Kuan-Pin Lin’s research while affiliated with Portland State University and other places

This paper considers a hierarchically spatial autoregressive and moving average error (HSEARMA) model. This model captures the spatially autoregressive and moving average error correlation, the county-level random effects, and the district-level random effects nested within each county. We propose optimal generalized method of moments (GMM) estimators for the spatial error correlation coefficient and the error components' variances terms, as well as a feasible generalized least squares (FGLS) estimator for the regression parameter vector. Further, we prove consistency of the GMM estimator and establish the asymptotic distribution of the FGLS estimator. A finite-scale Monte Carlo simulation is conducted to demonstrate the good finite sample performances of our GMM-FGLS estimators.

We propose a random effects panel data model with both spatially correlated error components and spatially lagged dependent variables. We focus on diagnostic testing procedures and derive Lagrange multiplier (LM) test statistics for a variety of hypotheses within this model. We first construct the joint LM test for both the individual random effects and the two spatial effects (spatial error correlation and spatial lag dependence). We then provide LM tests for the individual random effects and for the two spatial effects separately. In addition, in order to guard against local model misspecification, we derive locally adjusted (robust) LM tests based on the Bera and Yoon principle (Bera and Yoon, 1993). We conduct a small Monte Carlo simulation to show the good finite sample performances of these LM test statistics and revisit the cigarette demand example in Baltagi and Levin (1992) to illustrate our testing procedures.

We propose a multilevel spatial econometric model including spatially correlated error, spatially lagged dependent variable, county-level random effects and nested district-level random effects within a county. We construct joint Lagrange multiplier (LM) test to detect both the spatial effects and the random effects, LM tests to detect the spatial error correlation and/or the spatial lag dependence allowing for both types of random effects, and LM tests to detect the county-level random effects and/or the nested district-level random effects allowing for both types of spatial effects. A Monte Carlo simulation is conducted to show their good finite sample performances.

After more than 10 years since the establishment of the Shanghai Cooperation Organization (SCO), this paper studies how variances and correlations have evolved in four SCO countries' equity markets (China, Kazakhstan, Mongolia, and Russia) relative to Japanese, US and EU markets. We focus on the Mongolian equity market. We trace stock market co-movements of Mongolia and SCO member countries to the world market since the inauguration of the SCO in 2001. We also take into account the effects of recent world financial crisis in the region. As compared to the increasing global trend of market integration among countries, we don't find any significant contagion among SCO countries in the first half of sample periods. After 2005, market integrations among SCO countries are evident.

After more than 10 years since the establishment of the Shanghai Cooperation Organization (SCO), this paper studies how variances and correlations have evolved in four SCO countries' equity markets (China, Kazakhstan, Mongolia, and Russia) relative to Japanese, US and EU markets. We focus on the Mongolian equity market. We trace stock market co-movements of Mongolia and SCO member countries to the world market since the inauguration of the SCO in 2001. We also take into account the effects of recent world financial crisis in the region. As compared to the increasing global trend of market integration among countries, we don't find any significant contagion among SCO countries in the first half of sample periods. After 2005, market integrations among SCO countries are evident.

Lee and Yu (2010) propose spatial panel data models with one-way and two-way fixed effects. Debarsy and Ertur (2010) construct LM (Lagrange multiplier) and LR (likelihood ratio) tests in the one-way fixed effects model. He and Lin (2012) derive LM tests in the two-way fixed effects model. To guard against possible local misspecification, in this paper we apply Bera and Yoon (1993) principle, and construct locally adjusted (robust) LM tests for spatial dependence in both one-way and two-way fixed effects models. Monte Carlo experiment is carried out to show the advantage of using robust LM tests over the corresponding marginal and conditional versions.

In this paper, we define the spatial bootstrap test as a residual-based bootstrap method for hypothesis testing of spatial dependence in a linear regression model. Based on Moran’s I statistic, the empirical size and power of bootstrap and asymptotic tests for spatial dependence are evaluated and compared. Under classical normality assumption of the model, the performance of the spatial bootstrap test is equivalent to that of the asymptotic test in terms of size and power. For more realistic heterogeneous non-normal distributional models, the applicability of asymptotic normal tests is questionable. Instead, spatial bootstrap tests have shown superiority in smaller size distortion and higher power when compared to asymptotic counterparts, especially for cases with a small sample and dense spatial contiguity. Our Monte Carlo experiments indicate that the spatial bootstrap test is an effective alternative to the theoretical asymptotic approach when the classical distributional assumption is violated. KeywordsSpatial bootstrap test–Moran’s I–Monte Carlo–Size distortion–Power

After more than 15 years of Chinese equity markets, we study how variance, covariance, and correlations have developed in these markets relative to world markets, based on the dynamic conditional correlation (DCC) model of Engle [Engle, R., 2002. A dynamic conditional correlation: A simple class of multivariate generalized autoregressive conditional heteroskedasticity models. Journal of Business & Economic Statistics 20(3), 339–350.]. Chinese markets offer A-shares to domestic investors and otherwise identical B-shares to foreign investors. We find that the volatility of A-shares has declined over the past decade. We find no asymmetric volatility relative to world markets in China. Contrary to the global trend of increasing cross-country correlations, we find stationary correlations for China. A-share indices have never been correlated with world markets, and B-share indices exhibit a low degree of correlation with Western markets (0–5%) and a slightly higher degree of correlation with other Asian markets (10–20%). We interpret these findings using Gordon's growth model.

We introduce and apply bootstrap method for testing spatial correlation in a linear regression model. Given the consideration of a fixed spatial structure and heteroscedasticity of unknown form in the data, spatial bootstrap is a hybrid of recursive wild bootstrap method. Based on the Moran's index I, two versions of LM-Error and LM-Lag statistics, we demonstrate that the spatial bootstrap procedure can be used for model identification (pre-test) and diagnostic checking (post-test) of a spatial econometric model. With two empirical examples, the bootstrap method is proven to be an effective alternative to the theoretical asymptotic approach for hypothesis testing when the distribution assumption is violated.

Current Version: March 21, 2006; First Version: January 13, 2006 Using techniques of spatial econometrics, this paper investigates σ-convergence of provincial real per capita gross domestic product (GDP) in China. The empirical evidence concludes that spatial dependence across regions is strong enough to distort the traditional measure of σ-convergence. This study focuses on the variation of per capita GDP that is dependent on the development processes of neighboring provinces and cities. This refinement of the conditional σ-convergence model specification allows for analysis of spatial dependence in the mean and variance. The corrected measure of σ-convergence in China indicates a lower level of dispersion in the economic development process. This implies a smaller divergence in real per capita GDP, although convergence across regions is still a challenging goal to achieve in the 2000s. 21世紀COEプログラム = 21st-Century COE Program [revised]