Can Survival Bias Explain the 'Equity Premium Puzzle'?

Abstract

We show that the well-known model of market survival of Brown, Goetzmann and Ross (1995) fails to explain the "equity premium puzzle." The reasons are (1) the survival bias implied by the model is too small; (2) the model predicts rapidly declining of survival bias in the equity premium over the history of the survived market. We also demonstrate that other survival models are unlikely to succeed either, since to constantly generate high survival bias, the ex ante probability of long-term market survival has to be extremely small which contradicts the history of the world financial markets. Given that no theory in the existing literature predicts high survival bias in the U.S. equity premium, the current concerns for such bias are probably without grounds.

JEL Classification: C4, C5, G1