Risk-management criteria in the Latin-American stock markets: an assessment with a TGARCH model with a skewed normal distribution and autoregressive conditional asymmetry

Abstract

We build a TGARCH model with a skewed normal distribution and autoregressive conditional asymmetry. We use the model for modelling series of stock-market returns and for investigating some risk-management criteria prevailing in the Latin-American stock markets. The main results support the usefulness of the model. Particularly, they suggest that hedging and diversification practices among the markets may be useful for risk-management purposes. Moreover, they suggest that the most risk-averse investors are in Argentina and the least risk-averse ones in Colombia. Furthermore, they imply that the behaviour of investors may be more complex than the one postulated by the mean-variance paradigm.