Deterministic Effects of Volatility on Mixed Frequency GARCH in Means MIDAS Model: Evidence from Turkey

Abstract

Volatility is a key concept for understanding the dual relationships between the economic variables since it is inversely related to the stability of economies. Many models such as GARCH models have been constructed through time to understand which determinants and conditions can affect the volatility. These models mostly show the significant relationships between the volatilities generated by the low frequency macroeconomic activities and the high frequency financial variables in a stochastic way. However, it is required to check whether there exist deterministic effects of volatilities on high frequency economic variables. In order to reveal these deterministic effects, we developed a new component-wise model, namely GARCH-M MIDAS model. We formulate this model on stock prices and exchange rates, in which long run volatility is driven by consumer price and industrial production indexes in a separate way. Hence, our empirical analysis supports that both types of the volatilities have statistically significant deterministic effects on the asset pricing of high frequency financial variables. We also find that macroeconomic activities have a significant role on the asset pricing in long horizons.