Financial Distress

Financial Distress and Earnings Management: Effectiveness of Independent Audit Committees
Bikki Jaggi
Department of Accounting and Information Systems, Rutgers business school-Newark & New Brunswick Rutgers University School of Business, Piscataway-NJ 08854 Email: jaggi@rbsmail.rutgers.edu. Tel: 732-445-3539 And
Lili Sun
Department of Accounting and Information Systems, Rutgers business school-Newark & New Brunswick Rutgers University, Newark, NJ Email: sunlili@rbsmail.rutgers.edu. Tel: 973-353-5762.
Acknowledgement: the authors are grateful to the valuable comments and suggests provided by Ling Lei, Clive Lennox, and participants at the 12th annual mid-year auditing conference.

ABSTRACT This study empirically examines whether monitoring of earnings management by independent audit committees differs for financially distressed and non-distressed firms. Discretionary accruals are used as a proxy for earnings management, financially distressed and non-distressed firms are identified based on losses and negative cash flows for two consequent years, and audit committees are considered independent when all committee members are independent, as required under the Sarbanes-Oxley Act. The regression test results show that monitoring of earnings management by independent audit committees is more effective when firms are in financial distress. Additional, the analyses show that independent audit committees are especially concerned with the use of positive discretionary accruals by financially distressed firms to adjust the reported earnings upward. Our results are robust to alternative proxies for financial distress and earnings management. The findings thus provide support to the expectation that the requirement of independent audit committees protects investors interests by constraining managerial behavior of earnings management, especially when the firms are in a financial distress situation, which provides a strong motivation for managers to manipulate the reported earnings. Additionally, the results show that there is no significant difference in the monitoring effectiveness of audit committees when 100% of their members are independent, as required by the Sarbanes-Oxley Act, 70% or 80% are independent. Therefore, relaxation in the requirement of independence of 100% of membership may provide some flexibility to the firms, without compromising independence of audit committees. Key Words: Independent Audit Committee; Earnings Management; Financial Distress

I. INTRODUCTION Recent corporate scandals, such as Enron, Worldcom, etc. provided a strong incentive for re-evaluation of corporate governance boards of US firms, including independence of audit committees, so that investors interests could be safeguarded. In 2002, the Sarbanes-Oxley Act was passed, and this Act required that all members of audit committee should be independent. An important rationale for this requirement is to ensure an effective monitoring of accounting policies and systems so that reliable and unbiased information is provided to investors. The Securities and Exchange Commission (SEC) fully supported the requirement of independence of audit committee by adopting new rules, set forth in new Exchange Act Rule 10A-3 (SEC 2003). Recent studies have empirically tested whether independent audit committees would provide an effective monitoring of earnings management, and evaluated the association between independent audit committees and earnings management by using discretionary accruals as a proxy for earnings management (e.g. Klein 2002; Xie et al. 2003). The findings of these studies show that a higher percentage of independent members on audit committees is associated with lower discretionary accruals. These studies are based on all firms irrespective of the fact whether there is a motivation for managers to manipulate the reported earnings or not. It is, however, important to examine whether independent audit committees would provide an effective monitoring mechanism when there is a strong motivation for managers to manipulate the reported earnings. Furthermore, the need for an effective monitoring of earnings management is even greater in this type of situation because reliability and unbiasedness of reported
information is more critical for investors for proper evaluation of the firms future operating performance when discretionary accruals are used to send positive signals or to reduce the impact of negative signals. It has been well documented in the literature that managers of financially distressed firms would have a strong motivation to engage in earnings manipulation for sending positive signals or reducing the impact of negative signals emanating from financial distress (e.g. Sweeney 1994; DeFond and Jiambalvo 1994; DeAngelo et at. 1994; Burgstahler and Dichev 1997; Jaggi and Lee 2002). Thus, we focus in this study on financially distressed firms and conduct a comparative evaluation of the association between independent audit committees and discretionary accruals, a proxy for earnings management, for financially distressed and non-distressed firms. We hypothesize that monitoring effectiveness by independent committees will be stricter for financially distressed firms than for financially non-distressed firms. Additionally, we conjecture that independent audit committees would especially provide a stricter monitoring of upward adjustment of reported earnings using positive discretionary accruals by financially distressed firms because managers would have a stronger motivation to camouflage the firms poor operating performance by using positive discretionary accruals either to convert losses into small earnings or to meet the market expectations (e.g. Burgstahler and Dichev 1997). On the other hand, the use of negative discretionary accruals by financially distressed firms may not be of much concern to independent audit committees because their use could be justified as conservative accounting and may also not be of much concern to investors. Moreover, it has been pointed out that the use of negative discretionary accruals may not actually
reflect earnings management, but instead their use may be necessitated by the financial distress situation (e.g. Butler et al. 2004). The study is based on all firms that are covered by the 2002 corporate library data base, which contains information of independent audit committees. The firms are identified as financially distressed firms if they experienced either losses for two consecutive years, i.e. year 2000 and 2001, and/or their operating cash flows for two consecutive years was negative (e.g. Jaggi and Lee, 2002). This procedure resulted in 234 financially distressed firms. The remaining 990 firms covered by the data base in 2002 (excluding financial firms and the firms with missing data) are classified as financially non-distressed firms. We obtained financial data from the Compustat data base. Consistent with the existing literature, current discretionary accruals are used as a proxy for earnings management because current discretionary accruals can be easily manipulated (e.g. Ashbaugh et al., 2003; Xie, et al., 2003). We use the performanceadjusted current discretionary accruals model, as suggested by Kothari et al (2005) and Ashbaugh et al. (2003). In accordance with the Sarbanes-Oxley Act, audit committees are considered as independent when all committee members are independent. We conduct OLS regression tests to evaluate the association between current discretionary accruals, proxy of earnings management, and independent audit committees. Consistent with earlier studies (e.g. Ashbaugh et al. 2003, Butler et al. 2004), we use firm size, growth, leverage, litigation risk, and previous years current accruals as control variables. The findings show that there is a significantly negative association between absolute value of current discretionary accruals, proxy for earnings management and
independent audit committees. The negative association is significantly stronger for financially distressed firms compared to financially non-distressed firms. These findings thus show that independent audit committees perform a useful function in controlling managerial behavior of earnings management, especially when the firms are in financial distress. The findings also support the expectation that the negative association between discretionary accruals and independent audit committee is stronger when positive discretionary accruals are used by financially distressed firms to adjust the reported earnings upward. Additionally, we examine whether 100% threshold of independence of audit committee members, as required under the Sarbanes-Oxley Act, is a necessary condition to ensure independence of audit committees in monitoring earnings management effectively. Our findings do not detect any difference in the monitoring effectiveness of earnings management when the percentage of independent members is 100%, 80% or 70%. Thus, these findings show that the monitoring effectiveness of earnings will not be compromised if the percentage of independent members is reduced up to 70%. The study makes the following contributions to the literature. First, the findings provide evidence that independent audit committees are providing useful function in controlling earnings management, and enhancing the reliability of reported earnings. Second, independent audit committees ensure the reliability of reported information when it is critical for investors, i.e. when the firms are in financial distress and have a strong motivation to manipulate the reported earnings. Third, the monitoring effectiveness of independent audit committees is especially evident when managers in the financially distressed firms use positive discretionary accruals to adjust the reported earnings upward
to camouflage the firms weak operating performance. These findings can be interpreted to suggest that absence of independent audit committees is likely to result in biased information to investors, especially when the reliability of reported information is critical for investors decisions. Finally, the findings show that the monitoring effectiveness of audit committees is not reduced when the percentage of independent committee members is reduced to 80% or 70%. The remainder of the paper is organized as follows: In section two, we discuss the research design, including hypotheses for the study. Discussion of results is contained in part three, and part four contains summary and conclusion.
II. 1.
RESEARCH DESIGN Hypothesis
(a) Association between Independent Audit Committees and Earnings Management It is well documented in the literature that financial distress provides a strong incentive for managers to manipulate the reported earnings for different reasons, such as avoiding debt covenant violations, avoiding losses or declines in earnings, etc. Sweeney (1994) finds that managers may manipulate the reported earnings to avoid debt covenant violations and may use the choice of accounting methods to achieve this objective. DeFond and Jiambalvo (1994), DeAngelo et al. (1994) and Jaggi and Lee (2002) argue that managers would use discretionary accruals when they violate debt covenants as a result of financial distress. DeFond and Jiambalvo (1994) find that positive discretionary accruals are used to avoid debt covenant violations, whereas DeAngelo et al. (1994) detect the use of negative discretionary accruals to obtain better contract terms during
renegotiation of contracts. Jaggi and Lee (2002) present that the use of positive and negative discretionary accruals would depend upon the severity of financial distress. Burgstahler and Dichev (1997) also find that managers use positive discretionary accruals to avoid losses or meet market expectations, proxied by analyst forecasts or previous years performance. Manipulation of reported earnings would result in camouflaging the firms operating performance, which would reduce the reliability of reported earnings. In fact, manipulated reported earnings would provide biased information to investors. The policy makers and regulators would be concerned with biased information provided to investors because it would hurt them and consequently have a negative impact on the financial markets. These concerns have led to the development of regulations to ensure accuracy and reliability of reported information. These regulations may deal with external and/or internal monitoring of managerial behavior of earnings management. The external monitoring is provided by independent external auditors, whereas internal monitoring is accomplished through independent corporate boards, including independent audit committees. The passage of Sarbanes-Oxley Act in 2002 was motivated to deal with both of these issues. With regard to internal monitoring, this Act requires that all audit committee members be independent. This requirement is based on the premise that independent audit committees would be free of managerial influences, and thus they would be able to effectively monitor managerial behavior of earnings management. Empirical studies have been conducted to evaluate the association between the percentage of independent members on audit committees and earnings management, proxied by discretionary accruals (e.g. Klein 2002; Xie, et. al. 2003). The findings of
these studies report a negative association between the percentage of independent committee members and earnings management. These studies are based on all firms irrespective of the fact whether there is any motivation for managers to engage in earnings manipulations or not. We extend research on the association between independence of audit committee and earnings management by evaluating whether independent audit committees would effectively control managerial behavior of earnings management when there is a strong motivation for managers to engage in earnings manipulation. We focus on financially distressed firms, which are likely to provide strong motivation to manipulate the reported earnings. Financially distressed firms are likely to engage in earnings management to avoid losses or to meet the market expectations (e.g. Burgstahler and Dichev 1997). DeAngelo et al. (1994) document that financially distressed firms may adjust the reported earnings downward to obtain better terms during renegotiations on contracts. We conjecture that independence of audit committees would provide a deterrent to managers to engage in earnings management even when there is a strong motivation to do so as a result of financial distress. Since audit committee members of financially distressed firms face a higher likelihood of litigation risk, they would be stricter in monitoring the managerial behavior of earnings management. We conduct a comparative evaluation of the association between discretionary accruals and independent audit committees for financially distressed and non-distressed firms on the absolute value of positive and negative discretionary accruals. This analysis is based on both upward and downward adjustment of reported earnings because both are considered earnings management. The following hypothesis is developed to test this expectation:
H1: The negative association between independent audit committees and absolute current discretionary accruals is stronger for financially distressed firms compared to financially non-distressed firms. (b) Association between Independent Audit Committees and Upward Adjustment of Reported Earnings. Burgstahler and Dichevs (1997) analyses document that more firms fall in the range of small positive earnings, which may suggest that these firms use positive discretionary accruals to convert losses into small positive earnings. Additionally, it has been documented that managers may also use small positive discretionary accruals to meet market expectations, proxied by analyst forecasts or previous years reported earnings (e.g. Burgstahler and Dichev, 1997). On the other hand, negative discretionary accruals may be used to create cookie jar reserves (e.g. Jordan and Clark 2004; Reidl 2004), to achieve earnings smoothing (e.g. Kirschenheiter and Melumad 2002), or to obtain better terms in renegotiations on contracts (DeAngelo et al. 1994). An important question of interest to us in this study is whether independent audit committees would be equally concerned with upward adjustment of reported earnings using positive discretionary and downward adjustment of reporting earnings using negative discretionary accruals. We argue that independent audit committee would be more concerned with the use of positive discretionary accruals than negative discretionary accruals. As discussed above, the positive discretionary accruals are used either to convert losses into small positive earnings or to meet the market expectations. In the case of small positive earnings, the reported earnings camouflage the loss situation and send biased signals to investors, which would be of major concern to investors. Similarly, reporting of higher earnings to meet market expectations camouflages the
10
firms operating performance to show better performance than it actually is. This situation will also lead investors to wrong investment decisions, which may ultimately not be in the best interest of investors. We, therefore, conjecture that independent audit committees would be especially concerned with positive discretionary accruals used to adjust the reported earnings upward. On the other hand, the use of negative discretionary accruals, which is used either to create cookie jars for earnings smoothing or to exaggerate the losses to obtain better contract terms, may not be of much concern to audit committees because of the following reasons. First, whatever motivation, managers may justify the use of negative discretionary accruals as conservative accounting. Second, smoothing of earnings through cookie jar reserves or renegotiation of contracts may not be of immediate major concern to investors. Third, recent research (Butler et al. 2004) findings show that the firms may have to use negative accruals due to poor performance and this will not be earnings management to achieve any objective. Thus, we expect independent audit committees to provide stricter monitoring of positive discretionary accruals than negative discretionary accruals. We develop the following hypothesis to test this expectation: H2: Independent audit committees monitoring of earnings management in financially distressed firms is stronger for the use of positive discretionary accruals than negative discretionary accruals. 2 (a) Research Methodology Calculation of Current Discretionary Accruals Earlier studies have used total discretionary accruals, based on the Jones Model or Modified Jones Model, as a proxy of earnings management (e.g. Jones 1991; DeFond and
11
Jiambalvo 1994). Lately, it has been emphasized that current discretionary accruals are easy to manipulate than non-current accruals. Therefore, the use of current discretionary accruals is considered more appropriate to evaluate earnings management (e.g. Xie et al., 2003). Additionally, it has been pointed out that discretionary accruals are affected by the firms performance. In order to isolate the performance effect from earnings management, it has been suggested that the firms performance is taken into consideration for calculation of discretionary accruals, and two different techniques have been suggested in this regard (e.g. Kothari et al 2005; Kasznik 1999; Ashbaugh et al. 2003; Butler et al. 2004). First, a performance measure, i.e. ROA, can be included in the regression analysis for estimating normal accruals. Second, a portfolio technique with performance matching on the basis of industry and return on assets has been suggested. In this study, we use the portfolio technique, because it is considered to provide a better estimation of discretionary accruals (e.g. Kothari et al., 2005), and similar to Ashbaugh, et al. (2003), we term them these discretionary accruals as Performance-Adjusted Current Discretionary accruals (PACDA). In order to test the robustness of findings, we also use the other technique of including ROA in the regression analysis to estimate the parameters for calculating normal accruals, which has been termed as REDCA. In order to calculate PACDA, we partition the entire population of Compustat firms, excluding financial sector firms by two-digit SIC code (e.g. Ashbaugh, et al. 2003). Industries with fewer than 15 firms are deleted. We estimate the parameters for normal accruals for each two-digit SIC firms using the following equation: CA = 0 + 1 (1 / lag1asset ) + 2 ( Re v) (1)
12
Where : CA = Current accruals, reflected by net income before extraordinary items (Compustat date item # 123) plus depreciation and amortization (Compustat data item # 125) minus operating cash flows (Compustat data item # 308) scaled by the beginning of year total assets. Lag1asset= total assets at the beginning of the fiscal year t. Rev= net sales (Compustat data item #12) in year t less net sales in year t-1 scaled by the beginning of the year total assets.
All variables are winsorized at 1 percentile and 99 percentile. The parameters estimated from equation (1) are used to calculate the expected current accruals (ECA): ECA = 0 + 1 (1 / lag1asset ) + 2 ( Re v AR) Where: AR = accountings receivable (Compustat item #2) in year t less accounts receivable in year t-1, scaled by the beginning of year total assets. The current discretionary accruals (DCA) are calculated as follows: DCA = CA ECA (3) (2)
In order to obtain PADCA, we partition firms within each two-digit SIC code into deciles based on their last years return on assets (ROA), and obtain the median value of DCA for each ROA portfolio. The PACDA is then determined as follows: PADCA = DCA median DCA of the matching portfolio (b) Financial Distress (4)
13
Different stages of financial stress can be identified depending on the severity of financial distress (e.g. Lau 1987). Consistent with prior literature (e.g. Jaggi and Lee 2002), this study considers a firm financially distressed if its operating cash flows have been negative for two consecutive years and/or it incurred losses for two consecutive years. In order to test the robustness of our results, we also use the alternative proxies for financial distress as the Altmans Z-score (Altman 1968) and Zmijewskis risk score based on probit model (Zmijewski 1984). (c) Audit Committee Independence An analysis of the descriptive statistics on the audit committee size of our sample firms show that the mean of audit committee size is 3.78 members, with the range between 2 and 8 members. The median of the committee size is 3. While majority of the audit committees consist of 3 or 4 members, more than 20% of firms have audit committees with 5 members or more. Consistent with the Sarbanes-Oxley Act, we define the audit committee as independent committee when all of its members are independent1. We, however, also conduct tests by defining the audit committee as independent when 70% and 80% of its members are independent respectively. This classification scheme suggests that the audit committee will be identified as independent if 3 members out of 4 members are independent under the 70% threshold and 4 out of 5 members are independent under the 80% threshold. (d) Regression Model In order to capture both positive and negative discretionary accruals as earnings management, we first use absolute value of discretionary accruals as the dependent
14
variable, and regress it on independent variables of audit committees, financial distress, and control variables: absPADCA = + 1 AUIND + 2 FD + 3 LAG1CA + 4 LEVERAGE + 5 LITIGATION _ RISK + 6 LNMVE + 7 PBRATIO + (5)
We add an interaction between financial distress and independent audit committees in the regression analyses to test our hypothesis H1:
absPADCA = + 1 AUIND + 2 FD + 3 LAG1CA + 4 LEVERAGE + 5 LITIGATION _ RISK + 6 LNMVE + 7 PBRATIO + 8 FD AUIND +
(6)
We add a dummy variable to indicate positive discretionary current accruals, and also include an interaction among positive discretionary accruals, financial distress, and independent audit committees in the regression analyses to test our hypothesis H2:
absPADCA = + 1 AUIND + 2 FD + 3 LAG1CA + 4 LEVERAGE + 5 LITIGATION _ RISK + 6 LNMVE + 7 PBRATIO + 8 FD AUIND + 9 POSITIVE + 10 POSITIVE FD AUIND +
(7)
where: absPADCA= a proxy for earnings management. It is the absolute value of the performance-matched discretionary current accruals FD= an indicator variable which equals to 1 for firms in financial distress, 0 otherwise. (Expected sign +) AUIND= an indicator variable which equals to 1 for firms with independent audit committees, 0 otherwise. (Expected sign -)
15
FDAUIND=
the interaction term between financial distress indicator and audit committee independence indicator. It equals to 1 for firms in financial distress and with independent audit committee, 0 otherwise. (Expected sign -)
POSITIVE= 1 for firms with signed PADCA >0, 0 otherwise. POSITIVEFDAUIND= An interaction term among POSITIVE, FD, and AUIND. (Expected sign -) LAG1CA= Last years current accruals. current accruals equals net income before extraordinary items (Compustat data item123) plus depreciation and amortization (Compustat data item 125) minus operating cash flows (Compustat data item 308) scaled by beginning of year total assets. (Expected sign -) LEVERAGE= A firms total assets (Compustat data item 6) less stockholders equity of common shareholders (Compustat data item 60) divided by total assets. (Expected sign +) LITIGATION_RISK= = 1 for firms in a high litigation industry, and 0 otherwise. High litigation industries are industries with SIC codes of 28332836 (Pharmaceutical), 35703577 (Computer), 36003674 (Electrical and Telecommunication), 5200 5961 (Retailer and Wholesaler), and 73707374 (Programming and Software), 0 otherwise. (Expected sign +) LNMVE= Natural logarithm of a firms market value of equity. A firms market value of equity is calculated as its price per share at fiscal year end (Compustat data item 199) times
16
the number of shares outstanding (Compustat data item 25). (Expected sign -) PBRATIO= Price to book ratio. Market value at the fiscal year-end divided by book value of common equity. (Expected sign +)
3. Sample Selection and Data Collection We start the sample selection process by identifying the firms for which data on audit committees are available on the 2002 version of the Corporate Library data base, and we identify a sample of 1740 firms. As a second step, we delete the firms belonging to the financial sector with two digit SIC 60-69 (n=292), and the firms with missing financial and audit committee data (n=224). This process resulted in a sample of 1224 firms. In the third step, we classify the sample into financially distressed and nondistressed firms. The firms are identified as financially distressed firms if they experience losses for years 2000 and 2001 and/or negative operating cash flows for these two years. As a result of this process, we identify 234 financially distressed firms and 990 as non-distressed firms. The number of sample firms at different stages of the selection process is given in Table 1 (Panel A). ---------------------------Table 1 ---------------------------Distribution of the sample firms by financial distress and independent audit committees is provided in Panels B through D of Table 1. In Panel B, independence of
17
all committee members (100% independence) is assumed. The results show that 74.5% (n=912) of the sample firms have independent audit committees, whereas 25.5% (n=312) have non-independent audit committees. Distribution of financially distressed firms with independent and non-independent audit committee is 76.9% and 23.1% respectively, whereas it is 73.9% and 26.1 % respectively for the financially non-distressed subsample. In Panel C (Panel D), we provide distribution of firms by financial distress and independent audit committees where the audit committee is considered independent when at least 80% (70%) of its members are independent. Under these criteria, the number of independent audit committees increases slightly.
III. 1.
DISCUSSION OF RESULTS Descriptive Statistics. Descriptive statistics for the total sample as well as for the financially distressed
sample firms are provided in Table 2. We winsorize all variables at 1 percentile and 99 percentile. ---------------------------Table 2 ---------------------------Panel A shows that the average (median) discretionary current accruals for the full sample are -2.3% (-1.6%). The average market value is 1.23 billion; the mean leverage is 53.5%; the mean price to book ratio is 2.305. 19.1% of sample firms are defined as distressed; 31% are in the high litigation risk industries (i.e., Pharmaceutical, Computer,
18
Electrical and Telecommunication, Retailer and Wholesaler, Programming and Software), and 33.3% have positive discretionary current accruals. As expected, the financially distressed firms compared to the financially non-distressed firms have a higher magnitude of discretionary accruals, higher leverage, smaller market value of equity and smaller price to book ratio. (See Panel B and C of Table 2). 2. Univariate Test Results Table 3 contains the results of univariate analyses. We conduct t-tests on the means of absolute values of performance-matched discretionary current accruals between different groups. We also conduct nonparametric Wilcoxon rank-sum test to examine whether distributions of PACDA differs between the groups. ------------------------Table 3 ------------------------The t-test results are similar to the Wilcoxon rank-sum test results. Consistent with the findings of prior studies on financially distressed firms (e.g. Jaggi and Lee, 2002) we find that, on average, financially distressed have significantly higher current discretionary accruals (both positive and negative current discretionary accruals) than non-distressed firms (Panel A, Table 3). The comparative results on the distressed firms with independent and nonindependent audit committees (Panel B of Table 3) show that discretionary accruals are significantly lower for distressed firms with independent audit committees compared to the firms with non-independent audit committees. The results on the non-distressed firms also show that discretionary accruals are lower for firms with independent audit
19
committees compared to the firms with non-independent audit committees, but the difference is not significant. These results thus indicate that independence of audit committees plays an important role in deterring managers to manipulate the reported earnings when managers have a strong motivation for manipulation of reported earnings because of financial distress. These results are consistent with our hypothesis H1. 3. Regression Results We conduct OLS regression tests on the total sample of 1,224 firms using equations (5), (6), and (7). We include the control variables that may have an impact on the use of discretionary accruals in the regression analysis. The regressions results are presented in Table 4. ---------------------------Table 4 --------------------------The regression results of Model 1 (equation 5) show that the FD coefficient (FD is equal to 1 when the firm is financially distressed) is significantly positive, indicating that financially distressed firms are associated with higher discretionary accruals. The AUDIND coefficient for independent audit committees is significantly negative, indicating independent audit committees are more effective in monitoring earnings management than non-independent audit committees. The regression results of Model 2 (based upon equation 6) show that the coefficient for the interaction term FD and AUDIND is significantly negative, as expected. These findings thus show that the negative association between independent audit committees and discretionary accruals is
20
especially strong when the firms are in financial distress. These findings thus support our hypothesis H1. The regression results of Model 3 (based upon equation 7) show that the coefficient for the interaction term among FD, AUDIND, and POSITIVE (POSITIVE is equal to 1 when the firm has a positive discretionary current accrual) significantly negative, as expected. This suggests that independent audit committees monitoring of discretionary accruals in financially distressed firms is even stronger for positive discretionary accruals. This finding thus supports our hypothesis H2. Consistent with our hypothesis, the results suggest that independent audit committees are especially concerned with positive discretionary accruals and are not much concerned with the use of negative discretionary accruals. As discussed earlier, plausible explanation for their lower concern with the use of negative discretionary accruals is as follows: First, as documented by Butler et al. (2004), negative discretionary in financially distressed firm may not reflect earnings management. Instead, they may be due to poor performance, and poor liquidity-related transactions, e.g. write-off of nonperforming assets. Second, the use of negative discretionary accruals may be considered as conservative accounting and audit committees may not like to intervene in such managerial decisions. Third, the use of negative discretionary for obtaining better refinancing terms, as pointed out by DeAngelo et al. (1994), maybe perceived by audit committees as a proper action in the best interest of the firm and shareholders. The results on the control variables, consistent with our expectations and prior researchs findings, show that the last years current accruals (LAG1CA) has a significant negative coefficient, indicating the reversal of accruals. As expected, the firms in high
21
litigation risk industries (LITIGATION_RISK) have higher magnitude of discretionary accruals. The firms with larger size, measured in log of market value of equity (LNMVE), have higher absolute values of discretionary accruals. Finally, the firms with higher growth, measured by the price to book ratio (PBRATIO), have higher absolute values of discretionary accruals. Consistent with Butler et al. (2004), there is a significant negative association between LEVERAGE and absolute values of discretionary accruals for the full sample. Untabulated tests results for the sub-samples with positive and negative discretionary accruals indicate that the negative association is mainly driven by firms with negative discretionary accruals. A plausible explanation for this is that firms with higher leverage are less likely to use negative discretionary accruals to lower the reported income. 4. Tests Based on 80% and 70% independence of Audit Committee Membership The Sarbanes Oxley Act requires that all audit committee members should be independent in order to ensure independence of audit committees. We conduct additional test to examine whether 100% threshold for independence is important to ensure independence of audit committees in providing effective monitoring of earnings management by management. Our analysis of audit committee data indicates that the audit committee size varies between 3 and 7 members. Based on the 100% independence criterion, all committee members should be independent. If the threshold is reduced to 80 % or 70%, one or two members of audit committee may not be independent. We rerun the tests on the reclassified firms with independent and non-audit committees based on the lower independence threshold of 80% and 70% respectively to
22
evaluate their monitoring effectiveness of earnings management. The results are contained in Table 5 and 6. -----------------Tables 5 and 6 ---------------------The results contained in Tables 5 and 6 are consistent with the results presented in Table 4. These results thus also provide support to our hypotheses H1 and H2, and show that there is no significant difference in the results based on 70% or 80% audit committee independence compared to 100% audit committee independence. These findings suggest that some degree of relaxation in the requirement of independence of 100% of membership will not jeopardize the effectiveness of independent audit committees in monitoring earnings management. 5. Additional Tests We perform a number of sensitivity tests to evaluate the robustness of our findings. First, we use Altmans Z-score (Alterman 1968) and Zmijewskis risk score based on probit model (Zmijewski 1984) as alternative proxies for financial distress to test the robustness of our results. Altmans (1968) Z-score is obtained from the Compustat, which is defined as follows: Z-score = 0.012*(Working Capital/Total Assets) + 0.014*(Retained Earnings/Total Assets) + 0.033*(Earnings before Interest and Tax/Total Assets) + 0.006*(Market Value Equity/Book Value of Total Debt) + 0.999*(Sales/Total Assets). A company is classified in category of financial distressed firms if Altman Zscore is smaller than 1.812.
23
Under the criterion of Zmijewskis risk score, a company is classified in the distressed category if Zmijewski Probability is larger than 28% (Carcello and Neal 2000). Zmijewski Probability =
1 2
Z2 2
dZ , where y is calculated based upon the
following formula from Zmijewski (1984): y=-4.336-4.513*(Net Income/Total Assets)+5.679*(Total Liabilities/Total Assets)+0.004*(Current Assets/Current Liabilities). The results (untabulated) based on these proxies for financial distress are similar to those reported in the tables. Second, we use the alternative proxy of current discretionary accruals for earnings management. We use REDCA as used by Ashbaugh et al. (2003) and suggested by Kothari et al., (2004). This measure includes lagged ROA in the accrual regression to control for the firm performance. Procedures for calculation of REDCA are discussed in Appendix A of this paper. The results of this test are similar to those reported for PADCA. Therefore, the use of alternative proxy for discretionary accruals did not have any influence on our results. Third, we rerun the regression tests using log of the absolute value of PACDA as the dependent variable to avoid normality violations (e.g. Ashbaugh, et al. 2003). The results indicate no change in the findings tabulated in the tables.
IV. CONCLUSION This study has investigated the association between independence of audit committees and earnings management in financially distressed firms, whose managers have a strong incentive to manipulate the reported earnings. The study is based on a 24
sample of 1,224 firms including 234 financially distressed firms and 990 non-distressed firms contained in Corporate Library Database for year 2002. The firms are considered as financially distressed if they have negative operating cash flows, and/or losses for two consecutive years. Earnings management is proxied by the performance-adjusted current discretionary accruals model as suggested by Kothari et al. (2005) and used by Ashbaugh et al. (2003). Consistent with the Sarbanes-Oxley Act, audit committees are considered to be independent if all committee members are independent. The findings confirm our expectation that the monitoring effectiveness of independent audit committee of earnings management is stronger for financially distressed firms compared to financially non-distressed firms. The results also show that independent audit committees are more concerned with positive discretionary accruals used by financially distressed firms to adjust the reported income upward. In other words, independent audit committees provide stricter monitoring of upward earnings management in financially distressed firms because managerial behavior of earnings management in this case is motivated to camouflage the firms poor operating performance, which will make it difficult for investors to evaluate the firms future operating performance. In contrast, independent audit committees are less concerned with negative discretionary accruals because negative accruals can be due to the poor performance, or justified as conservative accounting. Our findings are robust to alternative measures for financial distress and current discretionary accruals. Similar results are obtained when an audit committee is considered as independent when 80% or 70% of its members are independent. This finding suggests
25
that a reduced level of threshold for independence is as effective as 100% independence threshold. The study does have its limitations. First, similar to other studies, it employs discretionary accruals as a proxy for earnings management. Interpretation of our findings is based on the assumption that discretionary accruals capture managerial behavior of earnings management. Second, validity of our findings depends on the accuracy of classification of firms into financially distressed and non-distressed firms. Though, we have used different proxies for this classification, caution is still warranted. The study has focused on managerial motivation for earnings management when the firms are financially distressed. There may, however, be other motivations for earnings management. Therefore, additional research is needed to evaluate the monitoring effectiveness of independent committees when there is some other strong motivation for earnings management, e.g. meeting analyst forecasts.
26
REFERENCE Altman, E. 1968. Financial Ratios, Discriminant Analysis and the Prediction of Corporate Bankruptcy. The Journal of Finance (September): 589-609. Ashbaugh, H., R. LaFond, and B. W. Mayhew. 2003. Do Nonaudit Services Compromise Auditor Independence? Further Evidence. The Accounting Review 78 (3): 611-639. Burgstahler, D. and I. Dichev. 1997. Earnings Management to Avoid Earnings Decreases and Losses. Journal of Accounting & Economics 24 (December): 99-126. Butler, M., A. J. Leone, and M. Willenborg. 2004. An Empirical Analysis of Auditor Reporting and its Association with Abnormal Accruals. Journal of Accounting and Economics 37: 139-165. Chen, K.C.W., and B.K. Church. 1996. Going Concern Opinions and the Markets Reaction to Bankruptcy Filings. The Accounting Review 71(1): 117-128. Clark, T., and M. Weinstein. 1983. The Behavior of the Common Stock of Bankrupt Firms. Journal of Finance 38 (May): 489-504. DeAngelo, H, L. DeAngelo, and D. J. Skinner. 1994. Accounting Choice in Troubled Companies. Journal of Accounting and Economics 17: 113-143. DeFond, M.L., and J. Jiambalvo. 1994. Debt Covenant Violation and Manipulation of Accruals. Journal of Accounting and Economics 17: 145-176. Eberhart, A. C., W. T. Moore, and R. Roenfeldt. 1990. Security Pricing and Deviations from the Absolute Priority Role in Bankruptcy Proceedings. The Journal of Finance 45 (December): 1457-1469.
27
Jaggi, B., and P. Lee. 2002. Earnings Management Response to Debt Covenant Violation and Debt Restructuring. Journal of Accounting, Auditing, and Finance: 295-324. Jones, J.J. 1991. Earnings Management During Import Relief Investigations. Journal of Accounting Research 29 (Autumn): 193-228. Jordan, C. and S. Clark. 2004. Big Bath Earnings Management: The Case of Goodwill Impairment under SFAS No. 142. Journal of Applied Business Research 20 (2): 6369. Kasznik, R. 1999. On the Association between Voluntary Disclosure and Earnings Management. Journal of Accounting Research 37: 57-81. Klein, A. 2002. Audit Committee, Board of Director Characteristics, and Earnings Management. Journal of Accounting and Economics. 33: 375-400. Kothari, S.P., A.J. Leone, and C.E. Wasley. 2005. Performance Matched Discretionary Accrual Measures. Journal of Accounting and Economics 39: 163197. Kirschenheiter, M. and N. Melumad. 2002. Can Big Bath and Earnings Smoothing Coexist as Equilibrium Financial Reporting Strategies? Journal of Accounting Research 40 (3): 761-96. Lau, A. H. 1987. A Five-State Financial Distress Prediction Model. Journal of Accounting Research 25 (Spring): 127-138. Reidl, E. 2004. An Examination of Long-lived Asset Impairments. The Accounting Review 79 (3): 823-52. Securities and Exchange Commission. 2003. Standards Relating to Listed Company Audit Committees. http://www.sec.gov/rules/final/33-8220.htm.
28
Sweeney, A.P. 1994. Debt-Covenant Violations and Managers Accounting Responses. Journal of Accounting and Economics 17: 281-308. Wertheim, P., and M. Robinson. 2000. The Effect of a Firm's Financial Condition on the Market Reaction to Company Layoffs. Journal of Applied Business Research 16 (4): 63-72. Xie, B., W.N. Davidson III, and P.J. DaDalt. 2003. Earnings Management and Corporate Governance: the Role of the Board and the Audit Committee. Journal of Corporate Finance 9: 295-316. Zmijewski, M. 1984. Methodological Issues Related to the Estimation of Financial Distress Prediction Models. Journal of Accounting Research (Supplement): 59-82.
29
Table 1 Sample Selection and Distribution by Independent Audit Committees

Panel A: Sample Selection No. of firms 1,740 292 220 4 234 990 1,224
Initial sample in Corporate Library database for year 2002 Minus: firms in financial sector (two-digit SIC 60-69) Minus: firms that have missing information for PADCA calculation Minus: firms that have missing information on audit committee Final sample Final stress sample Final nonstress sample Total final sample
Panel B: Sample Distribution When Audit Committee Independence Cutoff = 100% Financially Financially Nondistressed firms distressed firms Total % of independent members = 100% 180 (76.9%) 732 (73.9%) 912 (74.5%) % of independent members < 100% Total 54 (23.1%) 234 (100%) 258 (26.1%) 990 (100%) 312 (25.5%) 1,224 (100%)
Panel C: Sample Distribution When Audit Committee Independence Cutoff = 80% Financially Financially Nondistressed firms distressed firms % of independent members >= 80% 184 (78.6%) 777 (78.5%) % of independent members < 80% Total 50 (21.3%) 234 (100%) 213 (21.5%) 990 (100%)
Total 961 (78.5%) 263 (21.5%) 1,224 (100%)
Panel D: Sample Distribution When Audit Committee Independence Cutoff = 70% Financially Financially Nondistressed firms distressed firms Total % of independent members >= 70% 192 (82.1%) 835 (84.3%) 1,027 (83.9%) % of independent members < 70% Total 42 (17.9%) 234 (100%) 155 (15.7%) 990 (100%) 197 (16.1%) 1,224 (100%)
30
Table 2 Descriptive Statistics

Panel A: Full Sample ( n= 1,224) Variable Mean Continuous Variables PADCA -0.023 absPADCA LAG1CA LEVERAGE LNMVE PBRATIO Discrete variables AUIND FD LITIGATION_RISK POSITIVE Panel B: Stress Sample ( n= 234) Variable Mean Continuous Variables PADCA absPADCA LAG1CA LEVERAGE LNMVE PBRATIO Discrete variables AUIND LITIGATION_RISK POSITIVE Panel C: Non-Stress Sample ( n= 990) Variable Mean Continuous Variables PADCA absPADCA LAG1CA LEVERAGE LNMVE PBRATIO Discrete variables AUIND -0.018 0.043 -0.019 0.522 7.385 2.543 -0.045 0.086 -0.079 0.589 5.987 1.297 0.051 -0.030 0.535 7.118 2.305 Median -0.016 0.031 -0.017 0.545 7.058 1.787 coded as one 74.5% 19.1% 31.0% 33.3% Std Dev 0.076 0.064 0.084 0.232 1.565 2.612 Minimum -0.333 0.000 -0.406 0.078 3.325 -8.438 Maximum 0.228 0.345 0.187 1.193 11.475 15.630
Median -0.020 0.054 -0.047 0.572 5.932 1.083 coded as one 76.9% 44.4% 35.9%
Std Dev 0.114 0.092 0.118 0.287 1.525 2.251
Minimum -0.333 0.000 -0.406 0.078 3.325 -8.438
Maximum 0.228 0.345 0.187 1.193 10.208 15.630
Median -0.015 0.027 -0.013 0.537 7.261 1.927 coded as one 73.9%
Std Dev 0.063 0.053 0.069 0.215 1.452 2.636
Minimum -0.333 0.000 -0.406 0.078 3.325 -8.438
Maximum 0.228 0.345 0.187 1.193 11.475 15.630
31
LITIGATION_RISK POSITIVE PADCA absPADCA LAG1CA
27.8% 32.6% The signed discretionary current accruals measure controlling for performance using the portfolio match technique. Absolute value of PADCA. Last years current accruals. current accruals equals net income before extraordinary items (Compustat data item123) plus depreciation and amortization (Compustat data item 125) minus operating cash flows (Compustat data item 308) scaled by beginning of year total assets. A firms total assets (Compustat data item 6) less stockholders equity of common shareholders (Compustat data item 60) divided by total assets. Natural logarithm of a firms market value of equity. A firms market value of equity is calculated as its price per share at fiscal year end (Compustat data item 199) times the number of shares outstanding (Compustat data item 25). Price to book ratio. Market value at the fiscal year-end divided by book value of common equity. = 1 for firms with independent audit committees, 0 otherwise. An audit committee is considered to be independence if 100% of committee members are independent. = 1 for firms in financial distress, 0 otherwise. A firm is considered as distressed if its operating cash flow has been negative for two consecutive years, or it incurred losses for two consecutive years, or both. = 1 for firms in a high litigation industry, and 0 otherwise. High litigation industries are industries with SIC codes of 28332836 (Pharmaceutical), 35703577 (Computer), 36003674 (Electrical and Telecommunication), 52005961 (Retailer and Wholesaler), and 73707374 (Programming and Software), 0 otherwise. = 1 for firms with positive PADCA, 0 otherwise.
LEVERAGE
LNMVE PBRATIO AUIND
FD
LITIGATION_RISK POSITIVE
32
Table 3: Univariate Analyses Results

Panel A: Financially Distressed Firms vs. Non-distressed Firms (Full sample) T-test for means difference (pt-stat. value) Wilcoxon rank-sum test for distribution difference z-value (p-value)
Firms with positive PADCA Firms with negative PADCA Total
absPADCA-Mean [ absPADCA-Median] [No. of obs.] non-distressed firms distressed firms 0.059 0.039 [0.037] [0.020] [84] [323] 0.101 0.045 [0.068] [0.032] [150] [667] 0.086 0.043 [0.054] [0.027] [234] [990]
2.448
(0.02)
2.655
(0.01)
6.81
(0.00)
6.935
(0.00)
6.95
(0.00)
6.951
(0.00)
Panel B: Firms with Independent vs. Non-independent Audit Committees (full sample) T-test for means difference (pt-stat. value) Wilcoxon rank-sum test for distribution difference z-value (p-value)
Distressed firms
Non-distressed firms Total
absPADCA-Mean [ absPADCA-Median] [No. of obs.] Firms with Firms without independent independent audit committee audit committee 0.079 0.111 [0.048] [0.072] [180] [54] 0.041 0.048 0.028 0.025 [732] [258] 0.049 0.059 0.031 0.030 [912] [312]
-2.13
(0.04)
2.534
(0.01)
-1.65
(0.10)
0.085
(0.93)
-2.26
(0.02)
0.934
(0.35)
33
Table 4 Regression Results on the Association between Discretionary Accruals and Financial Distress Using 100% as Threshold for Audit Committee Independence
Model 1:
absPADCA = + 1 AUIND + 2 FD + 3 LAG1CA + 4 LEVERAGE + 5 LITIGATION _ RISK + 6 LNMVE + 7 PBRATIO +

Model 2:
Model 3:
N=1,224 Predicted sign Coefficient 0.090 + + + + ? -0.010 0.033 -0.073 -0.018 0.012 -0.005 0.001 Model 1 T-stat. 8.99 -2.62 6.52 -3.37 -2.19 2.99 -4.10 1.77 (p-value) (0.00) (0.01) (0.00) (0.00) (0.03) (0.00) (0.00) (0.08) Coefficient 0.087 -0.006 0.052 -0.070 -0.018 0.012 -0.005 0.001 -0.024 Model 2 T-stat. 8.66 -1.40 5.45 -3.19 -2.18 3.04 -4.17 1.87 -2.34 (p-value) (0.00) (0.16) (0.00) (0.00) (0.03) (0.00) (0.00) (0.06) (0.02) Coefficient 0.089 -0.006 0.051 -0.072 -0.016 0.013 -0.005 0.001 -0.012 -0.009 -0.034 Model 3 T-stat. 8.76 -1.47 5.47 -3.31 -2.00 3.22 -4.15 1.78 -1.06 -2.21 -3.39 (p-value) (0.00) (0.14) (0.00) (0.00) (0.05) (0.00) (0.00) (0.08) (0.29) (0.03) (0.00)
Intercept AUIND FD LAG1CA LEVERAGE LITIGATION_RISK LNMVE PBRATIO FDAUIND POSITIVE POSITIVE FDAUIND
34
Table 4 Cont. adjusted R2 absPADCA LAG1CA 11.4% 11.8% 13.5%
The absolute value of the performance-matched discretionary current accruals. Last years current accruals. current accruals equals net income before extraordinary items (Compustat data item123) plus depreciation and amortization (Compustat data item 125) minus operating cash flows (Compustat data item 308) scaled by beginning of year total assets. A firms total assets (Compustat data item 6) less stockholders equity of common shareholders (Compustat data item 60) divided by total assets. Natural logarithm of a firms market value of equity. A firms market value of equity is calculated as its price per share at fiscal year end (Compustat data item 199) times the number of shares outstanding (Compustat data item 25). Price to book ratio. Market value at the fiscal year-end divided by book value of common equity. = 1 for firms with independent audit committees, 0 otherwise. An audit committee is considered to be independence if 100% of committee members are independent. = 1 for firms in financial distress, 0 otherwise. A firm is considered as distressed if its operating cash flow has been negative for two consecutive years, or it incurred losses for two consecutive years, or both. = 1 for firms in a high litigation industry, and 0 otherwise. High litigation industries are industries with SIC codes of 28332836 (Pharmaceutical), 35703577 (Computer), 36003674 (Electrical and Telecommunication), 52005961 (Retailer and Wholesaler), and 73707374 (Programming and Software), 0 otherwise. An interaction term between FD and AUIND. = 1 for firms with signed PADCA >0, 0 otherwise. An interaction term among POSITIVE, FD, and AUIND.
LEVERAGE
LNMVE PBRATIO AUIND FD LITIGATION_R ISK FDAUIND POSITIVE POSITIVE FDAUIND
35
Model 1:

Model 2:
Model 3:
36
LEVERAGE
37
Model 1:

Model 2:
Model 3:
38
LEVERAGE
39
Appendix A Procedures for Calculation of REDCA The calculation of REDCA is started with estimating the following cross-sectional current accrual regression by each two-digit SIC code: CA = 0 + 1 (1 / lag1asset ) + 2 ( Re v) + 3 Lag1ROA Where : CA = current accruals, reflected by net income before extraordinary items (Compustat date item # 123) plus depreciation and amortization (Compustat data item # 125) minus operating cash flows (Compustat data item # 308) scaled by the beginning of year total assets. Lag1asset= total assets at the beginning of the fiscal year. Rev= net sales (Compustat data item #12) in year t less net sales in year t-1 scaled by the beginning of the year total assets. Lag1ROA= last years return on assets. (8)
Industries with fewer than 15 firms are deleted. The parameters estimated from equation (8) are used to calculate the expected current accruals (ECA): ECA = 0 + 1 (1 / lag1asset ) + 2 ( Re v AR) + + 3 Lag1ROA Where: AR = accountings receivable (Compustat item #2) in year t less accounts receivable in year t-1, scaled by the beginning of year total assets. Discretionary current accruals are calculated as: REDCA = CA ECA (10) (9)
40
Our inquiry from the Corporate Library indicates that the committee members are defined as independent if the board member is an independent outside director as defined under the Sarbanes-Oxley Act. Altman (1968) reported that firms with Z-score smaller than 1.81 clearly fall into the bankruptcy group.
41

Financial Distress

Enviado por

Dados do documento

Direitos autorais

Formatos disponíveis

Compartilhar este documento

Compartilhar ou incorporar documento

Opções de compartilhamento

Você considera este documento útil?

Este conteúdo é inapropriado?

Direitos autorais:

Formatos disponíveis

Financial Distress

Enviado por

Direitos autorais:

Formatos disponíveis

Financial Distress and Earnings Management: Effectiveness of Independent Audit Committees

Financial Distress and Earnings Management: Effectiveness of Independent Audit Committees

Financial Distress and Earnings Management: Effectiveness of Independent Audit Committees

RESEARCH DESIGN Hypothesis

dZ , where y is calculated based upon the

Table 1 Sample Selection and Distribution by Independent Audit Committees

Total 961 (78.5%) 263 (21.5%) 1,224 (100%)

Table 2 Descriptive Statistics

Std Dev 0.114 0.092 0.118 0.287 1.525 2.251

Minimum -0.333 0.000 -0.406 0.078 3.325 -8.438

Maximum 0.228 0.345 0.187 1.193 10.208 15.630

Std Dev 0.063 0.053 0.069 0.215 1.452 2.636

Minimum -0.333 0.000 -0.406 0.078 3.325 -8.438

Maximum 0.228 0.345 0.187 1.193 11.475 15.630

LITIGATION_RISK POSITIVE PADCA absPADCA LAG1CA

LNMVE PBRATIO AUIND

Table 3: Univariate Analyses Results

Firms with positive PADCA Firms with negative PADCA Total

Non-distressed firms Total

absPADCA = + 1 AUIND + 2 FD + 3 LAG1CA + 4 LEVERAGE + 5 LITIGATION _ RISK + 6 LNMVE + 7 PBRATIO +

Table 4 Cont. adjusted R2 absPADCA LAG1CA 11.4% 11.8% 13.5%

LNMVE PBRATIO AUIND FD LITIGATION_R ISK FDAUIND POSITIVE POSITIVE FDAUIND

absPADCA = + 1 AUIND + 2 FD + 3 LAG1CA + 4 LEVERAGE + 5 LITIGATION _ RISK + 6 LNMVE + 7 PBRATIO +

Table 4 Cont. adjusted R2 absPADCA LAG1CA 11.8% 12.2% 13.9%

LNMVE PBRATIO AUIND FD LITIGATION_R ISK FDAUIND POSITIVE POSITIVE FDAUIND

absPADCA = + 1 AUIND + 2 FD + 3 LAG1CA + 4 LEVERAGE + 5 LITIGATION _ RISK + 6 LNMVE + 7 PBRATIO +

Table 4 Cont. adjusted R2 absPADCA LAG1CA 12.0% 12.7% 14.3%

LNMVE PBRATIO AUIND FD LITIGATION_R ISK FDAUIND POSITIVE POSITIVE FDAUIND

Você também pode gostar