Nonlinear cointegration analysis of China’s money demand function stability

Abstract

ABSTRACTIn this study, an analytical framework of non-linear co-integration theory is applied to empirically analyze the Chinese money demand function data from Q1 1994 to Q3 2017. The results show that it neglects non-linearities in all variables and that M1 and M2 have long-run nonlinear equilibrium relationships with real GDP, interest rate, inflation rate, the effective exchange rate of RMB, and trade dependence. Further, an enhanced BP neural network is applied to estimate the long-run equilibrium equation, and it is found that the marginal coefficients of M2 are more stable than those of M1. It is suggested that although China was much less dependent on foreign trade since the 2008 financial crisis, the RMB-USD exchange rate and financial variables outside China had a slight impact on China’s money demand because of its closed financial market so that the financial crisis had little effect on China’s financial system.KEYWORDS: Non-linear co-integrationrank testrange testenhanced BP neural networkmoney demand functionstability Disclosure statementThe authors have no relevant financial or non-financial interests to disclose.Notes1. Data on money demand-related variables from 1994 Q1 to 2017 Q3, real and nominal effective exchange rates were obtained from the BIS website, GDP, interest rates, money supply, and RMB-USD exchange rate were obtained from the CEI database.Additional informationFundingThis work was supported by the 2023 Hunan Natural Science Foundation Joint Fund Project: Research on Key Technologies for Measuring, Enhancing, and Visualizing the Competitiveness of Hunan Huaihua International Inland Port under the RCEP Framework (2023JJ50459).Notes on contributorsXiaohui ShuXiaohui Shu is a professor of business administration at Huaihua University. His main research areas of interest is in Economic Statistics, Time Series Analysis, and Regional Economic Development. He has published many academic articles in peer-reviewed recommended journals and has led multiple fund projects.Jinqi SongJinqi Song is an associate professor at Jiangxi Normal University. His research interests include economic statistics, time series analysis, and their applications. He has published multiple academic articles in peer-reviewed journals and has also led several funded research projects.Qinli LeiQinli Lei is a professor in the Department of Statistics at the School of Economics, Jinan University. His primary research interests include sampling surveys, statistical analysis methods, and economic growth. He has published over 50 academic articles in peer-reviewed journals. Additionally, he has successfully led and completed multiple national and provincial research projects, and has also authored more than 10 books and university textbooks.Yangkuo LiYangkuo Li graduated from Jishou University in 2019 with a master's degree. His research interests include statistical modeling and big data analysis. He has done a lot of empirical research in the fields of economics, statistics, and related disciplines, and has published many research papers in academic journals.