Volatility persistence in metal returns: A FIGARCH approach

• Published in: Journal of economics and business Vol. 64; no. 4; pp. 287 - 305
• Main Authors: Cochran, Steven J., Mansur, Iqbal, Odusami, Babatunde
• Format: Journal Article
• Language: English
• Published: New York Elsevier Inc 01.07.2012; Elsevier Science Ltd
• Subjects: Economic crisis; FIGARCH; Financialization; Metal returns volatility; Metals; Rates of return; Securities markets; Studies; Uncertainty; Volatility; Volatility persistence; G12; Financialization; Metal returns volatility; FIGARCH; Volatility persistence
• ISSN: 0148-6195; 1879-1735
• DOI: 10.1016/j.jeconbus.2012.03.001

► Whether the volatility processes exhibit long-run temporal dependence. ► Whether the returns and conditional volatility of returns are affected by the uncertainty brought about by the financial crisis in September 2008. ► Whether the implied volatility in the equity market, as measured by VIX, plays a significant role in determining metal risk and return. This study examines the returns and the long-memory properties of the return volatilities of four metals – copper, gold, platinum, and silver. Daily returns for the January 4, 1999 to March 10, 2009 period are used. Three key issues are addressed: (1) whether the volatility processes exhibit long-run temporal dependence; (2) whether the returns and conditional volatility of returns are affected by the uncertainty brought about by the financial crisis in September 2008; and (3) whether the implied volatility in the equity market, as measured by VIX, plays a significant role in determining metal risk and return. The results show that VIX is important in the determination of metal returns and return volatility. The findings suggest that metal and equity returns are influenced by a common risk factor and failure to explicitly model this factor will yield less than optimal portfolio diversification. Events during the post-September 1, 2008 period contributed to increased return volatility for several of the metals. The interaction effect of VIX and a financial crisis dummy variable is also found to be significant. The results strongly suggest that VIX should be considered in any future modeling of metal returns and return volatility. FIGARCH (1,d,1) appropriately describes the volatility processes as all long-memory parameters are statistically significant.