Persistence in pension account returns: the impact of survivorship and reaction of asset flows

Abstract

This paper is the first to examine the effects of survivorship on persistence in the performance of U.S. equity pension accounts. Evidence of predictability of returns is stronger with the total sample of accounts than with the survivor only sample. The weaker result in the survivor only sample is consistent with the effects from a model where multiperiod performance evaluation determines whether accounts are terminated. Where other studies found predictability to be driven by the persistence of poor performers, we do not. We employ a time-varying conditional or dynamic alpha as our measure of past performance. We find the dynamic alpha predicts future returns across equity style groups, but not consistently within equity style groups. Given evidence of predictability, we examine the asset flow-performance relationship and find pension plan asset flows to equity accounts tend to follow three-year excess returns. We do not find a significant relationship between more sophisticated or risk-adjusted measures of performance and future asset flows. Since we do not study the hiring decision, pension plans may use sophisticated measures in the hiring decision, but not in the termination decision nor in determining intermediate asset flows. Asset flows follow performance more for small and large cap equity styles than for value or growth style accounts. Surviving accounts exhibit a stronger asset flow-performance relationship than the total sample.