|
PureBytes Links
Trading Reference Links
|
Gary,
> The problem with applying this logic to trading systems is that you
> assume each trade is totally independent from the previous ones.
> Many people say that's a valid assumption -- e.g. Monte Carlo testing
> is based on that assumption -- but in my opinion that's not always
> true.
Agreed. In fact I'd say never true. The market is made up of individuals
who have memories. What happens today affects what happens tomorrow.
Today's market is not independent of yesterday's market. People are also
irrational and inconsistent. Therefore, while the market may remember
yesterday, how it reacts today is irrational and inconsistent. Monte Carlo
testing provides some indication of a system's historical consistency. In
fact Monte Carlo simulations are about the only way to take the bias to be
successful out of testing. But that is all that MC simulation does. When
you start trading a system, all bets are off so to speak. A system with
good historical results will not work in reality unless it is based on
exploiting a market characteristic. If it doesn't, it's nothing more that a
mathematical oddity. That is why Mark's Oddball has such power. It
exploits a market characteristic that markets, once in motion, continue in
a direction until they overshoot and then they move back the other way.
Although Mark has indicated to me that stops are not necessary in Oddball's
case at least, we have different views of life and I, personally, cannot
accept that. Stops allow me to reconcile the irrational and inconsistent
nature of the market and my need to be rational and consistent.
Regards,
Mike
|