A long-run view on monetary policy - Part 2: Why are interest rates so low?

This is Part II of a series of articles on Swiss monetary policy I wrote jointly with Simon Schmid for Republik. They kindly agreed that I can publish an english version on my blog. Enjoy:

40 years ago, interest rates on 10-year Swiss government bonds stood at more than 7 percent. After the financial crisis in 2008 they fell to negative territory. At the same time commentators often suggest that central banks' policies are responsible for historically low interest rates. This article is about why this argument rests on a common misunderstanding

Monetary policy since the financial crisis is not the main reason why we have low interest rates today. More important were two key developments: a secular decline in inflation and productivity growth.

Let's look at some data first. The chart shows interest rates of Swiss government bonds since WWII (orange), averaged over five year intervals. In addition, it shows inflation (light blue) and nominal GDP growth per capita (dark blue). All three measures show a similar development: After WWII they increased but fell continuously since the 1970s.

Figure 1: Interest rates peaked 40 years ago

Note: Interest rates on long-term government bonds (orange), inflation (light blue) and growth in nominal GDP per capita (dark blue). Averaged over five year intervals and measured in percent. GDP growth is usually expressed in real (inflation-adjusted) terms to show how many products more were produced in a given year. Financial variables, such as interest rates, are usually available only in nominal terms (that is in Swiss francs, for example). To show the link between the nominal interest rate and economic growth we therefore show nominal GDP growth per capita. Source and construction of data: Daniel Kaufmann (2018): «Nominal stability over two centuries».

Interest rates seem to move with inflation and economic growth. But why? We will have a look at both factors in what follows.

Inflation affects the general level of interest rates

Let's start with an an example
  • Imagine, you put 100 Francs on your bank account. If you expect inflation to be zero in the coming year, you can buy the same amount of goods and services as today. The purchasing power of 100 Francs remains unchanged.
  • If you suddenly expect, however, that the general level of prices will increase by 5% over the coming year (hard to imagine, but this was often the case in the 1970s) you also expect that the purchasing power will fall. You will be able to buy 5% fewer goods and services with 100 Francs. Putting your money on a bank account with zero interest is actually a bad deal.
  • If the bank does not want to loose you as a customer, it has to compensate you for the expected increase in the price level and pay you around 5% interest. This inflation premium is the reason why interest rates increase.
The relationship between interest rates and expected inflation is a well-known phenomenon and termed the Fisher-effect, named after the US economist that first described the idea. The Fisher-effect contributed to the increase and subsequent decline in interest rates since the 1970s through three channels:
  • The supply of bonds: The lower expected inflation, the more the debtor has to repay in inflation-adjusted terms when issuing a bond (at a given interest rate). Therefore bonds become less attractive as a means to finance purchases and the supply of bonds declines; interest rates decline.
  • The demand for bonds: The lower expected inflation, the higher the inflation-adjusted return when holding a bond (at a given interest rate). The creditor receives more when buying a bond and therefore the demand for bonds increases; interest rates decline. 
  • The risk premium: The lower and more stable the inflation rate becomes, the lower is the risk that a sudden change in inflation changes the return or cost of issuing a bond. Therefore, investors ask for a lower compensation for this inflation risk (that is a lower interest rate).
Central banks were quite successful to stabilize inflation at a low level. Therefore, the level of interest rates declined as well. We can roughly guess how much the Fisher-effect contributed to this decline. In the 60s, 70s, and 80s, inflation fluctuated between 3 and 4%. Nowadays, inflation stands and just above 0%. We can therefore roughly attribute 3 percentage points to the decline in inflation; maybe even more

Slower growth pushes interest rates down

We cannot attribute the entire decline in interest rates to disinflation. Lower economic growth is a second important factor contributing to the decline. Why? another simple example:
  • Suppose an entrepreneur asks for a loan to invest in a new machine.
  • If she is relatively pessimistic about future sales of her products, she will take out the loan and buy the machine only if interest rate payments are quite low. Otherwise, she couldn't repay the loan.
  • If she is optimistic, she can afford higher interest rate payments because she expects to sell more products. Even if interest rates are high, she can easily repay the loan.
  • If all entrepreneurs expect that the economy will stagnate, banks have to offer loans at relatively low interest rates. The lower demand for loans will dampen interest rates. 
Maybe you realized that I shifted gears somewhat. Now, we are talking about real factors that affect real (that is inflation-adjusted) interest rates. This is important because there are several factors independent of monetary policy that in fact influenced the general level of interest rates.

One important development, and a major cause for the low growth rates, is a decline in productivity growth. In the following chart, we see that growth in labor productivity, that is the growth of output per hour worked, stood between 2% (Switzerland) and 4% in many developed countries (we averaged over 10 year intervals to focus on long-term trends). Then, we observe a long-lasting decline that continued during the financial crisis. Since 2010 labor productivity grew on average only by 0.4% (Switzerland) and around 1% in other countries.

Figure 2: Productivity has declined in advanced economies

Note: Annual growth in labor productivity in percent. Averages over 10 year intervals. Source: OECD

Measuring productivity is difficult. However, the decline in productivity growth appears to be a widely documented international phenomenon. Importantly, this decline started already before the financial crisis and many economists believe that low productivity growth is here to stay; as a consequence, also interest rates would remain low. Robert J. Gordon of Northwestern University argues that the era of useful new inventions has ended and Larry Summers of Harvard suggests that we are in a trap of persistently low demand (secular stagnation hypothesis).

The dire predictions are of course uncertain and not all economists share this view. On one thing, however, most economists surprisingly do agree: Monetary policy does not affect productivity growth in the long-term. As a consequence, the secular decline in productivity is unlikely the result of the actions of central banks.

Other developments that are largely independent of central banks' actions are particularly relevant for Switzerland:
  • Demographic factors, such as an aging population, increase the demand for safe assets pushing down the level of interest rates.
  • Asian countries, such as China, increased their foreign exchange reserves and hold them predominantly in safe government bonds. This phenomenon is termed global savings glut.
  • In Switzerland, the debt-to-GDP ratio has fallen from 48% in 2005 to 30%. The supply of safe government bonds has therefore declined and this pushes down interest rates further.
  • The Swiss franc is generally regarded as a safe haven because of the stable economic and political environment. The demand for assets in denominated in Swiss francs is therefore relatively high and leading to a Swiss interest rate bonus.
Monetary policy since the financial crisis is therefore, if we look at the long-run developments, not the main reason for the historically low interest rates. Even optimistic estimates on the impact of the first balance sheet expansion by the US central bank conclude that it has lowered interest rates at most by 1 percentage point. Most of the real factors that drove down interest rates are outside of the realm of monetary policy. And, the decline in inflation that actually is driven by monetary policy should be termed a restrictive rather than an expansionary policy.

Confusing communication

Why, then, do most commentators still believe that low interest rates are a sign of an expansionary monetary policy? We find the answer partly in the communication practices of central banks themselves. Usually, they communicate their policy stance using a short-term money market rate, in Switzerland the three-month Libor. This is an interest rate at which banks lend money to each other. 

The reason for using a short-term interest rate as a communication tool is that a change in the money supply leads to a change in the interest rate. In the textbook version, this works as follows:
  • A central bank surprisingly increases the money supply.
  • The recipient of this money, usually a bank, would like to get rid of it because it earns no interest. One possibility is to lend it to another bank that needs liquidity.
  • Because the supply of loans in the interbank market increases, it becomes easier for other banks to obtain a loan. Therefore, they have to pay a lower interest rate.
This mechanism is called the liquidity effect: A monetary expansion temporarily lowers the interest rate and stimulates economic activity. However, the textbook case is just a thought experiment - an artificial situation where all other factors affecting interest rates are held constant. In reality, of course, this is never the case. Demand and supply for loans and bonds change constantly; as we have seen, for example, because of changes in expected inflation and productivity growth, and not only because of unexpected changes in the money supply.

In reality, it is therefore challenging to judge whether a decline in interest rates stems from the liquidity effect (an expansionary monetary policy), the Fisher effect (a restrictive monetary policy) or long-run changes in productivity (independent of monetary policy). The difference, however, is key to judge how expansionary monetary policy actually is.

This insight is of course not new. Milton Friedman made the case in an article on the stagnating economy in Japan. He argued that interest rates are a misleading indicator of how expansionary monetary policy actually is:
Initially, higher monetary growth would reduce short-term interest rates even further. However, as the economy revives, interest rates would start to rise. That is the standard pattern and explains why it is so misleading to judge monetary policy by interest rates. Low interest rates are generally a sign that money has been tight, as in Japan; high interest rates, that money has been easy. 
Friedman puts the Fisher-effect first: After a restrictive monetary policy stance, lower inflation expectations lead to falling interest rates. The liquidity effect plays a secondary role because it affects the level of interest rates only temporarily.

Many commentators put the liquidity effect first and argue that low interest rates are a sign of an expansionary monetary policy (or even «ultraexpansionary» as journalists like to put it). However, we should rather view the low level of interest rates as a symptom and consequence of the general secular decline in inflation and productivity growth. Although non-conventional policies, such as large-scale bond purchases, certainly played a role, they are not the main reason (we will discuss the causes and impact of non-conventional policies in the next article).

Of course, low interest rates may still have negative repercussions. Investors may not distinguish between nominal and real (inflation-adjusted) interest rates. This kind of money illusion can lead to errors, financial instability, and maybe asset price bubbles.

Nevertheless, trying to increase interest rates is not the right remedy to ensure financial stability. In the long-run, this would lead to even lower inflation and therefore to even lower interest rates. It is better to take a breath and wait until the economy has recovered and inflation has risen to a sustainable level.

Comments

Popular Posts