Saturday, March 2, 2013

Gold vs U.S. debt linear regression -- correlation over?

The chart below is based on data since 2001. The adjusted R-squared is 0.97.


Best fit price is $1,789.
One sigma below is $1,706.
Two sigma below is $1,624.
Three sigma below is $1,536.

The correlation between the gold price and rising US debt has been very strong since 2001. The extremes in deviation have been slightly more than 2 sigma plus (September, 2011) or minus (September, 2008) the best fit linear regression line. The gold price as of March 1, 2013 is testing the bottom end of the range at about 2.5 sigma below the best fit price.

This price either signals an incredible buying opportunity or the end of the twelve year correlation. If gold closes below $1,536 in March, I would suggest it means the end of the correlation. The odds of the correlation remaining intact with a three sigma variation is 99.7% against.

Whether this is the end of the gold vs. debt correlation or not, I am convinced that U.S. debt will continue to accumulate at a faster than average pace. Apologists for the administration like to point to the second derivative of the deficit, saying the rate of the improvement, or reduction, in the deficit as a percentage of GDP is better than ever. However, the actual deficit for 2012 was a very high 8.5% compared to the average from 1970-2008 of less than 3%. If the sequester remains in effect through 2013, the estimated deficit would still be a higher than average 5.5%. Total credit market debt remains near all time highs.

No comments:

Post a Comment