I read somewhere (and i'm just typing this out of my head) that for a lognormal model: Expected return does not change over time because investment return over non-overlapping periods are assumed to be independent of each other, and therefore the model is consistent with weak EMH as knowing past patterns cannot help you predict future ones Then I also read that: The lognormal model generates equity prices by directly using past prices, unlike Wilkie's model that involves indirect calculation The former claims the model does not depend on past information, but the latter says otherwise. Did I read it correctly?
It is the INCREASE in the log share price that is independent of the past, not the share price itself John