Breaking Conventional Wisdom on Permanently Lower Interest Rates

Summary

• A new Fed working paper attributes the long-term decline in the natural rate of interest to errors in the beliefs of central banks and the private sector – a brand new theory.
• They refer to this effect as the ‘hall of mirrors’, where each party incorrectly infers information from the signals they are giving.
• This effect has reduced the effectiveness of monetary policy and led to a lower inflation and interest rate environment.
• To overcome the problem, the Fed must communicate its long-run trajectory of interest rates far better.

Introduction

One of the leading hypotheses for the persistently low interest rates in many developed countries over the last decade is secular stagnation. It stipulates that factors such as falling trend productivity growth, population ageing, higher demand for safe assets and rising inequality have driven r-star persistently lower. But could there be another explanation?

Yes, according to a new Fed working paper. The answer lies in beliefs and informational feedback. Specifically, a lack of communication between central banks and the private sector over the long-term trajectory of the real interest rate has compounded errors in the estimate of r-star. According to the authors, this is why interest rates and inflation remained permanently suppressed post-GFC, and it can explain the reduced effectiveness of expansionary monetary policy since.

What Is R-Star?

The natural rate of interest, or r-star, is the real interest rate that keeps economic growth at its long-term trend and maintains stable inflation. Its role in economics is as a theoretical construct – we cannot observe it, and we can only estimate it by taking a view on the ‘correct’ model of an economy. Nevertheless, important relationships exist between it and the actual policy rate.

For example, r-star gives the Fed an indication of whether its monetary stance is accommodative, neutral or restrictive. If the policy rate is above r-star, monetary policy is restrictive. This means economic slack will tend to increase – economic growth will tend to slow, the unemployment rate should tend to rise, and the rate of inflation should be more muted or decline.

We can also think of r-star as the interest rate that balances the supply of savings and demand for investment in the economy. Like how prices adjust to the supply and demand of goods, the interest rate will fall if there is an excess supply. Greater aggregate savings are one reason secular stagnation theorists believe r-star has fallen over time (Chart 1).

Fed economist Thomas Laubach and NY Fed President John Williams produce a popular estimate of r-star. It uses real GDP, core PCE inflation, oil prices, import prices and the federal funds rate as inputs. These are key macroeconomic variables, which are themselves determined at least partly by events in the private sector. A rough market measure of r-star is the 5y2y real swap yield.

Why Beliefs About R-Star Matter

In periods of economic stress, i.e., recessions, the Fed may choose to focus on a short-term estimate of r-star to help steer the policy rate. However, it is much more common to think of r-star as a long-term measure. In other words, it gives the long-term direction of the economy’s real interest rate. And it is this signalling function on which the paper predominantly focuses.

The intuition is as follows. Say economic activity has fallen. The Fed does not know whether the drop in activity is because of cyclical or permanent factors. It therefore assigns some weight to the possibility it was the latter and so drops its estimate of the r-star and reduces the interest rate.

The private sector knows the Fed uses r-star for its long-term view on the direction of the real interest rate. Therefore, it sees this policy rate decline as a signal that the r-star has fallen. The private sector consequently adjusts its consumption-saving behaviour according to this more pessimistic outlook.

The behavioural change then leads to an output and inflation fall. The central bank mistakenly interprets this demand shortfall as indicating the r-star has fallen further. It therefore needs to drop the policy rate further, which then feeds back into the private sector in a vicious cycle.

The authors call this phenomenon the ‘hall-of-mirrors’ effect. It builds on the fact that the central bank’s expectations reflect the private sector’s expectations and vice versa. Empirically, this is true – monetary policy affects market expectations of interest rates over very long horizons (Chart 2). The forward rate (9y1y) is closely correlated with the one-year nominal bond year.

The Problem

Now we understand the mechanism, how does the cycle lead to a bad outcome? Well, a bad equilibrium requires the Fed to get its estimate of r-star from the private sector and the private sector to get its estimate of r-star from the central bank.

Crucially, the Fed cannot distinguish between whether a decline in macroeconomic activity is because of cyclical or permanent factors. And the private sector cannot differentiate whether a drop in the policy rate is because the central bank has reduced its estimate of r-star or not.

If we have the above, a long-term decline in r-star occurs when both parties are overly optimistic about the information the other party holds. The central bank overestimates the private sector’s knowledge of economic fundamentals, and the private sector likewise overestimates the central bank’s ability to estimate r-star independently of macroeconomic factors (Chart 3b).

But if they both had common knowledge of each other’s beliefs, the private sector would know how much of the policy drop was related to r-star, and the Fed would know that a change in macroeconomic activity was unrelated to a fall in r-star. Both parties understand the drop in demand was impermanent, and there would be no decline in r-star over time (Chart 3a).

Explanation of chart: In Chart 3, the red lines give the central bank reaction function, and the blue lines gives the private sector reaction function. The flatter the red line, the less the central bank relies on learning r-star from the private sector. The steeper the blue line, the less the private sector relies on the central bank to get its estimate of r-star.

Also, the drop in the red line occurs first, and it falls because of a shock to u_h, which is a negative demand shock. The blue line then shifts right because of a shock to u_c, which is the shift in policy rate. In panel b, the shift right is insufficient to offset the decrease in r-star expectations (unlike in panel a). This is because the private sector mistakenly thinks that the central bank has very good private information and will not react much to private sector expectations.

The Macroeconomic Implications

With the hall-of-mirrors effect, an aggressive/expansionary policy strategy is less effective at reviving spending. And worse, it exacerbates the very problem policymakers are trying to solve. Why?

First, the central bank estimates that the r-star has fallen in response to a drop in economic activity (Chart 4a, yellow). This causes them to drop the policy rate by more than if they had known that r-star had not fallen (Chart 4b, yellow vs blue). The private sector then revises down their r-star beliefs and by more than the central bank (Chart 4a, red vs yellow).

This downward revision of r-star is equivalent to the private sector expecting a worse long-term economic outlook because a lower r-star implies a lower long-run real interest. Businesses then reduce their consumption habits in preparation for a worsening output. This decline in activity then offsets the typical increase in activity that a policy rate decline usually has (Chart 4c, blue vs red).

In turn, the smaller-than-expected increase in output leads to a smaller-than-expected increase in inflation (Chart 4d, blue vs red). This then feeds back into the central bank’s estimate of r-star, which now stays persistently lower than the level at which it started (Chart 4a, blue vs yellow).

In sum, the policy rate cut not only lowers the short-term interest rate but also endogenously prompts a decline in the natural rate through the private sector. This then reduces the expansionary effect of the policy rate decline, which leads to a permanent fall in r-star.

It is worth emphasising how this differs from the full information case. If the central bank knew that a decline in economic activity was purely due to cyclical factors, it would assign no weight to permanent factors, and so r-star would remain the same (blue). They then need only reduce the policy by a little to get a large output and inflation response. This is the conventional monetary policy response.

Bottom Line

The paper gives a new perspective on why interest rates, inflation, and growth remained persistently lower in the aftermath of the GFC. The central bank’s reliance on information garnered from the private sector makes it susceptible to the private sector’s knowledge of the economy. If businesses are unclear on the trajectory of long-term interest rates, one can rationalise an environment with a lower r-star, less effective monetary policy, and lower inflation.

The key to rectifying the issue is either (a) improving communication of r-star between the central bank and private sector, or (b) estimating r-star independent of variables endogenously determined by the private sector. Until that is achieved, aggressive policy easing will only lead to a continuously declining natural interest rate.