Journal of accounting, auditing finance tập 27, số 01, 2012 1

Most of the firms in our sample about 68% appear only once,indicating that most sample firms are probably not attempting to manage earnings infor-mation strategically across their prelim

Ó The Author(s) 2012 Reprints and permission: sagepub.com/journalsPermissions.nav DOI: 10.1177/0148558X11409006

http://jaaf.sagepub.com

Earnings Revisions in SEC

Filings From Prior Preliminary

announce-Keywords

market efficiency, asset pricing, earnings announcement, SEC filings, earnings revisions

Regulators and prior research have shown that investors have suffered significant losses asmarket capitalizations have dropped by billions of dollars due to restatements of auditedfinancial statements (Levitt, 2000; Palmrose, Richardson, & Scholz, 2004) Less discern-able restatements occur between the preliminary earnings announcement and the immedi-ately subsequent Securities and Exchange Commission (SEC) filing (henceforth calledearnings revisions); however, very little is known in the academic literature about theserevisions This study first examines the characteristics of firms most prone to earnings revi-sions and the type of income statement components that are most likely to be affected

We then assess the market’s response to the additional earnings surprise (in the SEC filing)and the total earnings surprise (i.e., the preliminary earnings surprise plus the additionalearnings surprise at the time of the SEC filing) to determine whether the market’s reactions

to the mere act of earnings revision is negative, irrespective of its content

Typically, companies issue a press release to report preliminary earnings to the market

26 days after the fiscal quarter-end for the median company, with 1% of firms reporting in

revises its earnings from the preliminary earnings announcement to the subsequent SECfiling for a variety of reasons For example, a firm’s auditor may require the filing of differ-ent earnings figures with the SEC than those previously released to investors Subsequentinformation revealed after the preliminary announcement may also cause firms to modifytheir earnings before the SEC filing Although an infrequent phenomenon (1.98% of oursample observations), some of these revisions are quite large On average, the absolutevalue of the revision is 2.9% of the market value of equity for our sample of 3,337 firm

earnings revisions are greater in absolute value than 6.2% of the market value of equity atquarter-end (mean market value for sample firms is US$4.864 billion) Although these earn-ings revisions are economically significant at the time of their revision, unlike typical earn-ings restatements, virtually none of the earnings revisions are preceded by a formalannouncement or a press release, nor are companies required to file an amendment such as10-K/A or 10-Q/A Furthermore, in many cases, firms do not even explain the reasons forthe earnings revisions in their SEC filings Still, the phenomenon of earnings revisions isimportant for academic research and practitioners because it indicates cases of potentialproblems with earnings quality that have not previously been documented in the academicliterature Furthermore, a study by Hollie, Livnat, and Segal (2005) shows that market parti-cipants seem to react adequately to the new earnings surprise contained in SEC filings offirms that made earnings revisions However, they do not study, as we do here, whether thetotal market reaction to both sources of earnings surprise indicate that market participantsseem to treat firms with earnings revisions as if they have an inferior quality of earnings

In our sample, approximately 64% of all significant earnings revisions are incomedecreasing, and a large portion of the revisions (about 50%) occur in the fourth fiscalquarter, consistent with both audit work at year-end and a longer period before SEC fil-ings in which subsequent events may require an earnings revision We find that the mostfrequently revised components of net income are recurring items (e.g., sales; cost ofsales; Selling, General, and Administration [SG&A]; and depreciation) and that earningsrevisions are more likely to occur for firms that are more complex in nature, are morefinancially leveraged, have greater earnings volatility, have losses, and have switchedauditors We also find that the likelihood of an earnings revision is inversely related tothe persistence of earnings changes These variables are also shown in the literature to beassociated with lower earnings quality and subsequent earnings restatements However,contrary to expectations, we do not find that significant earnings revisions occur in high-litigation risk industries Most of the firms in our sample (about 68%) appear only once,indicating that most sample firms are probably not attempting to manage earnings infor-mation strategically across their preliminary earnings announcements and SEC filings in

In regression results, we also find that market reactions to the total earnings surprise forfirms that revise their earnings (i.e., the associations of abnormal returns from a day beforethe preliminary earnings announcement through a day after the SEC filing with the com-bined preliminary earnings surprise and SEC filing earnings surprise) do not differ signifi-cantly from those of non-Revisers However, in matched-sample tests, we mostly findevidence consistent with weaker market reaction to total earnings of Revisers than those of

non-Revisers Thus, we cannot unambiguously conclude that Revisers are associated withlower earnings quality.

Our empirical findings are particularly relevant to academics, financial analysts, tors, and investors The results of this study are important to academic researchers because itfocuses on cases of potential lower earnings quality, which have not been studied in the aca-demic literature We do find inconclusive results regarding whether firms with earnings revi-sions have lower earnings quality, as indicated by weaker market reactions to the totalearnings surprise These cases may be of particular interest to the auditing profession, wherethe identification of firms with a higher audit risk is extremely valuable The evidence in thisstudy also suggests that financial analysts, investors, and regulators alike should pay closeattention not only to an earnings surprise at the preliminary earnings announcement date butalso at the SEC filing date to determine whether a subsequent earnings surprise occurs.The remainder of this article is organized as follows The section titled ‘‘Background,Hypotheses, and Research Design’’ reviews the related literature and outlines our hypoth-eses and methodology The section titled ‘‘Data and Results’’ describes the sample and pre-sents the empirical findings The section titled ‘‘Summary and Conclusions’’ summarizesand concludes the article

regula-Background, Hypotheses, and Research Design

Most firms disclose their preliminary earnings for the quarter or year through a pressrelease, following it with an SEC filing several weeks later Easton and Zmijewski (1993)report a median lag between the balance sheet date and the preliminary earnings announce-ment of 28 days and a median SEC filing lag of 45 days for Forms 10-Q Our sampleshows a similar pattern with a median preliminary earnings lag of about 25 to 26 days inrecent years Some firms issue a press release to discuss earnings after their SEC filings(Stice, 1991), and others do not issue any press release at all, relying on the informationavailable in the SEC filings alone (Amir & Livnat, 2006) When firms issue their prelimi-nary earnings release prior to the SEC filing, investors implicitly assume that these earningswill be identical to the SEC-filed earnings However, as this study shows, a nontrivial por-tion of firms file significantly different earnings figures with the SEC than the one previ-

Consider the following two examples, which are highlighted in Appendix A R&B, Inc.(NASDAQ: RBIN) reported net income of US$2.8 million in its second quarter preliminaryearnings release on July 31, 2002, but revised it upward to US$4.115 million on August 8,

2002, in its SEC filing, an increase of almost 150% from its preliminary earnings Anexamination of news items regarding R&B, Inc (using LexisNexis) reveals that no pressreleases were issued between the preliminary earnings announcement and the SEC filingdate Hence, no public information seems to have been available between the earningsannouncement and the SEC filing date that would have indicated that an upward revision

in SEC-filed earnings was forthcoming MEMC Electronic Materials, Inc (NYSE: WFR)issued its second quarter preliminary earnings release on August 10, 2001, reporting anUS$88.1 million loss, followed by an SEC filing (August 13, 2001) reporting a US$355.3million loss, reducing the previous earnings figure by approximately 400% just days later.There are no other news reports in LexisNexis between the preliminary earnings announce-ment and the SEC filing date to suggest that a downward revision was forthcoming Asthese two examples illustrate, some earnings revisions between the preliminary earnings

release and the SEC filing date are upward revisions, whereas others are downward

Plausible explanations (not necessarily specific to the two examples above) for earningsrevisions between the preliminary announcement and the SEC filing are the uncovering ofissues not known at the preliminary earnings release date due to audit work or new auditorswho are more likely to compel clients to revise previously issued earnings under the priorauditor, information that becomes known subsequent to the preliminary earnings announce-ment, or discovery of accounting errors before the filing date For example, several studiesdocument the existence of accounting errors that are discovered by auditors and corrected

about 50% of our firm-quarter observations existing in the fourth quarter, when a full audit

is required (see Panel A of Table 2), and where the window between the preliminary ings announcement and the SEC filing is typically longer, leading to greater opportunitiesfor discovering material subsequent to information that requires earnings revisions

earn-Characteristics of Firms With Earnings Revisions

Absent a theoretical model to guide the selection of potential variables that are associatedwith the likelihood of earnings revisions in SEC filings, we use variables that may indicateinferior earnings quality

DeFond and Hung (2003) suggest earnings volatility and accruals magnitude as tives to disseminate cash flow forecasts when earnings are of inferior quality Chen,DeFond, and Park (2002) claim that firms with losses and firms with greater earnings sur-prises are more likely to provide balance sheet information in their preliminary earnings tostrengthen their weaker earnings quality We use these variables, as well as the persistence

incen-of earnings surprises, to show that firms with a lower quality incen-of earnings are more likely torevise their earnings in the SEC filings

Consistent with subsequent corrections during audit work (or review) between the nary earnings announcement and the SEC filing, we conjecture that the firm’s complexity(measured by the number of its operating segments), size, number of analysts following, andauditor changes are all positively associated with the likelihood of an earnings revision.However, it may be argued that larger firms with a greater analyst following are managed

profit-able firms (lower return on assets) are more likely to have earnings revisions because ers in such firms have greater incentives to manage preliminary earnings and because lowerprofitability may indicate operational problems and potentially weaker accounting controls.Studies of restatements in the literature (e.g., Richardson, Tuna, & Wu, 2002—annualrestatements—and Livnat & Tan, 2004—quarterly restatements) assert that financial lever-age is positively associated with the likelihood of restatements due to management’s desire

manag-to inflate earnings initially as a way manag-to avoid debt restrictions The literature also asserts thatgrowth firms are more likely to have earnings restatements because of their desire to showcontinued earnings growth and ability to beat analyst expectations Firms with high financialleverage may also be subject to additional scrutiny by auditors and creditors, leading togreater chances for earnings revisions Thus, we expect financial leverage to be positivelyassociated with the likelihood of earnings revisions and the earnings-to-price ratio (an indi-cator of growth) to be negatively associated with the likelihood of earnings revisions

Market Reactions to Preliminary Earnings Announcements and SEC Filing DatesPrior research shows that the market responds to earnings surprises included in preliminaryearnings announcements and that stock prices incorporate this information instantaneously(see Kothari, 2001, and Lev, 1989, for summaries of these studies) However, there is sub-stantially less agreement about the incremental information content in 10-Q/Ks beyond pre-liminary earnings announcements For example, Foster and Vickery (1978), as well asWilson (1987), document that 10-Ks have information content beyond earnings announce-ments, whereas other studies suggest that the market fails to react to earnings informationcontained in SEC filings (e.g., Chung, Jacob, & Tang, 2003; Cready & Mynatt, 1991;Easton & Zmijewski, 1993; Foster, Jenkins, & Vickery, 1983, 1986; Stice, 1991).

Easton and Zmijewski (1993) examine whether the 10-Q/K filing dates are associatedwith abnormal returns, using squared market model prediction errors to avoid any predictionsabout the direction of expected returns around the SEC filing dates Their results show abnor-mal market returns that are significantly different from zero around preliminary earningsannouncements but show no significant difference for market reactions to SEC filings, except

in those cases where only the 10-Q dates are known but no preliminary earnings ment dates are available on the Quarterly Compustat File These results seem to imply thatSEC filings contain no incremental information beyond preliminary earnings announcements.Stice (1991) examines whether the information content of an earnings announcementcan be affected by the method in which earnings are announced, concentrating on firmsthat file their 10-Q/Ks several days before the earnings announcement Stice finds that SEC

Chung et al (2003) corroborate Stice’s findings and show that some firms in their samplebehave as if they manage earnings

Balsam, Bartov, and Marquardt (2002) investigate whether investors in firms that aresuspected of engaging in earnings management are able to more rapidly incorporate theinformation about accruals available in 10-Q filings and whether institutional investorsseem to respond even before the SEC filing dates They find evidence consistent with noinvestor reaction to the managed accruals around the preliminary earnings announcement(with event windows up to 9 days later), with market reactions to discretionary accruals byfirms with at least 40% institutional investors during the window spanning 10 days afterthe preliminary earnings announcement through 2 days before the SEC filing date, andwith market reactions to discretionary accruals in the window from a day prior to the SECfiling date through 15 days afterward for firms with fewer institutional investors Theirinterpretation is that institutional investors seem to find the information necessary toreverse accruals faster than other investors and prior to the SEC filing dates

Qi, Wu, and Haw (2000) suggest that prior research’s inability to detect little, if any,information content around the SEC filing date may be due to the SEC paper filing system

in place at the time of prior studies Their study compares SEC paper filings with SECelectronic filings to test whether the information content of 10-Ks has changed because ofelectronically available SEC filings In contrast to most of the prior research, Qi et al pro-vide evidence that 10-K filings through the Electronic Data Gathering, Analysis, andRetrieval (EDGAR) system provide incremental information content that did not exist forthe paper filings However, they study the years 1993 to 1995, in which the EDGARsystem was still voluntary (becoming mandatory in May 1996) In addition, their study islimited to firms with available Association for Investment Management and Research ana-lyst rankings

In a recent study covering the period 1996 to 2001, Griffin (2003) finds that SEC filingshave abnormal market returns that are significantly different from zero, where abnormalreturns represent the absolute value of excess returns around the filing date compared withthe excess returns in a prior period He finds greater market reactions to 10-Q/Ks, tosmaller firms, to firms with lower proportions of institutional holders, to firms that report

on days with many filings by other firms, and to firms with delayed filings In addition, inmultivariate results, Griffin finds evidence of stronger market reactions to filings made inrecent periods and to delayed SEC filings

Asthana, Balsam, and Sankaraguruswamy (2004) show that small trades increase in the5-day period around the 10-K filing after EDGAR as compared with the pre-EDGARperiod, but not large trades, implying that small investors are better able to use the informa-tion in SEC filings in the post-EDGAR period They also show that small investors seem toincorporate better the information content of the 10-Ks in the post-EDGAR period thanbefore and provide evidence consistent with an erosion of the information advantage thatlarger traders have as compared with small traders in the post-EDGAR period

Callen, Livnat, and Segal (2006) show that earnings news is important in explaining thevolatility of unexpected returns around both the preliminary earnings announcements andthe SEC filing dates, with the cash flows and accruals components of earnings significantlyassociated with the volatility of unexpected returns around the SEC filing dates

Hollie et al (2005) show that firms with significant earnings revisions in SEC filingsexperience significant market reactions to the new earnings surprise in the SEC filings.They also show that market reactions to a dollar of earnings surprise in the preliminaryearnings announcement are not statistically different from the market reactions to a dollar

of earnings surprise in the SEC filing for their sample with material earnings revisions.Their study, although important in highlighting that market participants are able to identifyand react to the new earnings surprises in SEC filings, does not investigate whether thetotal market reaction to the total earnings surprise (preliminary and revision) is weaker due

to the potential lower earnings quality

Research Design

Figure 1 provides the time line of events in this study and highlights the range of periodsover which abnormal returns are calculated This figure identifies the short windowsaround the preliminary earnings announcement and SEC filing dates, the long windowbetween the preliminary earnings announcement and SEC filing, and the long window from

1 day before the preliminary earnings announcement through 1 day after the SEC filing date.For some of our tests, we examine differences between two samples: a sample of firms withearnings revisions (Revisers) and a sample of control firms from the same Fama and French(1997) industry, closest in size (market value of equity) and in the same quarter (non-Revisers) For other tests, we use the entire population of non-Revisers

We first assess the characteristics of firms that have earnings revisions between nary earnings announcements and SEC filings We use a logistic regression model to deter-mine the likelihood that a firm is a Reviser

where REVISER is a dummy variable equal to one if a firm-quarter has an earnings sion and zero otherwise The variables we use to discriminate Revisers and control firmsare as follows: (a) Earnings volatility (EARNVOL) is the absolute value of the ratio of thestandard deviation of Earnings per share (EPS)/Price over the most recent 12 months to theaverage EPS/Price over the same period It is winsorized to a maximum of 5; (b)Persistence of earnings changes (PERSE) is estimated as the first autocorrelation betweenscaled earnings surprises in the prior eight quarters The earnings surprise is earnings in thequarter minus earnings of the same quarter in the preceding year, scaled by market value atthe beginning of the quarter; (c) Debt/Assets (DEBT) is estimated as short-term plus long-term debt divided by total assets at the end of the quarter; (d) Auditor change (AUC) is adichotomous variable obtaining one if the firm’s auditor changed from the prior year(mergers of audit firms such as Coopers and Lybrand and Price Waterhouse are not counted

revi-as an auditor change); (e) The number of segments (SEGNUM) is from the CompustatSegment file and is a surrogate for operating complexity; (f) ROA is the ratio of earnings

to total assets at quarter-end and proxies for firm profitability; (g) LOSS is a dummy able obtaining one if earnings for the quarter are negative and zero otherwise; (h) EP is thecurrent earnings to market value at quarter-end; and (i) ACCPROP is the proportion ofaccruals divided by sales over the prior four quarters

vari-Market Reactions to Earnings Revisions

As portrayed in Figure 1, this study focuses on four buy-and-hold abnormal returns

the raw buy-and-hold return from the Center for Research in Security Prices (CRSP) minusthe buy-and-hold return on the portfolio of firms with the same size (the market value of

CARtotal

CARprelim

CARaf

CARfile–1

the preliminary earnings announcement date The abnormal return is the raw return minus the age return on a same size, book-to-market (B/M) portfolio (six portfolios), as provided by professor

SEC filings It controls for information released to the stock market between the two time periods;

filing date of 10-Q/10-K; CARtotal= the buy-and-hold abnormal return from 1 day before the nary earnings announcement through 1 day after the SEC filing date

prelimi-equity) and B/M ratio The daily returns (and cutoff points) on the size and B/M portfoliosare obtained from Professor Kenneth French’s data library, based on classification of the

abnormal returns in the top and bottom 0.5% of the population

We then examine the stock returns around the preliminary earnings announcements toprovide initial evidence on the effect of earnings surprises in preliminary earningsannouncements Consistent with prior research (e.g., Kothari, 2001; Lev, 1989; Penman,1987), we expect a positive and significant relation between earnings announcements andstock market returns This analysis is intended only to show the consistency of our samplewith the results in prior studies For this test, we examine the relation between the first

CARprelim it5b01b1FSURPit1eit ð2Þ

preliminary earnings announcement date (Date 0) and FSURP is the preliminary (or first)earnings surprise, calculated as preliminary Compustat earnings at Quarter t minusCompustat earnings at Quarter t 2 4, scaled by market value of equity at the end of

coefficient (ERC) for the preliminary earnings surprise

We then examine the stock market reaction to the additional earnings surprise in SECfilings We measure the additional earnings surprise (ASURP) as the SEC-filed earnings(As-First-Reported [AFR] Compustat earnings at Quarter t) minus the preliminaryCompustat earnings, scaled by market value at quarter-end In a similar manner toCARprelim, we define CARfileas the 3-day (21 to 11) buy-and-hold abnormal returns cen-tered on the SEC filing date (Date 0) To control for additional news possibly obtained bymarket participants between the preliminary earnings announcement and the SEC filing

preliminary earnings announcement through 2 days before the SEC filing date It isassumed that changes in stock prices capture all news during this event period.Accordingly, we investigate the relationship between ASURP (additional surprise), FSURP,

CARfile it5b01b1FSURPit1b2ASURPit1b3CARaf it1eit ð3Þwhere ASURP is defined above Our equation for total earnings surprise (TSURP) for non-Revisers and Revisers is provided below Because non-Revisers do not have an additionalearnings surprise, ASURP, by definition, will always be equal to zero for the non-Revisercontrol group of firms

For Revisers and for non-Revisers:

5½Preliminary Compustat earnings at Quarter t minus AFR

equity at the end of Quarter t1½AFR Compustat Earnings in the

The coefficient b1 captures the market reaction at the time of the SEC filing to the

insig-nificantly different from zero if stock prices have already fully impounded the earnings

represents the market reaction to the additional earnings surprise contained in the SECfiling beyond earnings reported in the preliminary earnings announcement It is expected to

be positive and statistically different from zero if the SEC filings are noticed by investorswho react to the additional earnings surprise, as is documented in Hollie et al (2005) We

con-trol variable for information released to the market between the two event dates This ysis is intended to show that our sample produces similar results to those of Hollie et al.Next, we estimate the market reactions to the total earnings surprises of Revisers ascompared with those of control firms through a system of simultaneous equations.Essentially, this method consists of the following steps: a first stage regression to get fittedvalues for the probability that the firm is a Reviser and a second stage to run the regres-sions using these fitted values The two endogenous variables are therefore REVISER and

AUC, and LOSS), identified in the logistic analysis previously, are used as instrumental

announcement date through 1 day after the SEC filing date

CARtotal it5b01b1REVISERit1b2TSURPit1b3REVISERit3TSURPit1eit ð4Þ

CARtotal it5b01b1REVISERit1b2TSURPit1b3REVISERit3TSURPit1b4ROAit

where TSURP is the sum of ASURP and FSURP as defined above We assign each quarter into 10 deciles according to their TSURP We then substitute the deciles rank forTSURP, divide it by 9 and subtract 0.5, as is often done in the postearnings-announcementdrift literature The coefficient on TSURP measures the return on a hedge portfolio that

that the market views an earnings revision in a negative light, likely due to weaknesses inaccounting or control systems or perceived lower earnings quality, we expect the coeffi-

This would indicate that a dollar of earnings surprise in the case of Revisers contributesless to the market reaction than a dollar of earnings surprise of non-Revisers

We further test whether the total market reactions to the total earnings surprises forRevisers are different from those of the matched control firms by comparing the returns ofRevisers and matched control firms in the same quarter The matching is done by the totalsurprise decile and size (market value of equity); total surprise decile and industry; totalsurprise decile, size, and industry; and total surprise decile, size, and number of segments

We also match the firm to itself in another quarter with no earnings revision but when thetotal earnings surprise was in the same decile rank If the market reacts negatively to theact of earnings revisions, we should observe lower total market reactions for Revisers thanfor the matched control firms

Finally, we examine whether these earnings revisions are possibly a result of strategicbehavior by management In this study, we examine two specific strategic behaviors, whichare (a) whether a firm initially meets or beats earnings forecasts but misses the forecastafter the earnings revisions and (b) whether it appears that the firm has initially taken a bigbath falling below its forecast but subsequently revised earnings upward above the forecast.

We use the previously described approaches to evaluate these noted behaviors

Data and Results

Original Compustat Quarterly Data

Compustat’s database data entry procedure is as follows: When a firm releases its preliminaryearnings announcement, Compustat takes as many line items as possible from the preliminaryannouncement and enters them into the quarterly database The preliminary earnings data aredenoted by a code of 2, until the firm files its Form 10-Q/K with the SEC or releases it tothe public, at which point Compustat updates all available information with an update code

of 3 The Compustat quarterly database is further updated when a firm restates its previouslyreported earnings For example, if a firm engages in mergers, acquisitions, or divestitures andrestates previously reported earnings, Compustat inserts the restated data into the databaseexpunging the previously reported numbers Similarly, if the annual audit requires a restate-ment of its previously reported earnings, Compustat updates the quarterly database to reflectthese restated numbers—again expunging the previously reported earnings numbers

Charter Oak Investment Systems, Inc (Charter Oak) has compiled the weekly original CDsthat Compustat distributed to its PC clients to create a database similar to Compustat that con-tains all original and updated information without expunging any previously recorded earningsinformation This database contains, for each firm in the Compustat quarterly database, threenumbers for each Compustat line item in each quarter: (a) preliminary earnings announcement(Compustat update code of 2), (b) the AFR figure when Compustat first changed the updatecode to 3 for that firm-quarter, and (c) the number that exists in the current version ofCompustat, which is what most investors and researchers utilize Therefore, the Charter Oakdatabase permits us to establish whether an earnings revision has occurred in any particularquarter by comparing the preliminary announced earnings to the first-reported 10-Q/Kearnings

or where the buy-and-hold returns for any of the windows is in the top or bottom 0.5% ofthe observations, and obtain 168,436 observations Out of these observations, we have3,337 firm-quarters with differences between preliminary earnings and first-reported earn-ings in excess of US$100,000, which meets the additional criteria described below This isnot a trivial proportion; about 1.98% of the relevant population files a different earningsfigure with the SEC than the one disclosed in the preliminary earnings announcement just afew days/weeks earlier We tested 100 of these earnings revisions to determine whether theCompustat preliminary earnings announcement and SEC-filed earnings as captured byCharter Oak is reliable We were able to corroborate all cases we checked by hand to thesource documents.

To examine the characteristics of firms that materially revise their earnings on the SECfiling after publicly disclosing a different earnings figure, we matched our sample ofRevisers with a control sample of non-Revisers from the same industry (Fama & French,1997; based on 48 industries), where the initial earnings surprise decile is the same (earn-ings for the quarter minus earnings for Quarter t 2 4, scaled by market value), and wherethe Revisers and non-Revisers are no more than one size (market value of equity atquarter-end) decile apart If we cannot find a matching firm from the same industry, weeliminate the Reviser from our sample

In addition, the following requirements must be met for Revisers and control firmsalike: (a) the absolute value of earnings to market value of equity (E/P) and to total assets(ROA) at quarter-end is below one—this is intended to eliminate extreme cases of E/P orROA, (b) the absolute value of the scaled earnings surprise at the preliminary earningsannouncement is available and is below one, and (c) the preliminary earnings announce-ment is available on Compustat and is prior to the SEC filing date (eliminating observa-tions subject to the Stice, 1991, effect) Thus, our sample selection criteria yield a finalsample of 6,674 observations (firm-quarters) for Revisers matched by control firm observa-tions In addition, we examine the full Compustat and matched sample as part of our sensi-tivity analysis throughout this study

Descriptive Statistics

Panels A, B, and C of Table 1 contain descriptive statistics about Revisers and Revisers (i.e., control firms) It shows that Revisers have greater earnings volatility, propor-tion of accruals, debt, number of segments, frequency of auditor changes, analyst following,and losses than control firms Preliminary earnings surprises are more negative for Reviserfirms The table also shows that Revisers have lower E/P ratios, persistence of earnings sur-

lower earnings surprise on that date The return for the window between the preliminary

group but negative (2.003) for Revisers, which is consistent with more negative newsduring that period for Revisers

Panel C of Table 1 provides univariate tests of mean differences for Revisers versuscontrol firms for all variables As can be seen from the table, earnings volatility, proportion

of accruals, financial leverage, frequency of auditor changes, number of segments, number

of analyst following, and losses are significantly larger for Revisers than control firms Theearnings-to-price ratio, persistence of earnings surprises, profitability, and buy-and-holdreturns for the preliminary earnings as well as through the SEC filing periods are all signif-icantly larger for control firms than Revisers Consistent with our matched-pair design, we

Table 1 Summary Statistics on Variables for Sample and Control Firms

Panel A: Descriptive Statistics for Sample Firms

10thpercentile

25thpercentile

50thpercentile

75thpercentile

90thpercentile

25thpercentile

50thpercentile

75thpercentile

90thpercentile

aver-6 COREOCF = correlation of quarterly earnings and OCF, estimated over the eight previous quarters.

7 PERSE = persistence of earnings surprises Estimated as the first autocorrelation between earnings surprises in the previous eight quarters Earnings surprises are earnings at Quarter t minus earnings at Quarter t 2 4, scaled

by market value of equity at the beginning of the quarter.

8 DEBT = debt/assets Estimated as short-term plus long-term debt divided by total assets at the end of the quarter.

9 AUC = auditor change A dichotomous variable obtaining one if the firm’s auditor was changed from the prior year.

10 MARKET VAL = market value of equity in millions of dollars as of quarter-end.

11 SEGNUM = number of segments Taken from the Compustat segment file and is a surrogate for operating complexity.

12 ROA = ratio of earnings to total assets at quarter-end.

13 IBESN = number of analysts on Institutional Brokers Estimate System (IBES) A measure of coverage and mational environment.

infor-14 LOSS = dummy variable obtaining one if earnings for the quarter are negative.

15 BIG = an indicator variable equals to one if the firm is audited by a big auditor (currently the Big 4).

16 FSURP = The First (Preliminary) Surprise Calculated as the Compustat preliminary quarterly earnings nic IBQ) minus the As-First-Reported (AFR) Compustat quarterly earnings for t 2 4, scaled by market value at quarter-end.

(mnemo-17 ASURP = additional surprise Calculated as earnings in the SEC filing (AFR Compustat earnings) minus earnings reported in the preliminary earnings release (preliminary Compustat earnings), scaled by market value at quarter- end By definition, it is zero for control firms.

do not find significant differences for core operating cash flows, firm size, and whether afirm is audited by a big audit firm.

Panel A of Table 2 reports the number of observations by year and fiscal quarters There

is an indication of an increasing trend in the number of Revisers across years through 2008,likely due to more extensive coverage of Compustat firms than a real increase in the propor-tion of Revisers to total Compustat firms A noticeable trend is that there are about threetimes as many Revisers in the fourth fiscal quarter as in other quarters This may be attrib-uted to revisions required by the auditor as part of the year-end audit work It may also beattributed to a longer period between the balance sheet date and the SEC filing for the fourthquarter, when material errors may become known requiring a revision in earnings

Panel B of Table 2 shows that about 64% (2,132 out of 3,337) of the earnings revisionsfrom the preliminary earnings announcements to the SEC filings are downward revisions,resulting in lower filed earnings This is not only consistent with strategic reporting bymanagement that is intended to inflate earnings initially but also with additional uncoverederrors that are more likely to cause downward revisions due to the conservative nature ofaccounting Panel B also indicates that about 54% of all Revisers (1,800 of 3,337) had anegative initial earnings surprise, as compared with about 61% (2,043 of 3,337) for thecontrol firms Earnings revisions (like restatements) may indicate problems in managerialinformation systems and potentially other performance problems These problems are likely

to be manifested in future periods as well However, unlike restatements that may be closed many months after the problems occurred, the problems may surface very quicklyafter earnings revisions are disclosed Hence, untabulated tests show that Revisers have astatistically significant higher proportion of firms (6% more than the matched non-Revisers) with negative earnings surprises in the subsequent quarter after the earnings revi-sion Revisers, on average, have 8% more negative surprises over the subsequent four quar-ters than the matched non-Revisers (a statistically significant difference between the twogroups) Thus, earnings revisions imply a higher likelihood for earnings disappointments inthe immediate subsequent quarters

dis-Panel C of Table 2 shows the distribution of earnings revisions across firms The ity of firms, 68%, have only one revision during the 16-year (64 quarters) period Another18% have two earnings revisions during the sample period, and fewer than 14% have three

major-or mmajor-ore earnings revisions during the sample period Although the overall frequency ofearnings revisions is low (about 2%), we find that 25% of all firms in the Compustat data-base (with the same data requirements as our sample firms) have at least one revisionduring the study period, indicating the breadth of this phenomenon In sensitivity tests, we

Table 1 (continued)

_

18 CAR prelim = the buy-and-hold abnormal return from Day 21 through Day 11, where Day 0 is the preliminary earnings announcement date The abnormal return is the raw buy-and-hold return minus the buy-and-hold return

on a same size book-to-market (B/M) portfolio (six portfolios), as provided by professor French.

19 CAR file = the buy-and-hold abnormal return from Day 21 through Day 11, where Day 0 is the SEC filing date

Table 2 Sample Composition

Panel A: Observations by Year and Fiscal Quarters

Panel B: Upward and Downward Earnings Revisions and Initial Earnings Surprise

Year

Upward

revisions

Downwardrevisions

PositiveFSURP

NegativeFSURP

PositiveFSURP

NegativeFSURP

report on the robustness of our results to using only firms with just one earnings revisionduring our sample period.

Untabulated results indicate that business services, telecommunications, trading, andutilities have greater sample representation (proportionately more Revisers) than otherCompustat industries, where the first two industries are mentioned in prior studies ashaving a greater litigation risk In contrast, insurance, retail, pharmaceutical products, steel,consumer goods, and medical equipment industries are represented less often in theRevisers sample Thus, one cannot strongly conclude that industry membership can explainsatisfactorily earnings revisions However, it seems that more stable industries are slightlyunderrepresented in the Revisers sample, whereas industries that are more volatile tend to

be slightly overrepresented

Table 3 reports summary statistics about the components of earnings that are revisedbetween the preliminary earnings release and the SEC filing This table summarizes revisionsonly for those items that had been disclosed in the preliminary earnings release and subse-quently revised; if Compustat could not find an earnings component in the preliminary earn-ings release, a component revision is not recorded for that firm, although the earnings

Table 2 (continued)

_

Panel C: Frequency of Revisions per Reviser-Only Sample

Notes: FSURP = first earnings surprise.

1 Sample observations are for firm-quarters during 1994 to 2009 with earnings revisions between preliminary announcements dates and subsequent Securities and Exchange Commission (SEC) filings of at least US$100,000.

2 Control firms are from the same Fama and French (1997) industry (48 industries) that are closest in size to the revising firm in the same quarter but without an earnings revision.

3 Compustat firms are those observations that fit the initial sample selection criteria during the sample period The sample selection criteria require market value at quarter-end of at least US$1 million, price per share at quarter-end in excess of US$1, sales and total assets at the beginning of the quarter in excess of US$1 million, availability of a preliminary earnings announcement date on Compustat that is prior to the SEC filing date, availabil- ity of preliminary earnings and As-First-Reported earnings on Charter Oak, and availability of the number of shares used to calculate earnings per share (EPS).

component may in fact have been revised for that firm We classify revisions to sales, cost ofsales, SG&A, and depreciation as recurring items if there are no revisions to nonrecurringitems for these firms We classify revisions to nonoperating gains and losses, special items,and extraordinary items and discontinued operations as nonrecurring items if there are no

to the most revisions followed by cost of sales and sales components of net income

Determinants for Revisers

Table 4 provides the results of the logistic regression analysis, where the dependent variable

is one for Revisers and zero for control firms We provide two sets of tests In the first, wereport all significant variables from the univariate tests In the second, we exclude all insig-nificant variables from the first test Our regression model has high explanatory power with

a significant likelihood ratio Approximately 61% of the observations are predicted rectly All variables in the model are statistically significant and in the predicted direction,except proportion of accruals, E/P in the full matched sample, and earnings persistence inthe abbreviated matched-sample logistic regression In particular, we find that the morecomplex a firm (indicated by the number of segments), the greater the likelihood of a subse-quent earnings revision Similarly, the likelihood of a revision is higher for firms with losses

cor-or mcor-ore financial leverage, possibly because these firms are mcor-ore likely to have operationaland accounting control problems and greater incentives to report strategically on the prelimi-nary earnings announcement date Firms with greater earnings volatility are also more likely

to have subsequent earnings revisions, possibly because subsequent information after thepreliminary earnings announcements is more likely to become available Firms with greaterprofitability (ROA) have a lower likelihood of a revision Firms with auditor changes aremore likely to have a subsequent earnings revision, possibly because new auditors are more

Table 3 Revisions to Components of Earnings

3 Recurring items are sales; cost of sales; selling, general, and administration (SG&A); and depreciation expenses Firms are classified as recurring when they have a revision in a recurring item and there is no revision to nonoperat- ing gains and losses, special items or extraordinary items, and discontinued operations Nonrecurring is the opposite.

likely to uncover issues not known at the preliminary earnings release date Firms with ahigher persistence of earnings surprises are less likely to have subsequent revisions, proba-bly because of their greater earnings stability Thus, firms with earnings revisions are thosethat have operations that are more complex, greater earnings volatility, higher leverage, andlosses, and those that experienced an auditor change Profitable firms and firms with persis-tent earnings changes are less likely to have an earnings revision.

Market Reactions to Earnings Revision

Panel A of Table 5 presents the results of regression equations for the full Reviser sample,where the dependent variable is the market reaction and the independent variables includevarious earnings surprises for Revisers Column 2 provides information about market reac-tions to the preliminary earnings announcement Consistent with prior studies, there is a sig-nificantly positive association between the scaled earnings surprise (FSURP) and the 3-day

Table 4 Logistic Regression to Predict Earnings Revisions Using a Matched-Pair Sample and a FullSample

Matched pairsample

Matched pair

Parameterestimate

Parameterestimate

Parameterestimate

Parameterestimate

Table 5 Market Reactions to Earnings Surprises in Preliminary Earnings Announcements and SECFilings

Panel A: Market Reactions to Earnings Revision

Reviser sample

Panel B: Market Reactions to Upward/Downward Earnings Revision

Full reviser sampleEarnings revision:

Dependent Variables

FSURP = The First (Preliminary) Surprise Calculated as the Compustat preliminary quarterly earnings (mnemonic IBQ) minus the As-First-Reported (AFR) Compustat quarterly earnings for t 2 4, scaled by market value at quarter-end ASURP = additional surprise Calculated as earnings in the SEC filing (AFR Compustat earnings) minus earnings reported in the preliminary earnings release (preliminary Compustat earnings), scaled by market value at quarter- end By definition, it is zero for control firms.

earnings announcement date The abnormal return is the raw buy-and-hold return minus the buy-and-hold return

on a same size, book-to-market (B/M) portfolio (six portfolios), as provided by professor French.

CAR file = the buy-and-hold abnormal return from Day 21 through Day 11, where Day 0 is the SEC filing date of 10-Q/10-K.

CAR af = the buy-and-hold abnormal return between preliminary earnings announcements and SEC filings It trols for information released to the market between the two time periods.

con-buy-and-hold return centered on the preliminary announcement date (0.0367, p \ 0001).Column 3 presents the results of regressing 3-day buy-and-hold excess returns around theSEC filing date where the earnings surprise includes also the additional earnings surprise(ASURP) We also control for the information flow to the market between the preliminary

between the two dates, as an independent variable in the regression We find that the additionalsurprise on the revised earnings on the SEC filing date, ASURP, has a positive and signifi-cantly different from zero association with the 3-day abnormal return around the filing date(0.0225, p = 0023) Thus, consistent with the findings of Hollie et al (2005), investors react tothis new earnings information on the SEC filing date As expected, we find that there is no sta-tistically significant market reaction to the already-known initial earnings surprise (FSURP).Also, note that the information flow to the market between the preliminary earnings release

In untabulated results, we also examine the lag between preliminary earnings and

We find that Revisers have an average lag increase of about 1 day more than the matchedsample in each of the next four quarters after the revision (all differences between Revisersand controls are statistically significant), indicating that managers of Revisers take moretime in future quarters to disclose preliminary earnings This indicates both an economicconsequence of the earnings revision (less timely information to market participants) and aunique feature of earnings revision as compared with restatements Similarly, we examinethe lag between the preliminary earnings announcement and the subsequent SEC filing forRevisers and non-Revisers in Quarter t 1 4 (because many revisions occur in the fourthfiscal quarter) The lag had decreased by more than 2 days on average for Revisers thannon-Revisers (statistically significant), indicating that managers of Revisers delay the vol-untary disclosure of preliminary earnings closer to the SEC filing

We also investigate whether the market reactions differ for Revisers with upward versusdownward earnings revisions In Panel B of Table 5, we find that the significant marketreactions to the additional earnings surprises in SEC filings, ASURP, are only significantwhen earnings are revised downward When earnings are revised upward, there is an insig-nificant association between the SEC filing returns and the additional earnings surprise,ASURP This may be related to market participants’ suspicions about the strength of theaccounting and reporting system of firms with an earnings revision, even if the revision issupposed to convey ‘‘good’’ news to the market Hence, there is a positive and significantassociation between market returns around the SEC filings and the additional earnings sur-prise when firms revise earnings, although this association is mainly concentrated in firmsthat revise earnings downward

Table 6 presents the two-stage least squares (2SLS) parameter estimates in column 2and the ordinary least squares (OLS) parameter estimates in columns 3 and 4 As a firmcan appear several times during the sample period, we use robust standard errors clustered

by firm to control for the dependence among the regression residuals for the OLS mates The dependent variable in this table is the total return from a day before the prelimi-nary earnings announcement through one day after the SEC filing date This windowcaptures the effects of the preliminary earnings surprise and the new information thatappears in SEC filings For non-Revisers, the independent variables include the earningssurprise, TSURP, which will be equal to the preliminary earnings surprise For Revisers, it

esti-is equal to the sum of the preliminary earnings surpresti-ise, FSURP, and the additional ings surprise in the SEC filing, ASURP If the market does not penalize Revisers, the

earn-Table 6 Market Reactions to Total Earnings Surprises From the Preliminary Earnings AnnouncementThrough the SEC Filing

Variables

Full sample(2SLS)

Full sample(clustered by firm)

Full sample(clustered by firm)

2004 TSURP is the total earnings surprise, consisting of FSURP plus ASURP FSURP, the First (Preliminary) Surprise, is calculated as the Compustat preliminary quarterly earnings (mnemonic IBQ) minus the As-First- Reported (AFR) Compustat quarterly earnings for t 2 4, scaled by market value at quarter-end ASURP is the Additional surprise, calculated as earnings in the SEC filing (AFR Compustat earnings) minus earnings reported in the preliminary earnings release (preliminary Compustat earnings), scaled by market value at quarter-end By defini- tion, it is zero for control firms p values (in italics) are shown below each estimate Bolded entries are significant below 5%.

All firms are ranked according to TSURP each quarter and are assigned into 10 deciles TSURP in the table sents the decile rank divided by 9 minus 0.5 Thus, the coefficient on TSURP measures the abnormal buy-and-hold return on a ‘‘hedge’’ portfolio holding long (short) positions in the most positive (negative) decile of total surprises TSURP 3 REVISER is an interaction variable of TSURP and REVISER, and is used to assess the additional market reaction to the total surprise of Revisers as compared with that of other firms For the definition of other control variables, see the notes to Table 1 The table reports the results of a two-stage least squares (2SLS) where REVISER is an endogenous variable and is estimated by the control variables and is then used as an independent variable in the market reactions regression, as well as two separate ordinary least squares (OLS) regressions, with and without controls.

repre-coefficient on the interaction term TSURP 3 REVISER should be insignificantly differentfrom zero If we find that the coefficient on TSURP 3 REVISER is negative and signifi-cant, it indicates that the market reacts less strongly to a dollar of earnings surprise forRevisers, suggesting a penalty for their lower earnings quality The regression also controlsfor variables that were previously shown to affect earnings revision.

We find that the coefficient on REVISER is negative and significant The coefficient onTSURP 3 REVISER is statistically insignificant, which suggests that the total market reac-tion to the total earnings surprise does not differ significantly when an earnings revisionoccurs This is consistent with no penalty for earnings revisions Our OLS regression resultsare consistent with our findings using a 2SLS model

Table 7 presents the average differences in excess total returns (preliminary earningsannouncement through SEC filing) for firms with earnings revisions and matched controlfirms We find a negative and significant (t statistic = 23.35, 24.06, and 22.79) differencebetween Revisers and control firms, when matched on (a) total earnings surprise and indus-try, (b) total earnings surprise and size, and (c) earnings surprise, size, and number of seg-ments, which suggests that firms that had earnings revisions have weaker market reactionsthan firms without earnings revisions This weaker market reaction indicates that marketparticipants likely interpret the earnings revision as a weakness in the quality of earnings orthe firms’ reporting and control systems We find no significant difference between the twowhen matched on earning surprise, size, and industry, although the returns are lower forRevisers than non-Revisers When we compare a Reviser with itself in another quarter with

no revision and with the same earnings surprise decile rank, we find that market reactionsduring the earnings revision quarter are significantly weaker (t statistic = 24.16), likelydue to the penalty imposed on Revisers for their lower earnings quality These results areinconsistent with those in Table 6, which compare the Revisers with all non-Reviser firms

Sensitivity Analyses

We delete all Revisers with more than one revision This is likely to eliminate cases whereCompustat did not find sufficient data in the preliminary earnings release but was able toreport a different income figure when the SEC filings became available The main resultsare virtually the same; there is a positive and significant market reaction to ASURP, with

no market reaction to FSURP after controlling for information between the preliminary

the total earnings surprise of Revisers than to those of non-Revisers

We examine the results in the pre- and post-Regulation Fair Disclosure periods Firmsmay behave differently after the year 2000, where private channels of communication withspecial groups of investors are restricted Nevertheless, our main results hold in the periodthrough 2000 and during the years 2001 to 2009

We repeat our main tests separately for earnings revisions in the fourth quarter, when anaudit is required, and all other quarterly earnings revisions as one group For fourth quarterobservations, the logistic regression coefficients are all in the same direction as our mainresults, but the debt, size, and persistence of earnings are insignificant The additional earn-ings surprise is positively associated with abnormal returns around the filing date, but themarket reaction to the total earnings surprise, although weaker than for non-Revisers, is notstatistically different For earnings revisions in other quarters, the logistic regression coeffi-cients are all in the same direction and significant as our main results, except for earningsvolatility, which is insignificant The additional earnings surprise for other quarterly

revisions is positively and significantly associated with abnormal returns around the filingdate, and the market reactions to the total earnings surprises of Revisers are significantlyweaker than for non-Revisers.

Next, all Revisers’ public announcements between the preliminary earnings ment and the subsequent SEC filing dates were examined to determine whether the firmissued a press release that announced the upcoming earnings revision in the SEC filing.Only 102 Revisers with sufficient return information issued such revision announcements,with 68 (66.7%) of them having a negative additional SEC earnings surprise (as comparedwith 62.2% for all nonannouncing Revisers) Forty-two revision announcements occurred

results for the nonannouncers are very similar to those reported for the main sample Even

Firms (CARtotal2matched)

Total earnings surprise and

Total earnings surprise, size, and

number of segments

Same firm in a different quarter,

matched by total earnings

2 CAR total , CAR total2matched : the buy-and-hold abnormal return from 1 day before the preliminary earnings announcement through 1 day after the Securities and Exchange Commission (SEC) filing date.

3 Matching techniques (a) matching by earnings surprise decile and industry; (b) matching by size (size difference is capped at 10%) and earnings surprise decile; (c) matching by size, earnings surprise decile, and the number of seg- ments; (d) matching by size, earnings surprise decile, and industry; and (e) matching the Reviser with itself in another quarter that has the same total earnings surprise decile rank.

4 Excess returns are measured in the 3-day period (21, 11) around the release of preliminary earnings Excess buy-and-hold returns are calculated as the buy-and-hold return from Center for Research in Security Prices (CRSP) minus the buy-and-hold return on the portfolio of firms with the same size (market value of equity) and book-to-market (B/M) ratio Daily returns and cutoff points on the size and B/M portfolios are obtained from pro- fessor Kenneth French’s data library, based on classification of the population into six (two size and three B/M) portfolios Observations in the top and bottom 0.5% of excess return are deleted from the sample to ensure that our results are not driven by outlying returns Portfolio returns are computed each quarter The table presents average quarterly returns and t statistics based on a Fama and MacBeth (1973) approach.

TSURP is the total earnings surprise, consisting of FSURP plus ASURP FSURP, the First (Preliminary) Surprise, is calculated as the Compustat preliminary quarterly earnings (mnemonic IBQ) minus the As-First-Reported (AFR) Compustat quarterly earnings for t 2 4, scaled by market value at quarter-end.

ASURP is the Additional surprise, calculated as earnings in the SEC filing (AFR Compustat earnings) minus earnings reported in the preliminary earnings release (preliminary Compustat earnings), scaled by market value at quarter- end By definition, it is zero for control firms.

All firms are ranked according to TSURP each quarter and are assigned into 10 deciles.

for those firms that announced an upcoming earnings revision in the SEC filing, weobserve a positive and significant market reaction to the additional SEC earnings surprise.

We examine the sensitivity of our results to the calculation of the earnings surprise asbeing either based on Institutional Brokers Estimate System (IBES) forecasts or on time-series forecasts when the firm is not followed by analysts Our main results are stronger forthe subsample of firms followed by at least one analyst; the market reaction to the addi-tional earnings surprise is positive and significant and there is a significantly weakermarket reaction to the total earnings surprise of Revisers as compared with non-Revisers

We examine the robustness of our results to whether revisions are to recurring or curring items As expected, when the revisions are to recurring items the reduction inmarket reactions to the total earnings surprise of Revisers as compared with non-Revisers

nonre-is more pronounced than for nonrecurring items

We examine the sensitivity of the results to the various reporting lags between Revisersand control firms, with no significant role for any reporting lag Specifically, we find thethat Revisers have a mean (median) lag of 32 (28) days between the quarter-end and thepreliminary earnings announcement, as compared with 29 (27) for control firms Similarly,the lag from the preliminary earnings announcement to the SEC filing has a mean (median)

of 32 (28) for Revisers and 24 (20) for control firms However, in spite of the greater lagsfor Revisers when compared with control firms, these lags are not statistically significant inany of the return regressions we use in the study

Summary and Conclusions

In this article, we provide evidence on the occasions where earnings revisions are morelikely to occur, whether the market reacts to these additional earnings surprises and whethermarket reactions to the total earnings surprises are weaker, likely due to indications oflower earnings quality or imperfections in the firm’s reporting and control systems

Contrary to what might be expected, the sample of earnings revisions we use does notcontain just minor, insignificant corrections to the previously issued earnings figure Themedian company reports an earnings revision of more than 2.9% of market value of equity atquarter-end More than 25% of our revisions are above 0.7% of market value, and more than10% are above 2.6% of market value These are not trivial numbers and potentially can (and

in fact do) cause significant market reactions We find conflicting results about the marketreactions to the mere act of an earnings revision When we compare the total market reac-tions with the preliminary and subsequent additional surprise of Revisers and the full popula-tion of non-Revisers, we find no significantly weaker market reactions to Revisers However,

in matched-pair designs (including using the revising firm as its own control in other ters), we find significantly weaker market reactions, consistent with potentially lower quality

quar-of earnings We find that earnings revisions are more likely to occur for firms that are morecomplex in nature, are more financially leveraged, have greater earnings volatility, and havelosses, and where an auditor change has occurred We also find that most of the revisionsoccur in the sales, tax, and cost of sales component of earnings

Our results contribute to the literature on earnings quality, showing another tion of a potentially deficient accounting/control system within a firm Researchers thatstudy market reactions around SEC filings should specifically exclude cases of earningsrevisions, which likely contaminate the sample as a whole (e.g., Callen et al., 2006) Thisstudy goes further than Hollie et al (2005) by showing that mangers cannot ‘‘game thesystem’’ by initially reporting higher earnings that are subsequently revised in SEC filings;

manifesta-market participants may actually attribute lower earnings quality to such firms Moreover,the results of this study point out a potentially useful requirement for regulators to consider:The SEC may require companies that revise their earnings from the preliminary earningsrelease to identify it more prominently, possibly through a Form 8-K filing just like restate-ments Although we find that the market reacts to the new information in the SEC filings,some investors may benefit from this information at the same time.

Appendix A

Examples of Upward and Downward Earnings Revisions

Upward Revision

Firm: R&B, Inc (NASDAQ: RBIN)

Period: Second quarter ended June 30, 2002

Preliminary Earnings Announcement Date: July 31, 2002

Earnings Announcement: R&B, Inc., announced financial results for the second quarterand 6 months ended June 29, 2002 For the second fiscal quarter ended June 29, 2002,sales increased 6.6% to US$55.5 million from US$52.0 million in the same period lastyear Net income in the second fiscal quarter was US$2.8 million, or US$0.31 per fullydiluted share, compared with net income of US$1.6 million, or US$0.18 per fully dilutedshare, excluding goodwill amortization in the prior year For the 6 months ended June 29,

2002, sales increased 8.5% to US$106.5 million from US$98.2 million in the same periodlast year Net income in the 6 months ended June 29, 2002, was US$4.9 million, orUS$0.55 per fully diluted share, compared with net income of US$2.0 million, or US$0.24per fully diluted share, excluding goodwill amortization in the prior year

SEC Filing Date: August 8, 2002

10-Q Reported Earnings: In the filing, the company’s second quarter 2002 net income hasbeen revised to US$4.115 million There is no indication in the 10-Q why earnings wererevised upward from preliminary earnings

Downward Revision

Firm: MEMC Electronic Materials, Inc (NYSE: WFR)

Period: Second quarter ended June 30, 2001

Preliminary Earnings Announcement Date: August 10, 2001

Earnings Announcement: The company reported a net loss of US$65.8 million, orUS$0.95 per share, before the restructuring charge of US$22.3 million, compared with anet loss of US$17.6 million, or US$0.25 per share, for the quarter ended March 31, 2001.Net loss for the quarter ended June 30, 2000, was US$15.5 million, or US$0.22 per share.The higher net loss in the 2001 second quarter compared with the 2001 first quarter is pri-marily due to significantly lower product volumes, moderately lower pricing, and the dis-continued recognition of income tax benefits related to the company’s U.S net operatingloss carryforwards effective April 1, 2001

The company incurred an operating loss of US$51.5 million in the 2001 second quarter,excluding restructuring charges, compared with an operating loss of US$6.6 million in thequarter ended March 31, 2001 In the year-ago period, the company incurred an operatingloss of US$9.4 million.

SEC Filing Date: August 13, 2001

10-Q Reported Earnings: In the filing, the company raised the valuation allowance related

to deferred tax assets in the amount of US$267 million As a result of this noncash ment, the company’s second quarter 2001 net loss has been revised to US$355.3 million, orUS$5.10 per share

adjust-Acknowledgments

The authors gratefully acknowledge the preliminary and original Compustat quarterly data provided

by Charter Oak Investment Systems, Inc The authors are also grateful for the contribution ofThomson Financial for providing forecast data available through the Institutional Brokers EstimateSystem These data have been provided as part of a broad academic program to encourage earningsexpectations research The authors also gratefully acknowledge the SEC filing dates data provided byCompustat The authors benefited from comments made by Eli Amir, Benzion Barlev, Paul Griffin,Nicole Jenkins, Michael Kimbrough, Chandra Seethamraju, Dan Segal, Eyal Sulganik, KirstenAnderson (AAA conference discussant), and Anne Beatty (JAAF conference discussant), and seminarparticipants at Case Western Reserve University, Fordham University, New York University,Pepperdine University, University of California2Davis, University of Colorado–Boulder, University

of Houston, University of South Florida, Washington University in St Louis, the 2005 AmericanAccounting Association Annual Conference, Finance and Accounting in Tel-Aviv 10th AnnualConference, and the 2005 Journal of Accounting, Auditing and Finance Conference

Authors’ Notes

JEL classification: G34 Data Availability: All data sources are identified in the article

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/

or publication of this article

2 The filing date is 45 days for 10-Q and 90 days for 10-K Beginning with fiscal years ending on orafter December 15, 2003, some companies are subject to a phased-in accelerated filing schedule.

millions of dollars with one decimal in their preliminary announcement and then provided amore exact number in the SEC filing

with cases where firms ‘‘forced’’ their auditors to go along with the preliminary earningsannouncement, at least until the Form 10-K is filed at year-end

5 The preliminary earnings in this study should not be confused with ‘‘street’’ or proforma ings This study uses Generally Accepted Accounting Principles (GAAP) earnings as disclosed

earn-in the prelimearn-inary earnearn-ings release

the fourth quarter that ends on September 30, 2000 On November 21, 2000, it issued anotherpress release that due to a revenue recognition error, the preliminary revenues and Earnings pershare (EPS) had been overstated On December 1, 2000, a class action was filed against Lucentalleging, among other things, that Lucent knew (or was reckless in not knowing) about this over-statement at the time of the preliminary announcement Lucent is the only firm we found where

an earnings revision resulted in a class action suit

Leitch, and Neter, 1981; Kinney, 1979; and Wright et al., 1989

8 The degree of analyst following may also be positively associated with the likelihood of an ings revision due to the desire to meet or beat analyst forecasts in the preliminary announcement

earn-9 Stice (1991) conducts this study at a time when SEC filings were not as readily available (e.g.,online and other media) as they are today Chung, Jacob, and Tang (2003) examine the same issuewhen filings were available on Electronic Data Gathering, Analysis, and Retrieval (EDGAR), butuse only a handful of quarters from the beginning of the EDGAR database Their findings seem tosuggest that Stice results hold true even with the availability of the SEC EDGAR database

winsoriz-12 We do not classify taxes as recurring or nonrecurring because it may include both components

13 In untabulated results, we ran the regression reported in Panel A without the buy-and-hold abnormalreturn between preliminary earnings announcements and SEC filings (CARaf) and without the prelim-inary earnings surprise (FSURP) In all of these regressions, the coefficient on the additional earningssurprise ASURP is positive and significantly different from zero around the SEC filing date

14 In untabulated results, we ran additional analysis on the lags

provisions, responses to new SEC guidance, and admitted accounting errors and auditor-forcedrevenue recognition changes

References

Amir, E., & Livnat, J (2006) The effects of disclosure medium, timing, and detail on market reactions

to earnings announcements (Working paper) New York University, New York City

Asthana, S., Balsam, S., & Sankaraguruswamy, S (2004) Differential response of small versus largeinvestors to 10-K filings on EDGAR Accounting Review, 79, 571–589

Balsam, S., Bartov, E., & Marquardt, C (2002) Accruals management, investor sophistication, andequity valuation: Evidence from 10-Q filings Journal of Accounting Research, 40, 987–1012.Callen, J., Livnat, J., & Segal, D (2006) Information content of SEC filings and information environ-ment: A variance decomposition analysis Accounting Review, 81, 1017–1043

Chen, S., DeFond, M., & Park, C W (2002) Voluntary disclosure of balance sheet information inquarterly earnings announcements Journal of Accounting & Economics, 33, 229–251.

Chung, K., Jacob, R., & Tang, Y B (2003) Earnings management by firms announcing earningsafter SEC filing International Advances in Economic Research, 9, 152–162

Cready, W., & Mynatt, P (1991) The information content of annual reports: A price and tradingresponse analysis Accounting Review, 66, 291–312

DeFond, M., & Hung, M (2003) An empirical analysis of analysts’ cash flow forecasts Journal ofAccounting & Economics, 35, 73–100

Easton, P., & Zmijewski, M E (1993) SEC form 10K/10Q reports and annual reports to ers: Reporting lags and squared market model prediction errors Journal of Accounting Research,

reac-Hylas, R E., & Ashton, R (1982) Audit detection of financial statement errors Accounting Review,

28, 2000 Retrieved from http://www.sec.gov/news/testimony/ts152000.htm

Livnat, J., & Tan, C (2004) Restatements of quarterly earnings: Evidence on earnings quality andmarket reactions to the originally reported earnings (Working paper) New York University, NewYork City

Palmrose, Z., Richardson, V J., & Scholz, S (2004) Determinants of market reactions to restatementannouncements Journal of Accounting & Economics, 37, 59–89

Penman, S H (1987) The distribution of earnings news over time and seasonalities in aggregatestock returns Journal of Financial Economics, 18, 199–228

Qi, D., Wu, W., & Haw, I (2000) The incremental information content of SEC 10-K reports filedunder the EDGAR system Journal of Accounting, Auditing & Finance, 15, 24–46

Richardson, S., Tuna, I., & Wu, M (2002) Predicting earnings management: The case of earningsrestatements (Working paper) University of Pennsylvania, Philadelphia

Soffer, C L., Thiagarajan, S R., & Walther, B (2000) Earnings preannouncement strategies Review

buy-ÓThe Author(s) 2012 Reprints and permission: sagepub.com/journalsPermissions.nav DOI: 10.1177/0148558X11409143

http://jaaf.sagepub.com

The Impact of Nonaudit

Services on Capital Markets

Abstract

The framers of the Sarbanes–Oxley Act (SOX) presume that nonaudit services (NASs)lower the quality of financial statements, so they have prohibited auditors from offeringmost NASs In addition, regulators believe that NASs may cause the auditor to be perceived

as ‘‘dependent’’ in appearance, thus increasing information risk, even if they have no impact onthe quality of financial statements The authors investigate two hypotheses using pre-SOXdata First, they ask whether the proportion of NASs fees to total fees has a positive or nega-tive association with the ability of financial statements to predict a firm’s future cash flows,which can be considered a measure of the quality of the statements Second, they ask whetherthe proportion has a negative or positive association with the cost of capital (COC) and thebid/ask spread, controlling for the predictability The COC and the bid/ask spread serve asproxies for information risk Contrary to the proponents of prohibiting NAS, the authors findthat the proportion of NASs fees to total fees has a positive association with predictability.Controlling for the quality of financial statements, NASs still have a negative association withthe COC and the bid/ask spread as proxies for information risk These findings suggest thatNASs improve the quality of financial statements and reduce information risk

Keywords

nonaudit services, quality of financial statements, capital markets

The issue of nonaudit services (NASs) and their effect on auditor independence has beendebated for more than 50 years In 1990, 71% of the Big Five revenues were derived fromaccounting and auditing; by 1999, that percentage had declined to 48%, whereas consultingrevenue grew from 12% to 32% of total Big Five revenue That is, the ratio of auditing rev-enues to consulting revenues dropped from 6 to 1 in 1990 to 1.5 to 1 in 1999 Responding

to this trend and to the suspicion that the increased proportion of revenue derived from sulting might compromise auditor independence, the Sarbanes–Oxley Act of 2002 (SOX)

the effects of NAS on audit or financial statement quality includes both support for and tique of the NAS ban (see the appendix for a summary of a critical subset of the literature).

cri-A majority of the studies—conducted before the Securities and Exchange Commission(SEC) started diverting its attention to NAS in 2000—failed to find evidence that proxiesfor auditor independence were adversely affected by NAS In fact, some research (e.g.,Kinney, Palmrose, & Scholz, 2004) suggests that the provision of tax services might bepositively associated with proxies of audit quality potentially implying beneficial ‘‘knowl-edge spillovers’’ between some NASs and the audit However, regardless of whether thebundling of services (audits and nonaudit) compromises independence, the appearance ofconflict may well matter both to policy makers and investors Indeed, policy options wouldhave to deal also with the perception problem

This study contributes to the debate by focusing on the ‘‘real’’ and the ‘‘perceived’’effects of NAS on two important dimensions hitherto unexplored in the literature: first, theoverall quality of financial statements proxied by the ability to predict cash flows—thestated objective of financial statements in the Financial Accounting Standards Board’s(FASB) Conceptual Framework (FASB, 1978) and analysts’ forecast accuracy; and second,the perceived credibility of financial statement information—possibly compromised by thereal or perceived adverse effects of NAS on auditor independence—as proxied by measures

The first dimension, financial statement quality, is chosen because what matters most inassessing the desirability of NASs is in how it affects the overall quality of the financial state-ments, not just the quality of the audit In fact, as it happens, audit quality is difficult if notimpossible to measure: Most proxies used for audit quality include components of both auditquality and the quality of unaudited financial statements As a proxy for the overall quality ofthe financial statements, we use the ability to predict cash flows based on financial statements,specifically income statement numbers This choice is predicated on the prediction of cashflows being the major objective of financial statements (Statement of Financial Concept No 1:Financial reporting should provide information that helps present and potential investors andcreditors and other users in assessing the amounts, timing, and uncertainty of prospective cashreceipts): Financial statements are of high quality if they attain their major objective

The second dimension, the perceived credibility of financial statement information,overlaps the first (financial statement quality) but only partially: Even if financial state-ments are of high quality, supporters of the ban claim that the appearance of auditor inde-pendence must be considered as well Specifically, if the auditor is perceived to benonindependent, investors would attach little credibility to the financial statements he orshe attests to Diminished credibility would increase the perceived risk of and the uncer-tainty surrounding the use of the information, thus resulting in a higher COC, higher bid/

have investigated the impact of NAS on various proxies for audit quality, this is the firststudy to our knowledge that examines the impact of NAS on capital market indicia such asthe COC and the bid/ask spread (see also Figure 1)

Using NAS data for the years 2001 to 2003, the study tests two hypotheses: (a) NASsare positively (negatively) associated with the predictability of cash flows and with ana-lysts’ forecast accuracy and (b) controlling for the predictability of cash flows, NASs are

NAS fees to total fees is the metric used to proxy for the intensity of NAS For proponents

of the ban on NAS, the metric reflects the degree to which audit independence is

compromised For opponents of the ban, this metric represents the potential synergybetween audit and NASs: Engaging in NAS helps the auditor become less dependent oninformation from the client, hence contributing to fewer errors or biases in the financialstatements.

Figure 1 The impact of nonaudit services on the quality of financial statements and capitalmarkets—Different views

We find that NASs are positively associated with predictability and analysts’ forecast racy, and negatively associated with the COC and bid/ask spread Hackman’s two-step procedurewas also employed to address potential self-selection bias, yielding similar results Although it isgenerally difficult to derive normative prescriptions from empirical findings, our results nonethe-less seem to reinforce the opponents of the ban: NAS appears to improve the predictability ofcash flows based on financial statement information and analyst forecast accuracy—contributingtoward attaining the major objective of financial statements as well as enhancing the credibility

accu-of the financial statements as manifested in the COC and bid/ask spread

A caveat is in order Data on profits of audit firms are not available; hence, we and allresearchers before us had to resort to data on revenues The problem is that presumably,audit firms are more affected by profit than revenue The latter is only a good proxy forprofit if costs make up a fairly constant proportion of revenue; this is not an innocuousassumption, as it is less than likely that profit margins remain constant across engagements,

com-parable with an additional dollar of audit services revenue

The rest of the article is organized as follows The section titled ‘‘Recent Literature’’provides background Hypotheses are developed in the section titled ‘‘Hypotheses,’’ andthe research design is laid out in the section titled ‘‘Research Design.’’ The ‘‘Sample andDescriptive Statistics’’ section describes the data, and univariate tests are presented in thesection titled ‘‘Univariate Tests.’’ The section ‘‘Multivariate Results’’ describes the mainresults Robustness tests are described in the section titled ‘‘Robustness.’’ The

‘‘Conclusion’’ section concludes the article

Recent Literature

Controversy over NAS escalated as fees for such services rose dramatically in the late 1990s.This, coupled with the growing incidence of restatements over the 1990s, gave birth to anSEC rule requiring disclosure of both the audit fees and fees for Financial InformationSystem Design and Implementation (FISDI) and other NASs (SEC, ‘‘Final Rule,’’ 2000).Using the newly available data as a result of the rule, numerous studies followed with ambig-uous results DeFond, Raghunandan, and Subramanyam (2002) found no association betweenNAS and going concern audit opinions However, Frankel, Johnson, and Nelson (2002)found NAS to be positively associated with beating analysts’ forecast by a very small marginand with discretionary accruals, but at the same time, Chung and Kallapur (2003) failed tofind any association with abnormal accruals for the Big Five audit firms Antle, Gordon,Narayanamoorthy, and Zhou (2006) and Ashbaugh, LaFond, and Mayhew (2003) were notable to find an evidence of earnings management, although Larcker and Richardson (2004)documented a positive association with the absolute value of accruals Kinney et al (2004),although unable to find an association between FISDI and the incidence of restatements,documented a positive (negative) association between other NAS (tax services) and the inci-

association between restatements and unexpected NAS, fee ratios, and total fees Turning toperceived independence, Lindberg and Beck (2004) found in a survey that Certified PublicAccountants considered NAS other than FISDI more harmful to audit independence and theappearance of independence after the accounting scandals

The SEC in 2002 placed new restrictions on the provision of NAS Subsequently, SOX

in 2002 made it unlawful for accounting firms to offer a list of specific NAS to audit ents The audit committee must now approve the utilization of all NAS that were not

cli-explicitly banned, including tax services.7Despite this ban, a vigorous debate has continued

Opponents of the ban point to the difficulty researchers have had in establishing a linkbetween NAS and loss of auditor independence (Romano, 2005) and to the potential reduc-tion of economies of scope that could improve the quality of both audits and managementconsulting Supporters of the ban claimed that NAS might inhibit audit independence byaffording management a way to exert pressure on auditors in ways that are invisible to themarket (e.g., Coffee, 2002, 2003; Davis & Hollie, 2004) and that the threat of cutting off

actually impairs independence, supporters of the ban insist that the appearance of auditor

Collectively, the studies failed to offer a clear-cut answer to the question of whether NASdegrades audit independence, whether in substance or in perception Notably, in the absence

of visibility of audit quality, the studies resorted to proxies on overall financial statementquality, which is a joint product of audit quality and the clients’ own quality of financialreporting controls As explained in the introduction, overall financial statement quality is theappropriate criterion to judge the desirability of NAS, as the former is ultimately what mat-ters However, even as proxies for overall financial statement quality, the variables usedconvey ambiguities: As mentioned, restatements may reflect either good quality of financialstatements (in that they are ultimately corrected) or bad (in that they needed to be restated inthe first place); going concern qualifications may reflect not the quality of audit or financialreporting controls but simply adverse business conditions that give rise to the qualification;and, as already indicated, audit fees may signal conflicting scenarios (see Note 5 in the Notessection) In this study, we focus on (hitherto unexamined) unambiguous proxies for overallfinancial statement quality and credibility: predictability of cash flows, accuracy of analystforecasts, COC, and bid/ask spread

Hypotheses

In light of the ambiguity inherent in the findings of past studies, our first hypothesis isintended more definitively to ascertain whether NAS improved or degraded financial state-ment quality as manifested in the ability based on financial statement information to predictcash flows (the objective of financial statements) utilizing both a mechanical predictionmodel and analyst forecasts An improvement could result from synergies in joint contribu-tion of audit and NASs; degradation could result from impairment of independence and theconsequent injection of biases and errors into the financial statements Controlling for pre-dictability, we then ask whether NAS has a residual effect on measures of information risk(including information asymmetry; see, for example, Diamond & Verrecchia, 1991)induced by diminished credibility of the financial statement information stemming fromeither real or perceived impairment in auditor independence; specifically, as measures ofinformation risk we use the COC and bid/ask spread As we cannot ex ante predict whetherNAS has a baneful or baleful impact, the hypotheses formulated in the alternative form are

as follows:

Hypothesis 1 (H1): Predictability

a NASs have a nonzero association with cash flow prediction errors

b NASs have a nonzero association with analysts forecast errors

Hypothesis 2 (H2): Information risk

a NASs have a nonzero association with the COC

b NASs have a nonzero association with bid/ask spread

Measures of Predictability

Cash Flow Projection Errors

To predict cash flows, several researchers such as Dechow, Kothari, and Ross (1998) andBarth, Donald, and Nelson (2001) regressed Cash Flow From Operations (CFO) on priorCFO and components of accruals in a pooled cross-sectional model that assumes constantcoefficients in cross-section to generate within-sample predictions (The appendix sum-marizes the extant literature on the prediction of cash flows.) For this study, we use instead

an ordinary income-based, out-of-sample predictor (Ronen & Sadan, 1981) computed asfollows:

Net income before extraordinary items

Predictability is measured by the absolute deviation of the ordinary income-based

This estimation of prediction error has two major advantages: (a) the prediction is firmspecific and (b) the resulting prediction error is, on average, smaller than that of other,

Analysts Forecast Errors

Analysts often use sophisticated, firm-specific models to predict a firm’s future profitbased on financial statement information and impound nonfinancial statement information

as well in their forecasts If auditors’ real or perceived lack of independence degrades thequality of the information embedded in the financial statements and/or diminishes the cred-ibility analysts attach to the statements, the result would be reliance on biased/noisier infor-mation and/or diminished reliance on otherwise relevant information This in turn wouldresult in a higher forecast error Our metric for analysts forecast error is AFE defined as

Actual EPS t11 Analysts0Consensus Forecast t11

Actual EPS t11



Measures of Information Risk or Noncredible Information—COC

We estimate the implied COC derived from a reduced form of Ohlson and Nauroth (2005; which assumes a temporally constant dividend payout ratio) as follows:

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffig1

where EPS1is expected Earnings Per Share (EPS) 1 year ahead and EPS2is expected EPS

2 years ahead g is the perpetual growth rate of firm, plus one, and in this study, g is set to1.04

Bid/Ask Spread (SPREAD)

Information risk resulting from real or perceived auditors’ lack of independence can giverise to perceived information asymmetry, which, in turn, widens the bid/ask spread (e.g.,Copeland & Galai, 1983; Glosten & Milgrom, 1985) We use the quoted percentage bid/

1 (ASK 1BID)

NAS Fees

The presented results in this article utilize the metric used in studies such as DeFond et al.(2002) and Frankel et al (2002), namely, the proportion of NAS fees over total fees(NASRATIO) as a measure of the intensity of NAS, where total fees are the sum of auditfees and NAS fees Our results do not change qualitatively when we use other measuresused in prior research such as the ratio of NAS fees over audit fees or a rank based onNASRATIO (Frankel et al., 2002) and the ratio of NAS fees to the audit firm’s total U.S.revenue (Chung & Kallapur, 2003)

Research Design

Test of H1a: Cash Flow Projection Errors

We estimate the following model:

an interactive variable of the indicator and the ratio, respectively We expect BTM to loadpositively but a negative ratio to load negatively in the regression STDROE is the standarddeviation of ROE included to capture the difficulty of predicting future performance whenpast performance is more volatile and is expected to be positively associated with AbsDEV/ TA (using the standard deviation of earnings per share or the growth in earnings per shareyields qualitatively similar results) LEVERAGE could either reflect fewer risky projects(Myers, 1977) implying enhanced predictability but at the same time higher financial risk;hence, we cannot predict the sign of this variable’s association with AbsDEV / TA.AbsDACC is the absolute value of discretionary accruals scaled by total assets derived

from the modified-Jones model The effect of this variable is ambiguous: Although it couldreflect opportunistic earnings management that complicates the task of predicting cashflows, it could also be used by management to signal future (genuine) expectations (seeBadertscher, Collins, & Lys, 2007) SEGMENT is the number of segments in a firm used

as a proxy for operating complexity expected to be associated positively with AbsDEV /

TA NUMEST is the number of analysts following the firm proxying for the informationenvironment expected to be negatively associated with AbsDEV / TA LOSS is an indicatorfor when earnings after extraordinary items are negative; Hayn (1995) and Ertimur (2004)implied that firms with losses are characterized by a higher degree of information asymme-try expected to be positively associated with AbsDEV / TA Big5/4 is an indicator variablefor firms hiring Big 5/4 audit firms, expected to be negatively associated with AbsDEV /TA; there is evidence that the market perceives the audit quality of the Big5/4 to be higher(Krishnan, 2003b; Teoh & Wong, 1993), that the Big 5/4 audit clients have higher qualityearnings because the auditors are more likely to issue modified audit opinions (Francis &Krishnan, 1999, 2002), and that these clients have smaller discretionary accruals (Becker,Defond, Jiambalvo, & Subramanyam, 1998; Francis, Maydew, & Sparks, 1999; Krishnan,2003b) Due to Arthur Andersen’s collapse in 2002, the Big 5 became Big 4; as our samplecovers the period between 2000 and 2002, the code Big 5/4 refers to Big 5 prior to 2002and to Big 4 in 2002 AUDITOPINION is an indicator variable that equals one in the casethat the auditor issues a clean opinion and zero in the case of a going concern qualification,expected to be negatively associated with AbsDEV / TA; a going concern qualification

TA of firms in the same industry (in which the firm represented by the dependent variableoperates), excluding the firm itself, controls for industry effects: More stable industriessuch as utilities afford better prediction A complete definition of all the variables is pro-vided in Table 1

Test of H1b: Analysts Forecast Errors

The estimated model is as follows:

1b4INEGBTM it3BTMit1b5STDEVit1b6LEVERAGEit

LEVERAGE, AbsDACC, SEGMENT, NUMEST, LOSS, Big5/4, and AUDITOPINION) areincluded as in Equation 1a Instead of STDROA in Equation 1a, we include STDEV, the

ana-lysts’ forecasts error of all other firms in the same industry (to which the firm represented

by the dependent variable belongs) to control for an industry effect

Table 1 Definitions of the Variables

total fees paid to auditor

Market capitalization t



the ordinary income-based CFO predictor

Total Assets t



the audit opinion includes a going concern opinion

of Cost of capital

represented by the dependent variable, excluding the firm)

or postretirement plan in Year t, 0 otherwise

or debt during the Fiscal Year t, 0 otherwise

operations in Year t, 0 otherwise

\ 0, 0 otherwise

extraordinary items \ 0

acquisitions during the Fiscal Year t, 0 otherwise

Year t

NASDAQ, 0 otherwise

fees are greater than 0, 0 otherwise

year starting from 3 months after the end of the Fiscal Year t

(continued)

Test of H2a: COC

The estimated model is as follows:

The dependent variable, COC, is the COC based on the Ohlson and Juettner-Nauroth

Fiscal Year t 1 1

Among the independent variables, we include AbsDEV / TA to control for the predictability

of cash flows so as to gauge the residual effect of NAS If information risk decreases because

of NASs after the effect on financial statement quality (predictability of cash flows) has beenaccounted for, the coefficient on NASRATIO in Equation 2a would be negative We expect thepredictability (AbsDEV / TA) to be negatively associated with the COC (positive coefficient onAbsDEV / TA): greater predictability injects smaller information risk The remaining indepen-dent variables are documented determinants of COC: risk factors (Gode & Mohanram, 2003),information environment (Botosan, 1997; Botosan & Plumlee, 2002), and industry effect(Gebhardt, Lee, & Swaminathan, 2001) Specifically, as risk factors we include (a) beta(derived from an ordinary least squares [OLS] regression over the previous 60 months); (b)UNSYST, the standard deviation of the residual of the market model over the past year

Table 1 (continued)

special items at Year t, 0 otherwise

1 (ASK1BID), measured over 1 yearstarting from 3 months after the end of the Fiscal Year t

represented by the dependent variable, excluding the firm)

where the unsystematic return is the residual of the market model

starting from 3 months after the end of the Fiscal Year t