Recently there has been a debate about whether low interest rates counterintuitively actually leads deflation. Narayana Kocherlakota, President of the Minneapolis Fed, made such an argument a couple of years ago (but seems to have changed his mind now) and it seems like BIS’ Claudio Borio has been making an similar argument recently. Maybe surprisingly to some (market) monetarists will make a similar argument. We will just turn the argument upside down a bit. Let me explain. 

Most people would of course say that low interest rates equals easy monetary policy and that that leads to higher inflation – and not deflation. However, this traditional keynesian (not New Keynesian) view is wrong because it confuses “the” interest rate for the central bank’s policy instrument, while the interest rate actually in the current setting for most inflation targeting central banks is an intermediate target.

The crucial difference between instruments, intermediate instruments and policy goals

To understand the problem at hand I think it is useful to remind my readers of the difference between monetary policy instruments, intermediate targets and policy goals.

The central bank really only has one instrument and that is the control of the amount of base money in the economy (now and in the future). The central bank has full control of this.

On the other hand interest rates or bond yields are not under the direct control of the central bank. Rather they are intermediate targets. So when a central bank says it is it is cutting or hiking interest rates it is not really doing that. It is intervening in the money market (or for that matter in the bond market) to change market pricing. But it is doing so by controlling the money base. This is why interest rates is an intermediate target. The idea is that by changing the money base the central bank can push interest rates up or down and there by influencing the aggregate demand to increase or decrease inflation is that is what the central bank ultimately wants to “hit”.

Similarly Milton Friedman’s suggestion for central banks to target money supply growth is an intermediate target. The central bank does not directly control M2 or M3, but only the money base.

So while interest rates (or bond yields) and the money supply are not money policy instruments they are intermediate targets. Something central banks “targets” to hit an the ultimate target of monetary policy. What we could call this the policy goal. This could be for example inflation or nominal GDP.

When you say interest rates will be low – you tell the markets you plan to fail

Why is this discussion important? Well, it is important because because when central banks are confused about these concepts they also fail to send the right signals about the monetary policy stance.

Milton Friedman of course famously told us that when interest rates are low it is normally because monetary policy has been tight. This of course is nothing else than what Irving Fisher long ago taught us – that there is a crucial difference between real and nominal interest rates. When inflation expectations increase nominal interest rates will increase – leaving real interest rates unchanged.

The graph below pretty well illustrates this relationship.

The correlation is pretty obvious – there is a positive (rather than a negative) correlation between on the one hand interest rates/bond yields and on the other hand inflation (PCE core). This of course says absolutely nothing about causality, but it seems to pretty clearly show that Friedman and Fisher were right – interest rates/yields are high (low) when inflation is high (low).

This is of course does not mean that we can increase inflation by hiking interest rates. This is exactly because the central bank does not directly control interest rates and yields. Arguably the central bank can of course (in some circumstances) in the short decrease rates and yields through a liquidity effect. For example by buying bonds the central banks can in the short-term push up the price of bonds and hence push down yields. However, if the policy is continued in a committed fashion it should lead to higher inflation expectations – this will push up rates and yields.  This is exactly what the graph above shows. Central bankers might suffer from money illusion, but you can’t fool everybody all of the time and investors, consumers and labourers will demand compensation for any increase in inflation.

This also illustrates that it might very well be counterproductive for central bankers to communicate about monetary in terms of interest rates or yields. Because when central bankers in recent years have said that they want to keep rates ‘low for longer’ or will do quantitative easing to push down bond yields they are effectively saying that they will ensure lower inflation or even deflation. Yes, that is correct central bankers have effectively been saying that they want to fail.

Said, in another if the central bank communicates as if the interest rate is it’s policy goal then when it says that it will ensure low interest rates then market expectation will adjust to reflect that. Therefore, market participants should expect low inflation or deflation. This will lead to an increase in money demand (lower velocity) and this will obviously on its own be deflationary. This is why “low for longer” if formulated as a policy goal could actually lead to deflation.

Obviously this is not really what central bankers want. But they are sending confusing signals then they talk about keeping rates and yields low and at the same time want to “stimulate” aggregate demand. As consequence “low for longer”-communication is actually undermining the commitment to spurring aggregate demand and “fighting” deflation.

Forget about rates and yields – communicates in terms of the ultimate target/goal 

Therefore, central bankers should stop communicating about monetary policy in terms of interest rates or bond yields. Instead central bankers should only communicate in terms of what they ultimately want to achieve – whether that is an inflation target or a NGDP target. In fact the word “target” might be a misnomer. Maybe it is actually better to talk about the goal of monetary policy.

Lets take an example. The Federal Reserve wants to hit a given NGDP level goal. It therefore should announces the following:

“To ensure our goal of achieve 5% nominal GDP growth (level targeting) we will in the future adjust the money base in such a fashion to alway aiming at hitting our policy goal. There will be no limits to increases or decreases in the money base. We will also do whatever is necessary to hit this goal.”

And lets say Fed boss Janet Yellen is asked by a journalist about the interest rates and bond yields. Yellen should reply the following:

“Interest rate and bond yields are market prices in the same way as the exchange rate or property prices. The Fed is not targeting either and it is not our policy instrument. Our policy goal is the level of nominal GDP and we use changes in the money base – this is our policy instrument – to ensure this policy goal. We expect interest rates and yield to adjust in such a fashion to reflect our monetary policy. My only advice to investors is to expect us to alway hit our policy goal.”

Said in another way interest rates and yields are endogenous. They reflect market expectations for inflation and growth. So when the Fed and other central banks in giving “forward guidance” in terms of interest rates they are seriously missing the point about forward guidance. The only forward guidance needed is what policy goal the central bank has and an “all-in” commitment to hit that policy goal by adjusting the money base.

Finally, notice that I am NOT arguing that the Fed or any other central bank should hike interest rates to fight deflation – that would be complete nonsense. I am arguing to totally stop communicating about rates and yields as it totally mess up central bank communication.

PS Scott Sumner and Tim Duy have similar discussions in recent blog posts.

PPS Mike Belongia has been helpful in shaping my view on these matters.