Three Essays on Macroeconomic Information, Volatility Persistence and Structural Change

Abstract

This thesis explores the relationship between macroeconomic information and stock market volatility. Firstly, we employ the GARCH-MIDAS model to investigate the impact of fundamental macroeconomic variables on long-term stock market volatility across three developed countries: U.S, UK and Japan. Secondly, we further investigate the impact of macroeconomic information on long-term persistence and structural changes in volatility. Thirdly, we introduce a discrete-time realized volatility option pricing model which incorporates macroeconomic variable for option valuation. This thesis consists of five chapters. In the first chapter, we make a brief introduction of my thesis. In the second chapter, we explore the relationship between macroeconomic information and long-term stock market volatility across three developed countries, U.S, UK and Japan. We employ the two-component GARCH-MIDAS model to carry out international analysis and observe the relationship between macroeconomic variables and stock market volatility changes over time, both in magnitude and significance. This time-varying relationship might largely attributes to the time-varying expectations of market participants towards forthcoming monetary policy changes and macroeconomic uncertainty. In the third chapter, we extend the Heterogeneous Autoregressive Realized Volatility-type models (HAR and Tree-HAR models) that allows macroeconomic information to explain long-term persistence and structural changes in stock volatility, simultaneously. We find that both macroeconomic information and its uncertainty have prominent impacts on stock volatility. Strikingly, macroeconomic information helps to deliver a more elaborate regime-switching structure for U.S stock volatility, which infers a tight link between macroeconomic information and potential structural changes in stock volatility. In the fourth chapter, we employ our extended HAR model for option pricing domain. We aim to examine whether macroeconomic information, through its influence on conditional volatility, can affect corresponding option prices? Root mean squared errors for both put and call S\&P500 Index options show that adding macroeconomic information, in particular unanticipated information, into option pricing process increases option pricing accuracy and mitigates implied volatility biases, relative to traditional Black-Scholes model and GARCH model. In the last chapter, we make a conclusion and possible future directions.

Date of Award	6 May 2020
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Ian Garrett (Supervisor) & Alex Taylor (Supervisor)