If we analyse monetary policy as a threat strategy, then how do we make sure that the threat is credible? According to Nick Rowe, “The Fed needs to communicate its target clearly. And it needs to threaten to do unlimited amounts of QE for an unlimited amount of time until its target is hit. If that threat is communicated clearly, and believed, the actual amount of QE needed will be negative.” In essence, this is a view that a credible threat will cause market expectations to adjust and negate the need for any actual intervention in markets by the central bank.

The current poster-child for this view is the SNB’s maintenance of a floor on the EURCHF exchange rate at 1.20. The market-expectations story argues that because the SNB has credibly committed to maintaining a floor on EURCHF, it will not need to intervene in the markets at all (See Evan Soltas here and here, Scott Sumner, Matthew Yglesias and Timothy Lee). And indeed the SNB did not need to intervene at all…..that is, until May, when they were required to intervene to the tune of CHF 66 billion within the span of just a month in order to defend the floor from euro-crisis induced safe-haven flows.

Even when the central bank wants to hit a target as transparent and as liquidly traded as an exchange rate, it seems that actual intervention is needed sooner or later. Therefore when the transmission channel between central bank purchase of assets and the target variable is as blurred as it would be in regimes such as NGDP targeting, it is unlikely that the central bank will get away with just waving the magic wand of market expectations.

Comments

Nick Rowe

Ashwin: whether you will actually need to carry out your (conditional threat depends on two things: 1. How credible your threat is, and whether you need to start doing it to make it credible. 2. The model. What exactly it is you are targeting, and how that thing depends on everything else. For monetary policy and NGDP, theory and experience tells us that central banks balance sheets are abnormally large right now, because the demand for base money is abnormally high when expected future NGDP growth is abnormally low and risky. So if the threat were credible, and expected NGDP growth increased, and became less uncertain, I'm pretty sure the demand for base money would fall, and banks would have to do reverse QE. That (actual being the reverse of what's threatened) won't always be true, for all target variables, for all possible shocks. E.g.: if I'm leading a bull by the nose, I will need to actually turn in the same direction I threaten to turn, even if my threat is credible and the bull knows it. But if I'm driving a cow from behind, I will need to actually turn in the opposite direction to what I threaten to turn. It depends. Off-topic: BTW, this concept of macroeconomic "resilience" looks interesting and important. I had a perusal of your blog a month or so back. Not fully sure I'm getting it yet. Canoes have "primary stability" and "secondary stability". Primary is when they don't tip much for small shocks. Secondary is when if you lean a lot the canoe firms right up and the gunwhale doesn't want to go underwater. In big waves, you want a canoe with secondary stability, even though it feels "tippy" in calm water. Related?

Ashwin

Nick - in a world with interest on reserves and no hot potato effect where the channel linking the threatened actions (such as asset purchases) and the target is so fuzzy, how can you be so certain that there won't even be a significant initial period where the threat needs to be carried out? I completely buy the argument that there may be a later reversal but even this is just unimportant when we're paying interest on reserves - the quantity of base in the system has no importance. On resilience, I know almost nothing about canoes but it sounds similar. The ecological concept of the stability-resilience tradeoff says that in many systems stability is inconsistent with resilience. So if you attempt to engineer a system with no disturbance you are more susceptible to collapse when the eventual unanticipated disturbance does hit. So your concept of secondary stability may be analogous to resilience. I've written lots of posts on the topic but probably the best places to start are my comparisons of macro-stabilisation to - forest fire suppression https://www.macroresilience.com/2011/06/08/forest-fire-suppression-and-macroeconomic-stabilisation/ - river flood prevention https://www.macroresilience.com/2010/10/18/the-resilience-stability-tradeoff-drawing-analogies-between-river-management-and-macroeconomic-management/ and psychotropic medication https://www.macroresilience.com/2011/12/14/the-pathology-of-stabilisation-in-complex-adaptive-systems/

srini

In the case of the SNB, there is a clear "mechanism" by which the "threat" can be enforced by the SNB. The SNB can print as many swiss francs as it wants and satisfy unlimited demand for swiss francs at the price ceiling (for the franc). What is the equivalent mechanism for NGDP? All I have seen from nick and others is hand waving, Chuck Norris, bazooka, guns, and no clear mechanism. I hear expectations--but modern economics has no settled theory of expectation formation. If you are going to hang your theory on the expectations fairy, you need to come up with a theory of how the expectations fairy really works. How do you make a credible promise? Sargent confidently said that a credible commitment to low inflation would not be costly. We learned otherwise during the Voclker era. Of course, monetarists can always claim that the commitment was not credible. If so, the question would be, why would they suddenly be credible now. Assuming that expectations can be toggled at will is begging the question.

K

Just because the SNB isn't buying a lot of Euros, that doesn't mean others aren't. At the current exchange rate, there's huge demand for Francs that's being filled by arbitrageurs who are betting that the SNB will succeed in its target. As those balance sheets fill up with Euros, the risk to them of a target failure will eventually become so great that they will stop buying. Then the SNB will have to step in to meet the demand or the target will fail. Period. And it doesn't get better when the SNB starts buying in serious quantities. Everyone knows the disaster the SNB will be facing if they end up with a balance sheet full of Euros just as Swiss inflation starts to rise, in which case the target will fail and the losses could be big. Independence of the CB is not consistent with big losses, and the SNB which prizes its independence and its record on inflation, is therefore weak against attacks. Which is why buying Francs could be a good bet. All central bank targets may require actual significant actions and all targets are vulnerable. I like Steve Roth's point that there's no such thing as turtles all the way down.

Ashwin

Srini - My point is just that even in such a transparent case the CB eventually has to put its money where its mouth is. My last post was an attempt to draw out the problems with the expectations-focused view which you criticise. K - Part of the problem is that every target is imperfectly credible, not least, as you note, because of the political pressures faced by the CB to avoid losses.

srini

Ashwin, Completely agree with you about the SNB case. Just to jog people's memory about another memorable case of managing expectations failing miserably, recall: “If you have a bazooka in your pocket and people know it, you probably won't have to use it.''—U.S. Treasury Secretary Hank Paulson, July 15, 2008 describing a proposed backstop authority for Fannie Mae and Freddie Mac at a Senate Banking Committee hearing Presently, the common perception of the events of 2008 is that Secretary Paulson waved a bazooka and, voila, order was restored. In reality, although Treasury Secretary Paulson wanted Congress to give him a bazooka big enough that he might not have to use it, it turned out that he did have to use his. Moreover, from the time Fannie and Freddie were placed into conservatorship on September 6, 2008 to the stock markets’ bottom in March 2009, it was a long six months and close to 50% decline in the S&P. I see that Nick has another post about "concrete steppes." Whenever he writes that, I know that he is back to faith-based arguments.

Ashwin

Srini - Yes, the Paulson-Freddie/Fannie example is an excellent one. And that quote is priceless!

SNB’s Swiss Franc Dilemma: A Solution at Macroeconomic Resilience

[...] I highlighted in my previous post, the honeymoon period for the SNB in its enforcement of the 1.20 floor on the EURCHF exchange rate [...]

Asymptosis » Turtles All the Way Down: Are Fed Promises “Actions”? How About Promises to Make Promises? Game Theory Edition

[...] makes The Case Against Monetary Stimulus Via Asset Purchases, and Nick Rowe responds. Ashwin replies, analyzing “monetary policy as a threat [...]