Interview with Narayana Kocherlakota
Narayana Kocherlakota, Lionel W. McKenzie Professor of Economics, University of Rochester; former President, Federal Reserve Bank of Minneapolis.
Has the experience of the crisis changed your view of the central bank policy toolkit?
Former President Kocherlakota: Yes. I’m going to focus on the U.S. experience—with which I am most familiar—aside from one comment later that relates to Europe.
I would divide the U.S. response to the crisis into two pieces. One is the liquidity interventions that the Federal Reserve undertook, largely under the rubric of Section 13(3) of the Federal Reserve Act, beginning in 2008 and then moving on into the fall of 2008 and 2009. I was not President of FRB Minneapolis yet, so my comments are really those of an economist observing those interventions from the outside.
I think that a number of really innovative things were done in that context. Even before March 2008, if you go back to late 2007, I think the Term Auction Facility (TAF) was really helpful in alleviating some of the stigma associated with use of the discount window. And, as the alphabet soup of interventions unfolded, I worried that they would end up being either ineffective or become permanent parts of the landscape. Using a broad brush, neither concern proved to be true.
I think you could criticize the Fed’s passivity in the March 2008 to September 2008 time frame. But, after September 2008, most academic economists with whom I’ve talked are quite approving of the Fed’s liquidity responses.
However, I think the communication surrounding these interventions was poor at the time. It was hard for the public and for Congress to understand what we were doing and why. And the legislative response to this poor communication has been to impose a number of restrictions on the use of these liquidity interventions.
Beyond that there were other interventions from other parts of the government that were very effective. These included the FDIC’s Temporary Liquidity Guarantee Program (TLGP) and the Treasury’s backing of money market mutual funds. As far as I understand, both of these have been taken off the table by Congress for use in future crises. As for single-firm interventions, as a member of the Fed, I was glad to get out of the business interventions like those undertaken with Bear Stearns and AIG. These create all sorts of complications for the institution.
That said, many people look back at the crisis and say, "Boy, why didn't the Fed do something about Lehman?" Well that would have been a single-firm intervention. And who’s going to do that next time? Is Congress really going to be the one to step up? If I were to draw the analogy of a fire department putting out fires, many Members of Congress have had the reaction, "If we could just have fewer firefighters on the streets, we won’t have any fires because people will be more careful in their homes." I think that is very short-sighted thinking about what happened in 2008. On the margin it is true: if you eliminate the fire department, people will take better care of their homes. But, I think we can all agree that that’s not the right approach in dealing with the systemic risk posed by fire.
Moving from the initial liquidity interventions to the monetary policy part where I was a policy maker, a host of tools were used, including large-scale asset purchases (LSAPs), or QE (quantitative easing) as it is sometimes called. I was initially skeptical about the effectiveness of LSAPs. My original thinking, based on lessons from the work of Neil Wallace, Mike Woodford, and others, was that swapping bank reserves for longer-term government securities really should not have much of an effect on asset prices. It is now clear from the data and the evidence that the Fed’s purchases did have an effect on longer-term yields.
But, I don’t think we know enough about how these interventions translate into real activity. As economists, we should do more research on how that reduction in long-term yields translates into real output. One countervailing view comes from the work of Arvind Krishnamurthy and Annette Vissing-Jorgensen, who argue that buying 10-year Treasuries takes a valuable liquidity instrument out of the hands of the private sector, so that while that will drive down the yield on the instrument, it actually has an adverse effect on output.
LSAPS do have limitations as a monetary policy tool. The big problem is that it always seemed that the Fed had some cap in mind. Now, that cap, and the relative comfort with large holdings of assets, evolved over time. But, I think, markets always got the message: “OK, the Fed is doing this, but at the end of the day there is some cap on what they are going to do.” The only time when that might not have been true was for a brief period between September 2012 and April-May 2013 when people talked about QE infinity. But as long as there’s a point beyond which you are not willing to go, QE has its limitations as a policy tool.
The other tool we used was forward guidance. Here, I have somewhat different lessons from others. I think that the use of quantitative, state-contingent, guidance was broadly effective and useful. The challenges were on the communication front. With the benefit of hindsight, we were a bit oblique in the way we communicated. Saying things like "We're going to keep the federal fund rate extraordinarily low at least until the unemployment rate gets to 6.5%" created in the minds of many observers the impression that once we reached 6.5%, the Fed would raise rates. We didn't actually say that. But we should have made clear exactly what we meant when the unemployment rate gets below 6.5%.
This is a long-winded way of saying that economists and lawyers, who are by and large the people around the FOMC table, aren’t necessarily the best group to figure out how to communicate effectively using this tool. There needs to be input from communications experts and others on how we can do a better job of shaping expectations using forward guidance.
Date-contingent guidance has been a challenge for the Fed that led to all sorts of problems. Even in 2015, a number of Federal Open Market Committee participants offered date-based guidance, saying in effect that 2015 was the year for hiking interest rates. They did not mean that as a commitment, but markets took it that way. So, I think the Fed was a little bit boxed in at the December 2015 meeting (where I did not take part).
That’s just the latest example of how date-contingent guidance is never meant as a commitment—it’s always a forecast. It’s meant to be Delphic rather than Odyssean, to use President Evans's language. But, the FOMC ends up in a box because conditions change and then the Committee is stuck with what it said. I remain a big fan of quantitative state-contingent forward guidance. But I believe that more qualitative, vague or date-contingent guidance works poorly and that the Fed should stay away from it going forward.
Another important aspect of the tool kit regards negative interest rates. In the past, I have written down economic models where the interest rate is constrained to be greater than or equal to zero. However, we have certainly learned from the experience in Europe that it’s not. And, I think that is something that should be explored by other major central banks, namely the Bank of England, the Fed, and the Bank of Japan. There are conversations about raising inflation targets to say 4%. But, if you can set the interest rate at -150 (“minus 150”) basis points, you've achieved much of the same result without creating the distortion that higher inflation brings. So, there is quite a bit to be gained from thinking about how low interest rates can go.
One final thing to say is that these unconventional tools would be more effective, the more they are treated as conventional. Treating them as “break the glass” instruments of policy robs them of their effectiveness because every time you bring them out, you will be signalling that things are going very badly. You want to treat your entire policy tool kit as conventional. For example, on negative interest rates, it should be made clear that there is no zero lower bound, and you should communicate ahead of time how low you are willing to go.
Where should we be looking for financial stability risks?
Former President Kocherlakota: This is a very interesting question. It really depends on who "we" is.
As a monetary policymaker, I worried that we are looking a little too broadly for financial stability risks. For example, we would worry about the impact that raising rates would have on corporate bondholders. But, even the perception of that can damage the public view of the institution. You want to be careful to always frame your decision-making in terms of the effects of policy on unemployment and inflation.
Once you do that, you've got to be looking for financial stability risks that affect large swathes of the population. Examples include housing in the 2005 period and stocks in the 1920's. Those are cases where, with the benefit of hindsight, you could see that broad ranges of the population were exposed to risks.
What you do about that is unclear. If you get the timing right, you raise rates at the right time and then lower them at the right time. But, that’s very hard. It could be that you're thinking there's a risk forming so you raise rates, and that ends up increasing the risk. That is very challenging.
I think people use the term financial stability in a lot of different ways. I have been struck by Gary Gorton’s work on liquidity transformation and how it is prone to multiple equilibria. So, I would be looking on a broad level for signs of instability in mechanisms for liquidity transformation. What you do about it, especially in the U.S. context, is very unclear.
More broadly, there has been talk about financial stability in the United States. But, frankly, Congress appears skeptical of having a group of technocrats interfering in markets to get them as stable as possible. And, without that political support, it is going to be very hard to drive things forward.
What do we need to do to preserve the benefits of global finance?
Former President Kocherlakota: This goes beyond finance, as there is quite a bit of skepticism about globalization in general. To many people, the answer is that there is a need for regulation of global finance. But that would require a global infrastructure for regulation, or international coordination at the minimum. Many Americans feel uncomfortable with this. For example, you hear members of Congress questioning the extent to which rules coming out of Basel should apply to the United States. To counter this, the Fed and other U.S. regulatory authorities are quick to say that we make our own rules.
The problem is that, as soon as you write down a model with global finance, it is good to have global coordination; if not just one single financial regulatory infrastructure. So, the benefits of global finance are being weighed against the cost inherent in the kind of coordination that appears to be needed. We are seeing a drumbeat of support for various kinds of capital controls. The hot money problems that affect emerging markets are really severe. It’s all well and good to say those countries should have better institutions than they do, but they don't. So, how should we deal with that?
I think the idea that open, cross-national markets are positive has been taken for granted by many, those of us who have worked on monetary policy, or economics more generally, for a long time. But it is not taken for granted by the public. And, when you talk to people, about how to resolve a failing institution, for example, the question of whether costs of global finance exceed the benefits comes up immediately.
So, I think we need more research on this. (I say this knowing that as an academic, my incentives are always to call for more research.) Here, we need to be a lot more careful about what we mean by global finance. And, whatever we do mean, the less need we have for global regulatory infrastructure the better, because that will get us more support in our home countries.