Monetary Policy

ENPC, 2nd year - SEGF

François Geerolf

OFCE/Sciences Po, CEPR

Introduction

Plan

• Income, not substitution effects ? The importance of adjustable vs. fixed rate mortgages, particularly housing markets. Institutional detail.

• Real vs. nominal rates. Problem with the “natural rate” view.

• “Sticky prices” or something else ?

“Monetary Dialogue” with ECB

• “Are inflation dynamics different in the euro area and the United States?” C. Blot & F. Geerolf, Monetary Dialogue Paper, November 2023, requested by the ECON committee, European Parliament. [pdf] [html]

• “Excess liquidity in the euro area? Assessment and possible ways forward.” C. Blot, J. Creel & F. Geerolf, Monetary Dialogue Paper, September 2023, requested by the ECON committee, European Parliament. [pdf] [html]

• “Is monetary tightening a threat to financial stability?” C. Blot, J. Creel & F. Geerolf, Monetary Dialogue Paper, June 2023, requested by the ECON committee, European Parliament. [pdf] [html]

• “The direct and indirect impacts of the war on inflation.” C. Blot, J. Creel & F. Geerolf, Monetary Dialogue Paper, March 2023, requested by the ECON committee, European Parliament. [pdf] [html]

• “Heterogeneity of inflation in the euro area: more complicated than it seems.” C. Blot, J. Creel, F. Geerolf & S. Levasseur, Monetary Dialogue Paper, November 2022, requested by the ECON committee, European Parliament. [pdf] [html]

What do we know ?

Effects on macroeconomy ?

Rudd (2024) - Chapter 7 excerpts

Title

Intro

Shocks: who cares ?

Monetary policy effects

Monetary policy in theory

Forward guidance puzzle

Does monetary policy work ?

John Kenneth Galbraith, in The Good Society:

John Kenneth Galbraith, in The Good Society:

Galí (2015) - Effects of monetary policy

Ramey (2016) - Effects of monetary policy

Nominal vs. real rates

IS Curve

The IS curve represents the relationship between the output gap and the real interest rate. It can be written as:

\[y_t = E_t[y_{t+1}] - \frac{1}{\sigma}(i_t - E_t[\pi_{t+1}] - r^n_t)\]

where:

• \(y_t\) is the output gap.
• \(E_t[y_{t+1}]\) is the expected future output gap.
• \(1/\sigma\) is the intertemporal elasticity of substitution.
• \(i_t\) is the nominal interest rate.
• \(E_t[\pi_{t+1}]\) is the expected future inflation.
• \(r^n_t\) is the natural rate of interest.

Phillips Curve

The Phillips curve describes the relationship between inflation and the output gap. It can be written as:

\[\pi_t = \beta E_t[\pi_{t+1}] + \kappa y_t + u_t\]

where:

• \(\pi_t\) is the inflation rate.
• \(\beta\) is the discount factor.
• \(\kappa\) is the slope of the Phillips curve.
• \(u_t\) is a cost-push shock.

Taylor Rule

The Taylor rule is a monetary policy rule that describes how the central bank sets the nominal interest rate in response to deviations of inflation from its target and the output gap. It can be written as:

\[i_t = r^n_t + \pi_t + \phi_{\pi} (\pi_t - \pi^*) + \phi_y y_t\]

where:

• \(\phi_{\pi}\) is the responsiveness of the nominal interest rate to the inflation gap.
• \(\phi_y\) is the responsiveness of the nominal interest rate to the output gap.
• \(\pi^*\) is the target inflation rate.

Context

Euler equation

• Euler equation (IS curve) of the NK model: real interest rates. Nominal rates \(\Rightarrow\) real rates in the NK model.

• Even outside of the NK model, we teach that nominal rates should take into account inflation in some way:

  – Realized,

  – Or expected

• Consequence: more realized / expected inflation \(\Rightarrow\) raise nominal interest rates to keep real rates unchanged.

• ECB has justified its actions in keeping real rates unchanged, so the stance of monetary policy.

ECB publishes this

IMF, WEO April 2023

IMF, WEO April 2023

IMF, WEO April 2023

IMF, WEO April 2023

Gita Gopinath (2023)

Gita Gopinath (2023)

Gita Gopinath (2023)

Mean Inflation expectations (CES)

Median Inflation expectations (CES)

Long-term inflation expectations

Measure of expected inflation (on average) over the 5Y period beginning 5Y from today.

Measuring the Natural Rate of Interest

Measuring the Natural Rate of Interest

Real rates ?

• Estimates of “natural rates” or “r-star” even appear in Fed minutes.

• As a consequence, financial markets have become intereted in the concept of a “natural rate”. (even if they are skeptical, since the Fed is probably the largest financial market participant)

• Even outside of the NK model, there is a view that only real rates should matter. We even teach this to undergraduates.

• It also has implications on monetary policy, at a very basic level.

• It’s probably one reason why central banks feel compelled to raise interest rates when inflation increases, whatever the source. Central bankers are told higher inflation makes the stance of monetary policy more expansionary.

Nominal rates matter !

• Nominal rates matter a lot for governments: through fiscal rules. Nominal rate increases \(\Rightarrow\) lower primary surpluses.

• Nominal rates also matter for households through adjustable rate mortgages. Adjustable rate mortgages in 2006-2009 as a case study.

• Revisiting “housing wealth effect papers”. (most well-known is Mian, Sufi but it also applied to also other papers on housing wealth effects, including some of my own…)

• Nominal rates also matter for households through fixed rate mortgages: borrowing constraints depend on nominal rates, not real rates \(\Rightarrow\) Effects of nominal rates, not real on economic activity, notably lending and also construction.

• In conclusion: the same real rates but with much more inflation are much less stimulative. Not necessarily a dovish comment: it was also true when inflation was lower. Monetary policy was much more accomodative than we thought.

Governments

Law of motion for public debt

• \(G_t\) : nominal government spending at period \(t\), \(T_t\) : nominal taxes in period \(t\), \((G_t-T_t)\) the (nominal) primary deficit in period \(t\), which is the excess of government expenditures over taxes levied by the government.

• Interest rate is \(i_t\), then the law of motion of government debt is given by:

\[B_{t}=(1+i_t)B_{t-1}+G_{t}-T_{t}.\]

• \(Y_t\) nominal GDP. Assuming that nominal GDP grows at rate \(g\) + \(\pi\), we have that: \[\frac{Y_t}{Y_{t-1}}=1+g_t+\pi_t.\]

• Therefore: \[\frac{B_t}{Y_t} = \frac{1}{1+g_t+\pi_t}\frac{B_{t-1}}{Y_{t-1}}+ \frac{i_tB_{t-1}}{Y_t} +\frac{G_{t}-T_{t}}{Y_{t}}.\]

Law of motion for public debt

• Therefore: \[\frac{B_t}{Y_t} \approx \frac{B_{t-1}}{Y_{t-1}} - \underbrace{\pi_t \frac{B_{t-1}}{Y_{t-1}}}_{\text{Inflation tax}} - g_t \frac{B_{t-1}}{Y_{t-1}} + \underbrace{i_t\frac{B_{t-1}}{Y_t}}_{\text{Interest payments}} +\underbrace{\frac{G_{t}-T_{t}}{Y_{t}}}_{\text{Primary Deficit}}.\]

• Whether one should include the “growth tax” term can be debated.

• However there is no question that bondholders receive nominal interest payments, but also pay an inflation tax by holding government bonds.

• Weird to take one into account (interest payments) and not the other one (interest payments).

• Yet this is what we do. So nominal interest rates matter !

Budget rules

• European Maastricht: Total Deficit \(\geq\) -3% of GDP.

• Total Deficit = Interest Payments Deficit + Primary Deficit.

• Higher nominal rates \(\Rightarrow\) Higher interest payments (% of GDP).

• Dividing interest payments by nominal GDP does not help.

• Problem is: inflation tax is not counted as a resource for the government, while it should. Debt / GDP are going down while total deficit / GDP is going up. 2% of GDP.

• Depends on maturity of the debt.

Back of the envelope

• Debt = 100% of GDP (we should say 1 year of GDP). Average nominal rate: 1%.

• 1 % increase nominal rates \(\Rightarrow\) average nominal rate: 2% => +1% of GDP interest payments

• 3% Maastricht rule: primary deficit goes from -2% of GDP to -1% of GDP.

• Raise taxes, reduce spending \(\Rightarrow\) Strong effects on aggregate demand, depending on your view of the multiplier.

• Not just theoretical !

• Restrictive monetary policy \(\Rightarrow\) Restrictive fiscal policy.

France, Germany, Italy

Current example

Well-known example

Well-known example

Adjustable rates: Revisiting housing wealth effects

Share of variable rates

Fixed rates - Real rates or nominal rates

• Fixed rates: constraints take the form of a mortgage payment / income constraint (say not more than 35%).

• Problem: this constraint binds much more in an environment of high inflation. (even pure = price and wage inflation)

• Intuition: with inflation, your current income represents less of your future income.

• Another: with time, your ratio of mortgage payment/income goes down as inflation increases your income.

• Financial constraints is much more binding with low inflation.

Consequences

What about corporations ?

Nominal rates also matter

• Nominal rates matter: True for governments, especially in Europe. Also true for households, both variable / fixed rates, through in different ways. Maybe true for corporations.

• Need for more regulatory-type data on types of loans more precisely.

• Perhaps need to give up on optimal contracting, and take these features as exogenous ? (for now ?) Extreme hetereogeneity across countries which seems very hard to explain.

• Increase in nominal rates can be recessionary even when real rates are falling, unlike what NK models say. (regardless of whether one looks at realized or expected inflation)

• Stance of monetary policy might in fact be much more restrictive than we think.

Sticky Prices / Wages ?

Introduction

• New-Keynesian economics rests on a stiky-price / wage assumption.

• Yet there is in fact very little direct evidence that prices / wages are that sticky.

• There’s even less evidence that the effects of demand on output are related to sticky prices, as the IS-LM view of the world would suggest. See Summers’ (1991) previously cited paper.

• So, what’s the evidence ?

Evidence for sticky prices?

• Evidence for sticky prices? Usual argument, from Romer’s Advanced Macroeconomics textbook:

Price rigidities

Evidence for sticky prices?

• This argument is often made: monetary policy has real effects, hence prices must be sticky.

• I think that this argument is incorrect. Here is another model with flexible prices, where monetary policy could have real effects:

  • Monetary policy redistributes disposable income between borrowers and lenders, especially through the mortgage market.

  • There is a paradox of thrift: more consumption, less saving leads to more investment:

    • Closed economy: increase in saving.

    • Open economy: increase in saving, and trade deficit, imbalance \(S-I\). (this is because there is a leakage of aggregate demand, so investment is also made elsewhere to increase capacity)

Frequency of price change

Response of real output and the price level to a one-time permanent shock to nominal aggregate demand in the Calvo model

Price series of Nabisco Premium Saltines (16 oz) at a Dominick’s Finer Foods store in Chicago

Monetary non neutrality

Transience of temporary sales

The expenditure weighted distribution of the frequency of regular price change

Inflation expectations and the Phillips curve

Rudd (2022)

Footnote 2

Another sceptic: (tarullo_monetary_2017?)

(coibion_how_2018?) - Firms don’t know/care

Households expectations correlated to oil prices

Price-wage spiral ?

• In my opinion, wages are indexed where inflation is regularly high. (cost of living adjustments)

• Sometimes automatic (legal) indexations.

• Outside of that there’s very little evidence for wage price spirals…

• In fact wages have been falling during this episode, just as you’d have predicted if you forget about the Phillips curve.

Additional data

M1, M2, M3

Loans to households

Loans to NFCs

U.S., France, Japan, Germany

U.S. 10-, 20-, 30- years

France - 30 ans, 10 ans, 5 ans

France, Germany, Italy, Spain, Greece

France, Germany, Italy, Spain

Europe, Germany, Portugal, Greece

Dates of monetary policy meetings

Fed, ECB main short rates

Fed, ECB main short rates

ECB

Lending spreads

U.S. TIPS

AAA, BAA - United States

Loans in France, Germany, Italy

U.S.: mortgage, Fed Funds

Lending for house purchases, old loans

France, Belgium, Germany, Netherlands

France, Germany, Italy, Spain, Netherlands

Estonia, Lithuania, Latvia, Portugal, Finland, Spain

Lending for house purchases

Germany

France

France, Germany, Netherlands, Belgium, Europe

France, Spain, Portugal, Estonia, Lithuania

France, Spain, Portugal

Poland, Australia, Sweden

Conclusions: what do we know ?

Negative

• Inflation expectations are way too important in the discussion.

• Same for price/wage spiral.

• No evidence that monetary policy effects have to do with sticky prices…

• Isn’t monetary policy some form of fiscal policy ? Should it then be left to independant central banks ?

Positive

• Monetary policy has huge effects on financial markets. No wonder you have Fed/ECB “watchers” who are trying to understand how central bankers think.

• Monetary policy works through nominal rates not just real rates.

• Effects push in both directions: as MMT rightly points out, raising rates means more revenue for bondholders. Then depends on how fiscal rules work.

• Substantial effect on the housing markets: know the macro-prudential policies, adjustable vs. fixed rates. Does not appear to be the result of an optimal contracting problem, so trying to model this should be probably abandoned.

• Same probably works for firms.

Bibliography

Galí, Jordi. 2015. Monetary Policy, Inflation, and the Business Cycle: An Introduction to the New Keynesian Framework and Its Applications, Second Edition. Princeton University Press.

Ramey, V. A. 2016. “Chapter 2 - Macroeconomic Shocks and Their Propagation.” In Handbook of Macroeconomics, edited by John B. Taylor and Harald Uhlig, 2:71–162. Elsevier. https://doi.org/10.1016/bs.hesmac.2016.03.003.

Rudd, Jeremy B. 2022. “Why Do We Think That Inflation Expectations Matter for Inflation? (And Should We?).” Review of Keynesian Economics 10 (1): 25–45. https://doi.org/10.4337/roke.2022.01.02.