Widely credited with being the seminal paper at the 2012 Jackson Hole conference and setting the scene for "threshold-based" policy, Michael Woodford discusses his views on the costs and benefits of "forward guidance" in this Goldman Sachs interview. The Columbia professor explains how he thinks about asset purchases versus forward guidance (it’s a mistake to think of asset purchases as a way to avoid having to talk about future policy intentions), and why the market and the Fed have seemed so disconnected at various points this year despite substantial attempts by the Fed to communicate more clearly (there were mistakes in communication, but that does not mean the situation would have been better if the Fed had instead kept its mouth shut, especially in such unprecedented times.) Ultimatley he warns, "by blinking when they did, I fear that they have made a negative reaction more likely in the future, because they are now back to square one, with people once again lacking a clear sense of how close the Fed is to tapering and thus vulnerable to surprise."
Goldman Sachs' Allison Nathan Interview with Michael Woodford,
Allison Nathan: Haven’t central banks always tried to influence interest rate expectations? What is so special about forward guidance today?
Michael Woodford: No, central banks have not tried to do things that were at all similar to this in the past. Until quite recently, all central banks were very reluctant to say things in advance about future policy decisions, and this reluctance remains to varying degrees at many banks. The forward guidance adopted by the Fed and other central banks, which tries to influence expectations by actually saying things about policy intentions, is therefore a new policy tool and indeed one that has become more important given the near-exhaustion of the most traditional policy tool – adjusting policy rates – as rates across the major economies already hover around their effective lower bound.
Allison Nathan: What are the benefits of forward guidance?
Michael Woodford: If policy expectations matter – and I think it is pretty clear that they are crucial to how longer-term assets end up getting priced – then there are two kinds of advantages of explicitly discussing future policy by the central bank. One advantage is that it can reduce misunderstandings about policy intentions, which, in turn, can reduce uncertainty for the central bank about the effect of its policy on the markets. In principle, talking directly about policy intentions would allow the use of more complex policies, which might not otherwise be pursued for fear that they would not be understood without explanation. The second general type of gain from explicitly talking about future policy is to help ensure that the policy committee itself will follow through with its commitments even though it may have motives to depart from them later on.
Both of these potential advantages are particularly clear when you reach an effective lower bound on policy rates. At that point, convincing people that the policy rate will remain “lower for longer” can help ease financial conditions today, providing additional stimulus to the economy when traditional tools no longer can. But talking about the intention of “lower for longer” is crucial because being at this lower bound is a very unusual situation, so there is little past experience that people can look to in order to anticipate how the central bank is going to respond. There is also a clear need for the central bank to commit itself in advance in order to achieve the stimulative benefit. That is because of course later – when the stimulus has worked and the economy is improving - the bank will have little motivation to actually keep rates low (the so-called “time inconsistency” problem) unless they committed to do so in advance. To overcome that problem, the central bank needs to make an explicit promise that would be difficult or embarrassing to just completely ignore later.
Allison Nathan: What are the dangers of forward guidance?
Michael Woodford: The most obvious danger, which has likely been the main reason for central banks’ reluctance to talk about future policy in the past, is the possibility that a policy commitment that looks sensible at some earlier time turns out to be unwise because things happen in the meantime that the central bank did not expect. Those costs can be reduced without losing all of the potential benefits of forward guidance if the central banks think carefully about what kind of commitments about future policy should be made. It makes sense to avoid unnecessary specificity about things that do not need to be specified too precisely in order to achieve the desired change in expectations. For example, in the case of a commitment to keep the federal funds rate low for longer in order to stimulate the economy today, the central bank could make a very specific commitment about the path of the policy rate over time. But there would be much more likelihood of embarrassment in that case than if the bank instead committed to keep rates low until certain economic conditions arise, whenever that may be.
Allison Nathan: The BOE and the ECB have said that the intention of their shift to forward guidance has been to clarify their policies rather than to commit to “lower for longer”. Will this approach negate the benefits of the guidance?
Michael Woodford: Yes, to some degree. In the case of the Bank of England, the structure of their statement – with several so-called “knock-out” provisions - as well as their insistence that the statement was nothing more than a clarification of the BOE’s normal reaction function, has given people little reason to change their prior beliefs about how soon the Bank would raise rates. Because of this, the statement does not seem to have moved market expectations much and in the way that the BOE thought it should. Similarly, the ECB has taken small steps towards doing something that you might think of as forward guidance, but has also done so quite hesitantly; they are also inclined to deny that they are committing themselves at all about future policy. Given the aversion to talking about policy intentions in the past, this hesitation is not surprising, nor is the fact that even central banks that have decided that they should experiment with the policy do it in a way that simultaneously denies that they would ever do it, because it goes against their instincts. But that to some extent defeats the purpose of the policy. Their approach is quite different from that of the Fed, which has more clearly embraced the policy of “lower for longer”.
Allison Nathan: Is the Fed’s shift to outcome-based or “threshold” guidance from “calendar” guidance a good thing?
Michael Woodford: Yes, because threshold guidance is ultimately more credible. The problem with calendar-based guidance is that if there is a real promise to keep rates at a certain level until a certain point in time no matter what happens, it would be a pretty reckless policy. And because the policy would be reckless, it would ultimately be hard to believe. That would be the case unless the central bank restricted itself to a short horizon over which there could not be many surprises. But if the horizon is too short, the impact on future expectations would be small. So I think the possibility of making a commitment that extends far enough into the future for it to be news about future policy that would significantly matter to asset pricing is much more plausible if it is based on economic conditions rather than just based on a date.
Allison Nathan: There have been several instances when the use of forward guidance has had an opposite impact than the central banks intended – why?
Michael Woodford: The use of forward guidance is not some kind of magical tool where the mere fact that the central bank says something means that people will then think exactly that. A central bank needs to give people a reason to think something new or different about what it is going to do. A critical part of effective policy is therefore understanding what people will think they are learning about the bank’s policy. An example of this that I talked about in my Jackson Hole paper last year was the experience of the Swedish Riksbank in April 2009, when they cut their policy rate to 50 basis points and accompanied this with a statement and a published projected rate path that showed policy rates remaining at 50 basis points - the lowest level ever - until the beginning of 2011. To the Bank’s surprise, market forward rate expectations rose rather than fell following the announcement. Why? Because the big “news” of the statement was not the central bank’s lower projected rate path, which was in any case just a projection and not a commitment, but that the central bank was apparently regarding 50 basis points as a floor, which was higher than at least some market participants had previously guessed. That news shifted the markets’ most likely expected path of the future policy rate up rather than down.
Allison Nathan: How would you explain the violence of the bond market selloff in May/June, which came in response to a very small change in the Fed's message?
Michael Woodford: I am inclined to think that it indicated some mistakes in Fed communication prior to May that led to two possible types of misinterpretation about the Fed’s intentions. First, some people may have interpreted the start of “taper talk” as a signal that the Fed was trying to withdraw accommodation more broadly, and was also preparing to start raising interest rates. That was a surprise to the FOMC; they didn't think they were saying anything that would suggest they were preparing to raise rates. But they had left themselves open to that misinterpretation by failing to explain earlier the criteria that would determine the path of asset purchases in a way that sounded very different from the criteria that would determine the path of interest rates. The forward guidance about both asset purchases and interest rates focused on labor market conditions and sounded very closely related. I do not think that the Fed meant for the criteria to be the same, but they failed to sufficiently explain why they would not be. Second, there may have been a number of people who thought the purchases were going to continue at the current rate for a lot longer, and learned suddenly that they would not. If that was news to people, it was again a failure of communication because I doubt that the Fed had ever thought asset purchases would continue at the current rate beyond 2013. Despite these failures, it would be a mistake to conclude that the Fed should not have started speaking about tapering when it did; the problems would not have been avoided if the Fed had just kept its mouth shut, because the misinterpretations would still be there. And shutting your mouth is potentially setting you up for an even harder adjustment later when the misinterpretations must eventually be exposed.
Allison Nathan: Why was the market so surprised by the decision not to taper at the September FOMC?
Michael Woodford: It certainly seemed to me that during the summer the ground was being prepared for a slowing of the rate of purchases. As to why they did not actually do it, I think it was a reaction to the fact that the market had responded to those earlier hints more violently than expected. And there was evidently a decision that they could not risk a further unexpected negative reaction to an actual announcement of tapering. That was probably a mistake in judgment. By September, a modest reduction in purchases was widely expected, so I do not think there would have been a big negative reaction to that announcement. But, by blinking when they did, I fear that they have made a negative reaction more likely in the future, because they are now back to square one, with people once again lacking a clear sense of how close the Fed is to tapering and thus vulnerable to surprise.
Allison Nathan: Is there actually greater volatility and uncertainty in the markets as a function of this desire to communicate more, but not quite getting it right?
Michael Woodford: I don't think so, because the question is: what would be people's understanding of policy if the Fed had not tried to talk about it at all? There would be a lot of uncertainty if the Fed were adopting a policy of silence, especially given that we are in unprecedented territory. When conditions are unusual is exactly the time when trying to provide some explicit guidance is potentially most valuable, even if it is not a panacea.
Allison Nathan: How important are asset purchases as a signal of commitment to accommodative policy?
Michael Woodford: I think that a lot of the effects of asset purchases have been signaling effects. The advantage of purchases as a signal is that it is something that people see being done. It's not just talk, so it grabs people’s attention. And the fact that action is being taken gives some indication of where the majority of the FOMC stands. But the likelihood that purchases have had some signaling effect does not necessarily mean that they are the most effective way of providing the signals that the central bank wants to send. There has at times been a temptation to view asset purchases and forward guidance as two alternative means to providing further stimulus, so that we can avoid having to say more about future policy if instead we are acting to make additional asset purchases, and I think that is a mistake. To the extent that the main goal of purchases is to give a signal, then you should think consciously about what signal you are trying to give and be comfortable delivering that signal. Thinking about asset purchases as part of a coherent and consistent attempt to give signals about future policy is one thing. But it's very different from the idea that there will be a mechanical effect of purchases that allows you to avoid saying anything about future policy intentions.
Allison Nathan: What’s next for Fed communication?
Michael Woodford: It would be valuable for the Fed to provide more guidance about the process of policy normalization. When it is clear that they will begin slowing the rate of asset purchases -- which I think will have to be fairly soon, although not necessarily this year given that they did not do it in September -- the next obvious question will be how quickly the rest of the unusually easy policies will be unwound, and what the broader 'exit strategy' will look like. The last time they spoke about that was in 2011 and it’s pretty obvious that what they said then is no longer an operative strategy. They will need to say something about that at least by the time that they start tapering, because at that point it will be very clear that we are no longer in a period of just staying the course. But a likely reason not to make big statements right now is of course the imminent hand-off of the Chairmanship in January.