Earnings persistence over the macroeconomic cycle: Evidence from Korea

Sorah Park, Heejeong Shin

Research output: Contribution to journalArticlepeer-review

7 Scopus citations

Abstract

This paper examines whether the persistence of earnings components is affected by the macroeconomic cycle in Korea. To measure the macroeconomic cycle, we use the cycle variation value of Coincident Composite Index (CCI) data obtained from the Korea National Statistics Office. Results from a sample of 21,232 firm-quarter observations over the period 2002-2013 indicate that accruals (cash flows) are more persistent than cash flows (accruals) during expansions (recessions). Also, when going from an expansion to a recession, a decline in accruals persistence is greater than that in cash flows persistence. When total accruals are decomposed into non-discretionary and discretionary portions using the modified Jones model (Dechow et al., 1995), we find that non-discretionary accruals are most persistent than the other components during both expansions and recessions, and a decline in persistence is largest (smallest) for discretionary accruals (cash flows) when going from an expansion to a recession. Most of these results hold when we split the macroeconomic cycle into four phases including transitory periods. Taken together, we provide evidence on the differential effects of macroeconomic cycle on the persistence of individual earnings components in Korea. Our findings suggest that macroeconomic variables are needed to be considered in studies on earnings persistence.

Original languageEnglish
Pages (from-to)2147-2165
Number of pages19
JournalJournal of Applied Business Research
Volume31
Issue number6
DOIs
StatePublished - 1 Nov 2015

Bibliographical note

Publisher Copyright:
© author(s); CC-BY.

Keywords

  • Business cycle
  • Earnings components
  • Earnings persistence
  • Macroeconomic cycle

Fingerprint

Dive into the research topics of 'Earnings persistence over the macroeconomic cycle: Evidence from Korea'. Together they form a unique fingerprint.

Cite this