Correlation Of The U.S. Dollar To Commodities

I recently spoke at an ETF conference about commodities. During the panel discussion I mentioned that commodities includes a lesser known factor; currencies. An investor may view commodities partially as a currency play.

Many commodities are quoted in U.S. dollars (spot $DXY). In the short-term the direction of $DXY may not impact commodities (S&P GSCI spot), however, over the longer-term a trending $DXY may impact commodities. For example, if the dollar is moving higher, commodities become more expensive to the rest of the world. Thus potentially reducing demand and potentially having an impact to suppress commodity prices. If the dollar weakens, commodities become cheaper around the world relative to other currencies. This could increase demand for commodities and potentially increase prices.

The Long-term correlation of the $DXY to S&P GSCI is -0.3. However when analyzing the rolling 12-month correlation of the dollar to commodities we find the correlation has a maximum and minimum of 0.64 and -0.92 respectively. Over time the rolling 12-month correlation has gradually become more negatively correlated.

Chart1: 12 Month Rolling Correlation of $DXY to S&P GSCI. March 1985 to June 2016

Source: Bloomberg data

To eliminate some of the noise of a short-term correlation, I also examined the 36-month rolling correlation of $DXY to S&P GSCI (Chart 2) and found a stronger trend for a negative correlation. This would be when longer-term trends occur in both the dollar and commodities. With a maximum of 0.16 and a minimum rolling correlation of -0.76. Since 2008, the 36-month rolling correlation has sustained a strong negative correlation, ranging from -0.76 to -0.54. Implying when the $DXY becomes directional, there is an increased probability for the S&P GSCI to do the opposite.

By Mark Shore, Founder