Research

Publication:

“Transportation Infrastructure and Trade” with Zheng Han, Japan and the World Economy, Vol. 64, 2022, 101162.

This paper offers a variant of the Ricardian model able to structurally interpret the estimate of country specific variable—transportation infrastructure. Guided by this new theoretical framework, this paper shows that transportation infrastructure enhances international trade more than internal trade. Further quantitative analysis suggests 10% increase in transportation infrastructure induces about 4% increase in real income and more than 95% of the gains concentrate on the infrastructure improving country. This paper also suggests that transportation infrastructure improvement increases real income mostly through internal trade cost reduction. All the above results suggest that better infrastructure leads to sizable gains providing additional empirical support to policies aiming to improve transportation infrastructure.

Working Papers:

“Optimal Containment Policy with Private Protection” (Job market paper) (Under review at Economic Modelling)

In this paper, I develop a macro-SIR model to analyze how private protection changes optimal containment policy. Mask-wearing as a form of private protection reduces infection risk without requiring a proportional reduction in economic activity and, when sufficiently effective, qualitatively reshapes the optimal containment path. Optimal containment can be phase-shifted-remaining relatively flat during the infection upswing and tightening as infections decline. Under the benchmark calibration, mask-wearing reduces infections more strongly and causes a smaller recession than containment when each is considered in isolation. Although mask-wearing is effective in suppressing transmission, containment remains welfare-improving because it mitigates the infection externality generated by infected individuals’ economic activity.

“Lockdown Policy Rules with a Hospital Capacity Constraint” with Taisuke Nakata, Hiroki Sakamoto, and Hiroyuki Uneya (Under review at Journal of the Japanese and International Economies)

We analyze how hospital capacity affects health and economic outcomes during a pandemic using a macro-SIR model featuring a lockdown policy rule. The rule instructs the government to implement a lockdown when the number of ICU patients exceeds a trigger threshold—motivated broadly by a hospital capacity constraint—and to lift the lockdown when it falls below a lifting threshold. When vaccines are not available, we find that the government can reduce both COVID-19 deaths and economic loss by raising the trigger threshold in some situations. When the vaccine rollout is expected to begin in the near future, we find that the government can reduce both COVID-19 deaths and economic loss by lowering the trigger threshold.

“Real-Time Covid-19 Projections in Tokyo: Lessons for Future Pandemics” with Jianing Chu and Taisuke Nakata （Under review at Discover Public Health）

We examine the properties of five real-time COVID-19 infection projections in Tokyo. We find that projections tended to be (i) pessimistic, (ii) less accurate during the fifth infection wave, and (iii) optimistic before the peak and pessimistic after the peak. If policymakers and the public were to utilize real-time projections in future pandemics, it would be useful for them to be aware of the properties of these projections.

“Balancing Safety, Liquidity, and Yield under Geopolitical Risk: Model-Implied Evidence from BRIC Economies” with Changrong Lu, Bing Pang, and Fandi Yu (Resubmitted to Emerging Markets Review)

This study develops a counterfactual reserve-allocation framework to examine how geopolitical risk affects the feasible set of foreign-exchange reserve allocations in emerging markets. Using BRIC economies as the main sample, we derive model-implied reserve allocations from a mean-variance optimization model incorporating a COVOL-based geopolitical-risk constraint. The short-term debt (STD) ratio is used as a liquidity-pressure state variable, rather than as a proxy for central-bank preferences, to construct high-and low-adequacy counterfactual environments. The results show that geopolitical risk does not generate uniform diversification. In high-adequacy states, model-implied allocations adjust mainly along the USD-EUR-GBP margin, whereas in low-adequacy states, the relevant adjustment margin shifts toward substitution between JPY and CHF among lower-volatility safe-haven currencies. For highly exposed economies, geopolitical pressure may contract the feasible allocation set rather than merely shift optimal weights. The findings indicate that geopolitical risk affects not only model-implied reserve-currency weights but also the feasibility of reserve rebalancing.

“Does Comovement Reveal a Channel? Semiconductor Policy Shocks and Supply-Chain Transmission across Global Equity Markets” with Bing Pang and Yuanhao Wang (Submitted to International Review of Economics & Finance)

We examine whether comovement in global semiconductor equity markets around semiconductor policy shocks reveals supply-chain-specific transmission. We construct a global vector autoregression (GVAR) in which semiconductor policy shocks enter as exogenous drivers, using monthly semiconductor-sector returns and realized volatility for ten core economies from 2013 to 2025. We compare six alternative connectivity matrices, using semiconductor trade as the baseline and cross-border portfolio holdings as a nearly orthogonal financial-linkage matrix. Channel discrimination is based on out-of-sample forecast accuracy and adjudicated through the model confidence set. Our results show that direct pricing effects are limited. Although cross-market spillovers are strong, they do not appear to be supply-chain-specific. The semiconductor trade matrix does not outperform alternative connectivity structures out of sample and is statistically indistinguishable from the nearly orthogonal portfolio-holdings matrix. In the economy-event panel, neither network centrality nor exposure to China explains abnormal returns, while direct exposure to the United States shows only a suggestive negative effect in the shortest event window.