Kevin Spires, CFA, FRM

The NY Fed has researched the ability of the slope of the Treasury Yield Curve to predict recessions. I have reproduced one of their charts at the left which shows a 3.6% chance of the U.S. being in recession 12 months from now.
I have always liked the Yield Curve as a leading indicator because I think the slope of the curve is mostly driven by the relative tightness or easiness of the Fed's monetary policy. Unfortunately, I believe most recessions are precipitated by the Fed's lurching from overly easy money to overly tight money. Like a drunk continually falling off the wagon, easy money follows tight money as night follows day.
The main reason that I have blogged on this subject is not to take another gratuitous slap at the Fed (as fun as that may be), but rather to point out an improvement in the Fed's recession prediction toolkit. In a March 2012 working paper posted on the main Federal Reserve website entitled "A Dynamic Factor Model of the Yield Curve as a Predictor of the Economy", the Fed has improved upon the NY Fed's simple yield curve model by coupling the trend growth rate of Industrial Production to the slope of the yield curve. I simplify of course. Don't click through unless you have a basic understanding of Markov Processes, Conditional Probabilities, and Maximum Likelyhood Estimation.
The key innovations are an improvement in the ratio of correct predictions (to 100%) and a decrease in the number of false positives (to zero) coupled with an increased average lead time at predicting peaks and troughs in economic activity of 22 months. The end of the 2007-2009 recession was correctly predicted with a lead of 18 months. In other words, this new model predicted an end to the 2007-2009 recession on 12/31/2007 - before most Economists had called a recession in the first place. This is an astounding piece of research.
We all learned in Econ 101 that Monetary Policy has a long and variable lag. Now the Fed has quantified that lag at peaks and troughs - 22 months. I am in the process of reverse engineering their model, but let me give you a hint - there hasn't yet been a call for the next recession. If you have been paying attention to my blog posts, you would know that I am seeing an acceleration in activity that is normally associated with easy monetary policy and I am not surprised by the findings in this newish Fed Research.
I imagine this paper has been circulated to all the high muckity mucks and grand poobahs at the Fed, so why do we continue to see a dovish contingent calling for more Quantitative Easing? I think that the Fed Governors are, in general, no better than you or I would be at this Monetary Policy thingee. Actually, I think the incentives in place make them substantially worse. The Fed just "has to do something" or the whole reason for their being would be nullified. They can't stand still and let nature take its course. They over tighten and over ease to make the memory of their actions closer to the actual turn in the economy. 22 months is a long time to wait out the Economy and the vanity of Central Planning is such that the Fed must be seen to act in concert with the turn in the Economy. Occam's Razor and all that says this is so - although I bet they wish it could be avoided in the case of recessions.
I think the Fed has, once again, stayed too easy for too long. Sooner than most think, I expect the Fed to embark on a long tightening process that will lead us once again into recession - probably sometime 2015.
I have always liked the Yield Curve as a leading indicator because I think the slope of the curve is mostly driven by the relative tightness or easiness of the Fed's monetary policy. Unfortunately, I believe most recessions are precipitated by the Fed's lurching from overly easy money to overly tight money. Like a drunk continually falling off the wagon, easy money follows tight money as night follows day.
The main reason that I have blogged on this subject is not to take another gratuitous slap at the Fed (as fun as that may be), but rather to point out an improvement in the Fed's recession prediction toolkit. In a March 2012 working paper posted on the main Federal Reserve website entitled "A Dynamic Factor Model of the Yield Curve as a Predictor of the Economy", the Fed has improved upon the NY Fed's simple yield curve model by coupling the trend growth rate of Industrial Production to the slope of the yield curve. I simplify of course. Don't click through unless you have a basic understanding of Markov Processes, Conditional Probabilities, and Maximum Likelyhood Estimation.
The key innovations are an improvement in the ratio of correct predictions (to 100%) and a decrease in the number of false positives (to zero) coupled with an increased average lead time at predicting peaks and troughs in economic activity of 22 months. The end of the 2007-2009 recession was correctly predicted with a lead of 18 months. In other words, this new model predicted an end to the 2007-2009 recession on 12/31/2007 - before most Economists had called a recession in the first place. This is an astounding piece of research.
We all learned in Econ 101 that Monetary Policy has a long and variable lag. Now the Fed has quantified that lag at peaks and troughs - 22 months. I am in the process of reverse engineering their model, but let me give you a hint - there hasn't yet been a call for the next recession. If you have been paying attention to my blog posts, you would know that I am seeing an acceleration in activity that is normally associated with easy monetary policy and I am not surprised by the findings in this newish Fed Research.
I imagine this paper has been circulated to all the high muckity mucks and grand poobahs at the Fed, so why do we continue to see a dovish contingent calling for more Quantitative Easing? I think that the Fed Governors are, in general, no better than you or I would be at this Monetary Policy thingee. Actually, I think the incentives in place make them substantially worse. The Fed just "has to do something" or the whole reason for their being would be nullified. They can't stand still and let nature take its course. They over tighten and over ease to make the memory of their actions closer to the actual turn in the economy. 22 months is a long time to wait out the Economy and the vanity of Central Planning is such that the Fed must be seen to act in concert with the turn in the Economy. Occam's Razor and all that says this is so - although I bet they wish it could be avoided in the case of recessions.
I think the Fed has, once again, stayed too easy for too long. Sooner than most think, I expect the Fed to embark on a long tightening process that will lead us once again into recession - probably sometime 2015.