Geographic Lead-Lag Effects

• Parsons, Christopher A., Riccardo Sabbatucci, and Sheridan Titman. "Geographic lead-lag effects." The Review of Financial Studies 33, no. 10 (2020): 4721-4770.
• 关联度动量的秘密：分析师共同覆盖 - 因子动物园

1. Summary

This paper document lead-lag effects in stock returns between co-headquartered firms operating in different sectors.

• 投资组合构造：Every month, we rank each firm $i$ not by its own lagged return (as we would in a simple momentum strategy), but by $r_{c \notin j,t}$, the average lagged return of firms headquartered in the same region, but operating in different sectors. We use a one month horizon both for the sorting criterion (i.e., area-level stock returns are measured over a month) as well as the holding period (i.e., portfolios are reformed at the end of every month). Based on these rankings, we form value-weighted portfolios.
• Location beyond firm headquarters: the question we explore is whether regional predictability is stronger for more regionally concentrated firms (e.g., AutoDesk) compared to those with a more disperse presence (e.g., Whole Foods).
  • To obtain a more general measure of a firm’s geographical presence, Garcia and Norli (2012) utilize a text-based parsing algorithm that counts the number of unique state names mentioned in the annual reports of publicly traded firms from 1994-2008.
  • In all cases, the point estimates for the more regionally concentrated firms are somewhat larger compared to their less concentrated counterparts.