The interconnectedness between financial futures markets, commodities, and the macroeconomy: Empirical evidence from a time-varying approach

Abstract

This thesis contributes to the existing literature throughout its three empirical chapters. The first empirical chapter contributes new empirical evidence on the interconnectedness of oil, gold, and financial futures and their spillovers across the US, Europe, and Asia. The empirical results show a significant and changing relationship among oil, gold, and financial futures, especially during the global financial crisis. Dynamic analysis (rolling window and sub-samples) shows that crude oil markets become significantly more influential during the crisis, whereas gold turns from a net giver, in the pre- and post-crisis periods, to a net receiver during turbulent times. West Texas Intermediate and Brent provide valuable information about return dynamics, whereas S&P500 and FTSE100 play significant roles in volatility spillovers. Asian futures markets are strongly influenced by changes in the US and UK oil and stock futures markets. Finally, using different permutations of Cholesky orderings (Klobner and Wagner, 2012), confirm that the spillover index for both return and volatility are generally overestimated when the generalized forecast error decompositions are used. The second empirical chapter examines the connectedness, in return and volatility systems, among futures markets and macroeconomics variables from 1997 to 2015, covering two significant crises: the global financial crisis and the tech-bubble. We utilize a time-varying parameter VAR model (TVP-VAR), which is based on the recently developed connectedness approach by Diebold and Yilmaz (2014), Korobilis and Yilmaz (2018), and Koop and Korobilis (2013). Our results show a changing level of connectedness with an average of 70.5% for return and 75.6% for volatility. We find that macroeconomic variables are the main contributors to the overall forecast error variance, a result that holds at both return and volatility levels. For example, non-borrowed reserves and total reserves show the highest contributions among all other macro and finance variables. Overall, our findings are robust to Bayesian prior choice and reflect the rapid influences of both crises, which is essential to formulate policies that seek to achieve financial stability. Our final empirical chapter inspects the return and volatility interconnectedness between futures markets in the US, UK, Europe, and Asian markets (Japan and China) across different assets (commodities, indices, and exchange rates) using the connectedness approach. The main goal of this study is to examine the level of connectedness throughout a sample that covers the Covid-19 outbreak and vaccine rollout period. The empirical results demonstrate substantial evidence that the return and volatility connectedness are relatively high, mainly when the World Health Organization (WHO) categorized Covid-19 as a pandemic on 11/03/2020. Furthermore, the findings noticeably show that the spillover effects are mostly evident from the UK and Europe to Asian markets (Japan and China) and rarely from the US to Asian futures markets. The UK and European indices (FTSE100 and STOXX50) and exchange rates (GBP-USD and EUR-USD) are the net transmitters to the whole return system. In contrast, the volatility system results reveal spillover effects from GBP-USD, EUR-USD, and considerably from S&P500 to all other markets and regions. Finally, Bai and Perron’s (1980) test is applied to find the structural breaks and the results show three structural break dates that cover all key dates and provide valuable information. The results confirm that the return and volatility connectedness in the first sub-sample (the Coivd-19 period) surge to exceed 40% and 60%, respectively. More, the vaccine rollout sub-sample period documents a significant decline in connectedness for both systems.