With apologies to Mr. Krugman, I must correct some of what I wrote in my piece, “Pushing on a String? Credit Marches to its Own Drummer.” When one does statistical analyses, one needs to understand the limitations/features of the tools that one uses. Bloomberg’s regression function had a funny default that led me to make an error. Had I done it right, the R-squared over the full sample period would have been 64.8% (correlation 80.5%), with a beta of 0.614. Lagging the Fed funds target by one year, roughly the time it takes Fed policy to work boosted the R-squared to 77.2% (correlation 87.9%), with a beta of 67.1%.
But, here ‘s what is unusual. If one is looking at the last five years, the relationship has broken down. During that period, with no lag, the R-squared was 11.2% (correlation 33.5%), with a beta of negative 13.0%. Even with the lag, the R-squared was 3.8% (correlation 19.4%), with a beta of negative 3.7%.
My conclusion: given the unusual credit conditions in the 2000s, where we have had extremes of default and monetary policy, I would not rush to say that the Fed is pushing on a string, yet. That said, the debts of financial companies are a larger part of the index than they were five of ten years ago, and they are the ones in trouble at present, unlike the prior difficulties in industrials and utilities in 2001-2003. Because of that, the Baa index of Moody’s may lag longer than ordinary versus Fed funds… but Fed policy has been called impotent before, and usually just before it shows its bite, as in the tech bubble of 2000, or the liquidity rally of spring 2003.
To my readers: if you see something that might be amiss in my writings, post a comment. I owe it to all of you that I post corrections when I make mistakes. Thanks for bearing with me on this one. In the original piece, I sounded more certain than I should have, to my detriment…