Abstract:
In this study, we make a three-fold contribution to the literature on gold market analysis. First, we provide evidence for the predictive value of US Nonfarm Payroll (USNP) in the out-of-sample forecast of gold market volatility. Second, we extend our analysis to other precious metals and the US stock market index for robustness purposes. Third, we utilize mixed data frequencies based on the availability of data, thus, circumventing any bias or information loss due to the use of monthly (low frequency) USNP data and daily (high frequency) gold price data. The results show that the USNP, which reflects gain/loss in US non-farm jobs, is negatively related to gold return volatility, implying that deterioration (improvement) in the economy due to job losses (gains) raises (lowers) the gold market volatility as its trading improves (deteriorates) while the reverse is the case for US stocks. The out-of-sample predictive value of USNP in the return volatility of gold is also established as the model which includes the former offers better out-of-sample forecast gains than the benchmark model which ignores it. Additional analyses involving other precious metals, namely palladium, platinum, rhodium, and silver, show the same direction of relationship as gold, albeit with higher forecast gains for silver than the others. Our findings have useful implications for financial analysts and investors.
