Inflation Expectations and Monetary Policy Surprises*

• Published date: 01 January 2020
• DOI: http://doi.org/10.1111/sjoe.12350
• Date: 01 January 2020

Scand. J. of Economics 122(1), 306–339, 2020

DOI: 10.1111/sjoe.12350

Inflation Expectations and Monetary Policy

Surprises*

Snezana Eminidou†

University of Cyprus, 1678 Nicosia, Cyprus

eminidou.snezana@ucy.ac.cy

Marios Zachariadis

University of Cyprus, 1678 Nicosia, Cyprus

zachariadis@ucy.ac.cy

Elena Andreou

University of Cyprus, 1678 Nicosia, Cyprus

elena.andreou@ucy.ac.cy

Abstract

We estimate monetary policy surprises for European consumers over time, based on monetary

policy changes that were unanticipated according to consumers’stated beliefs. We find that such

monetary policy surprises have the opposite impact on inflation expectations from the impact

found when assuming that consumers are well informed. Relaxing the latter assumption by

focusing on consumers’ stated beliefs, unanticipated increases in the interest rate raise inflation

expectations before the 2008 financial crisis. This is consistent with imperfect information

theoretical settings where interest rate hikes are interpreted as positive news about the state

of the economy by consumers who knowthat policymakers have relatively more information.

Keywords: Beliefs; crisis; imperfect information; rational inattention; shocks

JEL classification:E31; E52; F41

I. Introduction

An unexpected rise in the monetary policy rate can have two different

effects on inflation expectations, as follows. (1) If individuals view

this as an unexpectedly contractionary policy, they will revise inflation

expectations downward. (2) An unanticipated increase in interest rates could

be interpreted by consumers as positive news about the state of the economy

*Weare grateful to Nicoletta Pashourtidou for the suggestion to extract beliefs-based surprises and

for pointing us to the European consumers’ expectations data, without whichthis project would

not have been possible. We thank Martin Geiger and Morten Ravn for extensive comments.

E. Andreou acknowledges support from Horizon2020 ERC-PoC2014g rant 640924.

†Also affiliated with the Central Bank of Cyprus.

C

The editors of The Scandinavian Journal of Economics 2019.

S. Eminidou, M. Zachariadis, and E. Andreou 307

if they are aware that policymakers have relatively more information,1

resulting in higher inflation expectations. In this case, the action of

policymakers merely reveals that they are no longer worried about deflation.

That is, if individuals initially possess less information than policymakers,

then they learn something new about economic fundamentals by observing

the realization of the central bank’s monetary policy, and they revise

inflation expectations accordingly. Thus, while “textbook channels” and a

neo-Keynesian approach, such as that in Garc´ıa-Schmidt and Woodford

(2019), associate contractionary monetary policies with a fall in inflation

and inflation expectations, imperfect information-based approaches, such

as Campbell et al. (2012), Del Negro et al. (2012), and Melosi (2016),

associate higher interest rates with a rise in inflation expectations.2

It is important to distinguish between the above two theoretical

mechanisms via which monetary policy might affect inflation expectations,

because inflation expectations drive inflation realizations.3As Yellen (2016)

points out “[t]heory and evidence suggest that the inflation trend is strongly

influenced by inflation expectations that, in turn, depend on monetary

policy”. Yellen also notes that “the broader question of how expectations are

formed has taken on heightened importance” in recent times. In this paper,

we empirically investigate the above-described theoretical propositions

by examining directly how monetary policy surprises affect inflation

expectations. This is what Cochrane (2016) describes as “the big question”.4

In order to answer this question, we use monthly data across 15 euro-area

1For example, Gurkaynak et al. (2005) find that market participants believe that central bank

announcements contain not previously known or anticipated information about future monetary

policy actions, and Campbell et al. (2012) find that market participants infer that unexpected

policy adjustments by the central bank are responses to non-public information about the future

state of the economy.

2Campbell et al. (2012) discuss how monetary policy (forward guidance) might influence

economic agents in the “Delphic” case, where it affects inflation expectations by enabling

individuals to predict economic activity based on the superior information set of policymakers,

revealed after the policymaker undertakes action rather than by its anticipated direct impact on

the economy.

3Forexample, the first theoretical mechanism suggests that a higher policy rate can have an indirect

negative effect on the economy, to the extent that inflation expectations affecteconomic activity

and future prices. Lower inflation expectations can lead to a fall in consumption of households

that postpone purchases in anticipation of lower prices.They can also affect expected wage rates

and real interest rates. For example, if a given hike in the nominal interest rate is not followed

by a similar movement in inflation expectations going in the same direction in a neo-Fisherian

manner, the expected real interest rate rises.This increases the real cost of bor rowingand acts as

a demand for loans suppressant, which adversely affects investmentand consumption decisions.

4In his words “[t]he big question is expectations. Will people read higher interest rates as a

warning of inflation about to break out, or as a sign that inflation will be even lower.” Similarly,

Garc´ıa-Schmidt and Woodford (2019) ask “is there reason ...thatacommitmenttokeepnominal

interest rates low ...willbedeflationary...?”Inanswer,“[t]here isoneway inwhichsuchan

C

The editors of The Scandinavian Journal of Economics 2019.

308 Inflation expectations and monetary policy

economies for the period 1985:1–2015:3 to obtain estimates of monetary

policy surprises under different assumptions. We then use these to explain

inflation expectations of different types of consumers before and since the

financial crisis of 2008. In an imperfect information setting, the impact of

monetary policy surprises could be different for consumers with potentially

different abilities in processing information. This is particularly so in crisis

periods during which both the incentive to obtain information and the ability

needed to obtain it are greater compared with tranquil periods.

Furthermore, if consumers have incomplete information, then they might

be surprised by a broader set of monetary policy changes, compared with

agents who have more information about macroeconomic fundamentals.

An important focus of our study is to assess whether monetary policy

surprises obtained under the assumption that consumers are well informed

about the state of the macroeconomy have different impacts on inflation

expectations than surprises obtained under the assumption that consumers

are only as informed as revealed by their stated beliefs about the economy.

We allow for the individuals’ information set to be revealed by their stated

beliefs about the macroeconomy rather than assuming that they observe

the histories of the complete set of macroeconomic variables. This allows

us to further assess the empirical relevance of imperfect information-

based theoretical mechanisms discussed in the above-mentioned papers.

Thus, surprises are estimated as changes in monetary policy that were

unanticipated according to the consumers’ type-specific stated beliefs about

the economy, without assuming that they necessarily observe past values of

a large set of macroeconomic variables. To our knowledge, this approach

to estimating monetary policy surprises has not been previously considered

in the literature.

We define the unpredictable change in interest rates as a monetary

policy surprise. The unpredictability of monetary policy changes – and their

subsequent interpretation as monetary surprises – depends on how much

information we assume individuals to have. A change in monetary policy is

a surprise to the extent that individuals have not observed the information

set based on which they could have forecast the policy change prior to

its arrival.5Thus, we consider monetary policy surprises that allow for

individuals to have imperfect information, in addition to surprises obtained

under the traditional assumption that individuals have an information set

outcome could easily occur, and that is if the announcement of the policy change were takento

reveal negative information (previously known only to the central bank) about the outlook for

economic fundamentals”, so that individuals change their inflation expectations accordingly.

5Moreover, it is a surprise relevant to particular types of individuals, to the extent that these

individuals also have the incentive and ability to pay attention to the shock once it ar rives.This

is a point we do not pursue further in this paper.

C

The editors of The Scandinavian Journal of Economics 2019.