Date of Award
Doctor of Philosophy (PhD)
Chapter 1 examines the impact of signed realized measures on forecasting realized volatility using a Time Varying Heterogeneous Autoregressive Model (HAR) framework.To account for this we propose an empirical methodology contribution considering a family of 3 extended models (Signed TV-HAR), where the disentangled continuous and jump signed variations are incorporated. We demonstrate empirically (using 3 major financial indexes for the period 2014-2020) and via Monte Carlo simulations that the proposed family produces significantly better forecasts than the standard Corsi (2009)’s HAR model and some of its classical extensions. The purpose of chapter 2 is to estimate and forecast the Mexican IPC (Quotes and Trade) Index. We introduce a new model that extends the Realized GARCH models of Hansen et al. (2012). Our model generalizes the original specification along three different directions.First, I adopt an autoregressivespecification for the volatility dynamics. Second, it features a time varying volatility persistence. That is, the response coefficient in the volatility equation is time sensitive. Finally, our framework allows to consider, in a parsimonious way, the inclusion of a jump adjusted realized measure in the measurement equation of the model. The forecasting performance of the model is evaluated obtaining gains relative to the traditional R-GARCH model. In chapter 3 the objective is to establish if there are Granger causality relationships between income inequality and trade in Mexico. What is the statistical relationship between globalization and income inequality in Mexico? Is there a causal relationship between inequality and trade volume in Mexico?
Chair and Committee
Gaetano Antinolfi, Ian Fillmore, Jose Figueroa Lopez, Robert Parks
Alcantara Lizarraga, Jose Angel, "Essays in Volatility Modelling and Applied Econometrics" (2021). Arts & Sciences Electronic Theses and Dissertations. 2389.