|
PureBytes Links
Trading Reference Links
|
Hello list,
Friday, December 21, 2001, 8:27:29 AM, you wrote:
ME> On Fri, 21 Dec 2001, grehert@xxxxxxxxx wrote:
>> I'd like to begin a discussion on the not very glamorous subject of "system
>> reports."
If this conversation is about testing a strategy, I would say that it
is better to always test with a fixed $ or share count. Applying any
dynamic method in this area may hide the true performance of the strategy
that is under test. Depending on how aggressive, probably in real
life unreasonable amounts, using MM can show a deceptive outcome to the test.
If you you talking about how to manage the correct # of $s or shares,
well that is another story which I do not know to much about.
I can tell you that I have played with a combination of a simple fixed ratio
+ volatility method which I kinda like simply because it seems to be
"reasonable" on how much and when to up and lower the bet size. I have
not done any risk of ruin kind of tests. I know that I can raise the Sharpe
ratio up by .5 to .7 of a point with this method using some
conservative or reasonable settings. Is it good? I really do not
know. This is all magic to me and only paper money.
--
Best regards,
Ernie mailto:ebonugli@xxxxxxxx
|