https://orcid.org/0000-0002-5306-058X
![Sample our Economics, Finance,Business & Industry journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5931/TOC-Subject-Sample-2021-Economics-Finance-Business-Industry.jpg"})
ABSTRACT
We investigate whether and how CEO knowledge background affects goodwill impairment and whether it mitigates investors’ risk assessment of goodwill impairment. We employ CEOs’ educational, financial, and overseas backgrounds to represent their knowledge background. The results, based on a sample of Chinese firms, show that CEOs with a higher level of knowledge background are more likely to test and record a goodwill impairment in a timely manner. In addition, we find that the negative effect of goodwill impairment on abnormal stock returns exists only in firms whose CEOs have a weaker knowledge background. The results of our cross-sectional tests and additional analyses are consistent with our main inference that a higher level of CEO knowledge background facilitates a deeper understanding of goodwill, which mitigates cognitive bias and thereby increases CEOs’ willingness to record a goodwill impairment in a timely manner and ensure investors’ trust.
Acknowledgments
We appreciate the helpful comments by Fabrizio Di Meo, Masahiro Enomoto, Junxiong Fang (ISAR discussant), Andrei Filip (Associate Editor), Keishi Fujiyama, Yuyao Gu, Hai Lu, Beatriz Garcia Osma, Yoshiko Shirata, Byron Song (Editor), Andrew Stark, Marco Trombetta, Shih-Bin Wu (AAA discussant), Junling Yao, Yong Yu, Yangxin Yu, Haiyan Zhou, two anonymous reviewers, and participants at the 18th International Symposium on Empirical Accounting Research in China (ISAR), the 2020 Annual Meeting of Japan Accounting Association, the European Accounting Association (EAA) Virtual Annual Congress 2021, and the 2021 American Accounting Association (AAA) Virtual Annual Meeting.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
2 Compared to CFOs who have a relatively homogeneous background in accounting and finance, CEOs’ backgrounds are considerably more diverse and likely to differentially affect goodwill impairment decisions.
3 In this study, “knowledge background” refers to the educational and practical experiences that collectively shape an individual's understanding and judgment.
4 For instance, we find that for our sample firms reporting unexpected negative goodwill impairment experience a 35.25 percent decrease in cumulative abnormal returns around the announcement day. This decrease suggests a significant negative impact on the firm's value due to the announcement of the unexpected negative goodwill impairment. For example, if a firm's market value is 16 billion yuan (our sample mean market value of Chinese listed firms is 16.3 billion yuan), it could lose 5.64 billion yuan of its market value when announcing an unexpected negative goodwill impairment.
5 In terms of goodwill impairment testing, both IAS 36 and ASBE 8 require firms to recognize goodwill impairment and prohibit the amortization of goodwill. As a side note, ASBE 8 prohibits the reversal of all impairment losses, whereas IAS 36 specifically prohibits the reversal of impairment only for goodwill.
6 In June 2001, the FASB began to change the goodwill estimation method from amortization (with “definite” life) to impairment (with “indefinite” life). Behind this qualitative change, the FASB argued that under SFAS 142, financial statements could better reflect the underlying economics of goodwill. Following FASB, IASB revised the standards on goodwill treatment (IFRS 3) in 2005. Although the aim of the regulatory change is to better reflect the underlying economics of goodwill, SFAS 142 and IFRS 3 have faced several critical issues. A stream of the literature shows the unintended outcomes of goodwill impairment, including managerial manipulation (Ramanna, Citation2008), adverse market reaction (Bens et al., Citation2011), and negative future earnings performance (Cready et al., Citation2012), among others.
8 The China's corporate governance code was issued on 7 January 2002.
9 For example, see the news article (in Chinese) for Hui Wang, the former CEO of Beijing Lanxum Technology Co., Ltd. (https://www.sohu.com/a/318922253_153054); see the news article (in Chinese) for Lanhang Xu, the former CEO of ChangYuan Technology Group Ltd. (https://www.sohu.com/a/332278818_114719); and see the news article (in Chinese) for Ke Xu, the former CEO of Meinian Onehealth Healthcr Hldngs Co Ltd. (https://www.thepaper.cn/newsDetail_forward_13576503).
10 There are three types of boards in Chinese stock market: main board, SME board, and tech board.
11 Regarding the timing of the goodwill impairment tests in China, ASBE 8 requires firms to conduct impairment tests on goodwill when there is an indication of impairment and at the end of each fiscal year regardless of whether there is an indication of impairment. The goodwill impairment information is disclosed in the annual report, and the date of the goodwill impairment announcement is the same as the annual report date.
12 It is computed by the market model as follows: FR = α0 + α1MR + ϵ, where FR is the firm's daily stock return and MR is the daily stock market return. Along with prior studies (Chen, Citation2019), the market model is estimated over the period [−200, −60] relative to the announcement day. The residual of the market model is the firm's daily abnormal return. We add the daily abnormal returns to measure the cumulative abnormal returns (CAR) during the three-day period [−1, +1] surrounding the announcement day.
13 The results of propensity score match method are shown in Appendix B. The Panel C of Appendix B shows the results of the logistic regression. In Panel C, we find that the signs of the coefficients are consistent with our prediction. Based on the regression results, we notice that the firms whose CEOs hold concurrent positions in other entities, with more foreign investors, with a larger size and more geographical segments are more likely hire a CEO with a higher level of knowledge-based background.
14 For comparison, as shown in the Panel D of Appendix B, regarding the unmatched sample, the mean differences in six matching variables in logistic regression model between the treatment group and nontreatment group are statistically significant.
15 According to the requirements of the Guidelines for the Industry Classification of Listed Companies (2012 Revision) (Announcement No. 31 [2012], CSRC).
16 The low correlations among CEO knowledge-related variables and other CEO-related variables suggest that our CEO-related variables capture the CEO characteristics from different aspects.
17 We partition the sample based on the median value of CEO_EDU. The high (low) CEO_EDU group indicates the observations with CEO_EDU higher than (lower or equal to) the median value of CEO_EDU.
18 One-to-one matching without replacement overlooks potentially useful observations.
19 AC_SCORE is estimated based on the adjusted accruals-cash-flows-based regression model (Ball & Shivakumar, Citation2005) by Lee et al. (Citation2015) (page 111). A higher AC_SCORE represents a higher conditional conservatism.
20 See the article “Should goodwill amortization be reintroduced?”, which notes that “Many have raised concerns that impairment losses on goodwill are recognized ‘too little, too late.’” (https://home.kpmg/xx/en/home/insights/2020/03/goodwill-and-impairment-dp.html).