at a more rapid rate.... based
>on your chart, that would be about "4".
>
>thus maximizing the delta return/delta deviation slope.
>
>between "4" and "6" the slope CHANGES rapidly downward.
Ralph Vince discusses this point, which he calls the "inflection point", in Chapter 5 of his latest book (along with several other possible points to the left of the Optimal_f point).
I think the best fixed ratio point according to economic theory it the point determined by your personal utility function, if you can determine it.
This assumes you are always investing a fixed percentage of your account value so that if you are making money, the size of your trades increase exponentially. The attached GIF shows the performance of a "decent" futures trading system over a 13-year period using compounding at Optimal_f. Note that it is almost a straight line on the logarithmic "Value" scale. The small glitch in about June of 1997 looks pretty minor on this scale. But, in fact, it is a 93% drawdown in our then $93 billion dollar account. I suppose it is OK to lose $27 billion every decade or so...
I real life, you could not trade even near Optimal_f in most markets since you would be limited by the margin requirements well before then.
Most traders I know do not compound their returns anyway. They tend to keep a nearly constant account size and withdraw profits to live on. So the whole subject of compounding indefinitely is somewhat academic.
Bob Fulks
Attachment Converted: "f:\eudora\attach\Compound.gif"
|