Conference marking the 20th Anniversary of the CNB and the Independent Czech Currency

(Stanley Fischer, Governor of the Bank of Israel)

Governor Singer, former Governor Tůma, members of the board of the Czech National Bank, ambassadors, among them I see the Ambassador of Israel, Mr. Yaakov Levy, fellow governors from other countries and ladies and gentlemen. I am sorry I wasn't here for the speech of the President Klaus but I would like to tell a little story about what happened to me in 1988 when I was Chief Economist of the World Bank. Somebody asked for a meeting and I asked around and I said, well you should see him. So in came a young man, I was also a young man in those days, and started telling me what was going to happen in Czechoslovakia and what the plans were for running the new economy and how the transition would be made. This was before the Velvet Revolution and so I said when is this going to happen. Very soon, which kind of, I didn't know that much but it wasn't what most people at the time thought. The plan sounded logical but I thought we were talking about something which would happen ten, fifteen years later but sure enough, it happened as advertised and the young man who walked into my office was now President Klaus and who became finance minister at the start of the transition. And some people have vision, one can say, and I will not forget the fact that here, in Russia, in other countries, people prepared for events which other people thought were very unlikely, and did it successfully.

I want to talk about recent developments in monetary policy but before doing that I’d like to say a word about the Czech central bank, Czech National Bank. This central bank has established itself in a very short time as a really outstanding central bank, a highly professional central bank. I had a long enough trip this morning from Israel to be able to read the most recent inflation report and the most recent financial stability report and there's no question just on the basis of those documents of the quality of the research that is done here and there's been no question over the years since I've had the privilege of following the work of the Czech National Bank with Governor Tošovský and Governor Tůma and now Governor Singer of the extraordinary high professional level at which a new central bank has been able to work. And as we look back, one should not also forget the description we've just had of the separation of a currency block, which was done almost flawlessly, at least it seemed from the outside, which is also another remarkable event.

So thank you very much for inviting me to this occasion and I will talk about recent developments in monetary policy. When the crisis, which still hasn't really ended in large parts of the world, particularly in Europe, began with the fall of Lehman Brothers in September 2008, many said the economic world would never be the same and that meant especially our world of monetary policy-making and of financial sector supervision. Well, there were two changes which were evident very quickly. One was the rise of the G20, relative to the G7 or the G8, and very soon after by November 2008 already the G20 had become the primary, we used to call the G7 in the IMF, the executive committee of the IMF, although in fact it had no official standing whatever, but when you had a big problem there would be consultation with the G7, it's now more likely to be with the G20. It was a trend that was beginning already in the late 1990s but it certainly accelerated. And secondly, the creation of the Financial Stability Board to replace the Financial Stability Forum which had been a talking show-up set up in about 1997 or early 98 was a very important step, as a country that doesn't belong to the Financial Stability Board, and I don't think the Czech Republic does either, basically because it consists of mostly members of the G20, I didn’t really appreciate the creation of this group, I thought it should have worked through the IMF, where all the countries it was proposing to operate on were members. But it's developed very well, it's done remarkably good work and we do have a new body which is gradually becoming a global body to try to coordinate supervision. It's not based on any treaty, it's not based on any formal agreement except its relationship with the G20. In that regard it sounds like a very British institution but I used to think it was run by the Governor of the Central Bank of Canada, turns out it's going to be run by the Governor of the Bank of England so it's becoming apparently a British type of institution; not a whole lot of constitutional validity but operating nonetheless.

Well, for several years after the crisis started it seemed that the two main changes in the lives of central banks and those who run them would be first, and I think here of also how the textbooks will change, how to deal with the situation in which the interest rate hits its zero lower bound. In textbooks of the late 80s, late 70s, late 80s, 90s I can tell you from private knowledge that the five or six pages on the liquidity trap which were in the 1979 edition of Dornbusch and Fischer gradually became two pages because everybody said this is not relevant to anything. Then the Japanese got into trouble and had to operate with a zero interest rate, so it went from two pages to four pages. Now, we're going to have a whole long section on how to deal with a zero interest rate and that will relate to quantitative easing and many variations thereof. The other major development was the invention of macroprudential supervision. As a result of the fact that allowing one medium size bank to fail, namely Lehman Brothers, we got ourselves into a – within two days – into a global financial crisis. People began to realize that you have to not look at each bank separately but understand how the financial system operates and look for system-wide effects of regulation and make sure that the system as a whole can operate rather than that each bank is managed well and regulated well.

Now there's an enormous amount going on with macroprudential supervision, the amount of paper it has generated is massive, the regulations come in every day. I pity my members of the bank supervision department for how much reading they do, I pity the bankers even more for how much reading they have to do as the regulations get translated into Hebrew or whatever language you use and amended for local conditions, and there's a whole lot going on. We need to recognize one thing about changes and regulations. We won't know until the next crisis whether these were the right changes to make or not. If I can take an example, in 1997 the British government decided to set up the FSA, the Financial Stability Authority, which concentrated all the regulators and it was, it's a lovely model on paper. And it got a lot of support, because it was the UK, the Bank of England is the oldest central bank, excuse me, members of the Riksbank who might be here, but it's the bank with the most developed history and most relevant to what goes on now. But there were some people who said just wait until there's a crisis before you decide that this is the way to go, and there was a crisis and for the first time in a hundred and fourty years the British suffered from bank runs which hadn't happened since the late 19th century. So you can do stress tests on paper and make them work out, but a good crisis is the best stress test. Unfortunately, you don't want to generate a crisis just to see, but it means that we'll have to be very careful, use all our logic and realize that we don't really know whether we've solved or whether we will have built a system which is much less liable to financial crisis than we had before.

Well, these were the two changes and I think that's what we said or what we would have said, or what textbooks all say until recently. But very recently, we've had announcements of two changes in monetary policy that are actually different and could be more significant over the longer run as we think about how we thought about monetary policy just four or five years ago. Those two changes are the Fed's announcement of a quantitative target for a real variable, namely the unemployment rate, and secondly, the Swiss National Bank's decision to set a lower bound for the exchange rate and to be willing to defend it with apparently unlimited intervention.

So, I will speak mostly about what the Fed has done with its announcement a few months ago of the quantitative target for unemployment as part of the definition of how long it will continue these policies. The Fed has announced that it will buy 85 billion dollars a month of both mortgage backed paper, I think that's 40 billion, and government paper, 45 billion. It comes to about a trillion dollars a year, which is large on a balance sheet of about three trillion, which is large relative to the 800 billion it was at the start of this crisis. So, we have a very, very expansionary policy put in place and the announcement is that it will continue in place until the unemployment rate comes down to 6.5% or until subject to the qualification, that it will stay in place, that if the expected inflation rate for one to two years ahead rises above 2.5%, they'll stop this policy. Now, we central bankers, and I'm sure including the Czech National Bank, have been very reluctant to define real targets of policy and the reason is a nice... something that sounds right, which is: We know that in the long run monetary policy can affect the inflation, right? We have only one instrument, the interest rate, and therefore we can have only one target and therefore it must be the inflation rate. Well, that's a very nice story. I won't take up the issue of how many instruments we have, but if you look at what central banks have been doing lately, you will consider that they have many instruments. In some sense similar, they are all related to the financial market, but in no sense absolutely identical.

That's not the argument I want to take up. I want to take up the argument of: We have only one instrument, therefore we can have only one target. That's the theorem in algebra which says in general if you have one to solve for two variables you need two equations, or three variables you need three equations, subject to a variety of conditions. And it's the so called Tinbergen rule in work that Tinbergen did soon after World War II, in which he argued that if you wanted to hit targets you needed as many instruments as you had targets. Well, that's true if you have to hit them exactly. But if you have a utility function, and that's what we learn in economics, you can put any number of variables in there and maximize it subject to constraints. That's what they teach us in micro; that's when you start learning what it is to be an economist because economist is a guy, or a woman, who talks about trade-offs. You can have less inflation and less output, or more inflation and for the short term more output, and so forth, and so central banks are not in practice choosing only the inflation rate at any minute, what we central bankers all of us have are models of short term dynamics, it's fashionable in most central banks to have DSGE models, but in every DSGE model that you use, as you calculate the expected inflation rate, the model gives you, just it's there, an associated part of output. And then you have to ask: Why do we always talk about the inflation path we expect and not very much about the unemployment and output path we expect? They come out at the same time and they're part of the same problem and they both react to monetary policy. In fact, if you look at the theoretical framework in which monetary policy has been studied and many of you have seen the book of Michael Woodford, many of you have read many papers by Lars Svensson and by one of my colleagues on our monetary policy committee Alex Cukierman. Their models do not justify the distinction between inflation and unemployment as targets of monetary policy because they actually have a utility function that includes both inflation and output in those models and they have a short-run trade-off between inflation and unemployment, which is the Phillips curve. We should just accept the fact that when we make a decision, we're choosing a time path and not a level of a variable. We're setting an interest rate to determine the time paths of inflation and output and that is the effect of choice that central banks make and all central banks take into account, the trade-off between inflation and output in the short run even though there is not such a trade-off in the long run. How do you know that? Well, it sort of came to my mind for the first time when I read about about how the Bundesbank was really the greatest of the monetary targeters in the 1980s. But you looked at the Bundesbank what it was doing and when you saw that it departed from its monetary target but they were facing a recession, they didn't immediately say: oops, we've got to cut the growth rate of money this morning because we've got to get back the money situation, or in terms of the approach which is now used, flexible inflation targeting, which we all know about. When inflation rate goes above target, you do not say I'm going to raise the interest rate by five percentage points so that as soon as possible we get back to our inflation target, you say: no, we'll take it gradually. Namely, we're choosing that path. Only what we do is we keep that path, the output path, for ourselves to decide and we present the inflation path. In fact, they're both relevant and we take both of them into account and we've solved the awkwardness of the fact that all our choices are determining paths of both variables that are important to society by inventing flexible inflation targeting.

Now, it sounds like I'm arguing that we should go back to the 1960s because then everything was about short-term trade-offs and Phillips curves that looked like that and you chose a point on the Phillips curve, but it's not quite because in choosing a path for the future you also have to take into account where that path is going and you have to take account of the fact that in the long run you cannot control the quantity variables through monetary policy, and I'll say to a first approximation because it's not strictly true.

So now we come to the Fed. What the Fed has said is, this new policy which they announced two or three months ago is just a technical change. It's technical because we always knew about this trade-off and we made our decisions on that basis and all we've done now is we put a number on the unemployment rate. Well, that's a very good story and technically it's true. But I think that it's going to have a major impact on what we do and, what is more important, what the politicians will demand from us in future. We now have a major central bank, the major central bank, announcing an output or unemployment target and the inflation target. Now, they have got in it what is the most important innovation about inflation targeting, which is the override, the lexical-graphic ordering that says we will run an expansionary policy until the inflation rate or unless the inflation rate hits its upper level. So it still says what inflation targeting says, namely there's an ordering, at least in our law and in many laws of central banks you can do other things as long as inflation is within the target range in our case, so below the inflation target, in other, depending on how the model is set up. And they still have that feature there, so they can indeed say, it's purely technical, we’ve just put numbers on it. The number they've put on the inflation, expected inflation, is actually above the inflation target but slightly, the inflation target is 2%, they've put it at 2.5%.

So is this technical? Yes, it's technical. Is it revolutionary? We won't know, but I think it is going to make a difference to our lives. I think that, we hope there are no politicians present, so that we can keep the secret to ourselves. At this stage in Israel, they say we're in this closed room, with five hundred people in the closed room, within this closed room we'd better recognize that we do have a responsibility, at least in the short run, for what happens to the unemployment rate or to the output rate and that we need to take that into account and that it may turn out that flexibility which we have given ourselves via flexible inflation targeting will not in the end stay with us. It may come back to us from the government side and then the question is: what should we do? Should we fight it like crazy and say it interferes with central bank independence, or what? Well, my view, but it's probably not a majority view among my colleagues, is: facts are facts and we'd better deal with the world as it is and not make up arguments like we have only one instrument, therefore we can have only one goal, which are not accurate in fact and not valid, to promote a policy viewpoint that we believe in and which is more useful, which is more easy for us. I think we have to deal with the world as it is, the world is difficult, the world is complicated, we wish it weren't and we all wish there was a simple rule we could all follow, money growth at 4%, you can go home every evening without worrying and watch television instead of trying to figure out what's going to happen to you tomorrow morning in the markets, but I don't think that's in fact the world in which we may be living sometime in the future. Now what are the dangers in this announcing both, or at least putting both those parts out, the dangers are that you may be told to do things, you may get targets set for you by the political, by those who set the monetary targets, and I know there are some central banks who set their own target inflation rates. In our country and in many others the government sets the target inflation rate. We may find ourselves given targets which are inconsistent. That's why the only feature of inflation targeting which I think is absolutely essential is the upper bound on the inflation rate. What was inflation targeting designed to deal with? It's a result of the problems of the late 1960 in the United States when we kept going down the Phillips curve and the inflation rate kept going up, but then it went up much faster than estimated Phillips curve had said, and of the developments of the 1970s which ended in the United States, partly due to supply shocks, with inflation rates in double digits: 13%, 14% early in 1980 which it took Paul Volcker to kill by setting the short-term interest rate very close to 20%, which is kind of hard to believe. We don't even remember it now but the United States did get to that position. The upper bound on the inflation rate prevents the very plausible argument that it's never worth another percent of inflation, to prevent unemployment rising by X it's never worth another one percent of inflation, you should just let the inflation go up by one percent, if you get to four, you say: well, we can have more output if we go to five, you get to five, we can have more output if we go to six and so forth, and that's in practice what happens, even though it sounds that it's not a mistake central banks would make, they did make that mistake. Fortunately, I was only a professor then so I can make all those accusations.

Now I want to go on to a related issue on monetary policy and that..., well sorry, I will leave for the fortieth anniversary of the Czech National Bank the discussion of whether the huge balance sheets that central banks now have or are going to make big problems somewhere down the road, I won't go into that. But the short answer I have is: no, central banks dealt with this crisis by buying lots of assets. If the situation starts getting out of hand and prices start going up they have lots of assets to sell and they can reverse these actions. Now people will say: it is going to be very complicated. Well, it is, it was very complicated in the other direction and central banks are pretty sophisticated.

Now, I want to just turn to two other things. Why not nominal GDP targeting? Mark Carney, the next Governor of the Bank of England, has said on a few occasions that he sees benefits in nominal GDP targeting and there's one huge benefit in nominal GDP targeting and that is that it gives you a trade-off automatically between supply side shocks and inflation, namely the trade-off is one for one between output and inflation in the nominal GDP targeting and on paper it has that advantage. It usually comes with an associated story, which is the story that with nominal GDP targeting you're targeting a path of GDP and past mistakes are not forgiven because you try to go back to the same path. But that is not an essential part of nominal GDP targeting. You could target an increase of three percent of nominal GDP per year or five percent or whatever you wanted without having to say we'll always go back to the original path just as you can also target price levels instead of inflation rates and get a different outcome. I think we shouldn't expect to do nominal GDP targeting, because at least in every country I know nominal GDP data are terrible. They come out and they're revised massively all the time, and what are you supposed to do about that. Oops, sorry. First we correct for the data correction and then we go on and make other changes, we put a lot of variation into this activity. I was relieved to see reading my Blackberry this morning that Mark has just given a speech which says what he'll do when he gets in is not nominal GDP targeting but he's going to try to do more quantitative easing, at least at the beginning.

Now the second major revolution that is very interesting now is the Swiss National Bank announcing that it is willing to buy infinite amounts of foreign exchange, of Euros in their case, in order to prevent the exchange rate from becoming more appreciated than 1.20 which is a number they fixed the exchange having declined to 1 at some point. Well, we all say you can't fight the market, that's a good story. You can fight the market, if the market wants your currency to appreciate, because if your currency is appreciating what the market wants is your currency. Namely, I'll talk in shekels because that's what I thought about. People want to come in to the country and buy the domestic assets, buy domestic currency and so you can produce shekels, we can produce shekels in very large amounts. We sterilize them, we sterilize the shekels by immediately selling short-term paper. But we have that, you can keep doing it forever. It will be very expensive when you've issued all this paper to neutralize it, but you can do it. And the Swiss, having tried everything else, got themselves into a position when initially the country was mad with them for losing money by having intervened in the foreign exchange market, and then was mad at them because the exchange rate kept appreciating when they didn't intervene, and so they decided then they would be able to intervene and they came up with this new policy. Now, they bought a lot of currency. They bought within a period of less than a year about 50% of GDP, that's a lot. I think it's the fastest, you know, when I was a boy, in the non-metric system there used to be how fast does a car accelerate from zero to 60 and there was always miles per hour, 0 to 100 kilometers an hour and there was always a car which had the record. Well, I think the Swiss National Bank has the record for the rate of increase as a share of GDP of its assets. Now, isn't this a big problem, isn't it losing lots of money, isn't the carrying cost much greater than the benefit? Well, the answer is the carrying cost is negative, Swiss interest rates are lower than interest rates in the EMU so they're not losing on the carry. Are they going to have a loss when the market eases up? Well, they'll need the Swiss franc to depreciate against the Euro for them to make money on the price. They believe that 1.20 is not the equilibrium rate, they believe the equilibrium rate is well above that and at least in the last few days the market has been working in the direction of their making a profit.

So this was a pretty amazing step. As you all know, the Swiss are extremely conservative, well this was not extremely conservative, this was extremely not conservative and so far, it's working. I'd like to add a footnote to the story. Several of the governors sitting here remember, listening in the BIS to the former governor of the Central Bank of Brazil Henrique Meirelles, who used to say, on paper, on my balance sheet, on the market basis, I always have a loss on foreign exchange reserves, whenever I intervene I make a profit. What is the meaning of that? When does he intervene? When the currency starts depreciating too much. And that's when you go out and you buy and that's the point at which you make a profit. So he said that the Central Bank of Brazil was showing losses but making profits. Now, I didn't check that but it's certainly logically valid, so that's another point you could take into account in thinking about this.

So I think what we've seen in the period of the last six months is two developments in monetary policy which may have an impact for a long time on central banks. I think the Fed’s giving numbers to both the inflation and the unemployment targets, albeit they are not permanent targets, is the more important of those changes and that over the years to come we may well see impacts of this in the way central banks’ tasks are defined and in the way the policy makers, the non-monetary policy makers, congresses, parliaments and governments relate to the central banks. So central banking is never static. Just when you think you found the perfect solution – inflation targeting – something comes along and tells you: well, it isn't quite perfect. And I have seen enough changes in the approach to monetary policy to basically give up on the hope that one day we're going to find the Holy Grail and then we can all relax because everybody will understand what to do, because there's always an element not only of the political versus the central bank game, there's also the element of the market versus the central bank, markets versus the central bank, that makes this a very very dynamic business and one that needs a quality central bank, and there's a very high quality central bank in Prague and I thank you all for this opportunity. Thank you.