Winner of the New Statesman SPERI Prize in Political Economy 2016


Saturday 18 January 2014

How the ECB could be radical by being old fashioned

In a recent post I wondered whether, in 20 years time, we might look back on this period with the same bewilderment that we now look back on monetary policy in the early 1930s or 1970s. After the 1929 crash the Federal Reserve was relatively slow to cut interest rates, and Milton Friedman argued the Fed was largely to blame for the subsequent depression. In the 1970s central banks failed to raise interest rates in response to rising inflation. As my previous post was ‘for economists’, let me spell out what central banks may be missing this time.

As we all know, short term nominal interest rates in the US, UK, Eurozone and Japan are as low as they can go, so it appears as if there is nothing more central banks can do with conventional interest rate policy. However that is not the case. What they can do is promise to keep future interest rates lower than they would otherwise be. This policy, first suggested by Paul Krugman for Japan and championed by the highly influential Michael Woodford, I will call ‘forward commitment’. It is not the same as ‘forward guidance’, which central banks are implementing.

The form of forward guidance that the US and UK operate involves giving the public some information about when interest rates might begin to rise. For example, if unemployment falls below some figure, each central bank may think about raising rates, but there is no commitment to do so. I believe the best way to understand this policy is that it is entirely conventional, targeting a combination of expected inflation and the output gap, but that it gives the public a bit more information about the trade off between these two goals.

The forward commitment policy is radically different. It promises to allow both inflation and the output gap to be above target in the future, so as to increase demand today. How does this work? Perhaps the easiest way to think about this is by considering long term interest rates. Long term (say 5 year) interest rates are mainly a combination of expected short rates from now until 5 years ahead. If you promise to have above target inflation tomorrow, the central bank must allow future short term interest rates to be lower tomorrow, which reduces long term rates today. Lower long term interest rates encourage additional consumption and investment today.

I call this the forward commitment policy because the central bank has to make the private sector believe it will carry it out. The problem is that the policy has a built in temptation for the central bank to cheat. They increase demand today by making the promise to allow inflation to be above target tomorrow, but once tomorrow comes they can change their mind - because who likes above target inflation? Yet if the private sector believes they will change their mind, the policy will not work today. So the central bank has to commit to allow inflation to be above target in the future, and get the private sector to believe in that commitment.

There are various ways it could do this. One is to have a target path for the level of nominal income (nominal GDP). The recession reduces nominal income, so nominal income has to grow more rapidly to catch up with its target path. That may well involve inflation rising above 2% for a prolonged period.

You might wonder whether this policy just helps one problem (lack of demand in the recession) by creating another - above target inflation after the recession is over. That is true, but if the recession is deep enough the net result is still positive on balance. However what is also true is that the policy is best implemented at the beginning of the recession. Once the recovery is underway, and the period at which nominal interest rates would normally be stuck at zero decreases, the net benefits of implementing the forward commitment policy decline.

Some people have interpreted forward guidance as forward commitment, because forward guidance in the UK and US allows inflation to go slightly above target if unemployment remains high. I think that is a mistake. Traditional inflation targeting allows inflation to go above target if unemployment is high (as we have seen in the UK and US). The distinctive feature of forward commitment is a promise to allow inflation to be above target when unemployment is low (or equivalently the output gap is positive, rather than negative as it is at present). No central bank has made this promise.

So what we may ask in 20 years time is why central banks did not try this forward commitment policy. In simple toy models, as my previous post showed (where conventional policy is called ‘discretionary’), it can lead to much better outcomes. Is it too late? In the US, with the recovery well under way, I suspect it is. In the UK, where we seem to have got very excited by the economy actually growing again, the same is probably (if regretfully) true. However the story might be very different in the Eurozone.

The Eurozone experienced a real double dip recession. Although positive growth has recently resumed, the OECD still expect the output gap to be -3.5% in 2015: much higher than they forecast for the US or UK. Inflation is currently below 1%, and my colleague Andrea Ferrero argues that there is a real risk of deflation. So the case for a forward commitment policy in the Eurozone remains strong. Furthermore, the ECB feels it cannot implement a Quantitative Easing programme of the type followed by the UK and US, so it may be more inclined to try forward commitment.

But, you may rightly say, isn’t the ECB also notoriously conservative? In particular, German central bankers and their allies would never allow a promise of above target inflation. I suspect this is right, but let me offer a glimmer of hope. Forward commitment could be sold not as a radical new policy, but a return to a very old one: money targeting. The one major central bank that did maintain a money targeting policy for more than a few years was the Bundesbank.

Why is money targeting like a forward commitment policy? Because the level of the money stock is closely related to the level of nominal income. So having a target path for the stock of money is like having a target path for nominal income. And as I suggested above, having a target for nominal income is one way of implementing a forward commitment policy.


So if I was ever in the position of advising the ECB (!) I would sell forward commitment this way. I suspect it would not work, partly because the Bundesbank in practice never rigidly targeted the level of the money supply. However I could reasonably argue that the inflation performance of the German economy in the 1970s was better than in the UK or US partly because expectations were anchored through money supply targeting. It would be worth a try.      

20 comments:

  1. I understand quite well how this policy might work due deviations from rational expectations, though once we get into that and start taking more bounded rationality/perceptions into account it's not clear decisions are that sensitive to interest rates in the current environment.
    But going back to understanding the rational expectations equilibrium in new keynesian theory: if the central bank anounces a lower interest rate forever, doesn't this just amount to anouncing a lower inflation target, with the same real interest rate? Why would it affect real interest rates outside the liquidity trap? I know this forward commitment works out in Werning's model, I know John Cochrane has shown there are other equilibria where behaviour would be closer to the flexible price equilibrium so a promise of lower interest rates in the long run would just get you lower expected inflation. Can you give an intuition in words for why promises of lower interest rates long into the future don't just create lower inflation expectations, unless prices are extremely sticky (again, ignoring all the recent debates about why the rational expectations equilibrium with long run monetary neutrality+Fisher relation may be unstable- though we do seem to observe over a long enough horizons that countries with higher inflation have higher nominal interest rates despite all the accusations in the blogosphere that people saying those things must hold at some horizon don't know macroeconomics)?

    ReplyDelete
    Replies
    1. The intuition is that agents believe the CB when it says its long run inflation target is still 2%. So far we have had five years of zero nominal rates, and I do not see central bankers worrying that the public thinks long run inflation expectations are below target.

      Delete
    2. But by the Fisher relation, the 2% long run inflation expectations may become contradictory with what could end up sounding like a 0-0.25% long run interest rate target, so you could end up with systematic downside violations of the official 2% target and look more like Japan in the 2000's (though in many ways Japan wasn't doing badly at all in 2000-2007 despite the very low inflation and most of its problems then and now are probably structural). Or the central bank could end up successfully convincing the public of a long run shift to negative real interest rates (a case of long run monetary non neutrality which would be somewhat controversial I'd think), but in that case it would also make them significantly more pessimistic in the short run about future economic prospects (since monetary policy risks convincing them that the secular stagnation people like Summers have been concerned about is the de facto baseline forecast of the central bank).
      I still remember the stability of the old style IS/LM under passive monetary policy (in the sense of any mechanisms for converging to the flex price equilibrium) as one of the most troublesome issues. It seems New Keynesian theory has similar unresolved issues. Trying to make decisions in real time rather than academic seminars/publications time, I'm not so surprised that people like Draghi or Carney or the FOMC are wary of implementing fully the prescriptions of people like Woodford and Werning. There's just too much regime/model uncertainty out there.

      Delete
  2. "I believe the best way to understand this policy is that it is entirely conventional, targeting a combination of expected inflation and the output gap, but that it gives the public a bit more information about the trade off between these two goals."

    is there a trade off between output gap and expected inflation in modern macro?

    ReplyDelete
    Replies
    1. The unemployment rate is an indicator of the output gap, which has a relationship with expected inflation.

      If the Central Bank commits to keep interest rates low for an extended period, then agents in the economy will anticipate lower unemployment and higher inflation in the medium term. However, as Wren-Lewis states, this commitment could lack credibility if the agents believe the Central Bank will raise interest rates before they initially indicated.

      If the Central Bank did "cheat" and decide to raise interest rates early, then it would have induced greater demand in the short run while also moving to lower inflation in the medium term, with a lower price level in the long run.

      Delete
  3. Fg is ok if the markets believe it. The rapid rise in yields on us govt debt suggests otherwise

    ReplyDelete
  4. Did Japan have monetary targeting in the 1970s, as they too came through that decade unscathed by stagflation? I say this to be unduly pessimistic.

    This ZLB recession needs a Fosbury flop moment, as you feel that if one country gets traction the others will follow.

    ReplyDelete
  5. Forward guidance is a defective policy for several reasons.
    1. A recent Fed study found that the relationship between interest rates and investment is tenuous.
    2. Demand can perfectly well be boosted NOW and without risking future excess inflation by having government plus central bank print money and spend it into the economy (as advocated by MMTers, Positive Money and others). That money can be withdraw via tax or public spending cuts if need be, if inflation looms.
    3. The optimum rate of interest is the free market rate, unless someone can prove market failure. So we shouldn’t interfere with it.
    4. The basic purpose of the economy is to produce what people want (both as determined by what people do with their disposable income, and the forms of public spending they vote for). Thus come a recession, people should be given more of the stuff that enables them to purchase what they want, i.e. money. And that’s what No.2 above achieves. Plus public spending should probably rise (though doing the latter is a semi-political decision and the pros and cons too complicated to deal with here).

    ReplyDelete
  6. The ECB has an explicit money-target: +4,5% increase of the M-3 stock of money in the medium run (i.e. 5 to 8 years). This is called 'the second pillar' of monetary policy. One might consult the Monthly Bulletin about this. It checks M-3 growth every month. See graph 5 and 6. http://www.ecb.europa.eu/pub/pdf/mobu/mb201401en.pdf

    ReplyDelete
    Replies
    1. Quite. So all you need to do is convert this to a reference path (not growth rates) for nominal GDP.

      Delete
    2. I do not see that this link actually says anything like +4.5% M3 target.

      Where?

      Delete
    3. You'll find it here, p. 85. Also, check out the box on p. 81, which shows how the ECB is failing its own 2006 standards. http://www.ecb.europa.eu/pub/pdf/other/ecbhistoryrolefunctions2006en.pdf?b06761d1bdc5f8356cdaddd57f5c5135

      Delete
  7. "Perhaps the easiest way to think about this is by considering long term interest rates. Long term (say 5 year) interest rates are mainly a combination of expected short rates from now until 5 years ahead. If you promise to have above target inflation tomorrow, the central bank must allow future short term interest rates to be lower tomorrow, which reduces long term rates today. Lower long term interest rates encourage additional consumption and investment today."

    Paging Scott Sumner...the problem with this logic is that an increase in the equilibrium interest rate (the goal of the policy) should force the central bank to raise rates, especially if the bank makes it clear that it won't do any unnatural rate smoothing. So explaining the policy in terms of lowering interest rates is misleading.

    ReplyDelete
  8. Anyone who thinks that the EU problems can be solved by adjusitng the temperature of thin air is in the best case not in touch with reality.

    ReplyDelete
  9. There are two problems with monetary targeting as a strategy to overcome the zero lower bound: 1) the link between money growth and inflation is a long-term link, and so as a commitment policy towards higher future inflation, it is not strong enough to convince consumers to move consumption out of the future into the present: it might take (lost) decades for the inflation to actually show up. 2) Monetary targeting is too hard for the public too understand. Announcing future inflation is the entire point of the commitment: blurring the message by calling it monetary targeting might blur the result you're looking for.

    ReplyDelete
    Replies
    1. I think money targeting is a hopeless policy. The idea is to convince the ECB that something like nominal income targeting, which may otherwise seem radical, is in the spirit of money targeting.

      Delete
  10. Interesting post and I basically agree. Lars Christensen has been arguing for a similar "pragmatic" approach. Prior to ECB Bundesbank kept M3 at some 5-6%. In that sense it is weird that they easily accept eurozone M3 growing in (much) "lower gear" since 2009. But considering all benefits over forward guidance on rates, I still think that even monetary targeting with an actual goal to "manage" income expectations can still hit a Bundesbank "wall". Weber opposed SMPs because he believed it was a form of financing PIGS. Also, isn't there an issue with Maastricht treaty, because, if ECB starts purchasing, as in Lars's proposal, GDP weighted basket of Euro area bonds, with a goal to boost m3/ngdp, Germans will interpret it as "rescuing" PIGS - financing governments. Treaty forbids that, right?
    They could buy private assets...but that is already allocative, and I think g.bonds would be more neutral approach.
    Id like to hear what you make of these issues. Thnx



    ReplyDelete
  11. ECB's M3 index increased by 0.3% over the past 12 months. Deposits from Euro area banks at the ECB are about 4.5% of Euro area GDP. Deposits at the Fed by US financial institutions are around 15% of GDP. Why? Asset purchase programme - $85bn (or now $75bn) a month causes excess reserves that come back to the Fed, which uses them to buy US Treasury and mortgage bonds. Keeps the ball rolling. But those reserves are now a problem. Got to get them down before that wall of money washes over the US. If the maturities on the bonds it holds are comfortably short, no problem. Otherwise, very tricky. ECB can't do an asset purchase programme. Against the rules. Can't keep the ball roilling the way the Fed has. Of course, the ECB could simply print a lot more bank notes, but M1 is already 58% of GDP (in USA about 15%), That may be dangeously high in itself. Monetary targeting may require some re-tooling of the ECB. Another US-Euro difference: M3 - M2 in the Eurone is about 10%; in the US probably much much higher, used to be about 27% before Fed stopped publising M3 data. Difference is that US business has a lot of cash, while EU business is critically short.

    ReplyDelete
  12. The ECB can still buy bonds under the OMT programme but with fiscal strings attached. However, your post raises a more fundamental question about the limits of monetary policy which I think are often overlooked. Apparently Bill Clinton once asked Greenspan to keep interest rates low, to which Greenspan replied 'We don't set interest rates, the market does'. Forward guidance and 'forward committment' (first time I've heard the term) may be different but if forward guidance hasn't worked in the US what hope for the implementation of forward commitment. The Fed hasn't changed its policy rate and there was no mention of winding down QE but long term interest rates ran away from them as the market started getting worried that there was only buyer around. As you say, the Fed is holding a huge amount of inventory purchased at above market rates which it can't offload anymore. The Fed didn't decide to wind down QE, the market decided for it. This is not necessarily a criticism of the Fed. I think they genuinely wanted to reduce unemployment. It's just a remark on the limitations of monetary policy, which are often overlooked in economic discussions.

    ReplyDelete
  13. Simon, as Petar notes above I have made a very similar suggestion.

    I think that the ECB should announce a 10% yearly M3 growth target and keep that until the euro zone output gap is closed. This would be a return the ECB's old monetary pillar, which mysteriously "disappeared" 6-7 years ago.

    See my posts on this suggestion here:

    http://marketmonetarist.com/2013/11/01/end-the-euro-crisis-now-with-a-10-m3-target/

    And here:

    http://marketmonetarist.com/2013/03/29/a-simple-monetary-policy-rule-to-end-the-euro-crisis/

    ReplyDelete

Unfortunately because of spam with embedded links (which then flag up warnings about the whole site on some browsers), I have to personally moderate all comments. As a result, your comment may not appear for some time. In addition, I cannot publish comments with links to websites because it takes too much time to check whether these sites are legitimate.