Working Paper: NBER ID: w23040
Authors: Ioannis Branikas; Harrison Hong; Jiangmin Xu
Abstract: Households hold nondiversified stock portfolios of firms headquartered near their city of residence. Explanations assign a causal role for proximity, either in generating an informational advantage or a familiarity bias. Empirical analyses assume households locate randomly, even though they optimally select a city. This selection is important since latent location factors might be correlated with latent demand for local stocks. Building on location choice models from urban economics, we develop a Heckman (1977)-style model to account for the effect of location choices on portfolio choices. Adjusting for selection significantly reduces local bias and the performance of local stock picks.
Keywords: location choice; portfolio choice; local bias
JEL Codes: G02; G11; R2; R22
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| proximity to firm headquarters (R30) | overweight local stocks (G31) |
| local stocks outperform distant stocks (G14) | overweight local stocks (G31) |
| selection bias (C24) | local bias in portfolio choices (G11) |
| distance (R12) | portfolio weight (G11) |
| random assignment to cities (C90) | less local bias (J15) |