Working Paper: NBER ID: w12500
Authors: John Ammer; Sara B. Holland; David C. Smith; Francis E. Warnock
Abstract: This paper investigates the underlying determinants of home bias using a comprehensive data set on U.S. investors' aggregate holdings of every foreign stock. Among those foreign stocks that are not listed on U.S. exchanges, which account for more than 96 percent of our usable data sample, we find that U.S. investors prefer firms with characteristics associated with greater information transparency, such as stronger home-country accounting standards. We document that a U.S. cross-listing is economically important, as U.S. ownership of a foreign firm roughly doubles upon cross-listing in the United States. We explore the cross-sectional variation in this "cross-listing effect" and find that the increase in U.S. investment is greatest for firms that are from weak accounting backgrounds and are otherwise informationally opaque, suggesting that the key effect of cross-listing is improvements in disclosure that are valued by U.S. investors. By contrast, cross-listing does not increase the appeal of stocks from countries with weak shareholder rights, suggesting that U.S. cross-listing cannot substitute for legal protections in the home country. Nor does the cross-listing effect appear to be driven simply by increased "familiarity"� with the stock or lowered cross-border transactions costs.
Keywords: home bias; crosslisting; US investors; transparency; accounting standards
JEL Codes: G11; G15; G3; M4
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| crosslisting (Y90) | increased US investor interest (F39) |
| crosslisting (Y90) | US ownership of a foreign firm (F23) |
| crosslisting (Y90) | improvements in disclosure (G38) |
| crosslisting does not increase appeal of stocks from weak shareholder rights (G34) | legal protections in home country are essential (P14) |
| increased familiarity or lowered cross-border transaction costs (F69) | crosslisting effect (C34) |