Talking Over Time ‐ Dynamic Central Bank Communication

• Published date: 01 August 2023
• Author: LAURA GÁTI
• Date: 01 August 2023
• DOI: http://doi.org/10.1111/jmcb.12934

DOI: 10.1111/jmcb.12934

LAURA GÁTI

Talking Over Time - Dynamic Central Bank

Communication

This paper studies the optimal dynamic communication strategy of central

banks using a Bayesian persuasion game framework. In a dynamic environ-

ment, nancial market participants and the general public have misaligned

interests because the present and future have different relevancein their op-

timization problems, leading to a novel trade-off for the monetary authority.

Compared to the static benchmark, I show that the central bank (CB’s) op-

timal dynamic communication policy should put a higher weight on talking

about the present state than the future. In addition, the CB should strategi-

cally send more noisy signals than in the static benchmark.

JEL codes: C73, D83, E58, E71

Keywords: Bayesian persuasion, dynamic communication, Delphic

forward guidance

W   (FM)  

intently to central bank (CB) announcements, this cannot be said of the general public.

As Coibion et al. (2020), Binder and Kim (Forthcoming), and many others show,most

rms and households do not incorporate information provided by the CB into their

expectations. A dynamic economic environment, however, introduces a discrepancy

between the interests of Wall Street and Main Street: the future economy is relevant

for the returns of the former, while nancial constraints and discounting of the future

limit the latter’s focus to the present. Talking over time thus presents a CB with a

novel trade-off: how should it communicate so as to serve the interest of the general

public while knowing that its messages only get through to investors?

The novelty of this paper is to characterize a dynamic CB communciation policy

that responds optimally to this trade-off. To highlight the novel trade-offs coming

I thank the editor, Sanjay Chugh, and one anonymous referee for helpful advice and suggestions. I also

thank Ryan Chahrour, Susanto Basu, and Peter Ireland, as well as Sylverie Herbert, Bartosz Ma´

ckowiak,

and Francesco Zanetti. The views expressed herein are my own and do not necessarily reect those of the

ECB or the Eurosystem.

L G is a Senior Economist at the Directorate GeneralResearch of the European Central Bank

(E-mail: laura_veronika.gati@ecb.europa.eu).

Received June 21, 2021; and accepted in revised form March 7, 2022.

Journal of Money, Credit and Banking, Vol. 55, No. 5 (August 2023)

© 2022 The Ohio State University.

1148 :MONEY,CREDIT AND BANKING

from dynamics, I contrast a dynamic communication problem with its static analogue.

The static and dynamic models involve a Bayesian persuasion-type communication

game between a CB and the FM about two economic fundamentals that represent

the current and future stance of the business cycle. Main Street cares about current

employment and is thus concerned with today’s business cycle, while Wall Street

sets today’s investment choices with an eyetoward future prots. Capturing the idea

that the CB has a mandate to stabilize current ination and employment, I assume

that the CB is concerned with the current business cycle, as opposed to the FM that

is maximizing future prots. At the same time, the monetary authority takes into

account that only a subset of the public listens to its messages—Wall Street.

In this environment, the CB sends the FM a noisy signal, which is a weighted

sum of the present and the future stance of the business cycle. Importantly, the

static and dynamic models are identical up to the correlation structure between the

two states. This allows me to isolate the role of dynamics for optimal communi-

cation design. In particular, I investigate two ways in which the CB can address

the trade-off between the interests of Wall Street and Main Street. I rst ask how

strongly the CB should weight the present against the future in its signal—a dimen-

sion of communication I refer to as “targetedness.” Second, I explore whether the

optimal precision of the dynamic signal differs from that of the static one. In other

words, does dynamic communication involve a different amount of noise than static

communication?

There are two key ndings. First, the CB’s signal is always more targeted toward

current conditions in the dynamic model than in the static one. In other words, in a

dynamic world, the CB prefers to talk more about the present than the future relative

to the static environment. On the one hand, this is because a temporal correlation

between output today and output tomorrow renders the two states more distinct than

variables that are correlated in the cross section. On the other, in a dynamic world,

the FM uses all the available information to learn about the evolution of the business

cycle. Therefore the CB has to weight the signal more heavily toward the present to

skew the available information over time in the right direction.

Second, in the dynamic model, the CB optimally communicates more noisily than

in the static model. The reason behind this is that because information is carried over

from one period to the next, providing too much information today lowers the CB’s

ability to persuade the FM tomorrow.Intuitively, communicating very precisely today

makes the FM too condent in its beliefs the next period, making it very hard for the

CB to convince it otherwise. Optimal precision is thus lower in the dynamic problem

than in the static one as the CB smoothes the information it provides the FM overtime.

The dynamic communication policy studied in this paper yields testable implica-

tions regarding the everyday communication decisions central bankers make. First,

the paper provides conditions for when the CB should give investors precise infor-

mation, and when it is preferable to communicate in noisy “Fedspeak” that leaves

some residual uncertainty around the current and future economic outlook. Second,

it also suggests that it is in the interest of the general public that the CB should talk