When forward guidance fails: the Fisher equation and the Swedish paradox

On 3 July the Swedish central bank, Riksbanken, cut its key policy rate by 50bp to 0.25%. Most analysts – and the markets – were taken by surprise by this decision. It was particularly surprising as Riksbanken’s governor Stefan Ingves had been voted down by a majority of Riksbanken’s board.

Most people – including myself – would say that when a central bank cuts it key policy rate more than expected, it is monetary easing, and it seemed that was how the market was interpreting Riksbanken’s move – the Swedish krona weakened significantly and Swedish share prices spiked. However, something was not as it should be – Swedish inflation expectations dropped (!) on the back of the rate decision, e.g. Swedish 2-year breakeven inflation dropped from around 0.85% before the rate decision to around 0.65% after the rate decision. This is a paradox – a Swedish paradox: when you cut rates you get lower inflation expectations. So judging from the inflation expectations Riksbanken had actually tightened monetary conditions rather than eased.

BE inflation expectations Sweden

The Fisher equation and focusing on the wrong target

So what went wrong? The answer in my view is that Riksbanken is focusing on the wrong policy target. Hence, the bank communicates in terms of interest rates rather than inflation expectations. And yes, the interest rates are an intermediate target.

Riksbanken controls the Swedish money base and it can use this to control money market rates – in the short term. However, the so-called Tinbergen rule also tells us that a central bank can only hit one target if it has one instrument.

Therefore, if Riksbanken targets interest rates it cannot at the same time effectively target inflation (expectations). Unless it uses an additional “instrument”, such as credibility. If the market believes that Riksbanken will always adjust monetary parameters to ensure that it hits its 2% inflation target, it will be able to move the money market rate (temporarily) away from the ‘natural’ interest rates.

Hence, if Riksbanken’s inflation target is fully credible, inflation expectations will basically be pegged at 2%. However, if the inflation target is not credible, the story is very different, and as inflation expectations are presently well below 2%, it is very clear that the 2% inflation target is presently not credible and has not been credible for years.

A way to illustrate this is to have a look at the so-called Fisher equation:

(1) i = r + pe

i is the nominal interest rate, r is the real interest rate and pe is inflation expectations. When we talk about money market rates we can also see i as the policy rate.

It follows logically from (1) that if the inflation target is fully credible – that is, if pe is ‘fixed’ – a cut in i will ‘automatically’ lower r. On the other hand, if inflation expectations are not well-anchored, a cut in i might as well reduce pe.

I believe this is exactly what happened in Sweden on the back of Riksbanken’s surprise cut.

Not only is Riksbanken communicating in terms of interest rates (rather than inflation expectations) but it is also communicating in terms of the interest rate path. Hence, Riksbanken is not only announcing rate decisions but it is also communicating about future expected changes in the policy rate.

In that regard it is important that Riksbanken actually lowered its expectations for interest rates in two years even more than it lowered its present key policy rate. In other words, Riksbanken flattened the money market rate curve. So for a given real interest rate Riksbanken is actually indirectly telling the market that it expects inflation expectations to decline even further in the coming two years.

Obviously this is not what Riksbanken meant to say (I hope) but when it chooses to focus on interest rates rather than inflation expectations, this is what the market will focus on as well. Riksbanken’s interest rate focus therefore ‘overruled’ the focus on inflation expectations. In fact, in Riksbanken’s statement there was no reference to the market’s inflation expectations.

Lesson: central banks should focus on the ultimate policy target rather than the intermediate one 

I think the lesson we can learn from thisis that central banks should not focus on intermediate targets – such as interest rates and the interest rate path – but should focus on the ultimate policy goal – in the case of Riksbanken expected inflation.

Imagine that Riksbanken had issued the following statement last week:

‘Inflation expectations are presently well below Riksbanken’s 2% inflation target. This is unsatisfactory and as a consequence the repo rate is now being cut by 0.5 percentage points to 0.25% and Riksbanken is fully committed to introducing further monetary easing if needed to ensure that market expectations will fully reflect its 2% inflation target. If needed the repo rate will be cut further and Riksbanken will actively intervene in the currency markets to ease monetary conditions through the FX channel until inflation expectations are at 2%’.

I think it is pretty clear that such a statement would have caused an immediate jump in (market) inflation expectations to 2%. This would obviously also have caused a significant drop in real interest rates – both as a result of the lower nominal rates AND, more importantly, through higher inflation expectations.

What a difference a few words make…

PS Riksbanken is not alone in terms of these problems. The ECB faces a similar problems, while the Fed and the Bank of Japan are focusing on the ultimate policy goal rather than on intermediate targets. However, during Operation Twist in 2011-12 the Fed was facing Riksbanken-style problems.

PPS In Swedish CPI there is an explicit (mortgage) interest rate component, weighing around 5% of total CPI (and hence, of course, incl when calculating breakeven inflation), implying that shorter breakeven inflation should indeed come down by some 0.3 p.p. if a full pass-through into mortgage rates from the cut. That, however, does not really change the point. The Riksbank is targeting CPI so it is really irrelevant why inflation is too low.


Related blog posts:

Committed to a failing strategy: low for longer = deflation for longer?
Riksbanken moves close to the ZLB – Now it is time to give Bennett McCallum a call
A scary story: The Zero Lower Bound and exchange rate dynamics


Leave a comment


  1. SvenR

     /  July 9, 2014

    Isn’t it actually possible that a lower interest rate leads to lower inflation because it really is an easing of monetary policy that leads to more output which in turn might lead to lower inflation?

    Or another channel how a monetary easing can lead to less inflation is by the fact that an easing (with its improvement of the unemployment situtation) gives less policy room for bad politics that harms the supply side, which again will lead to lower inflation.

  2. Tom Brown

     /  July 10, 2014


    This low-parameter-count model of the price level (P), based on this theory of price, predicts that for some economies the quantity theory of money (QTM) applies, while for others it doesn’t. Some economies will be somewhere in between those states. Here’s how the empirical data from multiple economies over several decades matches the theory (the theoretical curves are shown plotted on the right in blue to give a sense for the theoretical distribution that’s possible over a set of random economies, and the multi-colored empirical data is plotted on the left and was fit with a single parameter per country (M00)). The theory also makes specific testable predictions, and the author says he’s not much interested in patching his theory up with “epicycles” should those predictions fail. Where else are you going to hear something like that on an econ blog? Lol.

    So to answer your question, if this theory is true, then yes, what you say is possible (but not necessarily for the reasons you give). The theory makes no statement about individual human “agents,” their decisions, motivations, rationality, or their expectations, etc. It maintains a “maximum ignorance” stance regarding those issues. (Not to say those issues aren’t important for other reasons, BTW). But it does rely on there being lots of humans (an ensemble) doing all manner of possible things in an economy.

    I’m very interested to see if this new theory will continue to successfully make predictions and explain the observed data.

  3. Benjamin Cole

     /  July 10, 2014

    Print more money. If that does not work, then print even more. A feeble or timid central bank will convince no one of anything.

  1. Rezoniranje iz kamatnih stopa | Tko je John Galt?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: