Extracting information shocks from the Bank of England inflation density forecasts

• DOI: http://doi.org/10.1002/for.2501
• Author: Carlos Díaz
• Published date: 01 April 2018
• Date: 01 April 2018

Received: 24 January 2017 Revised: 25 October 2017 Accepted: 26 October 2017

DOI: 10.1002/for.2501

RESEARCH ARTICLE

Extracting information shocks from the Bank of England

inflation density forecasts

Carlos Díaz

University of Leicester School of Business,

Leicester, UK

Correspondence

Carlos Díaz, University of Leicester School

of Business, Astley Clarke Building,

University Road, Leicester LE1 7RH, UK.

Email: cdv7@leicester.ac.uk

Abstract

This paper shows how to extract the density of information shocks from revi-

sions of the Bank of England's inflation density forecasts. An information shock

is defined in this paper as a random variable that contains the set of informa-

tion made available between two consecutive forecasting exercises and that has

been incorporated into a revised forecast for a fixed point event. Studying the

moments of these information shocks can be useful in understanding how the

Bank has changed its assessment of risks surrounding inflation in the light of

new information, and how it has modified its forecasts accordingly. The vari-

ance of the information shock is interpreted in this paper as a new measure of

ex ante inflation uncertainty that measures the uncertainty that the Bank antic-

ipates information perceived in a particular quarter will pose on inflation. A

measure of information absorption that indicates the approximate proportion of

the information content in a revised forecast that is attributable to information

made available since the last forecast release is also proposed.

KEYWORDS

inflation forecasting, density forecasts, uncertainty, forecast revisions

1INTRODUCTION

Density forecasting has become a useful tool for

researchers and policymakers as a way to evaluate the

uncertainty surrounding the central path of key economic

and financial variables. In particular, inflation density

forecasts are commonly used by some central banks as

informative tools in the process of monetary policy imple-

mentation. The Bank of England periodically produces

inflation and gross domestic product (GDP) growth den-

sity forecasts for the following quarters, presented to the

public in the form of so-called fan charts.Thesefore-

casts are updated every quarter with new information

made available since the previous release. Any difference

between the density forecasts for a fixed point event pro-

duced in two consecutive forecasting exercises will be due

to the effect of the incorporation of this new information

into the revised forecast. The random variable that pro-

duces this change is what will be called in this paper an

information shock. Knowing the density and the moments

of this information shock can be helpful, among other

things, in evaluating the process of information incorpora-

tion into subsequent forecasts. For example, if this density

is nonsymmetric, this can be interpreted as a sign that the

Bank expects new information to have a bigger impact on

a particular side of the distribution than the one antici-

pated in the previous forecasting exercise. The skewness

coefficient of the information shock can be seen, then, as

a measure of how the Bank has changed its assessment of

risks affecting inflation in the light of new information.

Similarly, its variance can be interpreted as a measure of

the uncertainty that the Bank anticipates new informa-

tion will have on future inflation. The method presented

here to extract the density of this information shock just

316 Copyright © 2017 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/for Journal of Forecasting. 2018;37:316–326.