So why hasn't the pairs model fired a signal recently? That question has been asked by a few readers and here's what's going on over at MO2 central. The apirs signals are based on the Z-score algorithm which looks for disparities in the standard deviation of the target pair. We complicate the matter a bit by examining a divergent pair based on beta whereas a normal application of the program would be to look at convergent paits such as SPY and UPRO. (see tomorrow's post) We can see from chart A that there are clearly periods when the standard deviations converge and those are out signal points, but we are currently not in such an environment presently.