|
PureBytes Links
Trading Reference Links
|
In that last note I calculated the position size as follows:
value99=2000/(xaverage(truerange,50)*bigpointvalue);
A couple things:
1. My scaling was set for $50 per point to simulate e-mini contracts. The
above statement will buy enough eminis to swing $2000 per average true
range. If your SP scaling is set for $250 then you will need to use a number
bigger than 2000 or else your contract size will be less than 1 at some
point during the simulation.
2. In the 1980's the point moves in the SP were about 1/20 of what they are
now. So for a given account size you'd have traded 20 times the contracts
back then as you would now. The EL statement above will do this.
3. There are a number of commercial SP systems available, some from
contributors to this list, that justify their existence with a long-term
backtest. But not a single one takes this factor into effect! In my opinion
these systems are tragically FLAWED. The backtests do NOT represent realitic
statistics, especially regarding runups and drawdowns.
4. Having been made aware of this, any system vendors possessing a modicum
of integrity should immediately modify their systems as described and re-run
their backtests. I suspect things will CRUMBLE before your very eyes. Just
imagine how some of those losing trades from the past will look when you
trade 10 -20 times the contracts.
5. I have attempted to communicate privately with one such vendor about this
issue but the email silence was deafening. I am now considering taking it
public. This is a very major misrepresentation that needs to be exposed. You
would think that a professional system vendor would know better.
rgds phil
|