CMStatistics 2017: Start Registration
View Submission - CMStatistics
B0636
Title: Consistent test on semiparametric functional coefficient models with integrated covariates Authors:  Shan Dai - The Chinese University of HongKong (Hong Kong) [presenting]
Abstract: Cointegration is a powerful tool in studying long-run relationships among integrated time series and is widely used in macroeconomics and financial time series analysis. Nonetheless, empirical evidence often fails to support the existence of cointegrating relations with fixed cointegrating slope coefficients. Due to plausible source of variability into the cointegrating vectors, semiparametric functional coefficient cointegration model attracts wide attention recently. Based on the asymptotic result for both the regressors and covariates are I(1) variables, we propose consistent test for the instability of cointegrating parameters in the present framework. The asymptotic distributions of test statistics under both null and alternative hypotheses are studied. Monte Carlo simulations are conducted to examine the finite sample performance of the proposed test. Furthermore, we use the proposed test to help verify the test result for purchasing power parity hypothesis between U.S. and Canadian dollars, which is consistent with some recent result.