Resumen

TORRE-TORRES, Oscar V. De la. Investing with Markov-Switching GARCH models: A comparative test between Mexico and Argentina. Contad. Adm [online]. 2021, vol.66, n.1, 00016.  Epub 11-Oct-2021. ISSN 0186-1042.  https://doi.org/10.22201/fca.24488410e.2020.2657.

In the present paper I test the benefits, for active portfolio management purposes, of using two-regime Markov-Switching (MS) models with GARCH variance. This, with either a Gaussian or t-Student homogeneous likelihood function, in the Buenos Aires and in the Mexican Stock Exchanges. By performing 996 weekly simulations from January 2000 to January 2019 in each MS model, I tested the next investment strategy for a U.S. dollar based investor: 1) to invest in the risk-free asset if the probability of being in the high-volatility regime at t+1 is higher than 50% or 2) to do it in an equity index otherwise. The results suggest that the t-Student MS-GARCH model is the best option to generate alpha in Argentina and the constant variance gaussian one in Mexico. This, against a “buy and hold” investment strategy.

Palabras llave : Markov-Switching GARCH; Markov chain processes; Active portfolio management; Buenos Aires stock exchange; Mexican stock exchange; Frontier markets; Computational Finance, Risk management; C580; G11; G170.