Somewhere Between Rocket Science and Quantum Mechanics

Monetary economics is definitely harder to grasp than rocket science, but easier than quantum mechanics. One of the hard parts is breaking free from the notion that interest rate changes are reliable guides to the stance of policy.  And one thing that makes that difficult is that sometimes rate changes they are reliable. When rates rose on rumors of tapering last summer, it was in fact a reliable indicator that policy was expected to tighten, which meant that policy had already tightened by the time you read about it (another concept that is hard to grasp.)

But yesterday monetary policy clearly loosened  (based on the rather obvious stock market rally after the 2:15 announcement), but long rates were almost unchanged.  Yesterday I speculated that the zig zag of 10 year rates reflected the back and forth effects of tapering, forward guidance, and the Fisher/income effects.  The first two shouldn’t even be controversial.  Everyone knows:

1. A taper announcement by itself pushes up rates.
2. A taper announcement is the easiest part of the report to quickly ascertain. Traders with their finger on the “sell button” saw “$75 billion” flash on their retina at 2:15 and hit the button. Rates spiked. Stocks fell. That shouldn’t even be controversial.
3. Then they processed the more complex forward guidance info and rates quickly fell, as you’d expect with forward guidance.

I agree that reading market zig zags is usually a fools game, but I really don’t see how any reasonable person could disagree with this interpretation.  Remember that MMs don’t even have a dog in this fight; we have never claimed that tapering necessarily raises rates, indeed I was initially skeptical until I saw the effect last summer.

But here’s the clincher.  I did all this without having any idea what was going on in the fed funds futures markets.  I wouldn’t even know how to find that data on the internet.  My interpretation implies the rates should have been expected to stay lower for longer, but since 10 year yields finished the day unchanged the subsequent path of rates should have showed a steeper increase (income/Fisher effects).  Today kebko sent me the following comment:

Eurodollar futures started the day with an estimated first rate increase in Nov. 2015 with a slope of 35bp per quarter after that. At the end of the day, the date of the increase had moved back 1 month, and the slope had increased to 37bp per quarter. I think this basically reflects the Fed’s guidance today. They are going to try to keep the rate at zero for longer, which should increase inflationary pressure and lead to faster increases once they do. Rates in the 2-3 year range ended the day down, but rates beyond that ended the day up.

Exactly right.  Several commenters sent me a recent debate where John Taylor argued QE had failed because rates didn’t fall:

JOHN TAYLOR: No, I think — if you think about the purpose of the quantitative easing as stated was to lower long-term interest rates, and if — again, look at QE3. It began just in December of last year, September of last year.

And rates are higher now than they were then. So how you can say it helped? Low interest rates have not been the result.

It would appear Taylor forgot to account for the longer term effects (income and Fisher effects.) Taylor is certainly aware of those effects, but it seems to me that many people tend to overlook their importance in ordinary conversation.

So stocks soared yesterday as ten year yields were unchanged, perhaps because the “unchanged” hid large crosscurrents, which were both expansionary.  Lower yields on the expected future liquidity effect (forward guidance) and then a bounce back on the income and Fisher effects.  Both changes are expansionary.

BTW, Here’s one similarity to quantum mechanics.  A new monetary policy does not take effect until it is observed.  Yes, guidance must be backed up with future actions, but given that Yellen is a dove I see no reason why the markets would have been unusually skeptical about the guidance.  And remember that guidance has important effects even if markets only consider it 40% likely that it will be carried out.  In this case it’s far more than 40%.

Some readers misread my “negative multiplier” comment.  Yesterday was not a negative multiplier example.  It was an example of a market reaction that helps one to better understand how a negative multiplier could occur.  It proves nothing, just an analogy.  Let me explain in a different way.  Suppose I went to the Fed meeting and at the end told the FOMC; “your decision today will drive the Dow up almost 300 points.” I think it’s fair to say they would have been dubious.  They had decided to taper, which normally depresses stocks.  Yes they also slightly adjusted the guidance wording, but I don’t think any reasonable person would disagree with my claim that the rise in stocks was more than the FOMC might have expected for a combined contractionary/expansionary move.

My other claim was that if I am right that they underestimated the effect, it’s because the Fed (like most people) puts more weight on “concrete steppes” than forward guidance.  But they (and most other people) are wrong.  Instead Woodford/Krugman and MMs are right—forward guidance matters more than QE. Where I disagree with Paul Krugman (and perhaps Woodford) is that I believe the likely policy counterfactual in March 2009 if Congress had voted against stimulus was some really serious forward guidance, which would have had an impact that surprised even the Fed.   Maybe even producing a negative multiplier. Yesterday’s market reaction upped my subjective probability of being right about the negative multiplier from about 15% to 25%.  Which means I still think it unlikely, but slightly more plausible than before.