Nonlinear cointegration tests: An application on the phenomenon of price bubbles

Economic time series generally occur in nonlinear form as well as non-stationary. Because of these properties, the process of series being stationary is implemented. Performing the necessary analyzes ensures that analyzes are obtained in effective and without deflection. As a result of the cointegration analysis of non-stationary series, results showing the long-term equilibrium in stationary structure can be obtained. Considering the nonlinear properties of these variables, the use of nonlinear cointegration tests increases the reliability of the results. In this study, 24 indices in Borsa Istanbul were analyzed on a sectoral basis with monthly data for the period 1997-2019. The data in logarithmic form which has long-term equlibrium relationship were examined by nonlinear KSS (2006), Hepsağ (2019) cointegration test. In the result of study; cointegrated relationship was found on only "Menkul Kıymetler Y.O." in the indices analyzed, and price bubbles phenomenon was found in the other indices. Keywords: Time series, Stationary, Unit root, Cointegration, Price Bubbles, Financial time series, Nonlinear, KSS (2006), Hepsağ (2019)