I have noticed something odd which I am sure someone has run into.
Here is a simple strategy with IOG:
variables :a1(0), a2(0), a3(0);
a1= close data1; (ask on SPY)
a2= close data2; (ask on SPY) (yes, also ASK , a1=a2)
a3= (a1 + a2)/2;
so a3=a1, a3=a2, and a1=a2
Strat1
If a3>a3[1] then buy at market;
If a3<a3[2] then sell at market;
Strat2
If a1>a1[1] then buy at market;
If a1<a1[2] then sell at market;
What accounts for the dramatically different results when backtesting for the two strategies?
Does it add up?
-
- Posts: 176
- Joined: 02 Feb 2008
- Location: Lausanne, Switzerland
- Has thanked: 9 times
- Been thanked: 5 times
Re: Does it add up?
In MC User Guide for IntraBarOrderGeneration, there is a note stating the following:
I don't know if the user guide is up to date though.Notes
Intra-bar order generation is currently supported for single-data series charts only. Intra-bar order generation for multiple-data series charts will be implemented in the next release of the platform.
Attributes are applied at the time of compilation and cannot be changed at run-time.