A Comparison of Business Cycle Extraction Methods: Application to the UK. / Paramaguru, Kanya.

ESCOE, 2021.

Research output: Working paper

Published

Abstract

This paper seeks to expand the discussion surrounding the dating of UK business cycles. Two different time-series filters are applied to UK output time-series to investigate what they would imply for the creation of any official recession dates. The NBER has a business cycles dating committee that investigates the dates of turning points in US business cycle providing a consolidation of business cycle dates for the US. There is, at present, no analogous committee or consolidation on UK turning points dates. There is a broad definition adopted in the UK that defines a recession as two or more periods with negative growth. The Office of National Statistics (ONS) uses a series of real GDP (deflated GDP) and then observes two or more periods of negative growth to define a recessions. 2020 has certainly provided an interesting critique to this broad definition, whereby the second quarter of 2020 so a -20% in GDP but then the next quarter recovered by 12% therefore only categorising the first half of 2020 as a recession. However, the decline in economic output within the UK has been more notable than other business cycle recession in recent decades. The aim of this paper is to continue the discussion on defining the turning points of UK business cycles. This study looks at two filters and how they would define UK business cycles. Although the merits of the filters are discussed before being used in estimation. Based on the outcomes of the business cycle dates, the filters that produce the most reasonable results are defined as a better approach. Reasonable approach is defined as one that matches the theory as to how often peaks and troughs can reasonably expected to occur.
Original languageEnglish
PublisherESCOE
Publication statusPublished - 22 Dec 2021
This open access research output is licenced under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.

ID: 43989003