Chinese and Global ADRs: The US Investor Experience

Abstract

We study outcomes to ADR (American Depositary Receipt) investments between August 1954 and September 2020, with particular attention to ADRs associated with Chinese firms. Overall, ADRs improved investors’ wealth by $1.03 trillion, with more than a third of this amount attributable to ADRs associated with Chinese firms. A value-weighted portfolio of ADRs associated with Chinese firms earned 14.1% per year since the first Chinese ADR was created in 1993, as compared with 9.9% per year for the overall US stock market over the same period. These data are relevant to current policy discussions focused on Chinese firms listed in the United States.

Editor’s Note

Submitted 24 January 2021

Accepted 7 April 2021 by William N. Goetzmann

Declaration of Interests

Disclosure: The authors report no conflicts of interest.

Editor’s Note

Submitted 24 January 2021

Accepted 7 April 2021 by William N. Goetzmann

Acknowledgments

The authors thank William Goetzmann (the executive editor) and two anonymous referees for constructive and insightful comments and suggestions that have significantly improved the article. Te-Feng Chen and John Wei acknowledge financial support from the Research Grants Council of the Hong Kong SAR, China (GRF15505518).

Notes

¹ See, for example, Chiu and Yoon (2020) and Hulbert (2020).

² Most often, an ADR represents a claim on a specified number of shares in the underlying firm. Duggan (2017), however, described how some ADRs associated with Chinese firms represent a claim on an intermediary firm known as a “variable interest entity” (VIE). Typically, the VIE has a contractual claim on the profits of the Chinese firm but does not directly own its assets.

³ A broader issue is that ADR listings are not a random sample of non-US stocks, because the decision to create an ADR is endogenous. For that reason, the results here are not directly comparable to studies such as Bessembinder, Chen, Choi, and Wei (2020), who studied long-term shareholder outcomes for all publicly listed common stocks in 42 markets.

⁴ The sample start date corresponds to the ADR listing of Royal Dutch Petroleum. CRSP reports return data for a handful of ADRs prior to August 1954 but does not indicate the home market of the firms involved. We excluded six ADRs for which CRSP lists the United States as the market of origin. The CRSP database includes only stocks listed on the major US exchanges and, as such, does not include data for the “Level I” ADRs that are traded OTC, the most prominent of which is Tencent. In light of Tencent’s large market capitalization (more than $750 billion as of the end of March 2021) and stellar performance, it seems inevitable that our conclusions regarding the superior performance of a value-weighted portfolio of Chinese ADRs would only have been strengthened if we had been able to include Level I ADRs in our study.

⁵ The wealth ratio is the cumulative gross return to the portfolio of interest in units of the cumulative gross return to the benchmark. See, for example, Loughran and Ritter (1995). A wealth ratio greater (less) than 1.0 implies outperformance (underperformance) relative to the benchmark.

⁶ We relied on the wealth creation measure introduced by Bessembinder (2018), as described further in the next section. The measure is implemented on the basis of ADR shares outstanding, not the number of common shares outstanding. Thus, we measured wealth created from the viewpoint of those who invest in ADR shares, not the larger wealth improvement to all equity investors in the firms.

⁷ To assign ADRs to markets, we relied on the most recent location variable “Loc” in the CRSP/Compustat Merged Database Linking Table. However, we reassigned to China 10 ADRs that are identified in Loc as associated with Hong Kong SAR. These reassignments were based on information in the Compustat “Busdesc” field, which describes operations primarily in mainland China. The 10 ADRs reassigned from Hong Kong SAR to China for purposes of this study are China Unicom (Hong Kong) Limited, CNOOC Limited, China Mobile, Brilliance China Automotive Holdings Limited, Hutchison China MediTech, China Telecom Corporation Limited, iClick Interactive Asia Group Limited, Le Gaga Holdings Limited, Agria Corporation, and Focus Media Holding Ltd. These ADRs accounted for a total of $8.61 billion in shareholder wealth creation.

⁸ The CRSP data for ADR shares outstanding appear to contain errors in some cases. We corrected the ADR shares outstanding data as follows. First, we compared the ADR market capitalization implied by the CRSP ADR shares outstanding with the firm market capitalization implied by the Compustat Fundamental Quarterly firm shares outstanding data in those cases where the Compustat data were available. Working backward from the latest data, for any ADR for which this ratio exceeded 120% in any month, we adjusted shares outstanding in prior months so that all prior ratios were equal to or less than 100%. Second, we identified those cases where the market capitalization (share price times shares outstanding) of the ADR increased relative to the prior month by a factor of five or more times even while the return for the month was 200% or less, as well as those cases where the market capitalization fell relative to the prior month by a factor of five or more times even while the return for the month was –50% or less. For these cases, we adjusted the number of shares outstanding downward (either for preceding or succeeding months) in such a way that the market capitalization change was consistent with the return for the month.

⁹ The developed markets subsample includes all ADRs associated with firms from Austria, Australia, Belgium, Bermuda, Cyprus, Denmark, Finland, France, Germany, Greece, Hong Kong SAR, Ireland, Israel, Italy, Japan, Luxembourg, the Netherlands, New Zealand, Norway, Portugal, Singapore, South Korea, Spain, Sweden, Switzerland, Taiwan, and the United Kingdom. ADRs associated with firms from other sample markets are in the emerging markets subsample.

¹⁰ The assignment of markets to regions was as follows: Europe includes firms from Austria, Belgium, Denmark, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Luxembourg, the Netherlands, Norway, Poland, Portugal, Spain, Sweden, Switzerland, and the United Kingdom. The Asia-Pacific region includes Australia, China, Hong Kong SAR, Indonesia, Japan, Malaysia, New Zealand, Philippines, Singapore, South Korea, and Taiwan. Latin America includes Argentina, Bermuda, Brazil, the Cayman Islands, Chile, Colombia, the Dominican Republic, Honduras, Mexico, Peru, and Venezuela. The “other” category includes Cyprus, Ghana, India, Israel, Papua New Guinea, Russia, South Africa, Turkey, and firms without location information.

¹¹ The cumulative gross return is obtained by adding 1.0 to each monthly return to obtain gross monthly returns and then computing the product across months of the gross monthly returns. If the resulting cumulative gross return, CGR, is obtained from N monthly returns, then the geometric mean return is CGR^(1/N) – 1.

¹² The differential performance of ADRs as compared with the overall US market could be attributable, at least in part, to industry composition. To assess this possibility, we examined cumulative industry-adjusted returns. We assigned each ADR to 1 of 12 industries based on SIC codes provided by CRSP and industry definitions provided by Kenneth French (website: http://mba.tuck.dartmouth.edu/pages/faculty/ken.french/) and defined monthly industry-adjusted ADR returns as 1.0 plus the ADR return divided by 1.0 plus the value-weighted return to the matching industry portfolio, minus 1.0. The value-weighted portfolio that included all 1,162 ADRs generated positive industry-adjusted returns, reflected by an arithmetic monthly mean return of 0.14%, a geometric mean monthly return of 0.07%, and cumulative gross returns of 170%. Notably, the cumulative gross industry-adjusted return for Asia-Pacific firms was 437%.

¹³ The Sharpe ratio was computed as the mean monthly excess (over the T-bill) return divided by the standard deviation of monthly returns and was annualized by multiplying by the square root of 12.

¹⁴ In a study with conclusions that are complementary to our own, Lee, Li, and Zhang (2015) studied stock market performance for Chinese firms that went public via “reverse mergers” (in which a private firm merges with a publicly traded shell corporation rather than using an IPO). They documented that Chinese reverse mergers outperformed their matched US peers despite the inclusion in the sample of some firms accused of accounting fraud.