Earnings Management, Auditor Changes and Ethics: Evidence from Companies Missing Earnings Expectations

Abstract

Companies’ earnings are arguably their most important financial disclosure, and corporate tactics to manage earnings can be directly constrained by audit procedures. In this article, we examine whether there is a link between companies’ earnings performance and changes in the accounting firm they choose to audit their financial statements. We find that when companies’ reported earnings just miss analysts’ expectations, they are more likely to change their auditor. Consistent with an opportunistic auditor switch, we also find that these companies have lower levels of earnings quality after the change. Our results concentrate on companies with more incentive or ability to act opportunistically, including those with negative market reactions to earnings, higher levels of accruals flexibility, and weaker corporate governance. Our study advances research into the public’s interest by providing evidence of suspect auditor changes that increase auditors’ ethical conflict to maintain independence and objectivity.

Appendix: variable definitions

Variables	Definition
AC_SIZE	The number of audit committee members divided by the number of directors
AC_IND	An indicator variable coded 1 if all audit committee members are independent, and 0 otherwise
ACC	Net income less operating cash flows, net of cash flows for discontinued operations scaled by lagged total assets
ACC_FLEX	The ratio of net operating assets to sales at the beginning of the year, scaled by industry-year median
ACQ	Cash flows for acquisitions scaled by average total assets
AGE	The natural log of firm age
ALTMAN	Altman is calculated as:$\begin{aligned} AltmanZ = & 3.3 \times \frac{{Net\, Income}}{{Total\, Assets}} + 0.99 \times \frac{{Sales}}{{Total\, Assets}} \\ & + 0.6 \times \frac{{Market\, Value}}{{Total\, Liabilities}} + 1.2 \times \frac{{Curent\, Assets}}{{Total\, Assets}} \\ & + 1.4 \times \frac{{Retained\, Earnings}}{{Total\, Assets}} \\ \end{aligned}$
AUD_CHG	An indicator variable coded 1 if the client changes its auditor within one year of the earnings announcement date, and 0 otherwise
BD_IND	An indicator variable coded 1 if 60% or more of the directors are independent, and 0 otherwise (Consistent with DeFond et al., (2005))
BD_SIZE	The number of directors on the board
BTM	The ratio of the book value to its market value of equity
CASH	Cash and cash equivalents scaled by lagged total assets
CAR	Accumulated market-adjusted stock returns for the (− 1, 1) window around the earnings announcement date
CEOCFO_CHG	An indicator variable coded 1 if the company changed their CEO and/or CFO
DA	The absolute value of performance-matched discretionary accruals. Discretionary accruals are estimated as the residuals from the Jones (1991) model and performance adjusted based on the return on assets (Kothari et al., (2005)
DOWN	An indicator variable coded 1 if the client switches from a BigN auditor to a non-BigN auditor, and 0 otherwise
FEE	The natural log of total audit fees
FSD	Financial statement divergence is calculated as: ${\sum }_{1}^{9}(|AFi-EFi|)/9*100,$ where i = 1, 2, 3…9, and AF (EF) is the actual (expected) frequency of the first digit i from the financial statement items. If an item is smaller than one, then the next non-zero value after the decimal will be used as the first digit
GC	An indicator variable coded 1 if the company received a going-concern opinion, and 0 otherwise
GOV	Modified corporate governance index used in DeFond et al. (2005)
HERF	Herfindahl index capturing the variation in the number of audit firms present in a given market, as well as the distribution of audit fees across those firms. The sum of the squared audit fee market shares of all auditors in an MSA
IMP	The importance of the client to the auditor measured as the percentage of total audit fee revenue collected from the auditor’s client portfolio
INV_AR	Inventory plus receivables divided by total assets
ISSUE	The sum of debt or equity issued during the past three years scaled by total assets
JUST_MISS	An indicator variable coded 1 if actual earnings per share minus consensus analysts’ forecasted earnings per share falls within the range of [− 0.01, 0], and 0 otherwise
LAG_ACC	Total accruals in the prior year scaled by total assets
LEVERAGE	Total short-term and long-term debt scaled by total assets
LOSS	An indicator variable coded 1 if the company reported a loss, and 0 otherwise
MEET_OR_BEAT	An indicator variable coded 1 if earnings per share minus consensus analysts’ forecasted earnings per is greater than or equal to zero, and 0 otherwise
PSEUDO_MISS_FIRM	A pseudo just miss variable produced through random firm assignment using the quantity of just miss observations within the same industry-year
PSEUDO_MISS_IND	A pseudo just miss variable produced through random year assignment using the quantity of just miss observations within the same firm
RESTATE	An indicator variable coded 1 if the client announced a restatement during the prior-year to current-year earnings announcement date range, and 0 otherwise
RESTRUCTURE	An indicator variable coded 1 if restructuring costs are present, and zero otherwise
ROA	Net income before extraordinary items divided by total assets
SALE_G	Percentage change in sales from the prior year
SD_CASH	Volatility of cash flows calculated as the standard deviation of the operating cash flows in the last five years
SIZE	The natural log of total assets
TENURE	An indicator variable coded 1 if the current auditor has audited the company for more than five years, and 0 otherwise