Yeah, I've seen that image quite a bit, thanks. Surprise surprise, it's like a bunch of random noise blew differently through all those curves (which, yes, are generated by some sort of similar underlying human phenomena). And yet what you're basically doing is zooming in on the early part of one of them, where there's likely to be especially big noise, trying to fit a nice elegant little curve anyways, and then making investment decisions based upon that. Sometimes what gives the elegance-seeking part of our intellect great satisfaction just doesn't matter. This particular domain is one where qualitative fundamentals analysis actually seems better than quantitative attempts.