During the late 1990s, many institutional investors used a model where the expected equity return was a fixed constant estimated as the historical mean over the longest available sample. As markets rose, this model led to progressively higher expected return estimates and larger equity allocations. The primary flaw in this approach is best described as: