We build an autorregresive vectors model to forecast the volatility of the Índice de Precios y Cotizaciones of the Mexican Stock Exchange as a function of other market portfolios, and we test its ex-ante predictive power in the period January 1997-December 1998. One of the most surprising results is that our model would have been able to forecast the two contagion effects that ocurred in this time· interval: The Dragon Effect and the Vodka Effect.
