I'm having a look at the bollingerpenetration system to get an idea of how RE systems hang together. Once done I'll then be able to look at migrating some of my own code.
First of all, I notice that bollingerpenetration is long only but while I see how orders get entered the mechanism for how trades get closed isn't so clear. From what I see with my untrained eye it looks like the system should just keep buying until it runs out of money. Can someone explain how this system closes it positions?
Secondly, entries (and I assume exits) dont seem to be named.The position list report doesn't show the name of the entry that resulted in a new position. So, given a system that has multiple methods for entering or exiting how would i know which piece of system logic triggered that particular action?
Thirdly, what determines how many shares or contracts get bought. The simulation I ran started out buying about 50 shares per entry. Where are these position sizes configured? Perhaps there's a sample system that shows exposure management more clearly?
Is there a report, and an object that I can query in code, which shows account exposure by order and cumulative exposure aside from number of units taken?