Return spillovers and dynamic conditional correlations among US, China, and Japanese stock markets : a comparison between the administration of Obama and Trump

Abstract

The global nature of the financial markets and the capital flows facilitate institutions and portfolio management firms to profit from diversifications. The opportunity to invest in lucrative financial opportunities can also expose investors to different risks. This master’s thesis examines the return spillover effects and dynamic conditional correlations among the US, China, and Japanese stock markets. The goal is to understand and offer fresh perspectives to investors and portfolio managers to allocate decision making. The methodology of this work is based on the vector autoregressive (VAR) and dynamic conditional correlation-generalized autoregressive heteroskedasticity (DCC-GARCH) models to examine the return spillover effects and the dynamic conditional correlations among US, China, and Japan. During the Obama and Trump administrations, there were unidirectional return spillovers from US to Japan, but no return spillovers between US and China or between China and Japan. The DCC-GARCH model exhibits a lower short-term dynamic conditional correlation between US and China, but higher short-term dynamic conditional correlations between US and Japan and between China and Japan during the Trump presidency compared to that of Obama. The long-term dynamic conditional correlation between US and China was higher during the Trump presidency compared to the Obama administration. However, the long-term dynamic correlations were lower between US and Japan, and between China and Japan under the Trump administration.