Measuring the speed of convergence of stock prices: A nonparametric and nonlinear approach

This paper evaluates the speed of convergence across national stock markets employing a nonlinear, nonparametric stochastic model of the relative stock price. To estimate the persistence of the relative stock price, we employ an operational algorithm that is based on two statistical notions: the short memory in mean (SMM) and the short memory in distribution (SMD). Using MSCI stock price indices of the G7 countries, we obtain strong empirical evidence of convergence of national stock prices in France, Germany, and the UK vis-à-vis the US index. Also, we obtain much faster convergence rates from our nonlinear models in comparison with those from linear alternatives. On the contrary, our results imply very limited evidence of convergence for Canada, Italy, and Japan. Similarly weak evidence of convergence was obtained from non-G7 developed countries. Our simulation exercise for portfolio switching strategies overall confirms the validity of empirical findings in the present paper.