I fed into my PC’s statistical package five years’ worth of data for gold bullion and the Dow Jones FXCM Dollar Index (which represents the dollar’s value against a basket of the currencies of the U.S.’ largest trading partners). I was specifically interested in the extent to which changes in the dollar’s value led to changes in gold’s price.
As expected, I found an inverse correlation: Increases in the dollar’s value tended to correspond to decreases in gold’s U.S. dollar price, and vice versa. Crucially, however, I found that the ups and downs of the dollar were only able to explain about a quarter of gold’s gyrations.
(For the statistically minded among you: The r-squared for the correlation was never higher than 0.26, regardless of whether I focused on daily, weekly or monthly changes in the dollar index and gold.)
His finding makes sense, as the data he used included spaces when the U.S. dollar and gold were going up and down in tandem. The Eurocrisis stimulated demand for both safe havens.
Hulbert's results are presented as a soother for worried gold bugs fearing that the metal will plummet if the greenback snaps back. It's a nice gesture, but the negative correlation (qualitiatively) has gone up recently. Sad to say, a snapback in the greenback will hurt gold.