Managerial Auditing Journal

Emerald Article: Audit quality and earnings management for Taiwan IPO firms Ken Y. Chen, Kuen-Lin Lin, Jian Zhou

Article information:
To cite this document: Ken Y. Chen, Kuen-Lin Lin, Jian Zhou, (2005),"Audit quality and earnings management for Taiwan IPO firms", Managerial Auditing Journal, Vol. 20 Iss: 1 pp. 86 - 104 Permanent link to this document: http://dx.doi.org/10.1108/02686900510570722 Downloaded on: 07-04-2012 References: This document contains references to 34 other documents To copy this document: permissions@emeraldinsight.com This document has been downloaded 3133 times.

Access to this document was granted through an Emerald subscription provided by University of Pretoria For Authors: If you would like to write for this, or any other Emerald publication, then please use our Emerald for Authors service. Information about how to choose which publication to write for and submission guidelines are available for all. Additional help for authors is available for Emerald subscribers. Please visit www.emeraldinsight.com/authors for more information. About Emerald www.emeraldinsight.com With over forty years' experience, Emerald Group Publishing is a leading independent publisher of global research with impact in business, society, public policy and education. In total, Emerald publishes over 275 journals and more than 130 book series, as well as an extensive range of online products and services. Emerald is both COUNTER 3 and TRANSFER compliant. The organization is a partner of the Committee on Publication Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive preservation.
*Related content and download information correct at time of download.

The Emerald Research Register for this journal is available at www.emeraldinsight.com/researchregister

The current issue and full text archive of this journal is available at www.emeraldinsight.com/0268-6902.htm

MAJ 20,1

Audit quality and earnings management for Taiwan IPO rms

Ken Y. Chen
College of Management, National Cheng Kung University, Tainan, Taiwan

86

Kuen-Lin Lin
School of Management, Cheng Shiu University, Kaohsiung, Taiwan, and

Jian Zhou
School of Management, SUNY at Binghamton, Binghamton, New York, USA
Abstract
Purpose This paper investigates the relationship between audit quality (as measured by auditor size and industry specialization) and earnings management (as measured by unexpected accruals) for Taiwan IPO rms. Design/methodology/approach First uses unexpected accruals in the modied Jones model to measure earnings management in the IPO process. Then uses auditor type (big ve versus non-big ve) and industry specialist to measure audit quality. The hypothesis predicts that Taiwanese rms with higher quality auditors engage less in earnings management in the IPO process. The sample consists of 367 new issues between 1999 and 2002 from the Taiwan Economic Journal database. Findings It is found that big ve auditors are related to less earnings management in the IPO year in Taiwan. This shows that higher quality auditors constrain earnings management for Taiwan IPO rms. Research limitations/implications The nding shows that high quality auditors constrain earnings management and provide more precise information. This is important, given that management has incentive to engage in earnings management in the IPO process to garner greater proceeds and at-issue earnings management is negatively related to post-issue earnings performance and stock returns. Practical implications The research might be of interest to investors in IPO rms, given that at-issue unexpected accruals are opportunistic. Originality/value The study contributes to the literature in that it shows that audit rm size is an important determinant in earnings management for Taiwan IPO rms. Keywords Quality audit, Earnings, Taiwan Paper type Research paper

Managerial Auditing Journal Vol. 20 No. 1, 2005 pp. 86-104 q Emerald Group Publishing Limited 0268-6902 DOI 10.1108/02686900510570722

I. Introduction Going public is a major corporate event for a company to raise additional capital to fund its growth and enhance the entrepreneurs personal wealth, where an offering prospectus including externally audited nancial statements for up to a certain period (different among different jurisdictions) is needed to be prepared in the initial public offering (IPO) process. Accrual accounting provides management with discretion in the reporting of earnings, and thus provides opportunity for managers to engage in earnings management in an IPO. Earnings management in general and earnings management in the IPO process in particular, have attracted growing attention in the accounting research[1]. The current study is motivated by a study by Zhou and Elder (2003), which

nds that big ve auditors and industry specialist auditors constrain earnings management for IPO rms in the US. Given the different environments across countries and regions, it is useful to conduct an earnings management study for IPO rms besides the US. Therefore, the current study investigates whether the relationship between audit quality and earnings management in the scal year of IPO exists in Taiwan. Audit quality has become a concern after scandals such as Enron, WorldCom, and Ahold, etc. These scandals raised concerns about audit quality even among the big ve accounting rms, which are normally considered the premier accounting rms and associated with higher audit quality. Among all these scandals, the Enron scandal attracted the greatest attention partly because it was associated with the collapse of Arthur Andersen. We briey summarize the Enron and Andersen events as follows. On October 16, 2001, Enron announced that third-quarter earnings would include an unexpected nonrecurring charge of $1.01 billion after tax, which triggered the informal inquiry of SEC the next day. SEC changed its informal inquiry into a formal investigation on October 31. On December 2, Enron led for chapter 11 bankruptcy protection. On January 10, 2002, Andersen notied the SEC and the Department of Justice that the Houston ofce had shredded a signicant number of documents related to the Enron audit. On February 2, the Powers report provided by a special investigation committee of Enrons board of directors was released suggesting that the headquarters of Andersen were aware of the problems with Enron audit. Andersen was indicted for obstruction of justice on March 14 and was found guilty on June 15, 2002. Andersen ceased conducting audits of public companies by August 31, 2002. The increased lack of investor condence in nancial statement information resulting from these corporate scandals involving once well-respected companies such as Enron and WorldCom, and auditors such as Arthur Andersen served as a catalyst for the Sarbanes-Oxley Act that was signed into law on July 30, 2002. The primary goal of the Sarbanes-Oxley Act is to help restore investors condence in the integrity of nancial information, and auditors play an important role in the process. So it is important to investigate whether there is audit quality difference, especially during IPO process where rms are found to engage in opportunistic earnings management. We choose 1999-2002 as the sample period since IPOs are relatively clustered for this period. The average number of IPOs is around 90 each year for this period compared to the average number of 40 for 1991-1998. The Enron asco and the enacted Sarbanes-Oxley Act of 2002 may not affect the IPO market in Taiwan for our current study, because the regulation and procedures with regard to the IPO rms have been enacted before the Enron scandal. However, the demand for higher auditor quality can be reasonably expected for Taiwanese companies listed in three major stock exchanges in the post-Enron era. There are signicant differences in the IPO process between US and Taiwan with regard to the listing procedures and regulations. First, there exists procedural difference for a rm to go public and to be listed in the Taiwan Stock Exchange (TSE) compared to being listed in a US stock exchange. In the US, the listing stock exchange is already decided when a rm goes public. However, it is not the same in Taiwan. A rm goes public rst, then the rm les registration statements to the Securities and Futures Commission[2] (SFC, the SEC counterpart in Taiwan). If it meets the listing requirements of the Taiwan Stock Exchange, it will be listed there. Second, there are also regulation differences with regard to the audited nancial statements in the

Audit quality for Taiwan IPO rms 87

MAJ 20,1

88

registration statements. The auditor report in the registration statements to SFC in Taiwan is signed by two audit partners in addition to a signature representative of the audit rm, whereas only a signature representative of the audit rm is required in US. Third, prior studies suggest that in well-developed capital markets such as the US, there is information asymmetry between management and investors (Leland and Pyle, 1977), and between informed and uninformed investors (e.g. Rock, 1986; Beatty and Ritter, 1986), and such information asymmetry is a necessary condition for earnings management. The capital market is less developed and the average listed company size is much smaller in Taiwan than in the US major stock exchanges, so the information intermediary role played by the auditor is maybe even more important in Taiwan. The three areas of differences between the US and Taiwan make it interesting to examine whether audit quality constrains earnings management in Taiwan. Earnings management in the IPO process is of particular concern for several reasons. First, management has incentives to engage in income increasing earnings management to ensure that the issue is fully subscribed and/or priced higher to garner greater proceeds, because their compensation and/or reputation depend on the success of the IPO. Second, at the issuing stage, earnings management is found to be negatively related to post issue earnings performance (Teoh et al., 1998b) and post issue stock returns (Teoh et al., 1998a). As a result, at the issuing stage, earnings management has signicant resource allocation implication. Third, APB 20 allows IPO rms to change accounting principles in the prospectus as long as nancial statements of previous years are restated. This may give management an opportunity to engage in earnings management. Fourth, there is signicant information asymmetry between the owners-managers and investors (Leland and Pyle, 1977), and between informed and uninformed investors (Rock, 1986; Beatty and Ritter, 1986). Theoretical research shows that auditors play an important role in reducing the adverse impact of information asymmetry in the IPO process. Titman and Trueman (1986) develop a model in which the price of shares in an IPO is increasing in tandem with the quality of information provided by the offering company. Datar et al. (1991) nd that the information asymmetry in the IPO process is mitigated by the role of auditor and audit quality. The choice of auditor is made jointly with other decisions such as the percentage of retained ownership in the offering (Copley and Douthett, 2002). Empirical evidence indicates an increased demand for audit quality at the time of the IPO; companies frequently change to a big ve auditor at the time of an IPO (Carpenter and Strawser, 1971; Menon and Williams, 1991)[3]. Audit quality research has focused primarily on differences between big ve and non-big ve rms. Research in the Australian audit market (Craswell et al., 1995) indicates that industry specialist auditors receive a fee premium that represents a signicant portion of the premium to big ve rms in the Australian audit market[4,5]. Elder (1999) nds that IPO underpricing is lower for companies that use an industry specialist auditor. This evidence indicates that industry specialist auditors provide higher quality of audits compared with non-industry specialist auditors. Becker et al. (1998) nd that unexpected accruals are reduced when existing publicly-traded companies use a big ve auditor. They nd that clients of non-big ve auditors report unexpected accruals that are higher than unexpected accruals of clients of big ve auditors. They interpret this as indicating that lower audit quality is associated with greater accounting exibility.

Our study differs from Becker et al. (1998) in several respects. First, we use an industry specialist variable in addition to a big ve/non-big ve measure. Previous literature nds that audit rms devote signicant resources to the development of industry expertise (Craswell et al., 1995) and audit rms promote themselves as industry specialists. Second, we focus on a setting where the direction of earnings management is clear, specically, IPO companies have incentives to engage in income increasing behavior. In contrast, non-IPO companies may not always engage in income increasing behavior. For example, DeFond and Park (1997) nd that rms engage in income smoothing behavior because of managers job security concerns[6]. Third, there are signicant resource allocation concerns in the IPO process. Managers/owners can benet from higher IPO proceeds; however investors may make incorrect decisions based on reported earnings. Fourth, unexpected accruals have been found on average to be opportunistic for IPO companies. Teoh et al. (1998b) nd that earnings at the time of the IPO are high because of high unexpected accruals, and earnings after the issue are low because the earlier high unexpected accruals are not sustained. Teoh et al. (1998a) nd that at-issue earnings management is signicantly negatively related to post-issue stock returns. Using a sample of IPO rms from 1996-1998 from the December 1998 Compact D New Issue database, Zhou and Elder (2003) nd that unexpected accruals for IPO rms are lower when big ve auditors are used, suggesting that the big ve auditors are associated with reduced management discretion over earnings. They also nd that rms audited by industry specialist auditors engage in less earnings management. Similar to Zhou and Elder (2003), we seek to address these two research questions: (1) Do big ve auditors constrain earnings management in the IPO process? (2) Do industry specialist auditors provide higher quality of audits in the IPO process as evidenced by less earnings management? Using 367 IPO observations satisfying all the data requirements from 1999-2002, we nd that unexpected accruals for IPO rms are lower when big ve auditors are used, suggesting that the big ve auditors are associated with reduced management discretion over earnings. However, we do not nd that rms audited by industry specialist auditors engage in less earnings management. The remainder of the paper is organized as follows. The next section describes earnings management in the IPO environment and reviews research on audit quality and introduces the research hypotheses. Section III describes the research design. Sample selection and tests of the relation between auditor reputation and unexpected accruals are discussed in section IV. Section V is the summary and conclusion. II. Earnings management, audit quality and IPOs Earnings management and initial public offerings An extensive body of earnings management literature has developed (see Healy and Wahlen (1999) for a review of the literature). Most earnings management studies examine whether companies manage earnings in response to some economic incentive. One setting where management has an incentive to manipulate earnings is at the time of an IPO, since greater earnings may be reected in a higher offering price and greater proceeds to the company and offering shareholders. Whether a company benets from earnings management depends upon whether the market can see through the earnings manipulation. The IPO environment is

Audit quality for Taiwan IPO rms 89

MAJ 20,1

90

characterized by information asymmetry between management and investors (Leland and Pyle, 1977), and between informed and uninformed investors (Rock, 1986; Beatty and Ritter, 1986). Several analytical models demonstrate that the extent of earnings management increases with the level of information asymmetry. For example, Dye (1988) and Trueman and Titman (1988) demonstrate that the existence of information asymmetry between management and shareholders is a necessary condition for earnings management, because shareholders cannot perfectly observe a rms performance and prospects in an environment in which they have less information than management. In such an environment, management can use its exibility to manage reported earnings. Furthermore, managements discretionary ability to manage earnings increases as the information asymmetry between management and shareholders increases. Richardson (2000) provides empirical evidence consistent with this line of reasoning. He nds that the extent of information asymmetry, as measured by the bid-ask spread and the dispersion in analysts forecasts, is positively related to the degree of earnings management. The information asymmetry in the IPO environment creates an opportunity for management to engage in earnings management because it is difcult for related stakeholders (especially shareholders) to undo this behavior. Teoh et al. (1998b) evaluate whether accounting accrual choices during an IPO are informative to investors or opportunistic. They nd evidence consistent with opportunism. They nd that the net income of IPO rms are signicantly higher during the issuing year relative to subsequent years, and to non-issuing industry peers. IPO rms are able to report high earnings during the IPO by reporting unexpected accruals aggressively. The IPO rms earnings under-perform relative to their matched rms and non-issuing industry peers when high unexpected accruals at issue cannot be sustained in the following periods. More importantly, they nd that unexpected accruals explain the post-issue underperformance in earnings. In a related paper, Teoh et al. (1998a) nd that at issue earnings management is negatively related to post issue stock returns. A possible interpretation of their ndings is that at issue earnings management helps the company receive a higher issuing price. When at issue earnings performance attributable to earnings management cannot be sustained in the following periods, this is negatively reected in the stock price. Audit quality and earnings management Auditor rm size. In Taiwan, the big ve audit rms are T.N. Soong & Co. (member of Arthur Andersen), PricewaterhouseCoopers, KPMG, Deloitte and Touche, and Diwan Ernst & Young. Due to the Enron scandal and ceased operation of Andersen on August 31, 2002, T.N. Soong & Co. merged into Deloitte and Touche and became Deloitte and Touche on June 1, 2003. The clients of T.N. Soong were also transferred to Deloitte and Touche, which is different from the situation in the US, where Andersens clients chose other big ve or non-big ve audit rms as successor auditors. Given quality monitoring systems which operate across all accountancy forms, the clients demand for auditor reputation and auditor quality should not have changed, since all Andersens clients in Taiwan were transferred to Deloitte and Touche. Becker et al. (1998) nd that companies with non-big ve auditors (a proxy for lower audit quality) report unexpected accruals that signicantly increase income compared to companies with big

ve auditors. In Taiwan, non-big ve auditors, especially for those local or regional audit rms instead of national audit rms, are usually recognized as lower quality auditors when compared to big ve auditors, which can be evidenced by the clients selection of big ve auditors for their IPO process. They also nd that managers respond to debt contracting and income-smoothing incentives by strategically reporting unexpected accruals. In addition, companies with incentives to smooth earnings upwards (downwards) report signicantly greater income-increasing (decreasing) unexpected accruals when they have non-big ve auditors. Francis et al. (1999) argue that high-accrual rms have greater opportunity for opportunistic earnings management and have an incentive to hire a big ve auditor to provide assurance that earnings are credible. They nd that high accrual rms are more likely to hire a big ve auditor, but report lower unexpected accruals, consistent with big ve auditors constraining opportunistic reporting of accruals. The Becker et al. (1998) and Francis et al. (1999) studies provide evidence in non-IPO settings that higher quality auditors are associated with reduced levels of earnings management. Previous research suggests that the auditor can play a role in reducing information asymmetry at the time of the IPO. Balvers et al. (1998) and Hogan (1997) nd that big ve auditors are associated with lower under-pricing of the offering. Balvers et al. (1998) argue that a high quality auditor provides better information about earnings, which makes it easier for the investment banker to price the issue correctly and preserve reputation quality. This information argument is consistent with the model in Titman and Trueman (1986), in which the price of the shares in an IPO is increasing with the quality of the information provided by the offering company, which they argue is partially determined by the quality of the auditor. Evidence that big ve auditors are associated with lower unexpected accruals for IPO rms would further support the information hypothesis. Zhou and Elder (2003) provide evidence supporting the information hypothesis that audit quality provided by big ve audit rms is an important constraint in earnings management in the IPO process in the US. Therefore, we expect that IPO companies in Taiwan that use big ve audit rms will engage in less earnings management than IPO companies with non-big ve auditors: H1. IPO rms in Taiwan audited by big ve audit rms engage less in earnings management than rms audited by non-big ve auditors. Auditor industry specialization. Recent audit quality research has focused on the role of auditor industry specialization. Hogan and Jeter (1999) nd that measures of specialization have increased in both regulated and unregulated industries, consistent with returns to specialization. Craswell et al. (1995) argue that audit rms market themselves in terms of both a general reputation and industry expertise. In a test of audit fees in the Australian audit market, they nd that industry specialists receive a signicant fee premium, and that this fee premium is a signicant component of the fee premium received by big ve rms. Industry specialization is acknowledged in DeAngelo (1981) as one possible reason for the selection of big ve auditors by IPO companies. In the Titman and Trueman (1986) model, in which the pricing of an IPO is increasing with the quality of information associated with the expertise of the auditor, they suggest that industry knowledge is one element of auditor expertise. Elder (1999) nds that IPO

Audit quality for Taiwan IPO rms 91

MAJ 20,1

under-pricing is lower for companies that use an industry specialist auditor. Zhou and Elder (2003) nd that auditor industry specialists can be used to constrain earnings management in the IPO process in the US. Because of the expertise and experience of industry specialists, we also expect that industry specialists are likely to constrain earnings management in the IPO process. This leads into the second hypothesis: H2. Firms audited by industry specialist engage less in earnings management in the initial public offering process in Taiwan. The following section describes the research design, including the model used to estimate unexpected accruals and models used to test the research hypotheses. III. Research design Unexpected accruals We rst use unexpected accruals to measure earnings management in the IPO process. Unexpected accruals (also called discretionary accruals) are used in earnings management studies such as Jones (1991) and Subramanyam (1996). Dechow et al. (1995) provide evidence that the modied Jones model is the most powerful to detect earnings management among the alternative models to measure unexpected accruals. The model is estimated as follows: TACC it DCAit 2 DCASH it 2 DCLit 2 DSTDit 2 DEP it =TAit21 TACC it a1 1=TAit21 a2 DREV it 2 DREC it =TAit21 a3 PPE it =TAit21 1it where: TACC it total accruals for rm i in year t, dened as above. change in current assets for rm i in year t. change in current liabilities for rm i in year t. change in short-term debt for rm i in year t. change in depreciation for rm i in year t. change in revenue for rm i in year t. change in receivables for rm i in year t. net property, plant and equipment for rm i in year t. total assets for rm i in year t-1.

92

DCAit DCLit DSTDit

DEP it

DCASH it change in cash for rm i in year t.

DREV it DREC it
PPE it TAit21

This equation is estimated cross-sectionally each year for each two-digit Taiwan Economic Journal (TEJ) code industry using all available rms. At least ten rm year observations are required in a two-digit TEJ code. The residual from the regression is the unexpected accruals. Normal levels of working capital accruals related to sales are controlled through the changes in revenue adjusted for changes in accounts receivable. Normal levels of depreciation expense and related deferred tax accruals are controlled through the gross property, plant and equipment. In addition the total assets of the

previous period are used as a deator to control for potential scale bias. The cross-sectional model reects common industry factors applied to unexpected accruals. As a result, estimated unexpected accruals are more likely to reect managements choice rather than industry factors. Also, since the model is estimated year-by-year, changes in industry conditions are also factored in the model. Auditor industry specialization Craswell et al. (1995), and Ferguson and Stokes (2002) use the percentage of the audit rms share of total industry audit fees as an industry specialist measure. This measure incorporates size weighting into market share (weighted by audit fees). They dene an auditor to be an industry specialist if the auditor attains a 10 percent market share from either or both of these two measures. Since the audit fee data is not available to us, we cannot use this procedure and instead use a sales-based industry specialist measure. Since the Taiwan market is comparatively smaller than the US market, we require there to be at least ten rms in a two-digit TEJ code industry for an auditor to qualify as an industry specialist, rather than the 30 US rms used in Zhou and Elder (2003). Audit rm industry specialization is dened as follows:
J ik q X Aijk

Audit quality for Taiwan IPO rms 93

MS ik

j1 Ik X J ik q X Aijk i1 j1

where: Aijk total sales of client rm j in industry k audit by auditor i. i 1; 2; . . . ; I an index for audit rms. j 1; 2; . . . ; J an index for client rms. k 1; 2; . . . ; K an index for client industry. Ik J ik the number of audit rms i in industry k. the number of clients served by audit rm i in industry k.

Industry specialists are calculated yearly based on the rm-year observations from the TEJ database. When auditor js market share is greater than 15 percent in a two-digit TEJ code industry, the auditor j is treated as an industry specialist. Approach to testing Our hypotheses relate earnings management as measured by unexpected accruals to audit quality as measured by auditor type and industry specialization. Many other variables may play a role in managements unexpected accruals decision in the IPO process. Becker et al. (1998) nd that operating cash ows are signicantly different for rms audited by big ve versus rms audited by non-big ve rms. The absolute value of total accruals is used as a control variable because Becker et al. (1998) provide evidence that this is signicantly negatively related to unexpected accruals. Also

MAJ 20,1

94

Francis et al. (1999) nd that the likelihood of using a big ve auditor is increasing in rms endogenous propensity for accruals. Burgstahler and Dichev (1997) nd that rms manage reported earnings to avoid reporting earnings decreases and losses. Accordingly, loss and income change indicator are added as control variables to account for managers incentive to avoid earnings decreases and losses. Market-to-book value at the end of the IPO offering year is used as a surrogate for growth opportunity because information communication might play a role in some managers earnings management decisions, even though Teoh et al. (1998b) nd that managers use unexpected accruals opportunistically in the IPO process. Managers may try to signal the rms future prospects through unexpected accruals. The log of sales is used as an independent variable to control for the possible effect of size on earnings management in the IPO process. Large rms may have less incentive to engage in earnings management because they are subject to more scrutiny from nancial analysts and investors. Leverage may also be associated with earnings management in the IPO process. DeFond and Jiambalvo (1994) and Sweeney (1994) nd that managers use unexpected accruals to satisfy debt covenant requirements. The electronics industry (ELEC) is the largest industry in Taiwan, and the matter of earnings management is more likely to be a concern. Therefore, ELEC is also used as a control variable. Further, Taiwan Stock Exchange (TSE) is used as a control variable because of the requirements to be a listed rm in TSE is stricter than the rms in the over-the-counter market, and thus earnings management is more likely to be constrained in TSE. Inclusion of these control variables results in the following regression model: DAC it b0 b1 BIG5 b2 SPEC it b3 OCF it b4 ABSTAit b5 LOSS it b6 INCCHGit b7 MTBit b8 SIZE it b9 LEV it b10 ELEC b11 TSE 1it DAC it BIG5it SPEC it OCF it LOSS it unexpected accruals. 1 if the auditor is member of big ve; 0 otherwise. 1 if the auditor is an industry specialist; 0 otherwise. operating cash ow deated by lagged total assets. 1 if the rm incurs a loss; 0 otherwise.

ABSTAit absolute value of total accruals. INCCHGit 1 if this years income is greater than previous years income; 0 otherwise. MTBit SIZE it LEV it market to book ratio. log of total sales. leverage, dened as total liabilities over total assets.

ELEC it TSE it

1 if the rm is member of electronic industry; 0 otherwise. 1 if the rm is member of TSE; 0 otherwise.

The model is estimated using unexpected accruals (DAC) as the dependent variable. The main research variables of interest are BIG5 and SPEC. Consistent with the two research hypotheses, the coefcients on these two variables are predicted to be negative because rms audited by big ve auditors and industry specialists are expected to engage in less earnings management in the IPO process.

Audit quality for Taiwan IPO rms 95

IV. Sample selection and results The sample consists of 367 observations of new issues between 1999 and 2002 from the Taiwan Economic Journal (TEJ) database satisfying the following criteria: . IPO date, and the auditor for the IPO are available from the database; and . necessary data for calculating total accruals, unexpected accruals, industry specialist, market-to-book ratio and leverage are available from the database. Table I provides details about the sample selection. After deleting 21 nancial services and insurance companies (as in previous literature, the unexpected accruals does not apply to nancial industries) and 42 companies with missing data, the nal sample between 1999 and 2002 amounted to the 367 rm-year observations. Table II provides details about the sample distribution by year and by auditor type. The number of IPO companies is around 80 each year from 1999 to 2001, down to 56 in the year 2002. The big ve auditors audit more than 80 percent of the IPOs in any given year, except for the year 2002, which is around 90 percent of the IPOs. Table III provides the industry distribution of the sample rms. The sample includes 17 separate TEJ industry codes, indicating a wide distribution of industries. Electronics industry has the largest concentration of IPOs, with more than 66 percent of the total observations. The remaining sample rms are widely distributed across

1999-2002 IPOs Financial services and insurance Missing data Total

430 21 42 367 Table I. Sample selection

Note: Sample characteristics for 367 rms conducting initial public offerings during the period of 1999 to 2002 from December 2002 Taiwan Economic Journal (TEJ) database

1999 Year B5 NB5 Total Freq n 84 19 103 % 81.55 18.45 28.06 n 83 24 107

2000 % 77.57 22.43 29.15 n 77 18 95

2001 % 81.05 18.95 25.89 n 56 6 62

2002 % 90.32 9.68 16.90 n 300 67 367

Total % 81.74 18.26 100 Table II. Sample IPO rms by year and by auditor

MAJ 20,1

Industry Food Plastics Textiles Electrical and machinery Chemicals Steel and iron Electronics Constructions Transportations All others Total

TEJ codes 12 13 14 15, 16 17 20 23 25 26 11, 18, 21, 22, 27, 29, 40

Freq. 3 8 21 20 21 7 245 17 4 21 367

% 0.82 2.18 5.72 5.45 5.72 1.91 66.76 4.63 1.09 5.72 100.00

B5 2 7 16 14 14 3 218 10 3 13 300

NB5 1 1 5 6 7 4 27 7 1 8 67

SPEC 2 2 11 3 0 2 159 5 2 7 193

96

Table III. TEJ codes distribution

TEJ industry codes; no other TEJ industry code contains more than 6 percent of the sample rms. Table IV presents the distributions of auditors operating in different industries. As evident from Table IV, not every big ve auditors are industry specialist auditors. For example, all the big ve audit IPO rms in the electronics industry, Deloitte & Touche is not an industry specialist auditor. Table V provides descriptive statistics for the sample. The average total accruals are 0.047. The average unexpected accruals are 0.083 and the median of unexpected accruals is 0.062. Big ve auditors are used by 81.7 percent of the sample rms, and around 53 percent of rms use industry specialist as auditors. The mean and median operating cash ows are 0.046 and 0.058. The mean and median of absolute value of total accruals are 0.294 and 0.153 respectively. The median rm has positive net income and positive income change in the IPO year. The average market-to-book ratio at the end of the offering year is 2.251, and the average log of sales is 14.126. The mean and median leverage are 0.393 and 0.384, respectively. The sample companies from the electronics industry are 66.8 percentage of the sample, and there are 30 percentage of the sample companies listed in the Taiwan Stock Exchange.
Industry Food Plastics Textiles Electrical and machinery Chemicals Steel and iron Electronics Constructions Transportations B5 (Freq.) PWC, DT AA, KPMG, PWC, EY, DT AA, KPMG, PWC, EY, DT AA, KPMG, PWC, AA, KPMG, PWC, AA, KPMG, EY AA, KPMG, PWC, AA, KPMG, PWC, KPMG, PWC EY EY, DT EY, DT EY NB5 (Freq) Other MRI MRI, CH, other BDO, BDO, RSM, BDO, RSM, other other other other RSM, CH, MRI, other other SPEC (Freq) PWC, DT AA, KPMG AA, KPMG, DT KPMG, PWC AA, KPMG AA, KPMG PWC, EY AA, KPMG KPMG

Table IV. Auditors operating in different industries in Taiwan

Notes: AA (merged into Deloitte and Touche, June 1, 2003); KPMG; PWC; Diwan EY; Deloitte and Touche (DT); BDO Taiwan Union & Co.; RSM International; Chien Hsing (CH); Pyramid (MRI_Moores Rowland International); TC (MRI)

Var TAC DAC BIG5 SPEC OCF ABSTA LOSS INCCHG MTB SIZE LEV ELEC TSE

Mean 0.047 0.083 0.817 0.526 0.046 0.294 0.109 0.564 2.251 14.126 0.393 0.668 0.300

STD 0.503 0.504 0.387 0.500 0.170 0.410 0.312 0.497 2.321 1.295 0.155 0.472 0.459

Lower quartile 2 0.113 2 0.084 1 0 2 0.023 0.064 0 0 1.067 13.476 0.269 0 0

Median 0.036 0.062 1 1 0.058 0.153 0 1 1.443 14.042 0.384 1 0

Upper quartile 0.193 0.238 1 1 0.125 0.344 0 1 2.592 14.751 0.516 1 1

Audit quality for Taiwan IPO rms 97

Notes: TAC: total accruals, dened as DCAit 2 DCASH it 2 DCLit 2 DSTDit 2 DEP it =TAit21 ; DAC: unexpected accruals; BIG5: 1 if the auditor is member of big ve; 0 otherwise; SPEC: 1 if the auditor is an industry specialist; 0 otherwise; OCF: operating cash ow deated by lagged total assets; ABSTA: absolute value of total accruals; LOSS: 1 if the rm incurs a loss; 0 otherwise; INCCHG: 1 if this years income is greater than previous years income; 0 otherwise; MTB: market to book ratio; SIZE: log of total sales; LEV: leverage, dened as total liabilities over total assets; ELEC: 1 if the rm is member of electronic industry; 0 otherwise; TSE: 1 if the rm is member of TSE; 0 otherwise

Table V. Variable descriptive statistics (n 367)

Table VI shows the correlation among the dependent and independent variables. Total accruals and unexpected accruals are signicantly positively correlated with each other. The big ve variable is negatively related to total accruals and unexpected accruals. Industry specialist is insignicantly and negatively related to total accruals and unexpected accruals. This suggests that big ve auditors constrain earnings management in the IPO process, although the formal analyses are based on multivariate analyses. As expected, there is a positive correlation (0.50) between the big 5 and industry specialist variables. As described in Table V, 81.7 percent of the sample rms (300/367) audited by big ve rms, and around 53 percent of rms (193/367) use industry specialist as auditors. That is, there are 107 rms out of 300 big ve audited rms are not industry specialists. Tests of the research hypotheses using unexpected accruals in the scal year of IPO offering as the dependent variable are reported in Table VII. The rst two columns report the results using separate big ve and industry specialist auditor reputation measures, and the last column reports the results using both measures. The big ve measure is associated with lower unexpected accruals at the 5 percent level or better in both model specications. These results suggest that big ve auditors are associated with lower unexpected accruals in the scal year of IPO offering, and audit quality plays an important role in reducing earnings management. The coefcient on the industry specialist variable is insignicantly negative when the big ve variable is excluded, and is insignicantly positive when both big ve and SPEC variables are included. This is not consistent with the ndings in Zhou and Elder (2003) and Craswell et al. (1995) that industry specialization is an important element in auditor quality. The possible explanation is that the Taiwan audit market is comparatively smaller than the US, or the industry specialization is not well recognized

98

MAJ 20,1

TAC DAC BIG5 SPEC OCF ABSTA LOSS INCCHG MTB SIZE LEV ELEC TSE 0.08 0.42 0.22 0.00 0.03 0.04 0.00 0.51 0.24 0.24 0.62 0.00 0.77 0.38 0.15 0.63 0.01 0.26 0.58 0.00 0.36 0.06 0.53 0.02 0.15 0.11 0.02 0.72 0.00 0.85 0.55 0.01 0.01 0.00 0.40 0.00 0.47 0.37 0.33 0.09 0.09 0.24 0.86 0.81 0.00 1 2 0.38 2 0.12 2 0.21 0.05 0.10 2 0.11 1 2 0.05 0.09 0.09 2 0.06 0.01 0.01 2 0.24 0.00 0.02 0.00 0.34 0.06 0.03 1 2 0.03 2 0.14 0.13 0.23 0.04 2 0.26 2 0.04 0.05 1 0.10 0.03 0.12 2 0.08 0.08 0.12 2 0.02 0.35 2 0.01 1 0.30 0.23 0.04 0.16 0.17 1 0.50 2 0.02 0.05 0.07 0.03 0.14 0.06 2 0.03 0.27 0.05 0.00 0.00 0.40 0.00 0.00

Notes: TAC: total accruals, dened as DCAit 2 DCASH it 2 DCLit 2 DSTDit 2 DEP it =TAit21 ; DAC: unexpected accruals; BIG5: 1 if the auditor is member of big ve; 0 otherwise; SPEC: 1 if the auditor is an industry specialist; 0 otherwise; OCF: operating cash ow deated by lagged total assets; ABSTA: absolute value of total accruals; LOSS: 1 if the rm incurs a loss; 0 otherwise; INCCHG: 1 if this years income is greater than previous years income; 0 otherwise; MTB: market to book ratio; SIZE: log of total sales; LEV: leverage, dened as total liabilities over total assets; ELEC: 1 if the rm is member of electronic industry; 0 otherwise; TSE: 1 if the rm is member of TSE; 0 otherwise

Table VI. Pearson correlation matrix for dependent and independent variable (n 367)
BIG5 SPEC OCF ABSTA LOSS INCCHG MTB SIZE LEV ELEC TSE 1 0.06 2 0.18 0.26 0.27 0.29 0.00 0.00 0.00 1 0.35 0.17 0.20 0.00 0.00 0.00 1 2 0.06 2 0.16 0.27 0.00 1 0.05 0.39 1

TAC

DAC

1 0.97 2 0.09 2 0.03 2 0.07 0.18 2 0.11 0.11 0.17 0.06 2 0.07 2 0.03 0.02

0.00 0.10 0.54 0.17 0.00 0.03 0.04 0.00 0.26 0.20 0.50 0.74

1 2 0.09 2 0.04 2 0.06 0.20 2 0.11 0.11 0.16 0.03 2 0.06 2 0.06 0.03

DAC INTERCEPT (t-statistic) BIG5 SPEC OCF ABSTA LOSS INCCHG MTB SIZE LEV ELEC TSE Adj. R-sq. 2 0.225 2 0.134 2 0.447 0.222 2 0.066 0.050 0.038 0.035 2 0.365 2 0.121 2 0.011 8.4 2 0.754 2 1.966** 2 2.772*** 3.429*** 2 0.722 0.842 3.044*** 1.506* 2 1.896** 2 2.021** 2 0.171 2 0.293 2 0.002 2 0.441 0.216 2 0.076 0.052 0.036 0.034 2 0.360 2 0.146 2 0.014 7.4

DAC 2 0.982 2 0.039 2 2.692*** 3.314*** 2 0.829 0.869 2.896*** 1.432* 2 1.859** 2 2.340*** 2 0.226 2 0.191 2 0.167 0.060 2 0.473 0.220 2 0.071 0.058 0.039 0.033 2 0.367 2 0.137 2 0.009 8.4

DAC 2 0.634 2 2.186** 0.959 2 2.892*** 3.389*** 2 0.770 0.975 3.071*** 1.404* 2 1.905** 2 2.203** 2 0.148

Audit quality for Taiwan IPO rms 99

Notes: Signicant at the *0.10, **0.05, 0.01 level based on a one-tail test; DAC: unexpected accruals; BIG5: 1 if the auditor is member of big ve; 0 otherwise; SPEC: 1 if the auditor is an industry specialist; 0 otherwise; OCF: operating cash ow deated by lagged total assets; ABSTA: absolute value of total accruals; LOSS: 1 if the rm incurs a loss; 0 otherwise; INCCHG: 1 if this years income is greater than previous years income; 0 otherwise; MTB: market to book ratio; SIZE: log of total sales; LEV: leverage, dened as total liabilities over total assets; ELEC: 1 if the rm is member of electronic industry; 0 otherwise; TSE: 1 if the rm is member of TSE; 0 otherwise

Table VII. Regression of unexpected accruals on auditor size and industry specialization (using 15 percent industry specialization ratio) (n 367)

as an important element of audit quality by the IPO companies in Taiwan. The multivariate regression results in Table VII indicate that big ve auditors reduce earnings management for IPO rms in the offering year. Several control variables are signicantly related to unexpected accruals. Operating cash ow is found to be negatively related to unexpected accruals, which suggests rms with strong operating cash ow position are less likely to use unexpected accruals to increase earnings in the IPO offering year. Firm size is found to be positively related to earnings management, suggesting large rms engage more in income increasing earnings management in the IPO year. Leverage is found to be negatively related to earnings management, suggesting that these rms are not using earnings management to satisfy debt covenant requirements. This shows that earnings management in the IPO year is unlikely due to concern over debt covenants. ELEC is signicantly and negatively related to unexpected accruals, indicating electronics companies engage in income increasing earnings management more than companies in other industries in the IPO year. Additional analyses In Table VIII, we test for variable difference between big ve and non-big ve clients. We nd that the DAC for big ve clients is signicantly different from that of non-big ve clients, which further supports our nding in Table VII that big ve auditors reduce earnings management for IPO rms in the offering year. Interestingly, we nd that industry specialists for big ve clients is signicantly different from those of non-big ve clients, which indicates that industry specialist rms are highly recognized by big ve clients.

MAJ 20,1
DAC SPEC

BIG5 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0 1 0

n 300 67 300 67 300 67 300 67 300 67 300 67 300 67 300 67 300 67 300 67 300 67

Mean 0.061 0.181 0.643 0 0.044 0.051 0.302 0.254 0.120 0.060 0.570 0.537 2.401 1.582 14.162 13.967 0.391 0.403 0.727 0.403 0.310 0.254

STD 0.490 0.555 0.480 0 0.183 0.096 0.387 0.503 0.326 0.239 0.496 0.502 2.494 1.075 1.381 0.793 0.159 0.139 0.446 0.494 0.463 0.438

t-statistic 2 1.767* 10.963*** 2 0.290 0.872 1.432 0.487 2.630*** 1.117 2 0.552 5.260*** 0.907

100

OCF ABSTA LOSS INCCHG MTB SIZE LEV ELEC TSE

Table VIII. Test for variable difference between big ve and non-big ve clients

Notes: Signicant at the *0.10, **0.05, 0.01 level based on a two-tail test; TAC: total accruals, dened as DCAit 2 DCASH it 2 DCLit 2 DSTDit 2 DEP it =TAit21 ; DAC: unexpected accruals; SPEC: 1 if the auditor is an industry specialist; 0 otherwise; OCF: operating cash ow deated by lagged total assets; ABSTA: absolute value of total accruals; LOSS: 1 if the rm incurs a loss; 0 otherwise; INCCHG: 1 if this years income is greater than previous years income; 0 otherwise; MTB: market to book ratio; SIZE: log of total sales; LEV: leverage, dened as total liabilities over total assets; ELEC: 1 if the rm is member of electronic industry; 0 otherwise; TSE: 1 if the rm is member of TSE; 0 otherwise

In Table IX, we regresses unexpected accruals on industry specialization for big ve clients only, and subdivide big ve variable by using four dummy variables to examine whether specic big ve is recognized by its client as more likely to constrain earnings management in the IPO process in Taiwan. We nd that EY is the only big ve rm that is recognized by its client to constrain earnings management, although the coefcients of AA, DT, and PWC are insignicantly and negatively related to unexpected accruals. We also use total accruals as the dependent variable to test our hypotheses and conduct additional analyses, and the results (not reported) are qualitatively similar to the unexpected accruals. V. Summary and conclusions In this study, we examine whether auditor size and industry specialization are associated with lower earnings management (lower unexpected accruals) for IPO companies in Taiwan. We nd that auditor size is associated with lower unexpected

DAC INTERCEPT SPEC AA DT EY KPMG OCF ABSTA LOSS INCCHG MTB SIZE LEV ELEC TSE Adj. R-sq. (%) 2 0.286 0.065 2 0.943 1.048 2 0.195 2 0.073 2 0.077 2 0.112 0.033 2 0.571 2 0.079 2 0.072 0.085 0.037 0.036 2 0.544 2 0.098 2 0.046 5.7

DAC 2 0.628 2 0.828 2 0.892 2 1.236 0.393 2 3.415*** 2 1.041 2 0.744 1.314* 2.869** 1.524* 2 2.594*** 2 1.452* 2 0.664 2 0.198 0.015 2 0.074 2 0.067 2 0.106 0.030 2 0.577 2 0.079 2 0.073 0.088 0.037 0.036 2 0.544 2 0.103 2 0.046 5.4

DAC 2 0.636 0.181 2 0.835 2 0.655 2 1.109 0.360 2 3.388*** 2 1.038 2 0.752 1.323* 2.868*** 1.495* 2 2.590*** 2 1.408* 2 0.663

Audit quality for Taiwan IPO rms 101

2 0.606 2 0.072 2 0.086 0.088 0.038 0.037 2 0.510 2 0.118 2 0.058 5.8

2 3.591*** 2 0.960 2 0.895 1.339* 2.993*** 1.557* 2 2.458*** 2 1.685* 2 0.840

Notes: Signicant at the *0.10, **0.05, 0.01 level based on a one-tail test; DAC: unexpected accruals; SPEC: 1 if the auditor is an industry specialist; 0 otherwise; AA: 1 if the auditor is AA; 0 otherwise; DT: 1 if the auditor is DT; 0 otherwise; EY: 1 if the auditor is EY; 0 otherwise; KPMG: 1 if the auditor is KPMG; 0 otherwise; OCF: operating cash ow deated by lagged total assets; ABSTA: absolute value of total accruals; LOSS: 1 if the rm incurs a loss; 0 otherwise; INCCHG: 1 if this years income is greater than previous years income; 0 otherwise; MTB: market to book ratio; SIZE: log of total sales; LEV: leverage, dened as total liabilities over total assets; ELEC: 1 if the rm is member of electronic industry; 0 otherwise; TSE: 1 if the rm is member of TSE; 0 otherwise

Table IX. Regression of unexpected accruals on industry specialization for big ve clients only (n 300)

accruals, consistent with high quality auditors constraining earnings management and providing more precise information. This is important given that management has incentive to engage in earnings management in the IPO process to garner greater proceeds and at issue earnings management is negatively related to post issue earnings performance and stock returns. This is also important given investors have very little information about these rms prior to the IPO process to evaluate or undo the earnings management. Our research might be of interest to investors in IPO rms, given that at issue unexpected accruals are opportunistic, and Teoh et al. (1998a) nd that at issue unexpected accruals are negatively related to post issue earnings performance and stock return. Our study also contributes to the literature by showing that auditor quality constrains earnings management in Taiwan, thus complementing the ndings in Zhou and Elder (2003).
Notes 1. See Healy and Wahlen (1999), Beneish (2001) for a review of the earnings management literature, and Teoh et al. (1998b) and Teoh et al. (1998a) for a discussion of earnings management for IPO companies. 2. On July 1, 2004, Security and Future Commission has been renamed as Securities and Futures Bureau, which is directly governed by the Financial Supervisory Commission, Executive Yuan. Self-regulated bodies of the accounting profession include National Federation of Certied Public Accounts Association of the ROC, which govern the

MAJ 20,1

102
3.

4.

5. 6.

performance of professional services by the members. The Statements of Financial Accounting Standards and Statements of Auditing Standards are issued by the Financial Accounting Standard Committee and Auditing Standard Committee of the Accounting Research and Development Foundation. There are three stock markets in Taiwan: Taiwan Stock Exchange (TSE), Over-The Counter (OTC), and Emerging Stock. The IPO criteria such as net worth, protability, and diversication of stock ownership for securities listings in the three major stocks are different where the strictest criteria apples to the listed rms in Taiwan Stock Exchange. Other signals in addition to auditor choice, such as the percentage of retained ownership, can be used as signals to reduce the extent of information asymmetry. Copley and Douthett (2002) nd that auditor choice and retained ownership are substitutes that are jointly chosen to minimize the cost to the entrepreneur. CPA rms typically market themselves as industry specialists. For example, PricewaterhouseCoopers homepage indicates that we have organized ourselves to deliver our industry expertise to some 24 market sectors and have grouped these market sectors into three clusters consistent with effective delivery to the marketplace. Quote is available at: www.pwcglobal.com/gx/eng/about/ind/index.html Ferguson and Stokes (2002) do not nd strong support for the presence of industry specialist premiums in the post big eight/six mergers. Income smoothing is different from income increasing behavior because income smoothing also includes income decreasing behavior when current performance is relatively good and future expected performance is relatively poor.

References Balvers, R., McDonald, B. and Miller, R. (1998), Underpricing of new issues and the choice of auditor as a signal of investment banker reputation, The Accounting Review, Vol. 63, October, pp. 605-22. Beatty, R. and Ritter, J. (1986), Investment banking, reputation, and the underpricing of initial public offerings, Journal of Financial Economics, Vol. 15 No. 1/2, pp. 213-32. Becker, C., DeFond, M., Jiambalvo, J. and Subramanyam, K.R. (1998), The effect of audit quality on earnings management, Contemporary Accounting Research, Vol. 15, Spring, pp. 1-24. Beneish, M. (2001), Earnings management: a perspective, Managerial Finance, Vol. 27, pp. 3-17. Burgstahler, D. and Dichev, I. (1997), Earnings management to avoid earnings decreases and losses, Journal of Accounting and Economics, Vol. 24 No. 1, pp. 99-126. Carpenter, C. and Strawser, R. (1971), Displacement of auditors when clients go public, Journal of Accountancy, June, pp. 55-8. Copley, P. and Douthett, E. Jr (2002), The association between auditor choice, ownership retained, and earnings disclosure by rms making initial public offerings, Contemporary Accounting Research, Vol. 19 No. 1, pp. 49-75. Craswell, A., Francis, J. and Taylor, S. (1995), Auditor brand name reputations and industry specializations, Journal of Accounting and Economics, Vol. 20 No. 3, pp. 297-322. Datar, S.M., Feltham, G.A. and Hughs, J.S. (1991), The role of audits and audit quality in valuing new issues, Journal of Accounting and Economics, Vol. 14 No. 1, pp. 3-49. DeAngelo, L. (1981), Auditor size and auditor quality, Journal of Accounting and Economics, Vol. 3 No. 3, pp. 183-99.

DeFond, M. and Jiambalvo, J. (1994), Debt covenant effects and the manipulation of accruals, Journal of Accounting and Economics, Vol. 17 No. 1/2, pp. 145-76. DeFond, M. and Park, C. (1997), Smoothing income in anticipation of future earnings, Journal of Accounting and Economics, Vol. 23 No. 2, pp. 115-39. Dechow, P., Sloan, R. and Sweeney, A. (1995), Detecting earnings management, The Accounting Review, Vol. 70, April, pp. 193-225. Dye, R. (1988), Earnings management in an overlapping generations model, Journal of Accounting Research, Vol. 26, Autumn, pp. 195-235. Elder, R. (1999), Audit rm size, industry specialization and initial public offerings of common stock, working paper, Syracuse University, Syracuse, NY. Ferguson, A. and Stokes, D. (2002), Brand name audit pricing, industry specialization and leadership premiums post-big 8 and big 6 mergers, Contemporary Accounting Research, Vol. 19 No. 1, pp. 77-110. Francis, J., Maydew, E. and Sparks, H. (1999), The role of big 6 auditors in the credible reporting of accruals, Auditing: A Journal of Practice and Theory, Vol. 18 No. 2, pp. 17-34. Healy, P. and Wahlen, J. (1999), A review of the earnings management literature and its implications for standard setting, Accounting Horizons, Vol. 13 No. 4, pp. 365-83. Hogan, C. (1997), Costs and benets of audit quality in the IPO market: a self-selection analysis, The Accounting Review, Vol. 72, January, pp. 67-86. Hogan, C. and Jeter, D. (1999), Industry specialization by auditors, Auditing: A Journal of Practice and Theory, Vol. 18 No. 1, pp. 1-17. Jones, J. (1991), Earnings management during import relief investigations, Journal of Accounting Research, Vol. 29, Autumn, pp. 193-228. Leland, H. and Pyle, D. (1977), Informational asymmetries, nancial structure, and nancial intermediation, The Journal of Finance, Vol. 32 No. 2, pp. 371-87. Menon, K. and Williams, D. (1991), Auditor credibility and initial public offerings, The Accounting Review, Vol. 66, April, pp. 313-32. Richardson, V. (2000), Information asymmetry and earnings management: some evidence, Review of Quantitative Finance and Accounting, Vol. 15 No. 4, pp. 325-47. Rock, K. (1986), Why new issues are underpriced, Journal of Financial Economics, Vol. 15 No. 1/2, pp. 187-212. Subramanyam, K.R. (1996), The pricing of discretionary accruals, Journal of Accounting and Economics, Vol. 22, pp. 249-81. Sweeney, A. (1994), Debt covenant violations and managers accounting responses, Journal of Accounting and Economics, Vol. 17, May, pp. 281-308. Teoh, S.H., Welch, J. and Wong, T.J. (1998a), Earnings management and the long-run market performance of initial public offerings, Journal of Finance, Vol. 53, December, pp. 1935-74. Teoh, S., Wong, T.J. and Rao, G. (1998b), Are accruals during initial public offerings opportunistic?, Review of Accounting Studies, Vol. 3, pp. 175-208. Titman, S. and Trueman, B. (1986), Information quality and the valuation of new issues, Journal of Accounting and Economics, Vol. 8 No. 2, pp. 159-72. Trueman, B. and Titman, S. (1988), An explanation for accounting income smoothing, Journal of Accounting Research, Vol. 26, Supplement, pp. 127-32. Zhou, J. and Elder, R. (2003), Audit rm size, industry specialization and earnings management by initial public offering rms, working paper, Syracuse University, Syracuse, NY and SUNY-Binghamton, Binghamton, NY.

Audit quality for Taiwan IPO rms 103

MAJ 20,1

104

Further reading Dechow, P. (1994), Accounting earnings and cash ows as measures of rm performance: the role of accounting accruals, Journal of Accounting and Economics, Vol. 18 No. 1, pp. 3-42. Willenborg, M. (2002), Discussion of Brand name audit pricing, industry specialization, and leadership premiums post-big 8 and big 6 mergers, Contemporary Accounting Research, Spring, pp. 111-16.