|
PureBytes Links
Trading Reference Links
|
Glen,
On Tue, 4 Sep 2001 16:55:17 -0700, you wrote:
>Perhaps you could explain in a little more detail how you compute
>your position size -- maybe with an example. Since Optimal f
>bases position size on maximum loss, rather than "real investment"
>(is that another term for cost?), I'm having difficulty envisioning
>your problem.
My problem is not with Optimal f. (I'm quite familiar with probability
distributions and how to use them for risk minimization, and even if I
don't follow exactly the Optimal f procedure as discussed here several
months ago, my re-investment rates work fine with _long_ positions.)
My problem is, how to apply this to _short_ positions. - To give an
example from last month' trades:
My system opened a _long_ position in puts on Aug.06 at 59 /
contract. For this , my mmgmnt calculated the number of contracts in
accordance to my "working capital for trading" (not to be published
here) based on an "investment" of 59 / contract.
At Aug.08 this position was closed by the system at a "result" of
118.8 / contract, which makes about 100% win for the "investment".
(Remark: Not all of my long trades end like this.)
At the same time on Aug.06, my system opened a _short_ position in
calls at 60 / contract. For this, my mmgmnt calculated the appropriate
number of contracts, based on a "result" (!!!) of 60 / contract.
At Aug.15 this position was closed by the system at an "investment"
(!!) of 0.2 / contract, which makes about 29900% win for the
"investment". (Again: Not all of my short trades end like this.)
Do you see the "dilemma"?
For _long_ positions, the mmgmnt is done based on the _"investment"_
(as it should be done), whereas for _short_ positions mmgmnt is done
based on the _"result"_, which can (and should) be a big difference.
In case I could have done mmgmnt for the short position based on the
"investment" (as it should be done), the calculated number of
contracts would have been much higher.
Making the long story short: For _short_ positions the number of
contracts calculated by mmgmnt is too small ( for successful trades),
because mmgmnt is based on the _"results"_ and not on the
_"investment"_ (as it should be).
Any way out of this "dilemma"? - Any hint? - Every short position
trader imo is faced with this problem ...
mfg rudolf stricker
| Disclaimer: The views of this user are strictly his own.
|