Granger Causality Tests And Their Applications

In econometrics and finance filed,Granger causality tests have been developed into a set of useful tools to detect causality between time series and these tests have become the most popular and widely used approaches by scholars and practitioners In this paper,we will mainly focus on the theoretical development process of Granger causality tests and apply these tests to guide our real data analysis.Among the various tests of Granger causality,the Hiemstra-Jones(thereafter HJ)nonlinear test developed by Hiemstra and Jones[1]is the most cited by scholars and the most frequently applied by practitioners and there were over 1550 Google Scholar citations by June 2020,which illustrates its significance in the economics and finance literatures.In the last two decades there are numerous applications and theoretical extensions based on this pioneering work.Although HJ test is widely used,some counter-intuitive results are also obtained.For example,Diks and Panchenko[2][3]found that the HJ test is seri-ously over rejecting through simulation study.These findings motivate us to do these research.The paper is organized as follows.Chapter 1 is the preliminaries,which contains the literature review of the theoretical development process of Granger causality tests and introduces the research background of our investigation and the problems to be explored.In Chapter 2,we briefly review the linear and nonlinear causality tests in bivariate and multivariate test and pay much attention to the HJ test and the multivari-ate nonlinear Granger causality test(thereafter BWZ test)developed by B ai,Wong and Zhang[4],which plays an important role in detecting the dynamic interrelationships between two groups of variables.Furthermore,we try to point out some underlying reasons for the doubtable performance of the HJ test and BWZ test.In Chapter 3,we reinvestigate HJ's creative work in 1994 and found that their proposed estimators of the probabilities over different time intervals were not consistent to the target ones proposed in their criterion.To test the HJ's novel hypothesis on Granger causality,we propose new estimators of the probabilities defined in HJ's paper and reestablish the asymptotic properties which will induce new tests similar to HJ's ones.Some simu-lation work will also be presented to support our new findings.The similar study is also applied to the BWZ test.Chapter 4 shows our aims to explain how can we use linear and nonlinear Granger causality tests to do real data analysis.We study linear and nonlinear growth determinants of Mongolia as a small emerging economy,with a quantified assessment of the impact of China,based on cointegration analysis,the vec-tor error correction mechanism,and linear as well as nonlinear causality tests.They are of importance for policymakers in making decisions regarding the developmental path of Mongolia's economy and in assessing the possible impact of the OBOR ini-tiative.They are of importance for policymakers in making decisions regarding the developmental path of Mongolia's economy and in assessing the possible impact of the OBOR initiative.Last,we briefly introduce our future research and conclude the article.