PureBytes Links
Trading Reference Links
|
Thanks Volker,
That seems sound, to be wary of curve fitting at this layer as well as with
the strategy optimization. The layering effect can be a trap I think -
stick an overoptimized money management strategy on top of a perfectly good
entry/exit strategy and the entry/exit strategy becomes a waste of time.
The criteria I had in mind when testing MM runs like this;
A: Ensure that the same MM parameters provide similar benefit on the same
entry/exit strategy for a number of single symbols. This is what you are
getting at, yes?
B: Ensure that there is a smooth curve for a given metric amongst a range of
inputs for a single symbol.
"A" above is intended to be able to infer whether the effects of the MM
strategy are due to the strategy itself rather than the effects of MM
parameters specific to that particular curve. IE, keep all things equal,
change the symbol, if the effects are similar, then it's likely due to the
strategies in place.
You mention using related markets. What do think of using unrelated markets
with less chance for correlated price movements?
"B" above is intended to guage the robustness of the approach. If a minor
difference in parameters creates wildly different results, there can't be
much faith in a particular MM strategy/parameter combination.
What do you think? Are there further possibilities for avoiding
over-optimzation of a money management strategy? Are there any particular
MM strategies you are a fan or a skeptic of?
-----Original Message-----
From: Volker [mailto:volker@xxxxxxxxxxxxxx]
Sent: Wednesday, 12 October 2005 6:11 PM
To: 'ip'; omega-list@xxxxxxxxxx
Subject: RE: Money management/leveraging - varying position size to control
risk/enhance profit
Any MM approach on a single strategy on a single symbol rather than the more
complicated notion of portfolio management could be just another (hidden)
layer of over optimization.
Try it after you found your combination of method and MM let's say on the
S&P on other related markets like the DJ, DAX, EuroStoxx...
Kind regards,
Volker Knapp
(www.wealth-lab.com)
-----Original Message-----
From: ip [mailto:icp@xxxxxxxxxxxxxxx]
Sent: Wednesday, October 12, 2005 3:12 AM
To: omega-list@xxxxxxxxxx
Subject: Money management/leveraging - varying position size to control
risk/enhance profit
I would like to open a conversation on the topic of money management.
There seems to be a wide variety of algorithms, but generically, the issue
addressed is the criteria of when to and by how much to increase the size of
positions taken.
So, my questions are;
* What approaches to money management have you had a positive experience
with?
* What approaches to money management have you had a negative experience
with?
* What importance to you place on money management?
* Which do you prefer - simple or complex money management strategies?
* Which is more effective - simple or complex money management strategies?
I'm thinking here mostly of managing the size of short-term trades on a
single strategy on a single symbol rather than the more complicated notion
of portfolio management. However, all thoughts will be gratefully received.
Regards
Ian
|