|
PureBytes Links
Trading Reference Links
|
Size modulation based on TA criteria against a forcing function like
SharpeRatio.
This is a good concept. IE: You would use the systems internals to dictate
what leverage to trade that system. Note that DynaPort (I am not associated)
does
this using a family of Diff. E. Q's. In TradeStation you would have to brute
force
select the TA used in the modulation. Note that mean markowitz
or fixed fractional can't handle this type of analysis.
Still, carefully check OOS performance.
You might have just caught a small sample
of size adjustments.
Sincerely,
Mike Barna
President,
Trading System Lab
"TSL automatically designs Trading Systems
in a few minutes using an L-AIM-GP."
"Software that Creates Trading Systems with no coding"
www.TradingSystemLab.com
----- Original Message -----
From: "Mark Johnson" <janitor@xxxxxxxxxxxx>
To: <omega-list@xxxxxxxxxx>
Cc: "Gary Fritz" <fritz@xxxxxxxx>
Sent: Wednesday, August 19, 2009 2:38 PM
Subject: Re: Switch between systems
I was merely giving the Original Poster, Dave
Pyle, what he asked for. His message said,
"There must be a better way to switch between systems
than summing the last x bars. Suggestions to try?"
and I offered three suggestions to try: %K, LRslope, %B.
Will these perform well on Dave Pyle's systems?
I hope so; but there's a chance they won't.
Here's a fourth suggestion: replace the word "switch"
by the word "knob". Instead of restricting yourself
to only two possibilities:
(100% of system A, 0% of system B)
(0% of system A, 100% of system B)
allow yourself to imagine trading X% of system A and
(100-X)% of system B. Imagine a "knob" that you can
smoothly rotate, from (A=100, B=0) to (A=0, B=100).
Now imagine using technical analysis indicators to
choose an appropriate value of X%.
The natural way to implement the knob is in the
choice of position sizing, at trade initiation
and also throughout the lifetime of the trade
(if your software allows it).
Mark Johnson
At 01:24 PM 8/19/2009, Gary Fritz wrote:
Apply technical analysis "indicators" to the equity curves
of the two systems and switch to whichever one is "strongest"
based on technical analysis.
Mark, does this actually work for you? When I tested it, it failed
miserably. When a system went into a losing streak, you'd take all the
losses it took for it to get kicked out of contention. Then when it
pulled out of the losing streak, you'd miss all the wins it took for the
equity curve to get strong enough again. It only worked if the system was
prone to extremely long winning or losing streaks -- long enough so that
"missing most of the losing streak" made up for the other weaknesses.
(E.g. a simple long-only stock system might benefit from it, if you kept
it out of multi-year down markets, but I believe you can do a better job
of that within the system logic itself.)
Now I wasn't switching *between* multiple systems, just between "system"
and "no system." But I would think the dynamic would be the same.
Gary
|