Threshold Asymmetric Autoregressive Stochastic Volatility Strategies for financial markets

Since volatility and risk are intimately related in the financial markets, it is no surprise that volatility is one of the most controversial variables in the financial literature. This is the reason why an increasing number of volatility models have been developed since the introduction of the ARCH model by Engle in the early eighties. In this paper, we focus on the stylized fact known as the “leverage effect” in the financial literature. To explain the leverage effect, or the asymmetric answer of the volatility (good news and bad news have different effects on the conditional variance), we contribute to the literature with an asymmetric ARSV model: the Threshold Asymmetric Autoregressive Stochastic volatility (TA-ARSV) model initially proposed by So et al. in 2002 and eventually developed by García and Mínguez in 2009. The TA-ARSV strategy adds two new parameters to the ARSV model,measuring the effect of positive and negative returns on volatility. Among other advantages, the TA-ARSV model does not need to assume correlation between the innovations of the returns and volatility equations, as other ARSV models do, to capture the leverage effect. The power of our proposed strategy has been previously empirically demonstrated when it comes to explaining the above stylized fact for the prices of electrical power or for the level of pollution. In this article, we apply it in the American, Chinese and European financial markets, and compare the TS-ARSV yields with those obtained from the traditional GARCH and ARSV models. As expected, the results obtained confirm that in the financial arena the TA-ARSV model is also superior to the other competing models.