表題番号：2025C-232
日付：2026/03/31
研究課題The Distance Puzzle and Tax Havens: A Comparison of Portfolio Investment between China and USA
| 研究者所属（当時） | 資格 | 氏名 | |
|---|---|---|---|
| （代表者） | 国際学術院 大学院アジア太平洋研究科 | 教授 | 菊地 朋生 |
| （連携研究者） | Kwansei Gakuin University | Assitant Professor | Satoshi Tobe |
| （連携研究者） | The University of Texas at Austin | PhD Student | Xiaoyu Cheng |
- 研究成果概要
- This paper investigates the differential impacts of geographic distance and offshore tax havens on the international portfolio investments of China and the United States. Using a Pseudo-Poisson maximum likelihood method, we estimate the effects of distance and tax havens on international investment from both countries. Based on the stylized fact that China invests intensively in tax havens on a residency basis (Clayton 2023), and subsequently uses these offshore entities to route investments to other countries, one might predict a differential use pattern. We hypothesize that, holding other factors constant, China utilizes tax havens more extensively than developed economies. The rationale is that China's domestic institutional environment and stringent capital controls—the very frictions proxied by the distance effect—create a powerful incentive to use offshore intermediaries to bypass these constraints. Regarding the distance effect itself, we postulate a second hypothesis. We pre-suppose that China's investment patterns will demonstrate significant geographic sensitivity. Given that China's financial system is still developing and operates under strict regulations, the frictions that the distance variable captures (e.g., information asymmetry, institutional distance) should be more pronounced. Therefore, we predict that this distance-proxy effect will be more significant for China than for countries with more mature financial systems. Our gravity model estimation yields two principal findings, offering a nuanced answer to our hypotheses. First, the results do not support our initial hypothesis regarding tax haven use. Ceteris paribus, we find no statistical evidence that China utilizes tax havens more extensively than the United States. Conversely, our analysis provides strong support for the second hypothesis concerning the distance effect. The results confirm that China's portfolio investments are indeed significantly more concentrated in geographically closer countries compared to those of the U.S. This finding is particularly robust: the distance coefficient for China's bond investments is not only more negative than the U.S. coefficient but is also substantially more negative than the global benchmark estimate identified by Brei and von Peter (2018). This suggests a heightened sensitivity to the frictions captured by the distance proxy.