|
PureBytes Links
Trading Reference Links
|
If you backtest your trading system on enough data, you end up with the
profile of how your system handles price shocks. 10-15 years of data on
most markets has enough price shocks to tell you how your trading system
handles price shocks. Some exceptions are markets like gold where you will
have to test on older data. Do not use the TS Max DDown causes it hides
alot of information when you trade through a price shock. Use the
mark-to-market Drawdown. This is especially so for long term trading systems.
The point is the shorter your time frame, the less exposure you have to
price shocks while the longer your time frame, the more likelihood you will
be caught on a large profitable/unprofitable price shock. Nobody can be
immuned against price shocks even if your time frame is very, very short.
Price shocks are part of this game so accept them and live with them.
If you have a good robust trading system and you beleive that markets
exhibit infinite variance (i.e. stable paretian distributions (ala Paul
Levy and Benoit Mandelbrot)) then your real time trading results will
produce LARGER WINNING TRADES and LARGER LOSING TRADES than your backtest.
Interesting.
How do you minimise the effect on price shocks ?
QUICK ANSWER: Diversify.
LONG WINDED ANSWER:
There is a balance point between where you need to diversify the number of
markets you trade and reduce your betsize in each market so that price
shocks do not hit you too hard, but if you diversify TOO MUCH, then you end
up exposed in too many markets, your net trading profitability is
"diversified away" (i.e. your profits become bland because you diversified
the good profits away) and you also expose yourself to MORE and MORE price
shocks. I do not have the exact algorithm for this "balance point" yet but
I am working on it.
/*/*/*/*/*/*/*/*/*/*/*/*/*/*
Robert J Bianchi - Director
Venitia Pty Ltd
Email : r.bianchi@xxxxxxxxx
Phone : 61-7-3899 9998
Fax : 61-7-3899 8605
/*/*/*/*/*/*/*/*/*/*/*/*/*/*
|