McKinnon Tamamlayıcılık Hipotezinin ARDL Sınır Testi Yaklaşımı ile Analizi: Türkiye Örneği

Abstract

McKinnon (1973) and Shaw (1973) argue that financial liberalization practices in developing countries will stimulate domestic savings accumulation by raising interest rates, and the funds created will be a driving force for economic growth by financing the necessary investments. According to this hypothesis, known as the McKinnon's complementarity hypothesis in the literature, money is basically seen as a necessary tool for capital formation and the existence of a complementarity relationship between money demand and physical capital formation is emphasized. Our study, the relationship between real interest rates, demand for real money and the real physical investment by testing for Turkey investigates the validity of the hypothesis of complementarity. The study ARDL (Autoregressive Distributed Lag Bounds Test) test approach is used and Turkey Economy 2005: Q4-2019:Q4 period with quarterly data analysis is performed. In our study, unlike the relevant literature, Fourier ADF unit root test with non-linear structural break was used as well as standard unit root tests. ARDL bounds test analysis results; It indicates that there are negative and statistically significant relationships between real money demand and the ratio of investments to income and real deposit interest rate in the long run. When the short-term results are evaluated, it is seen that although there is a negative relationship between money demand variable and real deposit interest rates, there is no statistically significant relationship between real money demand and investment rate variables. In addition, while there is a positive and statistically significant relationship between the investment-to-income variable and real deposit interest rates in the long run, it is observed that this relationship is not valid in the short run. Considered together with findings obtained from the study of complementarity hypothesis in the case of Turkey the economy has reached the conclusion that partially valid