Effects on comparability and capital market benefits of voluntary IFRS adoption

Adopting firms with higher comparability withadopted firms have greater capital market benefits after adopting IFRS than adopting firms with lower comparability,and capital market benefi

Vol 3, No 1 DOI: 10.2308/jfir-52279Fall 2018

pp 1–22

Effects on Comparability and Capital Market Benefits of

Voluntary IFRS Adoption

Mary E Barth Stanford University

Wayne R Landsman Mark H Lang The University of North Carolina at Chapel Hill

Christopher D Williams University of Michigan

ABSTRACT:This study addresses whether voluntary IFRS adoption is associated with increased comparability ofaccounting amounts and capital market benefits We find that after firms voluntarily adopt IFRS (‘‘adopting’’ firms),their accounting amounts are more comparable to those of firms that previously adopted IFRS (‘‘adopted’’ firms) andless comparable to those of firms that apply domestic standards (‘‘non-adopting’’ firms) Adopting firms exhibitincreased liquidity, share turnover, and firm-specific information relative to adopted and non-adopting firms Neitheradopted nor non-adopting firms suffer capital market consequences Adopting firms with higher comparability withadopted firms have greater capital market benefits after adopting IFRS than adopting firms with lower comparability,and capital market benefits for adopting firms in countries with a relatively high percentage of firms that apply IFRSenjoy greater capital market benefits

INTRODUCTION

T he question this study addresses is whether voluntary adoption of International Financial Reporting Standards (IFRS) is

associated with increased comparability of accounting amounts and attendant capital market benefits In particular, wepredict that after firms voluntarily adopt IFRS (‘‘adopting’’ firms), their accounting amounts become more comparable

to those of firms that adopted IFRS before them (‘‘adopted’’ firms) and less comparable to those of firms that do not adopt IFRS(‘‘non-adopting’’ firms) We also predict that capital market benefits increase for the adopting firms relative to non-adoptingfirms and adopted firms Our findings are largely consistent with these predictions We link capital market benefits withcomparability by showing that the higher comparability is associated with greater capital market benefits

Many countries have mandated adoption of IFRS by listed firms Extant research provides evidence on the effects ofcomparability between U.S firms and non-U.S firms in these countries, as well as capital market benefits associated withmandatory adoption We focus on assessing the effects on comparability and capital market benefits associated with voluntaryIFRS adoption An advantage of voluntary IFRS adoption is the potential for increased comparability between the accountingamounts of voluntary adopters and those of other firms in the same industry applying IFRS.1A disadvantage is that voluntaryIFRS adoption may reduce comparability between accounting amounts of firms that apply IFRS and those that continue to

We thank seminar participants at the University of Graz, The University of Mississippi, the 2013 Workshop on European Financial Reporting, the 2013 American Accounting Association Annual Meeting, and an anonymous reviewer We appreciate funding from the Center for Finance and Accounting Research, The University of North Carolina Kenan-Flagler Business School, Stanford University Graduate School of Business, and the PricewaterhouseCoopers-Norm Auerbach Faculty Fellowship.

Editor’s note: Accepted by Richard A Lambert.

Submitted: March 2017Accepted: February 2018Published Online: September 2018

1

Another advantage is that voluntary IFRS adoption affords securities market regulators the opportunity to gain insight into the potential capital market

apply domestic standards Thus, we expect comparability of accounting amounts between adopting and adopted (non-adopting)firms to increase (decrease) after the adopting firms adopt IFRS because both groups of firms then apply (no longer apply) thesame accounting standards, i.e., IFRS (domestic standards) These expectations might not be borne out because it is unclearhow the incentives of voluntary adopters, regulation, and other institutional features influence voluntary implementation ofIFRS and, therefore, comparability Changes in comparability associated with voluntary adoption of IFRS also depend on thefact that domestic accounting standards are designed for domestic institutions, but IFRS are not, and differences betweendomestic standards and IFRS Further, even if voluntary IFRS adoption improves comparability, the extent of the improvement

is an empirical question

Following Barth, Landsman, Lang, and Williams (2012), we define accounting amounts as comparable if firms reportsimilar accounting amounts when they face similar economic outcomes (FASB 2010;IASB 2010) Thus, followingBarth et al

income and equity book value as accounting amounts, and stock price, stock return, and cash flow as economic outcomes.2Thefirst approach, whichBarth et al (2012)label accounting system comparability, is based on the extent to which the mappingfrom accounting amounts, e.g., earnings, to an economic outcome, e.g., stock price, of one firm is the same as that of the otherfirm The second approach, whichBarth et al (2012)label value relevance comparability, reflects the extent to which variation

in economic outcomes explained by accounting amounts is the same for two firms We find that accounting system and valuerelevance comparability between adopting and adopted firms increase significantly after adopting firms adopt IFRS We alsofind that both types of comparability between adopting and non-adopting firms decrease significantly after adopting firms adoptIFRS

Regarding capital market benefits, any increases (decreases) in comparability likely result in capital market benefits (costs).The extent to which changes in comparability result in capital market benefits depends on whether the benefits resulting fromincreased comparability with other firms applying IFRS dominate the costs resulting from decreased comparability with firmsapplying domestic standards Thus, we have four capital market benefits predictions First, benefits for adopting firms after theyadopt IFRS increase (decrease) if the benefits from increased comparability with other firms applying IFRS, including adoptedfirms, exceed (are exceeded by) the cost of decreased comparability with firms that continue to apply domestic standards,including non-adopting firms Second, benefits for adopting firms exceed those for non-adopting firms if the non-adoptingfirms’ benefits decrease because of decreased comparability with adopting firms or remain unchanged Third, adopted firms’benefits increase because of increased comparability with adopting firms or remain unchanged Fourth, adopting firms’ benefitsexceed the adopted firms’ benefits because the adopted firms have already realized any benefits from increased comparabilitywith other firms that adopted IFRS

Following prior research, we test for changes in three capital market benefits: greater liquidity, higher share turnover, andmore firm-specific information We use two measures of liquidity: theAmihud (2002)illiquidity measure, which captures theextent to which prices change as a result of trading; and the percentage of zero return days, which captures the extent to which afirm’s equity securities do not trade We base our measure of share turnover on the ratio of the number of daily shares traded tothe number of shares outstanding We measure firm-specific information as stock return synchronicity

We find that adopting firms enjoy relative capital market benefits after IFRS adoption in that they exhibit increasedliquidity, share turnover, and firm-specific information relative to adopted and non-adopting firms In addition, we find littleevidence of capital market consequences for adopted firms and little evidence that non-adopting firms suffer a decrease incapital market benefits Because a firm’s voluntary adoption of IFRS likely is not made in isolation of other complementarydecisions, e.g., improvements in financial reporting, governance, and auditing, the capital market benefits of voluntary IFRSadoption could be attributable to the joint effects of IFRS adoption and these other decisions

We link comparability with capital market benefits in two ways First, we provide evidence that adopting firms with highercomparability with adopted firms have greater capital market benefits after adopting IFRS than adopting firms with lowercomparability Second, we provide evidence that capital market benefits for adopting firms in countries with a relatively highpercentage of firms that apply IFRS enjoy greater capital market benefits than firms in countries with a relatively lowpercentage

We conduct all comparability and capital market benefits tests using two matched samples of firms, each of whichcomprises pairs of firms of similar size matched on country and industry The first comprises adopting firms matched withadopted firms; the second comprises adopting firms matched with non-adopting firms Our sample is drawn from the population

of firms domiciled in 27 countries that permitted voluntary IFRS adoption between 1996 and 2008, which includes the vastmajority of voluntary IFRS adopting firms Because more than half of the sample comprises German and Swiss firms, we also

2

We use the terms earnings and net income interchangeably.

test for changes in comparability and capital market benefits separately for German and Swiss firms and other firms Findingsfrom these additional analyses reveal that our inferences generally apply to both groups of firms.

The remainder of the paper is organized as follows The next section discusses related research and develops ourpredictions The third section explains the research design The fourth and fifth sections describe the sample and data andpresent the results The sixth section offers a summary and concluding commentary

RELATED RESEARCH AND PREDICTIONSRelated Research

Although many countries have mandated adoption of IFRS by listed firms, several countries permitted firms to adoptinternational accounting standards and IFRS voluntarily before adoption of IFRS became mandatory There is a substantialliterature examining the effects on the quality of accounting amounts that result from voluntary application of internationalaccounting standards and IFRS (Bartov, Goldberg, and Kim 2005; van Tendeloo and Vanstraelen 2005; Hung and

effects associated with voluntary adoption (Leuz 2003;Daske 2006;Kim and Shi 2012;Daske, Hail, Leuz, and Verdi 2013).3There also is a substantial literature examining the effects associated with mandatory application of IFRS on accounting quality

accounting amounts with those of firms in non-IFRS adopting countries (Cascino and Gassen 2015) and those of firms in IFRScountries before and after IFRS adoption (Yip and Young 2012) Barth et al (2012) find increases in comparability ofaccounting amounts and of accounting quality between firms adopting IFRS voluntarily and mandatorily and those of U.S.firms after the IFRS firms adopt IFRS

Although, collectively, these studies find evidence of an increase in quality of accounting amounts and capital market benefitsassociated with voluntary application of IFRS, the extent of the documented increase depends on the research design and samplefirms However, none of these studies address the question of whether voluntary adoption affects comparability of accountingamounts between firms adopting IFRS and those of firms in their own countries that adopted IFRS before them and that do notadopt IFRS In addition, it is not possible to infer changes in comparability of accounting amounts based on findings from studiesexamining changes in accounting quality For example, although it might be possible to infer that an increase in accounting qualityfor adopting firms leads to a loss in comparability of accounting amounts with non-adopting firms, there is no basis on which topredict how an increase in accounting quality for adopting firms affects comparability of accounting amounts with adopted firms.Although some prior studies address capital market consequences associated with voluntary adoption, none attempt to link capitalmarket consequences to comparability changes, or make comparisons of capital market consequences between firms adoptingIFRS and those of firms in their own countries that adopted IFRS before them and that do not adopt IFRS

Predictions

Comparability

Our first prediction is that comparability of accounting amounts between adopting firms and adopted firms increases afterthe adopting firms adopt IFRS We predict this because after the adopting firms adopt IFRS, both groups of firms claim to applythe same set of standards, IFRS Our second prediction is that comparability of accounting amounts between adopting and non-adopting firms decreases after the adopting firms adopt IFRS We predict this because after the adopting firms adopt IFRS, thetwo groups of firms no longer apply the same set of standards

Even though we make these comparability predictions, there are at least three reasons why they might not be borne out As

a result, whether our predicted changes in comparability obtain depend on whether and the extent to which these countervailingeffects offset our predicted effects The first is the incentives of voluntary adopters For example, a primary concern of the U.S.Securities and Exchange Commission (SEC 2012) is that firms may elect to apply IFRS for opportunistic reasons Such reasonscould lead them to adopt IFRS without making substantive changes in their financial reporting practices (Daske et al 2013); ifthis is the case, then there will be no change in comparability Alternatively, as noted by theSEC (2012), firms could applyIFRS opportunistically if it improves their reported performance, which could result in an increase or a decrease incomparability of accounting amounts with those of both firms that voluntarily adopted IFRS and firms that continue to applydomestic standards

3 The International Accounting Standards Committee (IASC), the predecessor body to the IASB, issued international accounting standards IFRS include standards issued not only by the IASB, but also by the IASC, some of which have been amended by the IASB.

The second is the fact that domestic accounting standards are designed for domestic institutions, but IFRS are not (Ball,

domestic institutions, there may be a decrease (no change) in comparability of accounting amounts with those of firms thatcontinue to apply domestic standards (already voluntarily adopted IFRS) The third is the extent to which there is a differencebetween domestic standards and IFRS (Bae, Tan, and Welker 2008; Barth, Landsman, Young, and Zhuang 2014) If thedifference is small, then it is likely that there will be little or no change in comparability after firms voluntarily adopt IFRS.Capital Market Benefits

Our four capital market benefits predictions follow from our comparability predictions First, regarding adopting firms afterthey adopt IFRS, we predict capital market benefits to increase (decrease) if the benefits from increased comparability withother firms applying IFRS, including adopted firms, exceed (are exceeded by) the cost of decreased comparability with firmsthat continue to apply domestic standards, including non-adopting firms.4Second, we predict benefits for adopting firms toexceed those for non-adopting firms if the non-adopting firms’ benefits decrease because of decreased comparability withadopting firms or remain unchanged Third, we predict the adopted firms’ benefits to increase because of increasedcomparability with adopting firms or to remain unchanged Fourth, we predict the adopting firms’ benefits to exceed theadopted firms’ benefits because the adopted firms have already realized any benefits from increased comparability with otherfirms that adopted IFRS before they did Underlying each of our capital market benefits predictions is that the comparabilitypredictions in the ‘‘Related Research and Predictions’’ section are descriptively valid To the extent that they are not, the capitalmarket benefits predictions are unlikely to be, as well

We focus on three capital market benefits commonly employed in the literature: greater liquidity, higher share turnover,and more firm-specific information (Leuz and Verrecchia 2000;Amihud 2002;Durnev, Morck, and Yeung 2004;Piotroski andRoulstone 2004;Daske et al 2008) A key assumption underlying our capital market benefits predictions is that when investorshave more comparable financial statement information, they are better able to compare a firm’s economic condition to that ofother firms when making their investment decisions Thus, for example, we expect that when an adopting firm adopts IFRS,investors are better able to compare the economic condition of the adopting firm to that of adopted firms, thereby making themmore willing to trade in the adopting firm’s equity This increased willingness to trade should manifest as an increase in theadopting firm’s liquidity and share turnover In addition, we expect the increased comparability to result in more firm-specificinformation being available to investors, which should manifest in lower stock return synchronicity

Research Design

We design our research to provide evidence on comparability and capital market benefits for voluntary adopters of IFRSrelative to other voluntary IFRS adopters and to firms applying domestic standards Our research setting enables us to provideinsights into the effects of voluntary IFRS adoption on comparability of accounting amounts and attendant capital marketbenefits

Matched Sample Design

We use a matched sample design to test our predictions developed in the ‘‘Related Research and Predictions’’ section Weconstruct separate matched samples to assess comparability between the adopting firms and adopted firms and betweenadopting firms and non-adopting firms to maximize the number of observations for each comparability test As a result, thenumber of observations underlying the two sets of tests differs

To construct the two matched samples, for each adopting firm we select an adopted firm and a non-adopting firm, each ofwhich is from the same country and industry as the adopting firm and, relative to other adopted or non-adopting firms, has theclosest equity market value (Barth et al 2008;Barth et al 2012).5FollowingBarth et al (2012), to identify a firm’s industry weuse the I/B/E/S industry classification from Worldscope; to identify the firm with the closest equity market value, we minimize the

4 Home bias and other informational constraints—particularly the use of domestic accounting standards—make it more difficult for investors outside a country to compare investments across countries ( Amiram 2012 ) To the extent that adoption of IFRS by a firm enhances comparability with foreign firms, foreign investors are more likely to include the firm’s stock in their investment portfolios This ability to attract foreign investors should augment any increase in capital market benefits Barth et al (2008) find that application of international accounting standards is associated with higher accounting quality than application of domestic standards However, if accounting amounts based on IFRS are of lower quality than those based on domestic standards, then capital market benefits associated with increased comparability could be offset by capital market costs associated with decreased accounting quality.

5 This procedure allows for the possibility that an adopting firm also can be an adopted firm, i.e., be matched with another adopting firm that adopts IFRS after it does.

sum of the equity market value differences between adopting firms and matched adopted or non-adopting firms at the end of theadopting firm’s adoption year We eliminate any matched pair for which the size difference exceeds 50 percent in absolute value.

To ensure that we properly identify an adopting firm’s adoption year, we require the firm to have data in the year it adopts IFRSand the year before Our analyses include all firm-years for which both firms in a matched pair have data For example, if anadopting firm has data from 1994 through 2000, and its matched adopted firm has data for 1995 through 2002, then our analysisincludes data from 1995 through 2000 for the adopting firm and its matched adopted firm Also, to ensure that our tests areunaffected by mandatory adoption, we eliminate all matched pairs after IFRS became mandatory in their country

Assessing Comparability

Following Barth et al (2012), we assess accounting system comparability and value relevance comparability based onmetrics using stock price, stock return, and cash flow as economic outcomes, and various combinations of net income andequity book value as accounting amounts The relations used to construct the comparability metrics associate earnings andequity book value with stock price, earnings and change in earnings with stock return, and earnings with cash flow

We first test whether comparability of accounting amounts between adopting and adopted firms increases after theadopting firms adopt IFRS, and whether comparability of accounting amounts between adopting and non-adopting firmsdecreases after the adopting firms adopt IFRS We also test whether the matched adopting and adopted firms’ accountingamounts, and the matched adopting and non-adopting firms’ accounting amounts, have comparable value relevance before andafter the adopting firms adopt IFRS.6

Accounting System Comparability

FollowingDe Franco, Kothari, and Verdi (2011)andBarth et al (2012), we define accounting systems as comparable if anestimated economic outcome, such as stock price, based on the mapping from accounting amounts, such as earnings, toeconomic outcomes from one system is the same as the estimated economic outcome based on the mapping of the other system.FollowingBarth et al (2012), we construct accounting system comparability metrics separately for the adopting/adopted andadopting/non-adopting firms matched samples.7

Construction of the metrics requires six steps In the first step, we estimate the relations between each economic outcomeand its associated accounting amounts For stock price, the accounting amounts are earnings and equity book value; for stockreturn, the accounting amounts are earnings and change in earnings; for subsequent year’s cash flow, the accounting amount isearnings In the second step, for each group of firms we calculate within-sample fitted economic outcomes

We conduct steps three through six separately for adopting firms and their matched adopted firms, and for adopting firmsand their matched non-adopting firms In the third step, we use the multiples from the other group of firms to calculate fittedeconomic outcomes For example, when constructing the comparability metric that relates to adopting firms and their matchedadopted firms, we calculate fitted stock price for each adopting firm based on the adopted firms’ multiples andvice versa

In the fourth step, we calculate the absolute value of the difference between the fitted economic outcomes that werecalculated in the second and third steps For example, when constructing the comparability metric that relates to adopting firmsand their matched adopted firms, for each adopting and adopted firm we calculate the absolute value of the difference betweenthe fitted economic outcomes based on adopting firm and adopted firm multiples In the fifth step, for each of the four groups offirms, i.e., the matched adopting and adopted firms, and the matched adopting and non-adopting firms, we average thedifferences in fitted economic outcomes calculated in the fourth step This results in two sets of average differences for eacheconomic outcome, i.e., one for the matched adopting and adopted firms and one for the matched adopting and non-adoptingfirms In the sixth step, we calculate the comparability metric for each economic outcome as the mean, median, and standarddeviation of the average differences calculated in the fifth step

To assess significance of mean (median) differences, we use a t-test (Wilcoxon Rank Sum Test).8To assess significance ofstandard deviation differences, we generate a distribution of the differences using a bootstrapping procedure.9To test for

no effect on our inferences.

differences in accounting system comparability after the adopting firms adopt IFRS, we use observations when the adoptingfirms applied domestic standards, i.e., before they adopted IFRS, and observations when they applied IFRS, i.e., after theyadopted IFRS.

Value Relevance Comparability

FollowingBarth et al (2012), we define value relevance of accounting amounts as comparable if they explain the samevariation in economic outcomes; when testing for differences in value relevance comparability, we use differences in theabsolute values of comparability metrics Because comparability relates to how close the metrics are to each other—notwhether one metric is larger than another—we do not use differences in signed values

PRICE, the first value relevance metric, is constructed using the adjusted R2from a regression of stock price,P, on netincome before extraordinary items per share,NI, and book value of equity per share, BVE:

Pit¼ b0þ b1BVEitþ b2NIitþX

jb3jCjþX

kb4 Ikþ eit: ð1ÞFollowingBarth et al (2012),P is stock price six months after fiscal year-end, CjandIkare indicator variables that equal 1for firms domiciled in countryj and industry k, and 0 otherwise; i and t refer to firm and year, respectively

RETURN, the second metric, is constructed using the adjusted R2from a regression of annual stock return,RETURN, onnet income and change in net income, deflated by beginning-of-year price,NIt/Pt1and DNIt/Pt1:10

RETURNit¼ b0þ b1NIit=Pit1þ b2DNIit=Pit1þ b3LOSSitþ b4LOSSit3 NIit=Pit1þ b5LOSSit3DNIit=Pit1

CASH FLOW, the third metric, is constructed using the adjusted R2from the following regression of cash flow on laggednet income:

CFitþ1¼ b0þ b1NIit=TAit1þX

jb2jCjþX

kb3 Ikþ eitþ1; ð3ÞwhereCF is net cash flow from operations scaled by lagged total assets, TA

We include the country and industry fixed effects in each equation to mitigate effects on our inferences of differencesacross countries and industries in mean stock price, stock return, and future cash flow from operations To ensure eachcomparability metric reflects only the explanatory power of the accounting amounts, we calculate each metric as the difference

in explanatory power of Equations (1), (2), and (3), and each equation’s nested equation that includes only the fixed effects

We estimate each full and attendant nested equation using those observations relevant to each comparison we make Forexample, when we compare value relevance of adopting and adopted firms after the adopting firms adopt IFRS, we estimate theequations using the combined sample of adopting firms and their matched adopted firms for years after the adopting firms adoptIFRS We test for differences in each value relevance metric using a bootstrapping procedure similar to the one we use for tests

of differences in accounting system comparability standard deviation

Capital Market Benefits

We test for capital market benefits—liquidity, share turnover, and firm-specific information—associated with adoptingfirms adopting IFRS by using a change-in-difference research design, using metrics adapted from prior research For eachcomparison we make, we test whether the change in a particular capital market benefit metric differs for the two groups of firmsbeing compared, i.e., adopting and adopted firms, and adopting and non-adopting firms The structure of the tests also permits

us to test for differences in capital market benefits for each group of firms and between two groups of firms before and after theadopting firms adopt IFRS

We use two metrics for liquidity The first is the Amihud (2002) illiquidity measure, which we denote asAMIHUD.AMIHUD is based on the ratio of the absolute value of the firm’s daily return (multiplied by one million) to its dollarvolume, averaged over the fiscal year For each firm, we calculateAMIHUD for the periods before and after the adoptingfirm adopts IFRS as the average of these yearly average ratios over the fiscal years in each period The second metric,

10

For ease of exposition, we use the same notation for coefficients and error terms in each equation.

ZeroRet, is based on the percentage of trading days for which a firm’s stock return is zero in each fiscal year (Daske et al.

2008) For each firm, we calculateZeroRet for the periods before and after the adopting firm adopts IFRS as the average ofthese yearly average percentages over the fiscal years in each period An increase in liquidity is reflected by a decrease inAMIHUD or ZeroRet

We base our measure of share turnover,TURN, on the ratio of the number of daily shares traded to the number of sharesoutstanding, averaged over the fiscal year (Leuz and Verrecchia 2000) For each firm, we calculateTURN for the periods beforeand after the adopting firm adopts IFRS as the average of these yearly average ratios over the fiscal years in each period As inprior research, we interpret stock return synchronicity as measuring the extent to which a firm’s stock return reflects firm-specific information, and measure stock return synchronicity,SYNC, as the adjusted R2from a regression of a firm’s daily stockreturn on the daily market return (Durnev et al 2004;Piotroski and Roulstone 2004) The lower is SYNC, the greater is theextent of firm-specific information, which we view as a capital market benefit

Specifically, when comparing the adopting and adopted firms (adopting and non-adopting firms), we estimate thefollowing regression separately for each group of firms before and after the adopting firms adopt IFRS:

R is the firm’s daily return; and RMrktis the daily market return for the country in which the firm’s stock is traded We require aminimum of 250 daily observations for each firm when estimating Equation (4) We test for differences in synchronicitybetween groups of firms, e.g., between adopting and adopted firms, based on a distribution of differences in adjusted R2obtained from a bootstrapping procedure that is similar to the one used for tests of significance of differences in value relevancecomparability metrics

SAMPLE AND DATAOur tests are based on a sample of firms that adopted IFRS between 1996 and 2008 The vast majority of voluntary IFRSadoptions occurred between 1999 and 2004 We extend the sample of adoption years to include the small number of voluntaryadopters in the immediately preceding and subsequent years Our tests require pre- and post-adoption sample years The pre-adoption sample years potentially range from 1992 through 2005, and post-adoption sample years potentially range from 1996through 2009, which provides at least four pre- and post-adoption years for most sample firms

We obtain our sample of IFRS firms from Worldscope, which identifies the set of accounting standards a firm uses toprepare its financial statements and its industry (similar toBarth et al 2012) Following prior research (Daske et al 2008;Barth

Applied data field are ‘‘international standards’’ and IASC or IFRS We limit IFRS firms to those that do not cross-list in theU.S to eliminate effects on the IFRS accounting amounts associated with the reconciliation requirement (Harris and Muller

1999;Lang, Raedy, and Wilson 2006;Barth et al 2012)

We obtain data used to estimate our equations from Datastream Using the full sample, we winsorize at the 1 percent and

99 percent levels all variables used to construct our metrics to mitigate the effects of outliers on our inferences The resultingsamples of adopting and adopted firms used in the comparability analyses comprise 106 matched pairs with 426 firm-yearobservations, of which 161 are before the adopting firm adopts IFRS and 265 are after; the resulting samples of adopting andnon-adopting firms comprise 204 matched pairs with 1,095 firm-year observations, of which 626 are before the adopting firmadopts IFRS and 469 are after That the number of matched pairs is larger for the adopting/non-adopting comparison reflects thefact that there were relatively fewer voluntary adopters of IFRS than non-adopters Capital market benefits tests are based onsomewhat smaller samples that reflect additional data requirements In particular, to constructZeroRet, we require each firm tohave 100 non-missing returns; to construct AMIHUD and TURN, we require trading volume, which is not available for allfirms

Table 1, Panel A provides a breakdown of sample firms by country Sample firms are from 27 countries, with the greatestproportion from Germany and Switzerland Panel B of Table 1 provides an industry breakdown Sample firms are from 34industries and, with the exception of Industrial Engineering comprising 14.29 percent of the adopting/non-adopting firm-years,there is no apparent industry concentration Panel C of Table 1 provides a breakdown by adoption year of the adopting firms.Table 2 reports descriptive statistics separately for the adopting/adopted matched pairs sample and the adopting/non-adopting matched pairs sample before and after adopting firms adopt IFRS Although we do not conduct significance tests fordifferences in means between the adopting and adopted firms and between the adopting and non-adopting firms, Table 2suggests that differences exist for several of our variables Because it is likely that such mean differences are, at least in part,attributable to country and industry differences, as described in the ‘‘Research Design’’ section, our value relevancecomparability metrics are constructed to exclude the explanatory power of country and industry indicator variables

Table 3 presents mean, median, and standard deviation accounting system comparability metrics; Panel A (B) presentsfindings for the full sample relating to comparability of accounting amounts of adopting and adopted (non-adopting) firms

As predicted, findings in Table 3, Panel A reveal that comparability between the adopting and adopted firms is greaterwhen the adopting firms apply IFRS than when they applied domestic standards, in that six of the nine accounting systemcomparability metrics for adopting and adopted firms decrease significantly after the adopting firms adopt IFRS The change indifferences in mean (median; standard deviation) PRICE, RETURN, and CASH FLOW are8.52 (1.46; 16.01), 0.13(0.21; 0.13), and 0.001 (0.000; 0.002) Neither of the two positive changes in differences, RETURN standard deviation andCASH FLOW median, is significant

As predicted, findings in Table 3, Panel B reveal that comparability between the adopting and non-adopting firms is lowerwhen the adopting firms apply IFRS than when they applied domestic standards, in that eight of the nine accounting systemcomparability metrics for adopting and non-adopting firms increase significantly after adopting firms adopt IFRS The change

in differences in mean (median; standard deviation) PRICE, RETURN, and CASH FLOW are 6.38 (2.07; 10.71), 0.06 (0.09;

0.04), and 0.002 (0.001; 0.002) The negative change in the RETURN standard deviation difference is insignificant

TABLE 1Sample Composition by Country, Industry, and Year of AdoptionPanel A: Composition by Country

Firm-YearObservations

(%) Firm-YearObservations

# ofFirms

% ofFirms

Firm-YearObservations

(%) Firm-YearObservations

# ofFirms

% ofFirms

Table 4 presents value relevance comparability metrics; Panel A (B) presents findings relating to comparability ofaccounting amounts of adopting and adopted (non-adopting) firms As predicted, and consistent with the accounting systemcomparability findings in Table 3, Panel A, the findings in Table 4, Panel A reveal that comparability between adopting andadopted firms is greater when the adopting firms apply IFRS than when they applied domestic standards, in that the difference

in value relevance metrics for adopting and adopted firms decreases significantly after adopting firms adopt IFRS In particular,the difference inPRICE, RETURN, and CASH FLOW decreased by 0.06, 0.11, and 0.02 Panel A also reveals that before theadopting firms adopt IFRS, their value relevance measured by PRICE and RETURN is significantly lower than that of theadopted firms However, after adopting firms adopt IFRS, their value relevance is higher than before they adopt IFRS and there

is no significant difference in value relevance between the two groups of firms

Also as predicted, and consistent with the accounting system comparability findings in Table 3, Panel B, the findings inTable 4, Panel B reveal that comparability between adopting and non-adopting firms is lower when the adopting firms applyIFRS than when they applied domestic standards, in that the differences in value relevance metrics for adopting and non-

TABLE 1 (continued)Panel B: Composition by Industry

Adopting/Adopted Sample Adopting/Non-Adopting SampleFirm-Year

Observations

(%) Firm-YearObservations

# ofFirms

% ofFirms

Firm-YearObservations

(%) Firm-YearObservations

# ofFirms

% ofFirms

adopting firms increase after adopting firms adopt IFRS, significantly so for RETURN and CASH FLOW In particular, thedifference inPRICE, RETURN, and CASH FLOW increase by 0.02, 0.05, and 0.02.

Taken together, the findings in Tables 3 and 4 are consistent with the prediction that voluntary adoption of IFRS increasescomparability between firms that adopt IFRS and those that adopted IFRS before them, but decreases comparability betweenfirms that adopt IFRS and those that do not adopt IFRS

Capital Market Benefits

Liquidity

Table 5 presents findings relating to tests of changes in differences in capital market benefits as reflected by liquidity; Panel

A (B) presents findings for adopting and adopted (non-adopting) firms The first (second) set of findings in each panel is based

on the Amihud (2002)illiquidity (zero returns) metrics, AMIHUD (ZeroRet) Regarding AMIHUD, the findings in Panel Areveal that when the adopting firms apply domestic standards, their shares are significantly less liquid than those of firms thatapply IFRS.AMIHUD means for adopted and adopting firms are 6.10 and 23.29; the difference of17.19 is significant Thedifference evaporates after the adopting firms adopt IFRS, with corresponding meanAMIHUD amounts of 4.66 and 4.50 foradopted and adopting firms In addition, consistent with the prediction that the increase in capital market benefits for adoptingfirms exceeds that for adopted firms, the change in the difference in mean AMIHUD, 17.35, is significantly positive, whichindicates that the increase in liquidity after adopting firms adopt IFRS is significantly greater for adopting firms than foradopted firms

RegardingZeroRet, the findings in Table 5, Panel A also reveal that when the adopting firms apply domestic standards,their shares are significantly less liquid than those of firms that apply IFRS.ZeroRet means for adopted and adopting firms are0.24 and 0.27; the difference of0.03 is significant The 0.02 difference after the adopting firms adopt IFRS is not significant

In addition, the change in the difference in meanZeroRet, 0.01, is significantly positive, which, as with the AMIHUD findings,indicates that the increase in liquidity after adopting firms adopt IFRS is significantly greater for adopting firms than foradopted firms

The comparison of adopting and non-adopting firms in Table 5, Panel B reveals that, as in Panel A, the adopting firms’shares increase in liquidity after they adopt IFRS—the mean decrease in illiquidity is significant for both AMIHUD andZeroRet,8.28 and 0.03 The non-adopting firms’ shares do not evidence a significant change in liquidity before and after theadopting firms adopt IFRS—the mean differences inAMIHUD and ZeroRet,5.85 and 0.00, are insignificant The changes inthe differences in meanAMIHUD and ZeroRet, 2.43 and 0.03, are significant, which is consistent with our prediction that theincrease in liquidity after adopting firms adopt IFRS is greater for adopting firms than for non-adopting firms Taken together,

TABLE 1 (continued)Panel C: Composition by Year of IFRS Adoption by Adopting Firms

TABLE 2Descriptive StatisticsPanel A: Adopting/Adopted Sample

Adopting Firms Adopted Firms Adopting Firms Adopted FirmsMean Std Dev Mean Std Dev Mean Std Dev Mean Std Dev

Panel B: Adopting/Non-Adopting Sample

Adopting Firms Non-Adopting Firms Adopting Firms Non-Adopting FirmsMean Std Dev Mean Std Dev Mean Std Dev Mean Std Dev

P ¼ stock price six months after fiscal year-end;

BVE ¼ book value of equity per share at end of fiscal year;

P/BVE ¼ P divided by BVE;

NI ¼ net income before extraordinary items per share;

RETURN ¼ annual stock return beginning nine months before and ending three months after fiscal year-end;

NI/P ¼ net income per share scaled by beginning-of-year stock price;

LOSS ¼ an indicator variable equal to 1 if NI/P is negative, and 0 otherwise;

CF ¼ net cash flow from operations scaled by lagged total assets;

NI/TA ¼ net income scaled by lagged total assets; and

D ¼ annual change.