Econometric and Statistical Methods, Financial Markets
Abstract
In this paper, the procedure developed by McQueen and Thorley (1991) to test for a random walk is applied, with some modifications, to the securities traded in the Mexican Stock Exchange. In order to, the daily returns of a portfolio made up of highly traded securities are examined by means of a second-order Markov chain avoiding the assumption of normality. The results of the estimation and the tests on the transition probabilities are presented and discussed. Also, it is investigated if the Markov chain is ergodic and therefore (covariance) stationary. Finally, forecasts based on the obtained estimations are provided.
