Accounting conservatism, earnings components and accounting losses

Valentinčič, Aljoša (2004) Accounting conservatism, earnings components and accounting losses. PhD thesis, University of Glasgow.

Full text available as:
[thumbnail of 2004valentincicphd.pdf] PDF
Download (22MB)
Printed Thesis Information: https://eleanor.lib.gla.ac.uk/record=b2224986

Abstract

This study provides evidence on accounting conservatism based on a large sample of publicly-quoted UK companies over the period 1969-2001. The effects of conservation accounting are studied both indirectly and directly by using earnings measures containing varying levels of accruals and by further decomposing earnings into its operating cash flows and distinct accruals components. The analyses are also separated according to the sign of earnings and earnings components, and account for the effects of asset-recognition rules. Even though conservatism is an accruals phenomenon, this is the first study to provide direct empirical evidence on the role of accruals in accounting conservatism.

The thesis addresses the following issues. First, under conservative accounting, earnings-decreasing changes in performance measures (reflecting economic losses) that contain more accruals mean-revert more and earnings-increasing changes (reflecting economic gains) are persistent. Working capital accruals and special items are particularly strongly mean-reverting when they are earnings-decreasing. Depreciation accruals are persistent.

Second, direct tests by earnings components show that operating cash flows exhibit low timeliness overall and, given that they contain no accruals, no asymmetry in reflecting bad news. Earnings figures with more accruals exhibit more asymmetry in reflecting bad news. Working capital accruals and special items are important in this asymmetry, but depreciation is not. Interestingly, good news results in a small earnings-decreasing charge, consistent with smoothing. Lagged tests on accruals reveal that bad news from as much as three previous periods is reflected in current earnings through special items, inconsistent with conservatism. Evidence indicates that conservatism is increasing through time. The sensitivity to good news has decreased over time. To capture these changes, higher-moments measures are developed.

Third, the analysis by the sign of “bottom-line” earnings does not reveal any differences in reflecting good/bad news for the profit/loss firms. Separating earnings observations by sign of cash flow also reveals no differences. In contrast, separating observations by the sign of accruals (other than depreciation) reliably shows that the asymmetric timeliness is significantly higher in the negative-accruals groups, as expected. The accruals components determine this asymmetry, rather than the operating cash flow (or, earnings by itself).

Finally, less conservative recognition rules lead to stronger responsiveness of earnings to bad news, as reflected in working capital accruals and special items. Asset-specific measures of conservative recognition rules reinforce these findings. A puzzling result is that operating cash flows reveal a significant asymmetric response to bad news in the group of observations where it is least-likely to be observed (low book to market).

A selection of other results by size, industry, extremity of news, methods, accounting year-ends, market-wide returns, yields, method of estimation, etc., not only corroborates, but generally strengthens the results obtained.

Item Type: Thesis (PhD)
Qualification Level: Doctoral
Subjects: H Social Sciences > HF Commerce > HF5601 Accounting
Colleges/Schools: College of Social Sciences > Adam Smith Business School > Accounting and Finance
Supervisor's Name: Garrod, Prof. Neil
Date of Award: 2004
Depositing User: Mrs Marie Cairney
Unique ID: glathesis:2004-1137
Copyright: Copyright of this thesis is held by the author.
Date Deposited: 11 Sep 2009
Last Modified: 10 Dec 2012 13:34
URI: https://theses.gla.ac.uk/id/eprint/1137

Actions (login required)

View Item View Item

Downloads

Downloads per month over past year