Sweden, Poland and Australia all managed the shock from the outbreak of Great Recession quite well and all three countries recovered relatively fast from the initial shock. That meant that nominal GDP nearly was brought back to the pre-crisis trend in all three countries and as a result financial distress and debt problems were to a large extent avoided.
As I have earlier discussed on my post on Australian monetary policy there is basically three reasons for the success of monetary policy in the three countries (very broadly speaking!):
1) Interest rates were initially high so the central banks of Sweden, Poland and Australia could cut rates without hitting the zero lower bound (Sweden, however, came very close).
2) The demand for the countries’ currencies collapsed in response to the crisis, which effectively led to “automatic” monetary easing. In the case of Sweden the Riksbank even seemed to welcome the collapse of the krona.
3) The central banks in the three countries chose to interpret their inflation targeting mandates in a “flexible” fashion and disregarded any short-term inflationary impact of weaker currencies.
However, recently the story for the three economies have become somewhat less rosy and there has been a visible slowdown in growth in Poland, Sweden and Australia. As a consequence all three central banks are back to cutting interest rates after increasing rates in 2009/10-11 – and paradoxically enough the slowdown in all three countries seems to have been exacerbated by the reluctance of the three central banks to re-start cutting interest rates.
This time around, however, the “rate cutting cycle” has been initiated from a lower “peak” than was the case in 2008 and as a consequence we are once heading for “new lows” on the key policy rates in all three countries. In fact in Australia we are now back to the lowest level of 2009 (3%) and in Sweden the key policy rate is down to 1.25%. So even though rates are higher than the lowest of 2009 (0.25%) in Sweden another major negative shock – for example another escalation of the euro crisis – would effectively push the Swedish key policy rate down to the “zero lower bound” – particularly if the demand for Swedish krona would increase in response to such a shock.
Market Monetarists – like traditional monetarists – of course long have argued that “interest rate targeting” is a terribly bad monetary instrument, but it nonetheless remains the preferred policy instrument of most central banks in the world. Scott Sumner has suggested that central banks instead should use NGDP futures in the conduct of monetary policy and I have in numerous blog posts suggested that central banks in small open economies instead of interest rates could use the currency rate as a policy instrument (not as a target!). See for example my recent post on Singapore’s monetary policy regime.
Bennett McCallum has greatly influenced my thinking on monetary policy and particularly my thinking on using the exchange rate as a policy instrument and I would certainly suggest that policy makers should take a look at especially McCallum’s research on the conduct of monetary policy when interest rates are close to the “zero lower bound”.
In McCallum’s 2005 paper “A Monetary Policy Rule for Automatic Prevention of a Liquidity Trap?” he discusses a new policy rule that could be highly relevant for the central banks in Sweden, Poland and Australia – and for matter a number of other central banks that risk hitting the zero lower bound in the event of a new negative demand shock (and of course for those who have ALREADY hit the zero lower bound as for example the Czech central bank).
What McCallum suggests is basically that central banks should continue to use interest rates as the key policy instruments, but also that the central bank should announce that if interest rates needs to be lowered below zero then it will automatically switch to a Singaporean style regime, where the central bank will communicate monetary easing and tightening by announcing appreciating/depreciating paths for the country’s exchange rate.
McCallum terms this rule the MC rule. The reason McCallum uses this term is obviously the resemblance of his rule to a Monetary Conditions Index, where monetary conditions are expressed as an index of interest rates and the exchange rate. The thinking behind McCallum’s MC rule, however, is very different from a traditional Monetary Conditions index.
McCallum basically express MC in the following way:
(1) MC=(1-Θ)R+Θ(-Δs)
Where R is the central bank’s key policy rate and Δs is the change in the nominal exchange rate over a certain period. A positive (negative) value for Δs means a depreciation (an appreciation) of the country’s currency. Θ is a weight between 0 and 1.
Hence, the monetary policy instrument is expressed as a weighted average of the key policy rate and the change in the nominal exchange.
It is easy to see that if interest rates hits zero (R=0) then monetary policy will only be expressed as changes in the exchange rate MC=Θ(-Δs).
While McCallum formulate the MC as a linear combination of interest rates and the exchange rate we could also formulate it as a digital rule where the central bank switches between using interest rates and exchange rates dependent on the level of interest rates so that when interest rates are at “normal” levels (well above zero) monetary policy will be communicated in terms if interest rates changes, but when we get near zero the central bank will announce that it will switch to communicating in changes in the nominal exchange rate.
It should be noted that the purpose of the rule is not to improve “competitiveness”, but rather to expand the money base via buying foreign currency to achieve a certain nominal target such as an inflation target or an NGDP level target. Therefore we could also formulate the rule for example in terms of commodity prices (that would basically be Irving Fisher’s Compensated dollar standard) or for that matter stock prices (See my earlier post on how to use stock prices as a monetary policy instrument here). That is not really important. The point is that monetary policy is far from impotent. There might be a Zero Lower Bound, but there is no liquidity trap. In the monetary policy debate the two are mistakenly often believed to be the same thing. As McCallum expresses it:
“It would be better, I suggest, to use the term “zero lower bound situation,” rather than “liquidity trap,” since the latter seems to imply a priori that there is no available mechanism for generating monetary policy stimulus”
Implementing a MC rule would be easy, but very effective
So central banks are far from “out of ammunition” when they hit the zero lower bound and as McCallum demonstrates the central bank can just switch to managing the exchange rates when that happens. In the “real world” the central banks could of course announce they will be using a MC style instrument to communicate monetary policy. However, this would mean that central banks would have to change their present operational framework and the experience over the past four years have clearly demonstrated that most central banks around the world have a very hard time changing bad habits even when the consequence of this conservatism is stagnation, deflationary pressures, debt crisis and financial distress.
I would therefore suggest a less radical idea, but nonetheless an idea that essentially would be the same as the MC rule. My suggestion would be that for example the Swedish Riksbank or the Polish central bank (NBP) should continue to communicate monetary policy in terms of changes in the interest rates, but also announce that if interest rates where to drop below for example 1% then the central bank would switch to communicating monetary policy changes in terms of projected changes in the exchange rate in the exact same fashion as the Monetary Authorities are doing it in Singapore.
You might object that in for example in Poland the key policy rate is still way above zero so why worry now? Yes, that is true, but the experience over the last four years shows that when you hit the zero lower bound and there is no pre-prepared operational framework in place then it is much harder to come up with away around the problem. Furthermore, by announcing such a rule the risk that it will have to “kick in” is in fact greatly reduced – as the exchange rate automatically would start to weaken as interest rates get closer to zero.
Imagine for example that the US had had such a rule in place in 2008. As the initial shock hit the Federal Reserve was able to cut rates but as fed funds rates came closer to zero the investors realized that there was an operational (!) limit to the amount of monetary easing the fed could do and the dollar then started to strengthen dramatically. However, had the fed had in place a rule that would have led to an “automatic” switch to a Singapore style policy as interest rates dropped close to zero then the markets would have realized that in advance and there wouldn’t had been any market fears that the Fed would not ease monetary policy further. As a consequence the massive strengthening of the dollar we saw would very likely have been avoided and there would probably never had been a Great Recession.
The problem was not that the fed was not willing to ease monetary policy, but that it operationally was unable to do so initially. Tragically Al Broaddus president of the Richmond Federal Reserve already back in 2003 (See Bob Hetzel’s “Great Recession – Market Failure or Policy Failure?” page 301) had suggested the Federal Reserve should pre-announce what policy instrument(s) should be used in the event that interest rates hit zero. The suggestion tragically was ignored and we now know the consequence of this blunder.
The Swedish Riksbank, the Polish central bank and the Australian Reserve Bank could all avoid repeating the fed’s blunder by already today announcing a MC style. That would lead to an “automatic prevention of the liquidity trap”.
PS it should be noted that this post is not meant as a discussion about what the central bank ultimately should target, but rather about what instruments to use to hit the given target. McCallum in his 2005 paper expresses his MC as a Taylor style rule, but one could obviously also think of a MC rule that is used to implement for example a price level target or even better an NGDP level rule and McCallum obviously is one of the founding father of NGDP targeting (I have earlier called McCallum the grandfather of Market Monetarism).
Matt Nolan
/ December 10, 2012Interesting stuff.
Up to the Asian financial crisis, New Zealand used an MCI – we ended up jumping off it and heading for focusing on interest rates due to issues with targeting a forward looking, unforecastable, exchange rate, such as the view that it creates unnecessary volatility in nominal interest rates. However, at the zero lower bound I think a clear nominal exchange rate policy pinned to a clear forward looking core inflation target makes sense.
What happens in Australia will be interesting, their bank has been relatively willing to respond to slowing domestic spending – consumption growth only became “weak” in the September quarter.
Lars Christensen
/ December 10, 2012Thanks Matt.
I am aware of the RBNZ’s use of a Monetary Conditions Index (MCI) in the 1990s. Bank of Canada has also experimented with the use of an MCI. However, I think we have to be careful in comparing those examples with what McCallum is suggesting.
I should, however, have mentioned that Reserve Bank of Australia at least indirectly uses the exchange rate as a policy instrument. However, at the moment I believe the the RBA is falling behind the curve and is taking too long to ease monetary policy and further the RBA seem to be unprepared for possibly hitting the zero lower bound.
kduda2
/ December 11, 2012Lars,
If I understand Sumner, his NGDPLT approach basically involves the CB buying whatever financial assets it can when NGDP is under target, particularly when interest rates are near the ZLB. Sumner usually talks about buying treasuries and MBS but surely foreign currency is just another asset that the CB could buy to increase the monetary base to drive up AD and NGDP along with it. (More M means more M*V!)
So I guess what I am wondering here is: do you think there is anything special about the CB buying forex as opposed to buying treasuries or MBS or anything else. Because to me, I read what you wrote above and my reaction was, “sure, that’s just one of many roughly-equivalent ways of providing monetary stimulus at the ZLB.”
Thanks,
Kenneth Duda
kjd@duda.org
Lars Christensen
/ December 11, 2012Kenneth, the short answer is no, There is no real difference and the purpose of all of it is to increase NGDP by increasing the money base and for the central bank to signal that it intent to do so until it has achieved its target.
That said for small open economies the exchange rate change seem somewhat more potent and other changes and furthermore, using the exchange rate might be more powerful in terms of signaling. And finally using the exchange rate channel is closer to most central banks “normal” operational procedures than NGDP futures, which might make it easier for central banks to understand.
kduda2
/ December 11, 2012Lars, thank you for the thoughtful response.
If I, as a non-economist, am able to understand using NGDP level futures as a better way to drive monetary policy than the nutty US regime of conducting OMO to push the federal funds rate to some target picked by some nutty Taylor rule to target inflation and employment… well, I just can’t believe the Fed’s staff economists would have any trouble understanding this too. Surely the real barrier to implementing NGDPLT is politics, not comprehension at the Fed.
My personal 18-month journey to understand macroeconomics and monetary policy has led me to the conclusion that NGDPLT would be a very good thing and a huge improvement over the nuttiness that prevails today. Any ideas on how to get the Federal Reserve to adopt NGDPLT would be greatly appreciated.
-Ken
Kenneth Duda
kjd@duda.org
Stefan
/ December 11, 2012Lars,
The big problem in Sweden is not the rule, it is Ingves.
http://www.riksbank.se/Documents/Tal/Ingves/2012/tal_ingves_121204_eng.pdf
Flexible IT plus the foolproof way seems very similar to a digital MC rule. For example, in 2009 Svensson suggested that the Riksbank could set a floor at 13 SEK to 1 EUR if the crisis got worse.
Also, Svensson doesn’t think a significantly negative policy rate would be a problem (see the minutes from 2009).
The policy has now been on the wrong track since early 2011:
http://www.scb.se/Pages/TableAndChart____30571.aspx
The riksbank was very active when the crisis hit. The total balance sheet went up to 350% of GDP in 2008. But after a couple of quarters of nice growth in 2010, they quickly halved it. See the last slide here:
http://people.su.se/~leosven/und/2012/120509%20Handels/LS%20120509%20Handels.pdf
/Stefan
Lars Christensen
/ December 11, 2012Stefan, you are certainly right.
This is Lars E. O. Svensson thinking and I believe that this is exactly what the Riksbank did in 2008/9, but is failing very badly to do right now.
In 2008/9 – as I note about – there was no doubt that the Riksbank clearly told the market that it would like to see a weaker krona to “support” the rate cuts. And the policy worked. Sweden had a massive recovery in 2009/10.
Unfortunately, since then Swedish monetary policy has been far from impressive…
Benjamin Cole
/ December 11, 2012Thoughtful blogging.
I do wonder about zero bound and Japan. It almost seems like an economy steps into quicksand when it hits zero bound. Is there a theoretical reason for this? It also seems like the entire Western world is on a trajectory to hit zero bound, if you look at sovereign yells for the last 10 years This is serious business.
That is why I call for very aggressive QE whenever inflation or interest rates get too low. That also monetizes debt, unburdening taxpayers.
Lars Christensen
/ December 11, 2012thanks Benjamin,
You are very right – it seems like most countries in Western world is headed for the zero lower bound. However, as McCallum (and I) stresses the zero lower bound is not the same as liquidity trap. Central banks are not helpless as the zero lower bound and monetary policy is exactly as potent as 0% interest rates as if the interest rate is at 5%. The problem is the central bank’s OPERATIONAL framework.
If the operational framework is change then “aggressive QE” and monetizing of debt and other “nuclear” (and in my view also somewhat disruptive) options will not be necessary. We can return to “normality” if the central bank forget about “interest rate targeting”. Well, they can actually maintain interest rate targeting as long as they have a clear and announced plan about what to do when they hit the zero lower bound.
I think there is a number of theoretical reasons why we will continue to hit the ZLB. First of all as inflation was brought down during the 1990s inflation expectations was brought down. Furthermore, I believe the negative demographic trends mean that trend growth is lower in many countries around the world. That means that the “natural rate” in many countries might be as low as 2-3% even when the economy is working at full capacity. Therefore even small negative demand shocks will bring interest rates close or below zero. As investors realizes this they will naturally fear deflation and central banks have not explained (or is mentally ready for) how operate monetary policy in a ZLB environment. The way to avoid that is to learn from a country like Singapore – McCallum has suggested.
scbstatistik
/ December 11, 2012Lars,
Do you have any comment on Lars’ view that significantly a negative policy would not be a problem.
From the July meeting 2009:
http://www.riksbank.se/Pagefolders/41730/ppp090701e.pdf
Page 8:
“Deputy Governor Lars E.O. Svensson began the discussion by pointing out that with regard to the lower limit for the repo rate there is a need to differentiate between what is possible and what is desirable from a monetary policy point of view. In the discussion about how far it is possible to lower the interest rate, something of a zero interest rate mystique has arisen which has exaggerated the problems relating to a zero interest rate. The zero interest rate bound does not apply to financial markets; they can handle negative interest rates if necessary. It is the relative prices between financial assets that matter. Interest rates are only one way of expressing relative prices; the price today of kronor tomorrow. If we take a 12-month treasury bill of a nominal value of SEK 100 000, there is no fundamental difference if today it has a price of SEK 99 000, 100 000 or 101 000, which would mean nominal interest rates of approximately 1, 0 or -1 per cent. In other words, there is nothing strange about negative interest rates. Moreover, it is real interest rates and not nominal interest rates that are important when making economic decisions. With an inflation rate of 2 per cent, the real price of 100 000 real kronor in the example above will be SEK 101 000, 102 000 and 103 000, that is the real interest rate will be -1, -2 or -3 per cent. The banks can also use charges to introduce what in reality are negative interest rates on transactional accounts.
The only reason why the concept of a zero interest rate has attracted so much attention is that there are banknotes. Banknotes yield zero interest. If households, companies and investors think that the rate of interest on their accounts is too low they can withdraw cash from these accounts and instead keep large amounts of banknotes in their safes, suitcases and closets. However, taking the handling costs into account, including crime-prevention measures, storage costs and so on, banknotes provide an actual yield that corresponds to a negative interest rate. The lower limit for the interest rate is thus not dependent on the financial markets but is determined by the interest rate at which households, companies and investors would begin to hoard large volumes of cash in the form of banknotes. It is probable that the interest rate could become negative before this happened, but we can not be certain where the lower limit lies. It is, however, possible to draw the conclusion that a policy rate of zero per cent would not necessarily entail any significant problems, especially in the light of experience in Japan, where the policy rate has been 0 and is now 0.1 per cent, in Switzerland, which has a target for the three-month Libor rate of 0.25 per cent and a repo rate of 0.05 per cent, and in the USA, where the federal funds rate is restricted to the interval between 0 and 0.25 per cent.”
At that time he couldn’t convince the others to go down to 0%.
And in February 2010 he got the following question from Wickman-Parak.
http://www.riksbank.se/Upload/Dokument_riksbank/Kat_publicerat/PPP/2010/ppp100224e_1.pdf
Page 11:
“Deputy Governor Barbro Wickman-Parak began by stating that she respects Mr Svensson’s way of reasoning, but that one can also approach the monetary policy decision in a different way. With regard to the zero interest rate mystery, Mr Svensson had on numerous occasions claimed that one could have a negative interest rate even if there was a limit to how far one could go. Ms Wickman-Parak wanted to ask Mr Svensson why he did not advocate an interest rate path with a negative interest rate, to achieve even better target fulfilment?”
Lars’ answer:
“In response to Ms Wickman-Parak’s question of why Mr Svensson did not argue for an even lower repo rate path with a negative repo rate, he replied that such a repo rate path could give an even better outcome than the low repo rate path he is now advocating. But the best should not become the enemy of the good. Mr Svensson considered it sensible to first lower the repo rate path to the level he advocates, and then in the new situation decide at the next monetary policy meeting whether it would be appropriate to make a further adjustment in any particular direction. Then there would also be time to gather more data and information and to decide whether any previously unforeseen problem had arisen in implementing any negative interest rates.”
/Stefan
Lars Christensen
/ December 11, 2012Stefan,
I am not a major fan of negative interest rates. It do certainly create some problems and I would much prefer central banks to use other much better tools for monetary easing such as simply buying governments bonds or using the exchange rate channel. In that regard it should of course be noticed that I in general is no fan of using interest rates (positive or negative) as a policy tool – instead I think interest rates should be market determined.
Lars Christensen
/ December 11, 2012Kenneth (see above),
I am certain that the clever economists at the fed understand the problem so you are absolutely right – the problem is probably not lack of knowledge of monetary policy and theory, but the politics of monetary policy. I have written numerous blog posts on that topic.
See for example here: http://marketmonetarist.com/2012/02/16/what-can-niskanan-teach-us-about-central-bank-bureaucrats/
And here: http://marketmonetarist.com/2012/04/21/central-bank-rituals-and-legitimacy/