was referring to Ex because it takes into account duration in addition to frequency, wich is more accurate imo to measure the stability of activity, than frequency isolated.
I'm thinking out loud, but one way to measure an indicator like that could be, to take the "Average Days per D-period" in Ex to make the benchmark (0), and plot each day the delta between "forecasted activity" vs. "realized activity" on a chart.
Plus, a range of x% acceptable deviation (as Rs does for the risk).
If the delta is negative (below 0) you are warned that the activity is decreasing.
If the delta is positive (above 0) you are warned that the activity is increasing.
If the delta extends below the x% acceptable deviation = (the score is impacted?)
If the delta extends above the x% acceptable deviation = warns overtrading (the score is impacted?)
(It leads to a lots of questions regarding to the calculation and sensitivity, or weight in the overall score etc... but it is not the topic)
imo it is a complementary information, not have to replace it in any way. The study we are talking about is independent from the investable attributes already in place.