Thursday, September 2, 2010

Do not forget that investors adapt to volatilities in money inventory

I finally read the new Hamilton paper, twice. I also looked that the recent Harless post.  Both conduct an experiment in altering the bankers yield curve, Operation Twist. Yeglesias is also playing the game.

All want to run an experiment which consists of the Fed injecting a stimulus which has greater bandwidth than the Yield Curve. The Hamilton paper evaluates the sudden dumping of Federal short term paper on the market and using the proceeds to buy long term Treasuries. Harless proposes the experiment of using printed money to bid up ten year bonds until the rate goes to 1.0. Yglesias wants simultaneous global printing.

The issue is will the yield curve adjust rapidly enough to accommodate the shock, or will we get short term disequilibrium? My claim is that the Yield curve adapts within six months, lets call that fast. The natural rate of interest identifies the bandwidth  of the yield curve.  Will investors respond fast enough to restore the natural bandwidth thus preventing the attempts at instability?

The problem is that the current economic bandwidth is about one year, anything below that is showing yields of near zero.  The actual capability of the investor network to adjust is about six months, twice the bandwidth.  Investors and bankers have idle capacity, they can gear up faster than any other part of the economy.

Hence, Jim Hamilton again points out the best policy operates at the natural frequency limit, the short end.

What surprised me about the Hamilton paper was the small effect they got in the two year after dumping the one year or less.   I have to look more closely, because I would expect the two year to show substantial effect.

No comments: