Do Peer Effects Influence Portfolio Choice? Evidence from 401(k) Allocations

Abstract

Financial researchers agree that allocating money to employer stock in a 401(k) plan is a poor strategy, yet many employees do so. Not only does this investment strategy bear unrewarded idiosyncratic risk, but it also correlates employees' retirement portfolios with their human capital. I find evidence that this selection is influenced by the investment preferences of peers, which includes both management at the firm and employees at other firms headquartered in the same city. Specifically, the percentage allocated to company stock by employees in the 401(k) plan is positively related to both the net open-market purchases of company stock by management and the average allocation to company stock in 401(k) plans by employees at other firms headquartered in the same city. Surprisingly, the allocation to company stock in 401(k) plans increases with investment in employee stock ownership plans (ESOPs), which are a dominant substitute since they often offer stock at a discount. To identify the cause of this relationship between employees and management, I provide additional support for the peer effects hypothesis through interaction tests. The relationship between employees and management is influenced by distance between peers and trust within the firm - two factors that alter group cohesion and, in turn, the effect of peer influence. These results are robust to size of the firm and also exist for the subsample of firms that have only made matching contributions in cash. Consistent with our understanding of the inertia that exists in employee investment decisions, employee allocations to company stock are influenced for many years by management's initial investment choices made near the time of going public. Lastly, alternative methodologies, including propensity score matched samples and panel regressions, yield the same conclusions.