Research

Publications

1). Xu, X., Wang, W., Shin, Y., & Zheng, C. (2021) Dynamic network quantile regression model. Journal of Business & Economic Statistics, Forthcoming

2). Chen, J., Shin, Y., & Zheng, C. (2021) Estimation and inference in heterogeneous spatial panels with a multifactor error structure. Journal of Econometrics, 229, 55-79. 

3). Li, H., & Zheng, C. (2018). Unit root quantile autoregression testing with smooth structural changes. Finance Research Letters, 25, 83-89. 

4). Li, H., Zheng, C., & Guo, Y. (2016). Estimation and test for quantile nonlinear cointegrating regression. Economics Letters, 148, 27-32. 

Working Papers

1). Estimation of dynamic quantile panel data with  interactive effects. (Job Market Paper)

Abstract: Although the Brexit is done, it has provoked a renewed interest in studying the effect of a EU membership (EUM), or a more general free trade agreement (FTA), on  bilateral trade. While researchers have not reached a consensus by utilising different econometric techniques, they generally assume the effect to be the same at different stages of a Business Cycle. The existing results are therefore unlikely to tell a complete story. In this paper, I developed a quantile panel model to characterise the distributional effects of EMU/FTA on bilateral trade, while admitting the high-persistence and strong cross-section dependence properties of the trade data. By estimating an international trade gravity model for 380 countries pairs, I discovered that the long run effect of EMU on bilateral trade is much higher at lower quantiles (118.72%) than at higher quantiles (27.40%), implying that the boosting effect of EMU is more significant during economic recession.

2). A Spatio-temporal autoregressive factor model of the global business cycle. (joint with Tomohiro Ando, Matthew Greenwood-Nimmo and Yongcheol Shin) 

Abstract: To study the synchronicity of national business cycles, we propose a new heterogeneous-parameter approach in which the global business cycle is modelled as a spatio-temporal autoregressive process with a common factor error structure. To achieve consistent estimation in the presence of parameter heterogeneity and endogeneity, we develop a modified quasi maximum likelihood estimation approach. We show that the resulting estimators are consistent and asymptotically normally distributed. We employ Monte Carlo simulations to demonstrate that their finite sample performance is satisfactory. Based on the proposed estimator, we further develop network analysis tools at both individual and regional level using diffusion FEVDs and multipliers. These tools are then applied to analyse the business cycle synchronisation covering 79 countries in the world over the period 1970-2019 (50 years).

3). Functional-coefficient quantile cointegrating regression with stationary covariates.  (joint with Haiqi Li and Jing Zhang)

Abstract: This study examines the estimation and inference of functional-coefficient quantile cointegrating regression. Firstly, a local linear quantile regression estimator is proposed to estimate the unknown coefficient function. Secondly, to alleviate the endogeneity problem, we propose a nonparametric fully-modified quantile regression estimator that is shown to be $n\sqrt{h}$ consistent and follow a mixed normal distribution asymptotically. Thirdly, we propose two Kolmogorov-Smirnov type test statistics for coefficient stability in a given quantile or across multiple quantile levels. Finally, to improve the finite sample performance, we propose a fixed regressor wild bootstrap procedure and establish its asymptotic validity. Monte Carlo simulation results confirm the merits of the proposed estimator and tests.

Working in Progress

1). A unified approach for panel data models with latent group structure. (joint with Jia Chen and Yongcheol Shin)

2). Panel data models with random interactive effects. (joint with Wenting Wang and  Yongcheol Shin)

3). Spatial panels with a multifactor error structure and multiple structural breaks. (joint with Haiqi Li  and Siqi Dai)

Research Project

Research Assistant for the ESRC project “New Cross-Sectionally Dependent Panel Data Methods for the Analysis of Macroeconomic and Financial Networks” (grant number ES/T01573X/1, in progress)