[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Limited life span of mechanical systems?



PureBytes Links

Trading Reference Links

carrslem wrote:
> 
> >
> > What would be the mechanics that would cause a system like OB to stop
> > working ?
> 
> If you discover the answer to this, please let us all know!
> 

Carrol,

It was me, not BF, who posed that question.

In any case, I would empirically quantify the fact that any system has
"stopped working", if it exceeds its previous depth and duration of
drawdown by a wide margin. If e.g. MaxDD during the last 5-months
increased to $32k vs $22k for the previous 51months, i'd say it stopped
working.

I keep an eye on about 10 different systems based on OddBall. The
variations I track include trend and gap filters of my own and all
"public" versions discussed over the last 2 years. Some of them had nice
smooth equity curves (which in retrospect I attribute to curve-fitting),
going back to 1996 until end of Sep-2001 (years 2000 and 2001 were
generally very good for OB-type systems, whereas 1999 was bad for most
of them).

My own OB-base system, produced very small profits and two 100+ point
DDs the last 5 months. But it had a good time in June-2002 where it came
ahead by about 180pt overall in U2, whereas Z2 is breakeven: TS says Net
AvgTrade (Win-Loss) is less than 1pt, so in reality it'd have produced a
mild loss.

MaxDD during this period was 120pt, which is still much lower than
historical MaxDD, so I wouldn't say OB has "stopped working". But I
think it's a very hard system to trade (I tried, but was unsuccessful).

I know others who reported good results of their own OB-versions. Care
to share how they performed the last half year?

Regards, Michael.