Sunday, June 2, 2013

Forward Guidance vs. Commitment

He: "Honey, I'm getting tickets for Sunday's football game. Do you want to come?"

Forward guidance.  She: "As things look now, I think I'll feel like coming when Sunday rolls around. Of course that might change. If my mother calls and wants to go shopping I might well feel differently."

Commitment. She: "Sure, honey, that sounds like fun. Get the tickets. I know my mom might call, and I'll regret it later, but we have to get the tickets now, so count me in."

 Commitment means declaring a plan, even a contingent plan, that you will follow, even if you will regret it later. Forward guidance means announcing now what you think you will feel like doing in the future, but not giving up any discretion to change your mind later. Obviously, to someone who has to plunk down money for tickets, commitment is useful.

These issues came up in the last year's fascinating discussions about monetary policy, and brought to the forefront again by Fed Chairman  Ben Bernanke's testimony on May 22, the subsequent question and answers, the FOMC meeting, and market gyrations and controversy surrounding these events.

The words that roiled the markets were, most briefly, "in considering whether a recalibration of the pace of its purchases is warranted, the Committee will continue to assess the degree of progress made toward its objectives in light of incoming information."  Recalibration? Says the  market.

Perceptions matter as much or more than actual statements here (this is the "managing expectations" game). The Wall street journal wrote
Wednesday's flurry of new information jostled markets, which moved up when Mr. Bernanke's congressional testimony was released in the morning, then pared triple-digit gains when he began taking questions and turned negative when the minutes were released in the afternoon....Taken together, the chairman's testimony before the Joint Economic Committee and the minutes suggested that Fed officials aren't yet near consensus on when to begin to wind down the bond buying but that a decision appears to be approaching in the months ahead. ...
"Rather we would be looking beyond that to seeing how the economy evolves and we could either raise or lower our pace of purchases going forward. Again that is dependent on the data," he [Mr. Bernanke] said.
The minutes of the most recent policy meeting said "a number of participants expressed willingness to adjust the flow of purchases downward as early as the June meeting if the economic information received by that time showed evidence of sufficiently strong and sustained growth."

But, the minutes added, "many [officials] indicated that continued progress, more confidence in the outlook, or diminished downside risks would be required before slowing the pace of purchases would become appropriate." 
The Economist wrote (my emphasis)  
One of the main points Mr Bernanke tried to make was that if QE slows, it will "not be an automatic mechanistic process." For example, if it drops from $85 billion to $65 billion, it need not drop to $45 billion at the next meeting. It could stay at $65 billion or if the data worsen, go back to $85 billion. This isn’t that surprising; the Fed always reserves the freedom to respond to the data and hates feeling boxed in by market expectations.
This all scores somewhere between total discretion (we'll do whatever we think is right given the data at the time) to a degree of forward guidance (here is an outline of what we think now we'll feel like doing, but of course we might change our minds.)

Why does this matter? It's an interesting denoument to a big discussion in academia, the Fed, and the broader evolution of central banking doctrine (I hate to say "theory" as it's all pretty loosey-goosey).

The idea was that the Fed can stimulate the economy by committing now to keep policy expansionary for longer than it will want to do ex-post.  I last wrote about this in "managing a liquidity trap." For the previous year, highlighted by a stellar speech by Mike Woodford at Jackson hole (see previous post), this idea was all the rage.

In the standard new-Keynesian model, consumption is low today because its future level is anchored, and a too-high path of real interest rates makes consumption grow too fast. Hence the current level of consumption is too low. That level can be raised by lowering expected future interest rates and hence expected future consumption growth just as effectively as by lowering today's interest rates and today's consumption growth. (If this all seems insane, read here.) So, goes the story, the key to stimulus when interest rates are zero is for the Fed to commit to keeping interest rates low, lower than than we and the Fed know it will want them to be when the time comes. 

 I expressed some doubts that the Fed would ever make such a commitment, or that people would believe it if it tried to do so. (I also expressed some doubts at the whole modeling approach, but that's not important now.)  These events seem to prove that conjecture in spades.

The vast market gyrations, and the Economist's trenchant quote are especially interesting. If the Fed had been committed to a path, or even to a rule (no change until unemployment falls below 6.5%), and most of all if people thought it had such a commitment,  then Mr. Bernanke's answers to questions from congressmen should have no effect.

If only commitment now to do things you will regret later were so easy as it is in our models. This is not a criticism of the Fed. Imagine the Fed chair explaining to Congress that he is keeping rates lower than everybody thinks they should be, with unemployment down to 5% and inflation heating up at 4%, because he made that commitment in order to stimulate the economy back in 2012. Commitments need more than words. It's easy in a model to write "the Fed commits to x," like a new inflation target or interest rate path. It's easy to write opeds that "the Fed should commit to x." Generating such a commitment -- which means, by definition, something that constrains your actions in the future -- is not so easy.

Benn Steil has a nice blog post and more media links.