Not sure why the study period matters as long as their sample is representative of the population, which it seems to be?
So for example, let's say tech starts to boom, and everyone good with tech will all of sudden have much better financial performance, although everyone were making their predictions based on historical performance.