The Dynamic Relationships between Oil Prices and the Japanese Economy: A Frequency Domain Analysis

This paper investigates the dynamic relationships between oil prices and the Japanese economy from a frequency domain perspective. Both the frequency domain causality test of Breitung and Candelon (2006) and the frequency dependent regression method developed by Ashley and Verbrugge (2009) are deployed in this research. The frequency dependent regression analysis indicates that nonlinear relationships exist between oil prices and the variables such as industrial production and consumer price index at the low frequencies, while the nonlinear associations at the high frequencies are merely detected between oil prices and unemployment rates. The results of the frequency domain causality tests suggest that oil prices have significant predictive power for industrial production, consumer price index and unemployment rates at the low frequencies. In addition, oil prices can predict industrial production and unemployment rates at some higher frequencies.