My question is about how to implement scaling-in for an Alpha in the Algorithm Framework. For example, I want to scale into a break-out like this: 25% on the first break of the entry level, then an additional 25% each time price breaks 0.5 x ATR further.

I think I could use one of the Insight parameters to indicate the relative position size, but I want to use them for their intended purpose. I understand magnitude refers to the expected price move, but what is the intended purpose of confidence and weight? Would it cause compatibility issues (e.g. for Alpha scoring) to use them for this relative position size purpose?

Alternatively, is it better to issue a new Insight for each partial entry, i.e. 4 separate Insights, with the Portfolio Construction module knowing that each Insight is only 25% of the maximum position size?

Any advice welcome.

Cheers, Tony