PDF · Last updated: Oct 2025
Abstract
Venture capital (VC) networks facilitate reverse knowledge spillovers from U.S. investments in foreign startups back to the United States. Using the first U.S. VC deal in a foreign company as a quasi-exogenous shock in a difference-in-differences design, I show that investees’ pre-existing patents become 18% more likely to be cited by U.S. entities, with no corresponding change in citations from other countries. Spillovers are concentrated among U.S. startups most closely connected to the investing VC through its network of prior syndication ties, while geographically proximate firms show no such effects. A similar design applied to domestic coast-to-coast VC deals confirms that the network channel is especially salient in cross-border settings. These knowledge flows translate into real outcomes: patent output rises by 10% overall and 22% in investee-related technologies; high-quality innovation increases by 10% overall and 34% in related technologies. Closely connected firms are also more likely to achieve successful exits through IPO or acquisition. The findings indicate that VCs transmit knowledge across borders through their syndication networks, enhancing innovation and performance beyond their direct portfolios.
with Zheyuan (Kevin) Cui · PDF · SSRN
Abstract
This paper finds that familiarity can generate both informational advantages and biases in international equity investments. Using a dataset we construct on managers’ countries of education as a proxy for familiarity, together with detailed holdings of U.S.-based international funds from 2000–2024, we find that funds over-allocate to equities from their managers’ education countries despite no abnormal country-level returns. At the equity level, we find that only the most confident equity positions in familiar markets generate significant excess returns, earning an annual alpha of 5.31%, while the remaining holdings track the benchmark returns. The outperformance persists over time but these positions account for only one-quarter of education-country holdings. This pattern indicates that familiarity provides a persistent informational edge, but only in a narrow set of most confident positions, while broader overweights reflect familiarity bias. Overall, our results highlight familiarity as both a source of private information and a driver of bias in global portfolio allocation.
In or Out? How the Inclusion of Currencies on a Payment Platform Impacts Multinational Firms
with Tommaso Mancini-Griffoli · Slides
Abstract
Research on the market impact of cross-border financial platforms remains at an early stage, and platform design and usage continue to evolve. This paper contributes to that agenda by examining how adding a currency to a cross-border payment platform affects multinational firms, and by offering evidence that can inform platform design and regulatory policy. We study how expansions of cross-border payment platforms affect incumbent multinationals. Our setting is the 2015 inclusion of the Hungarian forint (HUF) in the Continuous Linked Settlement (CLS) system: a shock that lowers HUF foreign exchange settlement risk and transaction costs, but can also facilitate entry by non-domestic firms and intensify competition in Hungary. Using a stock-market event study linked to Orbis subsidiary data on U.S. public multinationals, we find that firms with greater exposure to Hungary experienced more negative announcement-window abnormal returns. Specifically, a one percentage point in a firm’s Hungarian subsidiary share is associated with an 18 basis points decline in firm value. The effect is weaker in high-markup industries, consistent with lower entry threats where incumbents have greater market power. Overall, the net impact of cross-border payment infrastructure depends on market structure: while CLS expansion reduces transaction frictions, incumbent valuations decline when the competitive forces it triggers outweigh the direct cost savings.
