New Style Central Banking
An Interview with Ricardo Reis.
This is your first time visiting Tinbergen Institute. Did you know anything about TI before coming here to teach the TI Economics Lectures?
I didn’t know very much about TI; I had heard good things from a colleague who had taught at the Institute—that it was a very good experience and they had enjoyed interacting with the students. But beyond that, I knew very little.
How did you find the experience of lecturing in the TI environment?
I enjoyed it very much; it was fun. The students were certainly very eager and interested, and I thought it was overall a good experience.
Moving on to some economics-related questions, the main news of September 17th was the lack of news: the Fed still hasn’t tightened monetary policy. What is your opinion about this decision? When do you think the Fed might raise rates and what would be the conditions for this change?
First of all, I am going to object, just slightly, to your question. There was news, just no change. No change in action is news in economics, right? Relative to expectations, there was certainly a forecast error. But beyond that, I think it is clear from Fed announcements that interest rates will most likely rise in the next three to nine months. That is certainly where the Fed seems to see the economy going, but it is waiting for more evidence to accumulate.
The reason for this is that when one looks at real activity in the US right now, it seems plausible that either there is no output gap or that the output gap is quite small. After all, output is growing at its historical average (in terms of real output growth), and unemployment is now below its 2008 level. So, the economy seems to be operating either at full capacity or close to it. At the same time, though, inflation is well under target— and has been well under target for more than a year, which means that relative to the price level target there’s been an accumulation of undershooting the target for the Fed. The Fed is therefore now in a difficult position. On the one hand, they could use the old accumulated wisdom from the Philips curve that when the economy is operating at close to capacity, then (because we expect that infl ation is going to start picking up) we should raise rates early. (And why early? Because we think that with the lags of monetary policy we only see that in infl ation six, 12 or 18 months from now.) But on the other hand, the fact is that inflation right now is very low. Forecasts of infl ation, whether they come from surveys, or from financial markets, all point to little or no increase in inflation over the next year, or the next two or even three years. So, as a result, this news from infl ation all seems to point to not wanting to raise rates. Certainly, we see no signs of infl ation accelerating as a result of overheating in the economy. I think the challenge is to balance these two. In its last meeting, the Fed seemed to be more persuaded by the latter argument (we just don’t see inflation picking up), but at the same time it was clear that it saw the signs from the real economy that inflation could, perhaps (again, going through old Philips curve intuition), just be around the corner. As a result, the Fed decided to wait and see whether we see inflation pick up. This was the current trade-off, this was the current decision, and I think this is still going to be the trade-off in the next few meetings—barring any unexpected news.
Do you think that a change in the Fed’s policy would have severe consequences for emerging markets?
I am skeptical of that. It is true that this has worried emerging markets. Since forecasting how financial markets will always react is impossible, answering that question with great confidence is not going to be easy.
“Quantitative easing was only possible because we started paying interest on reserves… once you start doing that, much of the decade-long wisdom on monetarism becomes invalid.”
At the same time, it’s hard to see how slight increases in interest rates in the US (especially insofar as they come with no changes in the quantitative easing policy that the Fed has undertaken) could have such a large impact in financial markets in other countries. Again, of course, financial markets can be prone to panics and overreaction, and it is true that many emerging markets are currently in situations of some, let’s say, financial fragility—but I would say it doesn’t strike me as particularly likely that that would happen.
Will interest rates be permanently lower after the Great Recession? What do you think about the secular stagnation hypothesis coined by Paul Krugman and Larry Summers?
I think it’s important to remember the context of secular stagnation. The context was the presidential address of the American Economic Association in 1938, by an economist called Alvin Hansen, who after eight years of the Great Depression (by then, nine years) thought: “Well, maybe we’re just stuck in the Great Depression forever.” First, within two years he was proven to be spectacularly wrong. The economy started growing very quickly in 1940 and onward, and certainly for the next 20 years. So likewise, in 2015, it’s now been six to seven years since the peak of output preceding the recession of 2010—and perhaps unsurprisingly, you have once again some people arguing that we may be in a great stagnation forever. I would just add this note: “Let’s look at history, and realize how good or bad our ability is to forecast far away.” And I think there are certainly many reasons to be skeptical that this will be the case.
At the same time, it is the case that interest rates seem to be remarkably low all over the world. Now, it’s important to note that this is a claim about real, not nominal, interest rates. Given an inflation target, nominal interest rates will have to be consistent with a real interest rate and this given infl ation target, so this may require an average of nominal rates that is somewhat lower, but it does not need to be the case.
The Great Recession was addressed via unconventional monetary policy (quantitative easing mainly), due to interest rate policy hitting the zero lower bound. Do you think this type of monetary policy can fuel financial instability and lead to future bubbles?
That argument has been around for a very long time. I think that there are only a few models and empirical arguments that suggest that there may be some logic to it. There is still very scant empirical
evidence suggesting that low nominal interest rates by themselves are either a necessary or sufficient condition for the appearance of bubbles in financial markets (more generally, for large increases in prices in financial markets). Now, I find that to be a very interesting research hypothesis— and I appreciate students writing theses on this, investigating, looking for it in the data, characterizing the empirics. So, my academic “half” finds this to be a fascinating hypothesis, one we should consider. If I were to put a policymaker hat on, however, I would still remain very skeptical, because the evidence in favor of that channel is, in my view, still very small.
Are we witnessing a situation in which the overnight reverse repurchase agreements (RRPs) of the ECB are fueling a bond bubble? Do we see that yields on sovereign bonds are quite low relative to those justified by fundamentals?
As usual, the concept of bubbles is sometimes so ill-defined that it becomes hard to answer those questions with any precision. Let me make the attempt by first posing a couple of questions: Is it the case that the Fed (as well as many other central banks around the world) has kept nominal interest rates very low—at zero or close to it? The answer to this is ‘yes’. And have they announced a desire to keep them low for a long period of time? Again, a ‘yes’— although perhaps less so for the Fed, but probably for the ECB. On top of that, they have engaged in quantitative easing at different bond maturities, which has resulted in not only low future rates but also low money market bonds— which in turn have contributed to keeping medium- to long-term interest rates quite low… So, would I call that a bubble? Certainly not. That is exactly what monetary policy has been trying to achieve: lower interest rates as much at the short end of the yield curve, as at the medium point and the long end. After all, that was the whole purpose of the policy. Why? They reasoned that by keeping those rates low you would be stimulating the economy by discouraging savings and encouraging investment and consumption right now. Given that this is what they were trying to achieve, it is confusing to then say: “Well, you are achieving it so well that it is a problem.”
Now, you may say that in doing so, they are being overly successful: “their efforts to stimulate the economy have caused the yields to fall even lower than would be warranted, which has resulted in a bubble in financial markets.” That may be possible; it is an interesting academic hypothesis. From a policy perspective, I think there are many reasons to be skeptical about a bubble. After all, we are not observing a bubble in the transmission of monetary policy towards real activity, as refl ected in household purchases (durables) or investment. We do not see any kind of frenetic investment that would suggest, at least, the spillovers of a financial bubble towards real economic decisions that might lead to a crisis in the future. So I certainly don’t see the housing construction bubble that we saw a couple of years ago— no, not at all. If anything, investment by firms has been very low. I am extremely skeptical of that argument, because I just don’t see the real spillovers of that supposed bubble. And so, if you don’t see any real spillovers, then you end up stuck in the attempt to interpret financial markets—trying to say whether the value of bonds is fair or not, or correct or not… We have a decade-long experience in economics that suggests that those attempts at figuring out whether the prices are “right” or “wrong” tend to be doomed to fail.
Continuing to another topic, quantitative easing involves the purchase of financial assets from other financial institutions (in order to increase their price). This type of monetary policy led to a dramatic increase in the size of Central Bank balance sheets. Could you please explain the importance of this development?
I think that it is extremely important, and that we should care about it for many different reasons. Let me spell these out.
First of all, quantitative easing was only possible because we started paying interest on reserves. That is quite a big difference in the monetary transmission mechanism and it’s quite a big difference to how central banks conduct their operations. For instance, once you start paying interest on reserves, the monetary base stops being linked to the price level— and much of the decade-long wisdom on monetarism becomes invalid.
A second reason this is important: the increase in the balance sheet allowed the central bank to drastically increase the supply of safe assets in the economy, through bank reserves. That had quite a dramatic impact in terms of financial markets, in terms of the risk that was taken in those financial markets, and in terms of the ability to use safe assets in financial transactions.
That brings us to a third point: with regard to the asset size, the purchase of different bonds of different maturities suggests that central banks can try to interfere with credit conditions in the economy through more than one vehicle and more than one financial market— and in so doing they significantly enlarge their ability to try to control inflation.
Finally, on top of the opportunities that I’ve just discussed (on both the assets and the liabilities sides), this development is important in relation to the central bank balance sheet. The solvency of the central bank is currently endangered; its very large position— and especially the maturity mismatch between its very short-term liabilities (overnight reserves) and its assets (long-term bonds)— has created a vulnerability to potentially accumulate losses. Those losses can pose solvency considerations and could lead to the central bank losing its independence relative to the fiscal authority. The result could be very interesting from the perspective of the central bank’s ability to keep its mandate.
Therefore, it is very interesting to understand not only what allowed this situation to occur (interest on reserves, the expansion of liabilities and assets), but also what its impact will be on financial markets and on the ability of monetary policy to react to inflation, as well as the dangers and challenges are that it poses to the operations of the central bank.
Do you see any difficulties of interest rate increases due to the large balance sheet of the Fed?
In principle, no— but in practice, yes. The answer is twofold. First, if the CB only held short-termbonds on its asset side, then raising the interest paid on reserves would have zero- or close-to-zero effect. Why? Because while the increase in interest affects the CB’s liabilities, we expect that the interest paid on bonds will always be above that paid on reserves (as bonds are slightly riskier than reserves)— and as a result, the interest raised on assets will increase in tandem with that paid on liabilities, leading to no risk whatsoever for the central bank to record losses. Though, if the central bank buys long-term bonds, the way in which it moves short-term rates will have an impact on the yield curve (that is, on the long-term interest rates), and thus on the price of long-term bonds. Depending on the expected path of short-term rates and depending on how changing the rates affects that path, the central bank can make capital gains or losses on holdings of long-term bonds. And on those there is a risk for solvency, insofar as the bank can realize potentially large losses. So, an old-style central bank that only holds short-term bonds (say, central banks up to ten years ago), runs little risk from raising rates. It’s another story for a new-style central bank, such as the Fed now, which holds 0% assets with maturity below one year and 100% bonds that are more than one year— many of which are more than five years in maturity: for that bank there is indeed a risk.
If the Fed would like to reduce its balance sheet, how and at what pace do you think it would do it?
It could do so incredibly quickly— in a day. It would simply have to go out and sell the bonds it holds on its balance sheet and, by selling the bonds, receive the reserve payment from banks. So it can literally do this in 24 hours and get as much as they want. Now, in terms of what consequences this has for financial markets, I think that the Fed learned a year ago (and other central banks have also learned subsequently) that announcements of this kind can affect liquidity in financial markets, expectations in those markets— which means that a more gradual pace is suggested.
“Paying interest on reserves and using the balance sheet are very useful policy tools that central banks should use in the future to achieve their targets. Will they do so? That, I think, is much less clear.”
It will be important to understand exactly how the microstructure of those markets will be affected by these purchases. In that regard, we may want to be a bit more careful. But there is absolutely no hesitation as to whether a central bank could do it; the issue is the impact such a move would have on financial markets.
Do you see a permanent role for balance sheet policy in central banks?
Yes, I think that paying interest on reserves and using the balance sheet are very useful policy tools that banks should use in the future to achieve their targets, whether they are financial stability or inflation. Will they do so? That, I think, is much less clear. Central bankers are conservative by nature, and seem to want to go back to the good old days when they had a small balance sheet. This large balance sheet offers both opportunities and dangers, and my impression is that bankers will lean towards doing less of it. But at the same time, I think that there are a lot of exciting opportunities. Bottom line: I think that, in principle, balance sheet policy could and should be used.
There have been talks about a fiscal union in the Eurozone, but inter-member transfers could theoretically be carried out via the ECB as well, up to the fiscal limit. Is this desirable?
A fiscal union presupposes issuing bonds such that its members have joint liabilities; that they all answer for each other’s debt— and there have been calls for that. My take: the extent of that integration seems to be an ultimate step— one which raises skepticism in many people for simply political reasons. What is true is that transfers happen already in several other ways— through the use of structural funds, for example, which is something familiar to all Europeans. Those transfers tend to have a long-term goal. A second type of measure that has been proposed more recently is the possible creation of a state joint unemployment insurance scheme, which stipulates that, during a recession, countries that are doing worse will receive more than those that are not. Those instruments address macroeconomic stabilization and are interesting and potentially useful. Finally, keeping in mind the whole debate around the Greek debt forgiveness and the ESM, a third kind of transfer is connected with the extent to which we want to have transfers to deal with sovereign debt crises. Through the central bank’s balance sheet, one can indeed engage in a series of transfers—insofar as the central bank focuses on bonds of certain countries and not of others, or allows for the profit and losses from those bond holdings to be distributed in a certain way across members. Now, is this desirable? That I never said. It is possible? Yes… but if anything, I would alert you to the fact that given the desire of some people in Europe to engage in those transfers, one should be concerned— given that the central bank would be used and manipulated into engaging in those transfers for which there is no political desire. I think it is actually very dangerous because it would harm the legitimacy of the central bank. Transfers should be carried out using fiscal instruments.
What is your view on the main research directions in the field of monetary macroeconomics?
The field is extremely exciting right now because of the interesting questions stirred up by unconventional monetary policy. Ten or 15 years ago, the main question about inflation had to do with whether it is a little more or less persistent. Now we are again encountering the fact that the central bank is having a lot of trouble controlling infl ation; it is missing its target by a large amount. I think that understanding why that is, how we are affecting inflation, and what we are expecting inflation to be in the future— these questions are more exciting than our questions ten or 15 years ago. Second, when it comes to what a central bank does, 15 years ago all we talked about was the extent to which the interest rate should respond (a bit more, or a bit less?) to inflation or the output gap. Now the discussion is whether we should use the balance sheet, worry about bubbles in financial markets (like your question earlier), and be concerned about interest on reserves, the size of the balance sheet, how much it shrank, and the composition
of the balance sheet. In short, these questions seem to be much bigger and more exciting than those asked in the past. I think the field is very exciting right now, because there have been so many changes in central bank policy and because inflation is behaving in such an unexpected way.