UNIVERSITY OF CALGARY. Essays on the Impact of the Adoption of IFRS on Accounting Information Content,

Transcription

1 UNIVERSITY OF CALGARY Essays on the Impact of the Adoption of IFRS on Accounting Information Content, Liquidity and Institutional Investment in Canadian Stock Markets by Shahid Ali Khan A THESIS SUBMITTED TO THE FACULTY OF GRADUATE STUDIES IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY GRADUATE PROGRAM IN MANAGEMENT CALGARY, ALBERTA APRIL, 2016 Shahid Ali Khan 2016

2 Abstract In my thesis I examine the trade-offs of implementing International Financial Reporting Standards (IFRS) in Canada. My thesis consists of three papers. In my first paper, I examine whether IFRS adoption resulted in greater information content in earnings announcements for TSX and TSXV firms in Canada. I contrast the senior and junior exchange to evaluate the differential impact of IFRS within a single country as these two stock markets represent different information environments for local investors. I find that information content of earnings announcements, measured by abnormal volatility and abnormal volume during the announcement period, increased for TSX firms following the adoption of IFRS. These results indicate that IFRS earnings announcements increased the value-relevant information provided to investors possibly due to the more forward-looking measurement emphasis under IFRS. I find some evidence that information content of earnings was lower for TSXV firms in the post-ifrs adoption period. I cautiously suggest that this may reflect a greater reliance of TSXV investors on realized earnings information. In my second paper, I examine whether average liquidity for three groups of firms - Canadian, non-us international and U.S. firms - traded on Canadian stock exchanges increased or decreased after mandatory adoption of IFRS in Canada. I consider two competing forces affecting liquidity from IFRS adoption: enhanced comparability of firms within industries that span international boundaries and less tailoring of financial ii

3 reporting to satisfy local investor needs. I find that liquidity decreased for Canadian and U.S. firms, suggesting that the benefits of global comparability were not sufficient to offset the loss of local investor accommodation. However, I find some evidence that liquidity improved for non-u.s. international firms suggesting that enhanced comparability of firms within industries did span international boundaries. In my final paper, I examine whether non-u.s. international institutional investment, measured by percentage ownership by international investors and number of international investors, in Canadian firms increased in the post-ifrs adoption period. I used U.S. firms as my control sample. Results show that Canada did not benefit from higher non-u.s. international institutional investment in the Canadian-listed firms in the post-ifrs period compared to U.S. firms. iii

4 Acknowledgements I sincerely acknowledge the Almighty God for His Blessings in my life that I was able to finish my PhD. I wholeheartedly appreciate the supervisory guidance from my two wonderful supervisors, Dr. Mark Anderson and Dr. Hussein Warsame. Without their excellent support, I cannot imagine to have passed this hurdle. I sincerely appreciate from the bottom of my heart how much both of you have affected me both in my learning and as a person. I would also like to appreciate Dr. Michael Wright for his active involvement during all of my PhD tenure both as a teacher and later as a co-author of my papers in the thesis. I also would like to extend special thanks to Dr. Abu Shiraz Rahaman who played an important role in the start of my PhD program. Finally, special thanks to external examiners, Dr. Bruce McConomy of Wilfrid Laurier University and Dr. Eugene Choo of Department of Economics, University of Calgary. I also like to extend special appreciation to other faculty and staff of the Haskayne School of Business, particulary Lesley Dimarzo and Sylvia Fuchek. I also like to gratefully acknowledge the considerable financial support of the Haskayne School of Business, the PhD Award from the Society of Certified Management Accountants of Canada, and the PhD Award from the Canadian Securities Institute. iv

5 Dedication To my late father, Wing Commander Ghulam Haidar, whose guidance and passion for education made me to come this far; to my mother, Shaheen Jamal, whose support and affection has made me what I am. Most of all, my wife, Faiza, who has been the biggest blessing in my life and whose support made it all possible, and my two sons, Jamal Khan and Shan Khan, from whom I gain all my energy and direction. In the end, my special thanks to my late uncle, Qafait Khan, who played an important role in supporting my academic pursuits. v

6 TABLE OF CONTENTS Abstract... ii Acknowledgements... iv Dedication... v Table of Contents... vi List of Tables... viii Chapter 1: Introduction... 1 Chapter 2: Do IFRS based Earnings Announcements have more Information Content than Canadian GAAP based Earnings Announcements?... 8 Abstract Introduction Literature Review Hypothesis Development Methodology Sample Description Results Conclusions Bibliography Chapter 3: Did Mandatory Adoption of IFRS Increase Liquidity in the Canadian Stock Markets? Abstract: Introduction Literature Review Hypothesis Development Empirical Model Sample Description Descriptive Statistics Multivariate Statistics Australia and United Kingdom Liquidity tests Size tests Summary & Future Research Direction Bibliography Chapter 4: Has Mandatory Adoption of IFRS Increased Non-U.S. International Institutional Investment in the Canadian Stock Markets? Abstract: vi

7 4.1 Introduction Literature Review Hypothesis Development Research Design and empirical model Sample Description Descriptive Statistics Multivariate Statistics Summary & Future Research Direction Bibliography Chapter 5: Summary Appendix: Copyright permission documents vii

8 List of Tables Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table Table viii

9 Chapter 1: Introduction Canada s decision to adopt International Financial Reporting Standards (IFRS) involved many costs and benefits. In contrast to some countries, Canada already had high quality accounting standards before the implementation of IFRS in Canada was persuaded to replace domestic accounting standards with IFRS under the argument that IFRS would help increase access to global capital and increase investment efficiency. It is natural to ask whether IFRS adoption has indeed benefitted the Canadian stock markets. Overall my thesis considers two competing forces affecting Canadian markets from IFRS adoption: Canadian markets becoming more globalized markets and less tailoring of financial reporting in Canada to satisfy local investor needs. Local accounting standards evolved over time to meet local investors needs. IFRS represents one set of financial accounting standards for the whole world. Many countries adopted IFRS because IFRS makes accounting reports more comparable across the globe. However, replacing local standards with IFRS may ignore information needs of local investors hence I study some trade-offs between global accounting standards (IFRS) and local accounting standards. Without IFRS, global investors from different countries have to interpret the local accounting standards for each country before they invest in a company overseas. Therefore, local investors have an advantage, as they understand the local accounting standards and the culture. IFRS may help international investors, as they can avoid these costly reconciliations. In my first paper, I examine whether IFRS adoption resulted in greater information content in earnings announcements for TSX and TSXV firms in Canada.

10 The shift towards measurement of value changes in net assets based on future expectations from earning recognition based on more objective realizations, that is reflected in the balance sheet emphasis under IFRS, leads to the prediction that IFRSbased income measurement provides additional information content that increases divergence of opinion among investors. According to the measurement approach to decision usefulness, accountants take more responsibility for providing relevant information to the users of financial statements about changes in net asset values. This increase in value-relevance reflects more information about future expectations as opposed to information about realizations reflected in historical cost reporting. Such increases in provision of value-relevant information are likely to lead to greater divergence of investor opinion because investors have different expectations about future states and the performance of the firm in those states. Early evidence suggests that the volatility of earnings has increased under IFRS (Blanchette et al., 2011), supporting the notion that IFRS-based earnings reflect more information about expectations relative to realizations. I am investigating whether the quality of earnings information increased under IFRS. Investors rely on earnings to provide information about financial performance. For this purpose, I used the contrast between the senior and junior exchange in Canada to examine the differential impact of IFRS within a single country. Larger and mature firms are listed on the senior TSX exchange, while new and growth oriented firms are listed on the junior TSXV exchange. Institutional investors invest more in the TSX exchange while retail investors participate more in the TSXV exchange and there is more analyst and media coverage for TSX firms. 2

11 I find that the information content of earnings announcements, measured by abnormal volatility and abnormal volume during the announcement period, increased for TSX firms following the adoption of IFRS. These results indicate that IFRS earnings announcements increased the value-relevant information provided to investors in Canada in a manner that led to greater divergence of opinion, possibly due to the more forwardlooking measurement emphasis under IFRS. I find some evidence that information content of earnings was higher for TSXV firms relative to TSX firms in the pre-ifrs period and somewhat stronger evidence that the increase in information content for TSX firms post-ifrs was not shared by TSXV firms. I cautiously suggest that this may reflect a greater reliance of TSXV investors on realized earnings information. My first paper suggests that Canada s big exchange, TSX, may have benefitted in terms of more information content in the earnings announcements in the post-ifrs adoption periods but may have lost ground in terms of less information content in the IFRS based earnings announcements for TSXV firms in the post-ifrs adoption periods. TSXV exchange is one of the an important component of Canadian economic story. Lots of young, entrepreneurial and innovative Canadian firms obtain initial capital in this exchange. Hence Canada may have lost some ground when we consider the information content of IFRS based earnings announcements in the post-ifrs adoption periods. In my second paper, I examine whether average liquidity for three groups of firms - Canadian, non-us international and U.S. firms - traded on Canadian stock exchanges increased or decreased after mandatory adoption of IFRS in Canada. In other words, I investigate the overall quality of accounting information under IFRS and not just earnings information like I investigated in my first paper. I consider two competing forces affecting liquidity from IFRS adoption: enhanced comparability of firms within industries 3

12 that span international boundaries and less tailoring of financial reporting to satisfy local investor needs. There may be an important cost of losing the local language of business for local investors and entrepreneurs. Canadian capital markets are vastly influenced by the oil, gas, and mineral related industrial sectors, emphasizing the point that the Canadian business landscape is local in nature. It may be argued that domestic, non-u.s., and U.S. international firms would gain capital market benefits from more comparable financial information with firms in the rest of the world after the mandatory adoption of IFRS in Canada. It would be less costly for investors to compare financial reports across countries and even within countries with regards to international firms (Daske et al., 2008; Armstrong et al., 2007; Covrig, DeFond, and Hung, 2007). The enhanced comparability of financial information for international firms would help the capital markets price both domestic and international firms under IFRS. Further, investors would be able to differentiate low quality firms from high quality firms when comparing financial amounts leading to lower information asymmetries and/or lower estimation risk (Coffee, 1984; Dye, 1990; Lambert et al., 2007; Daske et al., 2008). This would lead investors to price the stocks of both domestic and international firms similarly, and estimation risk would be reduced overall in the market. Mandatory adoption of IFRS in Canada also means that domestic investors would become more skilled in analyzing IFRS-based financial statements. This would lead to reduction in potential bias by domestic investors to focus only on domestic companies against international firms reporting under IFRS in the pre-ifrs adoption period. Hence, I hypothesize that liquidity, measured by bid-ask spreads and zero-returns, would be higher in the post-ifrs adoption period for both international and domestic firms in 4

13 Canada (hypothesis 1). However, considering that Canada s decision to adopt IFRS meant moving away from its largest trading partner, U.S., and also that the Canadian economy is mostly focused in the extractive industries (the majority of its international firms are from the extractive industries), I may not see higher liquidity for international and local firms. This is why Ball (2006) argued that local forces and politics were too strong for the idea of international convergence of financial reporting standards to work as intended (Ball, Kothari, and Robin, 2000; Ball Robin, and Wu, 2003; Leuz, 2003, Ball and Shivakumar, 2005; Daske et al., 2008). I find that liquidity decreased for Canadian and U.S. firms, suggesting that the benefits of global comparability were not sufficient to offset the loss of local investor accommodation. However, I find some evidence that liquidity improved for non-u.s. international firms suggesting that enhanced comparability of firms within industries spanned international boundaries. To provide perspective on these results for Canadian exchanges, I also compare liquidity before and after IFRS adoption for all three groups of firms traded on the U.K. and Australian exchanges. Canada, the U.K. and Australia share the British influence on financial reporting historically and have similar legal and institutional settings. I expect that the benefits of global comparability would be relatively greater for the U.K. exchanges relative to the Canadian and Australian exchanges given the U.K. s proximity and commerce with other European countries. I find that liquidity increased for the U.K. exchanges but decreased for the Australian exchange, supporting the interpretation of the Canadian findings. Based on the overall results of U.K s, Canadian and Australian firms, I infer that diversified industry presence and more international firms are important factors to capture the comparability advantages of IFRS. The initial results show Canadian firms and 5

14 investors may have been disadvantaged in terms of lower liquidity by the adoption of IFRS in Canada. However, we do find supporting evidence that Canadian stock exchanges benefitted the non-u.s. international firms in terms of better liquidity. Hence there is evidence that Canadian stock markets may have become more global stock exchanges. In my third and final paper, I examine whether non-u.s. international institutional investment, measured by percentage ownership by international institutional investors and number of international institutional investors, increased in the Canadian stock markets in the post-ifrs adoption period. In this paper, I investigate whether non-u.s. international institutional investors find IFRS more useful than Canadian GAAP in their investment decisions. I use U.S. firms as my control sample. Improved comparability under IFRS is the key driving force that may lead to increased international institutional investment in the Canadian stock markets in the post-ifrs adoption period compared to the U.S. stock markets. Both the investment professionals and the academic literature support the argument that improved comparability under IFRS should lead to increased international institutional investment in the IFRS adopting country. Investment professionals consider cross-countries accounting reconciliations that clarify the differences in accounting standards as very time consuming and costly (Morgan Stanley Dean Witter, 1998). Other academic studies have pointed out the high costs of acquiring and processing information about international firms as a key reason why institutional investors are reluctant to invest inother countries (Kang and Stulz, 1997; Bradshaw et al, 2004; Chan et al., 2005; Covrig et al., 2007; DeFond et al., 2011). Hence I hypothesize that there will be a relative increase in international institutional investment in the Canadian stock markets as 6

15 compared to that in the U.S. stock markets after the adoption of IFRS in Canada. I measure international institutional investment by percentage share of ownership by international institutional investors and number of international institutional investors in the post-ifrs adoption period. Based on the empirical tests for two years, 2009 and 2013, I conclude that there is no evidence that Canada benefitted from the adoption of IFRS from the perspective of increased international institutional investment compared to that in the U.S. In fact, there is some evidence in the additional robustness test in propensity score matched sample results that Canada may have become disadvantaged as there are a lower number of international institutional investors who own stocks in Canadian firms on average in the post-ifrs adoption period compared to U.S. firms. Based on the results, I conclude that Canada was already an attractive place to invest for non-u.s. international institutional investors in North America even before the adoption of IFRS and that adopting IFRS did not make a noticeable change. In international circles among (non-u.s.) institutional investors, Canada already enjoyed a high reputation of high quality local accounting standards and established and reputable economic and legal institutions. 7

16 Chapter 2: Do IFRS based Earnings Announcements have more Information Content than Canadian GAAP based Earnings Announcements? Abstract Using an event study methodology, this paper investigates whether the information content of earnings announcements changed for firms traded on the Toronto Stock Exchange (TSX) and the Canadian Venture Exchange (TSXV) following mandatory adoption of international financial reporting standards (IFRS) in Canada. A priori, it may be argued that the information content of earnings would increase for both TSX and TSXV firms if IFRS earnings provided more value-relevant information than Canadian GAAP earnings. Increased value relevance of information provided by IFRS earnings would likely reflect increased measurement of changes in net asset values based on expectations as opposed to realizations. Because values based on expectations are subject to greater divergence of opinion than values based on realizations, greater value relevance is likely to be accompanied by higher abnormal return volatility and abnormal trading volume during announcement periods. Consistent with this argument, this study finds that abnormal volatility and abnormal volume during earnings announcement periods were higher in post-ifrs announcement periods than in pre-ifrs announcement periods for firms traded on the TSX. This study discriminates across the two exchanges in terms of information quality based on the mix of institutional and retail investors, analyst following, concentration in the oil, gas and mining sectors and size of firm. This study tests for a residual difference in information content based on the more speculative nature of the TSXV exchange and finds some evidence that divergence of opinion was higher for 8

17 TSXV firms than TSX firms in the pre-ifrs period but this residual difference does not carry through to the post-ifrs period. 2.1 Introduction The widespread adoption of International Financial Reporting Standards (IFRS) around the world has led to research on the economic and accounting effects of conversion to IFRS from domestic accounting. This research has produced mixed results for questions such as whether the adoption of IFRS has had positive or negative effects on accounting qualities such as value relevance or the information content of earnings. This study adds to this research stream by examining the effect of mandatory adoption of IFRS on the information content of earnings for companies listed on Canada s two major exchanges. This question is of interest to Canadian stakeholders including regulators, corporations and public accountants and potentially to U.S. stakeholders due to the similarity of financial reporting and capital markets across the two countries. Canada adopted International Financial Reporting Standards (IFRS) in While Canadian generally accepted accounting principles (GAAP) had been converging with IFRS prior to mandatory adoption (Bae, Tan and Welker, 2008), the adoption accelerated the trend towards greater measurement of asset values under the principles-based IFRS (PriceWaterhouseCoopers, 2007). With the adaption to a different conceptual framework under IFRS and specific standards that make different trade-offs between relevance and reliability of financial information, it is an interesting empirical question whether there is a difference in information content for IFRS based earnings announcements versus Canadian GAAP based earnings announcements. 9

18 O Brien (2009) stated the following: As 2011 approaches, many of us try to reassure ourselves that international financial reporting standards (IFRS) are not so different from our familiar Canadian generally accepted accounting principles (GAAP). At the same time worldwide convergence in standard setting is accelerating fundamental changes in the accounting landscape. These changes include emphasizing the balance sheet over the income statement, valuation over performance, and the hypothetical over the concrete. A concern expressed by O Brien (2009) is that inherent changes of a shift from earnings performance and reliability in earnings numbers to relevance of balance sheet numbers for valuation purposes may bring fundamental changes to the information analysis processes that analysts and other financial statement users apply. O Brien notes that the periodic measurement of performance and business activity using concepts like matching, stewardship and reliability are key to the financial analysis process. Financial statement users seek an understanding of a company s earnings process value driven activities when companies create or distribute something to its customers - from reading financial statements. The income statement was equipped, under Canadian GAAP, to reveal such information about the earnings process. O Brien (2009) argued that the proposed conceptual framework under IFRS reduces the focus on essential qualities such as matching, reliability and performance. The measurement approach to decision usefulness emphasizes relevance and places more emphasis on the balance sheet, increasing the role of accountants in helping investors understand firm value. The measurement approach opts for increasing the use of fair values within the financial statements proper (Scott, 2014). This approach to decision usefulness was developed due to increasing evidence that average investors were not as informed or rational as they were assumed to be under the efficient markets hypothesis. 10

19 For instance, average investors exhibit behavioural biases in decisions to buy and sell securities, for example representativeness bias, self-attribution bias, etcetera (Daniel, Hirshleifer, and Subrahmanyam, 1998). The measurement approach may help less sophisticated investors by increasing publicly available information about the firm. On the other hand, it may help sophisticated investors due to complementary information in the public earnings announcement news that supports private information gained from earlier research by sophisticated investors. One of the key goals of high quality accounting reporting standards is to reduce information asymmetry via a mechanism by which better informed insiders can provide information to less well-informed outsiders for investment decision making (Fields, Lys, and Vincent, 2001; Healy and Palepu, 2001). Arguments supporting IFRS as a higher quality set of accounting standards are based on its principled-based approach that allows management to use greater discretion in applying accounting principles (Scott, 2014; Blanchette, Racicot, and Girard, 2011). 1 On the other hand, Barth, Landsman, and Lang (2008) argue that IFRS reduces or limits accounting alternatives. Such limits may inhibit the firm's ability to report accounting amounts that provide more information about firm's economic condition and performance. Previous research has looked at differences in accounting quality resulting from IFRS adoption by comparing adopting versus nonadopting countries. One issue is that inter-country factors, such as changes in the enforcement or legal environment or auditing environment that occurred as IFRS were adopted, can affect the results of these studies (Landsman, Maydew, and Thornock, 2012;

20 Armstrong, Barth, Jagolinzer, and Riedl, 2010; Barth et al., 2008). Ball expressed his concerns: A deeper concern is that there inevitably will be substantial differences among countries in implementation of IFRS, which now risk being concealed by a veneer of uniformity. (Ball, 2006) Based on the experience in other countries, there is mixed evidence whether adoption of IFRS resulted in higher quality accounting numbers compared to the accounting numbers prepared using prior domestic accounting standards (Leuz and Wysocki, 2008; Dye and Sunder, 2001). A recent large-scale international study by Landsman et al. (2012) investigated whether the information content in annual earnings announcements, which is considered to be one dimension of accounting quality, increased in countries after the mandatory adoption of IFRS, relative to countries that continued to follow domestic accounting standards. Their findings suggest that the information content increased in 16 countries that experienced mandatory IFRS adoption, relative to 11 countries that continued to follow domestic accounting standards. This study finds evidence that abnormal return volatility and abnormal trading volume were higher in post-ifrs earnings announcement periods for TSX companies but not for TSXV companies, indicating that divergence of opinion resulting from earnings announcements, increased for companies trading on the TSX exchange but not for companies trading on the TSXV exchange. The next section provides a literature review. The following section develops this study s hypotheses. Then this study describes the methodology and provides descriptive statistics. The next section provides the results and the last section offers conclusions. 12

21 2.2 Literature Review Many studies link the capital allocation process and high quality reporting. Francis, Huang, Khurana, and Pereira (2009) investigated the effect of reporting quality on capital allocation. Their proxy for capital allocation was growth opportunities for firms. They argued that it would be easier to finance growth opportunities if a firm had higher financial reporting quality. Another study by Biddle, Hilary, and Verdi (2009) provides evidence that high quality reporting reduced both under and over-investment. Consistent with this fundamental market incentive linking capital allocation and high quality reporting, this study tests the prediction that IFRS would enable both TSXV and TSX firms to increase value-relevant information provided to investors, and it specifically tests for changes in the information content of earnings announcements for both sets of firms associated with IFRS adoption. Several recent studies have documented advantages of IFRS as a new accounting system. Barth et al. (2008) found that firms that adopted IFRS generally exhibited higher accounting quality as compared to firms that followed local GAAP. Another study found that, once firms adopted IFRS, the increase in accounting quality resulted in enhanced comparability of accounting amounts with high quality US GAAP amounts (Barth, Landsman, Lang, and Williams, 2012). Armstrong et al. (2010) investigated the market reaction to the adoption of IFRS in Europe. They found positive market reaction to the news of adoption of IFRS by those firms with lower pre-adoption information quality and higher pre-adoption information asymmetry, consistent with the argument that markets expecting net information quality benefits from IFRS adoption. However, they found 13

22 negative market reaction to the IFRS adoption news for those firms domiciled in countries where the enforcement of IFRS was doubtful. Using data from European countries, Beuselinck, Joos, Khurana, and van der Meulen (2010) found that IFRS disclosures revealed new firm-specific information in the IFRS adoption period, and reduced surprises from future disclosures. Many studies provide evidence that IFRS adoption resulted in capital market advantages including reduced cost of capital and enhanced liquidity (Daske, Hail, Leuz, and Verdi, 2008; Li, 2010), greater foreign investment (Amiram, 2012; Covrig, DeFond, and Hung, 2007; DeFond, Hu, Hung, and Li, 2011; Yu, 2010), higher levels of analyst following, and lower levels of analyst forecast dispersion (Byard, Li, and Yu, 2011; Horton, Serafeim, and Serafeim, 2012; Tan et al., 2009). Other studies document that fair value estimates are value relevant (Sharpe and Walker, 1975; Standish and Ung, 1982; Aboody, Barth, and Kasznik, 1999; Danbolt and Rees, 2008). Alternatively, contract theory and the need for outcome-based accounting provide arguments for financial statements standards with a mix of historical cost accounting and fair values based on conservative principles of accounting. Under these arguments, investors are assumed to be rational (O Brien and Scott, 2012). However, contract theory and the outcome-based accounting arguments inherently assume a system where the legal institutions and the accounting and auditing regime work effectively. This is why Ball s discourse emphasized that accounting standards take shape in response to economic and political forces (Ball, 2006). In Ball s words, 14

23 Financial reporting conveys an important economic role by accurately and independently counting actual outcomes, and hence confirming prior information about expected outcomes. (Ball, 2006) Ball calls it contracting inefficiency if more relevant information creates noise and increases the risk between contracting parties. Proponents of contract theory also argue that bringing more relevant information into the financial statements proper reduces the reliability of the financial numbers based on the relevance-reliability trade off argument. This also leads to reduced usefulness of financial statements in holding managers accountable for moral hazard problems (O Brien, 2011). But the asymmetric payoff between different parties to the contract, for example lender versus equity owners, owners versus managers etcetera, also creates a demand for conservative recognition of fair values that reflect losses, which increases the likelihood that financial statements can be used to predict financial distress (Watts, 2003a and 2003b; Basu, 1997; Beaver and Ryan, 2005). Some studies have found that higher quality financial reporting and better disclosures are associated with reduced adverse selection problems in markets and lower estimation risk (Verrechhia, 2001; Lambert, Leuz, and Verrecchia, 2007). Others argue that there are associations between information asymmetry and firm s disclosure and accounting policies (Welker, 1995; Healy, Huton, and Palepu, 1999; Leuz and Verrecchia, 2007). Daske et al. (2008) support the idea that the inherent flexibility as a result of more principles-based IFRS results in more opportunities for managers to manipulate earnings. 15

24 Several studies have found evidence that the adoption of IFRS did not result in improved earnings quality. Jeanjean and Stolowy (2008) investigated how mandatory adoption of IFRS affected the earnings management practice of firms in Australia, France, and the U.K. They did not find evidence of improvement (less prevalence) in earnings management in the post IFRS time period in those three countries. Similarly, Aussenegg, Inwinkl, and Schneider (2008) did not find any evidence of improvement (less prevalence) in earnings management in the U.K., Ireland (English legal origin) and Northern European countries (Scandinavian legal origin), although adoption of IFRS in Central European countries resulted in improved (less) earnings management. Kabir, Laswad, and Islam (2010) found evidence of increased earnings management in the post IFRS adoption period in New Zealand. Similarly, Doukakis (2010) found that the mandatory adoption of IFRS in Greece did not result in any change in earnings persistence or the information content of earnings. A concern with this previous research on IFRS implementation issues is that it focuses on information values using data from European or Asian countries, where, in many cases, IFRS implementation amounted to replacing lower quality local accounting standards with higher quality IFRS (Barth et al., 2008; Landsman et al., 2012). It is difficult to control for many inter-country local factors that may affect the results of such studies for instance changes in the enforcement or legal or auditing environment as IFRS was adopted (Armstrong et al., 2010; Barth et al., 2008). Hence it is difficult to attribute the increased benefits in the stock markets (in a market wide or macro level sense) in those countries to the implementation of IFRS as many other local regulatory changes were happening simultaneously. 16

25 Ahmed, Neel, and Wang (2013) provide evidence on the preliminary effects of mandatory adoption of IFRS on accounting quality of firms that adopted IFRS in Their carefully controlled tests revealed that IFRS firms from strong enforcement countries exhibited significant increases in income smoothing and aggressive reporting of accruals, and a significant decrease in timeliness of loss recognition. They argue that principle-based IFRS, on average, are looser than domestic standards. Their results indicate that, although IFRS provides greater flexibility in providing relevant information from a global perspective, IFRS does not necessarily result in informational value benefits for individual firms or countries. In summary, prior research has found mixed evidence of increased accounting quality in the post IFRS adoption period in many European and Asian countries. Many of these studies were inter-country comparisons, and the risks of country specific factors playing a role in the results cannot be ignored. This study s interest is in the Canadian experience with adoption of IFRS. This study addresses the question whether the information content of earnings announcements in Canada changed with adoption of IFRS. 2.3 Hypothesis Development (2001, p. 116): This study follows the notion of information content as expressed by Kothari If the level or variability of prices changes around the event date, then the conclusion is that the accounting event conveys new information about the amount, timing, and/or uncertainty of future cash flows that revised the market s previous expectations. 17

26 Lev (1989) used a similar definition in which information is considered useful if investors actions as a result of new information result in changes in stock prices or trading volumes. The key objective of standards setting bodies like IASB and FASB is to provide decision-useful information to investors. According to the proposed conceptual framework of IASB (2008), the purpose of financial statements is to provide financial information that is useful to present and potential equity investors, lenders, and other creditors in making decisions in their capacity as capital providers. Hence, some aspects of IFRS s decision usefulness can be assessed by observing changes in prices of stocks or trading volume as they reflect actions that investors take after incorporating new financial information into their investment decision making (Scott, 2014). Seminal papers of Ball and Brown (1968) and Beaver (1968) started the tradition of information content studies. Ball and Brown (1968) investigated both the long-window association between price changes and earnings as well as the shorter window relation between price change and the information in earnings announcements. While they were interested in the first-order relation between price changes and earnings information, this paper is primarily interested in the less noisy second-order relation between stock return volatility and information in earnings as a measure of information content. Beaver (1968) noted that price change around an announcement reflects the average change in investor s beliefs as a result of that announcement, and that trading volume reflects differences in individual interpretations of the announcements. In the context of rational investment decision-making among investors, several studies operationalize the Beaver (1968) measures of price change and trading volume in the event studies format (Kim and Verrecchia, 1991a, 1991b, 1997). Landsman et al. (2012) interpreted return volatility as 18

27 measuring the degree to which an announcement changes investors beliefs on average. They observe that the greater the content in the financial information, the more investors beliefs, on average, are likely to change. In other words, the more information content in announcements, the greater is the return volatility. They interpret trading volume as reflecting differences in the interpretations of the announcements (Kim and Verrecchia, 1991a, 1997). In other words, abnormal trading volume measures the degree of divergence of investor opinions as a result of an earnings announcement. According to the measurement approach to decision usefulness, accountants take more responsibility for providing relevant information to the users of financial statements about changes in net asset values. This increase in value-relevance reflects more information about future expectations as opposed to information about realizations reflected in historical cost reporting. Increases in value-relevant information are likely to lead to greater divergence of investor opinion because investors have different expectations about future states and the performance of the firm in those states. The greater emphasis on expectations over realizations is reflected in the new International standard IAS 37 that relaxes criteria for recognizing contingent liabilities. Also, IAS 37 emphasizes recognition of restructuring costs and asset retirement obligation as long as there is a constructive obligation to do so while Canadian GAAP requires such recognition of liabilities only if there is a legal obligation. Early evidence suggests that the volatility of earnings has increased under IFRS (Blanchette et al., 2011), supporting the notion that IFRS-based earnings reflect more information about expectations relative to realizations. 19

28 The shift towards measurement of value changes in net assets based on future expectations from earning recognition based on more objective realizations that is reflected in the balance sheet emphasis under IFRS leads to the prediction that IFRSbased income measurement provides additional information content that increases divergence of opinion among investors Hypothesis 1 The abnormal return volatility and abnormal volume for the annual earnings announcements of TSX firms and TSXV firms are higher in the post-ifrs time period than in the pre-ifrs time period. An interesting opportunity presented by the presence of both a senior and junior exchange in Canada is to investigate whether changes in information content between pre- and post-ifrs earnings announcements are similar across the two exchanges. The TSXV exchange is a venture capital exchange that is populated largely by explorationoriented companies in the extractive industries and innovation-oriented growth companies and caters primarily to retail investors. Companies on the TSX exchange are typically larger, have larger institutional investor bases and are followed by analysts more than companies on the TSXV exchange. Institutional investors own on average about 31.5% of this study s sample firms listed on the TSX and only 16.8% of this study s sample firms listed on the TSXV. Many smaller firms on the TSXV have no institutional ownership. TSX firms are covered by 4.86 analysts on average while TSXV firms are covered by 0.6 analysts on average. These differences reflect the different natures of the two exchanges. The TSX is a more traditional high-quality stock exchange regulated in the province of 20

29 Ontario. The TSXV is a more speculative venture-oriented exchange that was formed from combining local exchanges operating in Alberta and British Columbia. It caters more to oil, gas and mining companies who make up over 80% of the TSXV firms in this study s sample and about 50% of the TSX firms in this study s sample. The current debate 2 in Canada about the need for a single regulator is hotly contested in the western provinces and reflects the more entrepreneurial and speculative nature of the TSXV. In the analysis, this study controls for firm size, analyst coverage, institutional ownership and membership in the oil, gas and mining industry. While these factors are likely to influence the information content of earnings announcements in various ways, this study considers the possibility that there is a residual exchange effect that reflects the different trading and information environments across the two exchanges. This study does not have strong predictions ex ante other than to observe that income realizations may be more informative for firms in the more speculative TSXV trading environment, leading to higher information content on average from earnings announcements. Higher value relevance provided by IFRS earnings about future expectations may positively affect the information content of earnings but, on the other hand, the increased uncertainty associated with future expectations may negatively affect the information content of earnings for speculative TSXV investors who are more interested in realizations. This study s hypothesis is based on the former

30 2.3.2 Hypothesis 2 The change in abnormal return volatility and abnormal trading volume between the pre and post IFRS periods is greater for the TSXV firms than for the TSX firms. 2.4 Methodology Following the procedures used in prior research, this study measures the volatility of stock returns, AVAR, as the ratio of the mean of the squared abnormal return in the event period (5 days from t-2 to t+2), u it 2, to the variance of the abnormal return in the it 2 u it 2 it 2 non-event period,, that is, / (DeFond et al., 2007). This study applies (5 days event window due to the existence of smaller firms in the TSXV exchange. It believes that it is more likely to capture the abnormal return volatility and abnormal trading volume with 5 days windows than with 3 days windows for the smaller firms. The abnormal return is calculated using the market model. First, stock returns for the nonevent period are regressed on the relevant market index returns (separately for the TSXV and TSX market index) to get the coefficients α and β for each firm. Then this study calculates the expected returns for each trading day, and hence get the unexpected return for each day for each firm. The equation to calculate the unexpected returns is as follows: u it R it ( it i R mt ) where Rit is the stock return of firm i on day t and Rmt is the market index return for day t (DeFond et al., 2007). The non-event period covers all days from t-60 to t-10 and t+10 to t+60 where t represents the event dates, that is, annual earnings announcement dates 22

31 obtained from both the S&P Capital IQ and Thomson Reuters I/B/E/S databases. The annual earnings announcement dates are generally earlier than the actual SEDAR filing dates and represent the first time that the firms publicly announce the annual earnings. This study follows previous research and calculate the natural log of / to obtain AVAR (DeFond et al., 2007): u it 2 it 2 u it 2 it 2 AVAR = ln ( / ) This study uses the following measure for abnormal trading volume, AVOL, following Landsman et al. (2012) and DeFond et al. (2007). V it V i AVOL = ln ( / ) V it where stands for the mean of the event period trading volume and is the average estimation period trading volume for firm i. Following Landsman et al., this study scales both the numerator and the denominator by related daily shares outstanding. V i Information content of TSX and TSXV firms in the pre- and post-adoption periods This study pools all the TSX and the TSXV firm-year observations in the pre- and post IFRS periods to estimate regression equations for AVAR and AVOL similar to those estimated by Landsman et al. (2012). This study investigates whether abnormal return volatility and abnormal volume increase in both TSX and TSXV firms following mandatory adoption of IFRS in Canada by estimating the relevant coefficients for the variable POST (meaning post-ifrs). 23

32 AVARit = β0 + β1postit + β2exchangeit + β3exchange*postit + β4sizeit + β5numestit + β6reportlagit + β7levit + β8dlossit + β9 UEit + β10adrit + β11oil_minit + β12institutional_ownershipit + β13coverageit + β14 RET it + β15post*exchange* RET it + β16post * RET it + β17exchange* RET it + εit (1) AVOLit = β0 + β1postit + β2exchangeit + β3exchange*postit + β4sizeit + β5numestit + β6reportlagit + β7levit+ β8dlossit + β9 UEit + β10adrit + β11oil_minit + β12institutional_ownershipit + β13coverageit + β14 RET it + β15post*exchange* RET it + β16post * RET it + β17exchange* RET it + εit (1a) Here i refers to firm and t refers to year. POST is one if an observation belongs to either of the two years (fiscal years 2012 and 2013) of earnings announcements after IFRS was mandatory in Canada, and is zero if the observation belongs to either of the two years prior to the adoption of IFRS in Canada (fiscal years 2009 and 2010). This study excludes the 2011 fiscal year as that year represents the adoption year for the firms in the sample. EXCHANGE is one for firms traded on the TSXV exchange. Based on findings of previous research, this study includes other independent variables as controls that could potentially affect stock return volatility or trading volume. SIZE is the natural logarithm of the market value of equity calculated at the fiscal yearend of each of four years. There is no clear prediction for the sign of β2 as previous research has found mixed results for firm size (Bamber, Barron, and Stevens, 2011; Barron, Schneible, and Stevens, 2011). NUMEST represents the number of analysts that 24

33 follow a firm at any time during the firm s fiscal year, as reported by Capital IQ database for TSX and TSXV firms. (DeFond et al., 2007). A higher number of analysts indicates more information in price and higher analyst coverage is likely to increase the information in earnings announcements. Therefore, this study predicts the coefficient β3 to be positive (DeFond et al., 2007). REPORTLAG represents the time distance from the firm s fiscal year-end to the earnings announcement as reported by Thomson Reuters I/B/E/S database. The coefficient β4 for reporting lag is expected to be negative because more information is revealed before the report is issued when the report is delayed. LEV (LEVER) represents financial leverage and is calculated as total liabilities at fiscal year-end standardized by total assets at fiscal year-end. DLOSS is equal to one if the reported earnings per share (EPS) as per Capital IQ at fiscal year-end is negative, and zero otherwise. Prior research (Hayn, 1995) found that negative earnings are less informative, and hence this study expects the coefficient β6 to be negative. INSTITUTIONAL_OWNERSHIP represents the percentage of firm shares held by institutional investors in year t. The variable UE is the absolute difference between the actual annual EPS and the actual annual EPS in the prior year for TSX and TSXV firms, standardized by the corresponding closing stock price. The RET variable is the absolute value of the 5 days cumulative return during the event window. Based on prior research (Ahmed, Schneible, and Stevens, 2003), this study expects a positive sign for β14 and also expects a relatively high t-statistic for this variable. ADR is also measured as an indicator variable, and is equal to one if the firm is cross-listed on a U.S. exchange based on Capital IQ, and zero otherwise. The OIL_MIN variable is equal to one if the firm belongs to the oil, gas, or mineral industry, and zero otherwise. The COVERAGE 25

34 variable is equal to one when number of analyst is greater than zero, and zero otherwise. This study expects a positive sign for the coefficient of COVERAGE. In order to evaluate whether AVAR and AVOL are greater for firms in the TSXV exchange than for firms in the TSX exchange, this study uses an indicator variable, EXCHANGE, that equals one for TSXV firms and zero for TSX firms in equation 1 and 1a. The interaction of EXCHANGE with POST will give us the desired interpretation; if the coefficient of EXCHANGE * POST in equation 1 is positive, it means that the TSXV firms experienced a greater increase in information content in the post-ifrs period compared to TSX firms. 2.5 Sample Description First, this study obtained a comprehensive list of firms from S&P Capital IQ that are listed on the TSX and TSXV exchanges. In total, this study found 3859 firms listed on these two exchanges. Then, this study obtained price and volume data from Thomson s Datastream database and the number of firms dropped to 2,040 after excluding firms with missing data. The study lost another 200 firms when it removed those event windows with more than 3 days of zero returns in the 5 days event window. In the next phase, the number of analysts and other accounting and market data were obtained from S&P Capital IQ. Institutional ownership data were obtained from the Capital IQ database. When this study merged this database file with the Thomson s Datastream database file, the sample dropped to 1,605 firms. The study also excluded those firms listed on TSX and TSXV that adopted U.S. GAAP instead of IFRS and non- 12/31 year-end firms as 2012 was their first year of adoption of IFRS. This study 26

35 excluded insurance companies listed on these two exchanges. After applying these additional criteria, the study made sure that it had a constant sample respectively for both the TSX and TSXV exchanges (total firms remaining 405). The table 2-1 below summarizes the study s sample description above. Table 2-1 Sample Selection Sample Description Number of firms TSX and TSXV listed firms (from S&P Capital IQ) 3,859 Missing price and volume data (1819), liquidity threshold restrictions (200), and abnormal return and volume calculations (115). Merging 2 sets of databases (Capital IQ and Thomson Reuters Datastream) (2,134) (120) US GAAP firms, non-12/31 firms & Insurance firms (865). Missing information for constant sample (335) (1,200) Total number of remaining firms 405 Tables 2-2 provides descriptive statistics for AVAR, AVOL, and other variables used in equations 1 and 1a for TSX and TSXV firm during both pre- and post-ifrs adoption periods. The pre-adoption period includes fiscal years 2009 and 2010 and the post-adoption period includes fiscal years 2012 and For AVAR, there are 812 preadoption firm-year observations for TSX and TSXV firms and the same number of observations exists for post-ifrs adoption years. The overall mean and median AVAR (abnormal volatility) is 0.08 and 0.18 for the full sample of 812 firm-year observations in 27

36 the post-ifrs adoption period. Both the 0.08 mean and 0.18 median values are significantly greater than the mean and median in the pre-ifrs adoption period (p<0.01). The mean AVOL (abnormal volume) for both TSX and TSXV firms of 0.01 in the post-ifrs period is not significantly higher than the mean AVOL in the pre-ifrs adoption period. Similarly, the median AVOL of 0.07 in the post-ifrs period is not significantly higher than the median 0.00 in the pre-ifrs adoption period. The mean of REPORTLAG (represents lag between fiscal year end date and the earnings announcement date) variable decreased for all TSX and TSXV firms from days in the pre-ifrs adoption period to days in the post adoption period and the difference is statistically significant (p<0.01). Table 2-3 presents descriptive statistics for AVAR and AVOL for oil, gas, and minerals (OGM) firms and non-ogm firms listed in both TSX and TSXV during both pre- and post-adoption of IFRS. There are 450 pre-adoption firm-year observations for OGM firms and 362 observations for non-ogm firms. The mean AVAR and AVOL for the non-ogm firms are positive in both the pre- and post-ifrs periods. For non-ogm firms, the mean value of AVAR increases from 0.05 in the pre-ifrs period to 0.22 in the post-ifrs period and the difference is statistically significant (p<0.10). For non_ogm firms, the mean value of AVOL of 0.13 in the post-ifrs period is not significantly greater than the mean value of 0.05 in the pre-ifrs period.. Table 2-4 presents descriptive statistics for TSX firms during both pre- and postadoption periods. There are 644 pre- and post-adoption firm-year observations for TSX firms. The overall mean AVAR for TSX firms of 0.15 in the post adoption period is significantly higher than the mean in the pre-ifrs adoption period. The mean 28

37 AVOL for TSX firms of 0.09 in the post-ifrs adoption period is also significantly higher than the mean AVOL of in the pre-ifrs adoption period. The mean of REPORTLAG (represents lag between fiscal year end date and the earnings announcement date) variable decreased for TSX firms from days in the pre-ifrs adoption period to days in the post adoption period. The mean of NUMEST (representing the number of analyst covering a firm) increased from an average of 4.86 analysts in the pre-ifrs adoption period to an average of 6 analysts in the post-adoption period. The mean of INSTITUTIONAL_OWNERSHIP (representing the percentage of firm shares held by institutional investors in year t) also increased from an average of in the pre-ifrs adoption period to an average of in the post-adoption period. Table 2-5 presents descriptive statistics for TSXV firms during both the pre- and post-adoption periods. There are 168 pre- and post-adoption firm-year observations for TSXV firms. The mean AVAR for TSXV firms in the post-ifrs adoption period of is significantly higher than the mean in the pre-ifrs adoption period. However, the mean AVOL for TSXV firms of in the post-ifrs adoption period is lower than the mean AVOL of in the pre-ifrs adoption period. The mean REPORTLAG (represents lag between fiscal year end date and the earnings announcement date) decreased for TSXV firms from days in the pre-ifrs adoption period to days in the post adoption period. The mean of the NUMEST variable (representing the number of analyst covering a firm) increased from 0.58 analysts in the pre-ifrs adoption period to 0.71 analysts in the post-adoption period. Comparing TSX firms in table 2-4 with TSXV firms in table 2-5, this study notices that TSX firms are significantly larger in size with mean of 6.09 versus mean firm 29

38 size of 3.43 for TSXV firms (post-ifrs). There is also a greater analyst following of TSX firms with mean of 6 as compared to mean analyst following for TSXV firms of 0.71 (post-ifrs). The reporting lag is generally lower (mean 67.50) for TSX firms than for TSXV firms (mean ) in the post-ifrs adoption period. TSX firms are less likely to be affected by unexpected earnings UE (0.19 for TSX versus 0.49 for TSXV). TSXV firms have fewer institutional shareholders (15.29% for TSXV and 33.22% for TSX). Similarly, TSX firms are less likely to be loss firms (0.37 for TSX and 0.86 for TSXV firms). These results show that the underlying characteristics of TSX and TSXV firms are different, indicating different motivations for listing on the two exchanges. 2.6 Results Multivariate tests of information content pre- and post-ifrs adoption The multivariate results presented in table 2-6 support hypothesis 1 for TSX firms. The variable of interest POST represents the change in information content for TSX firms between the pre-ifrs and post-ifrs periods. POST is significantly positive for both AVAR and AVOL, indicating greater divergence of opinion for post-ifrs earnings announcements versus pre-ifrs earnings announcements for firms listed on the TSX exchange. The EXCHANGE variable represents the difference in information content between TSXV and TSX firms in the pre-ifrs period. The significantly positive coefficient for AVOL provides some evidence of greater divergence of opinion for TSXV investors in the pre-ifrs period relative to TSX investors. The coefficient on the interaction variable EXCHANGE*POST reflects differences in the information content of earnings for TSXV investors in the post-ifrs period relative to the pre-ifrs period. The significantly negative coefficient for EXCHANGE*POST for AVOL suggests that the 30

39 greater divergence of opinion for TSXV investors versus TSX investors is not sustained in the post-ifrs period (the sum of the coefficients on EXCHANGE and EXCHANGE*POST is not statistically significant). These results do not support hypothesis 2 and may reflect the more speculative nature of the TSXV exchange where information about future expectations may be less value-relevant to investors than information about past realizations. The results with respect to control variables are largely consistent with this study s expectations based on previous research. As predicted by the literature (e.g. Hayn, 1995), the DLOSS coefficient is significantly negative (p-value < 0.10) indicating that earnings announcements for loss firms have less information content. The findings for NUMEST (number of analysts) are consistent with the predicted signs in prior studies (e.g. DeFond et al., 2007), indicating more information content when more analysts follow the firm due to more precise earnings expectations. The coefficients for REPORTLAG are weakly negative, consistent with prior literature (e.g. DeFond et al., 2007), indicating a reduction in information content as the time between the year-end date and the reporting date increases. The weakly significant negative coefficient on ADR is a little surprising given the expectation that joint reporting on the U.S. and Canadian stock exchanges would lead to higher information content of earnings. Some of this effect may be washed away by other variables such as number of analysts, firm size and institutional ownership. As predicted by previous research, the absolute value of abnormal returns and related interactions are strong control variables (Ahmed et al., 2003; Hope et al., 2009). 31

40 2.6.2 Robustness tests As a robustness test, this study compared 2009 fiscal year observations with 2013 fiscal year observations instead of having two pre- and two post-ifrs years in the sample. Now POST is one if an observation belongs to fiscal year 2013 earnings announcements when IFRS was mandatory in Canada, and is zero if the observation belongs to fiscal year 2009 prior to the adoption of IFRS in Canada. In this sample, this study has included all non-december 31 fiscal year end firms observations that were excluded from the previous analysis because they switched to IFRS in This study also includes firms that have observations in both 2009 and 2013 without the more stringent requirement that they have observations in all four years. As a result, the sample size increases to 1,836 observations for the two years. The significantly positive coefficients on the POST variable for AVAR and AVOL presented in table 2-7 support hypothesis 1 for TSX firms. EXCHANGE is significantly positive for AVAR, providing some support for greater divergence of opinion resulting from earnings announcements for TSXV firms versus TSX firms in the pre-ifrs period. This difference is not sustained in the post-ifrs period (sum of EXCHANGE and POST*EXCHANGE is not significantly different from zero). For AVOL, the EXCHANGE variable is not significantly positive in the pre-ifrs periods but the interaction variable EXCHANGE*POST is significantly negative indicating a change in the divergence of opinion post-ifrs for TSXV investors. The combined value of EXCHANGE and POST*EXCHANGE is not significantly different from zero indicating that there is no significant post-ifrs difference across the two exchanges. 32

41 Due to the possibility that the aftermath of the financial crisis affected 2009 observations, this study also provides a comparison of 2010 and 2013 in table 2-8. Again, the POST variables are significantly positive in both the AVAR and AVOL estimations, consistent with hypothesis 1 that information content increased for TSX firms only. The EXCHANGE variable is not significantly different from zero in these estimations and the interaction POST*EXCHANGE is also not significantly different from zero in either regression. This study also runs the 2010 vs test and found that Post variable is not as strong for both AVAR and AVOL. In 2012, many firms were adopting IFRS for the first time, and it believes that 2013 is a cleaner post-ifrs adoption year for testing to mitigate the macroeconomic effects. To summarize these results, the significantly positive coefficients on the POST variable in all estimations (AVAR and AVOL and different samples) provides support for the hypothesis that divergence of opinion resulting from earnings announcements for TSX firms was higher in the post-ifrs periods than in the pre-ifrs periods. For TSXV firms, the significantly positive coefficients on the EXCHANGE variable in some estimations provides limited evidence of greater divergence of opinion resulting from earnings announcements relative to TSX firms in the pre-ifrs period. There is no evidence that this greater divergence of opinion across the two exchanges was sustained in the post-ifrs period. This study also conducted robustness tests using a 200 day non-event period. The main result for POST is significantly positive for both AVAR and AVOL estimations. However, the POST*EXCHANGE variable is not significantly different from zero for AVOL. As suggested in Bamber et al (2011), this study conducted the log versus no log 33

42 sensitivity test using alternative definitions for AVOL and found that the POST variable becomes weaker in the two years versus two years analysis, that is, 2009 and 2010 as pre- IFRS adoption years and 2012 and 2013 as post-adoption years. However, in the 2009 versus 2013 and 2010 versus 2013 comparisons, the POST variable remains significant and positive for both AVAR and AVOL. This study performs other robustness tests (not tabulated) that vary the screen for excluded observations based on zero price changes during the announcement period. Results of these estimations support the main finding that information content increased for TSX firms between the pre-ifrs and post-ifrs periods. The results with respect to EXCHANGE and POST*EXCHANGE are less stable across these different estimations making the conclusions of this study with respect to these variables less robust. 2.7 Conclusions This study examines whether IFRS adoption resulted in greater information content in earnings announcements for TSX and TSXV firms in Canada. This setting offers a unique opportunity to study the impact of IFRS implementation in a single country for firms traded on senior and junior exchanges. Similarities between the Canadian and U.S. settings mean that the findings may be useful to U.S. policy makers. This study s results consistently support a finding that information content of earnings announcements, measured by abnormal return volatility and abnormal volume during the announcement period, increased for TSX firms following the adoption of IFRS. These results indicate that IFRS earnings announcements increased the valuerelevant information provided to investors in a manner that led to greater divergence of 34

43 opinion, possibly due to the more forward-looking measurement emphasis under IFRS. This study finds some evidence that information content of earnings was higher for TSXV firms relative to TSX firms in the pre-ifrs period and somewhat stronger evidence that the increase in information content for TSX firms post-ifrs was not shared by TSXV firms. This study cautiously suggests that this may reflect a greater reliance of TSXV investors on realized earnings information. 35

44 Bibliography Aboody, D., M. E. Barth, and R. Kasznik Revaluations of fixed assets and future firm performance: Evidence from the U. K. Journal of Accounting and Economics. 26: Ahmed, A. S., M. Neel, and D. Wang Does mandatory adoption of IFRS improve accounting quality? Preliminary evidence. Contemporary Accounting Research. 30(4): Ahmed, A. S., R. A. Schneible, and D. E. Stevens An empirical analysis of the effects of online trading on stock price and trading volume reactions to earnings announcements. Contemporary Accounting Research 20(3): Amiram, D Financial information globalization and foreign investment decisions. Journal of International Accounting Research. 11(2): Armstrong, C., M. Barth, A. Jagolinzer, and E. Riedl Market reaction to the adoption of IFRS in Europe. The Accounting Review. 85(1): Ausseneg, W., P. Inwinkl, and G. Schneider Earnings management and local vs International accounting standards of European public firms. Working Paper, Vienna University of Technology. Bae, K. H., Tan, H., & Welker, M. (2008). International GAAP differences: The impact on foreign analysts. The Accounting Review, 83(3), Ball, R. and P. Brown An empirical evaluation of accounting income numbers. Journal of Accounting Research Ball, R International Financial Reporting Standards (IFRS): pros and cons for investors. Accounting and Business Research. 36(sup1): Bamber, L., O. Barron, and D. Stevens Trading volume around earnings announcements and other financial reports: Theory, research design, empirical evidence, and directions for future research. Contemporary Accounting Research. 28(2): Barron, O., R. Schneible, and D. Stevens What do we really Know about the Firm size Effect on Trading Volume Reactions to Earnings Announcement? A Re-Examination and Extension. Working paper. Barth, M., W. Landsman, and M. Lang International accounting standards and accounting quality. Journal of Accounting Research, 46(3),

45 Barth, M., W. Landsman, M. Lang, and C. Williams Are IFRS-based and US GAAP-based accounting amounts comparable? Journal of Accounting and Economics. 54(1): Basu, S The conservatism principle and the asymmetric timeliness of earnings. Journal of Accounting and Economics. 24(1): Beaver, W Alternative accounting measures as predictors of failure. The Accounting Review. 43(1): Beaver, W. and S. Ryan Conditional and unconditional conservatism: concepts and modeling. Review of Accounting Studies. 10(2-3): Beuselinck, C., P. Joos, I. Khurana, and S. van der Meulen Mandatory IFRS Reporting and Stock Price Informativeness. CentER Discussion Paper; Vol Tilburg University. Biddle, G, G. Hilary, and R. Verdi How does financial reporting quality relate to investment efficiency? Journal of Accounting and Economics. 48(2): Blanchette, M., F.-E. Racicot, and J.-Y. Girard The effects of IFRS on financial ratios: Early evidence in Canada. Certified General Accountants Association of Canada. Byard, D., Y. Li, and Y. Yu The effect of mandatory IFRS adoption on financial analysts information environment. Journal of Accounting Research. 49(1): Covrig, V., M. Defond, and M. Hung Home bias, foreign mutual fund holdings, and the voluntary adoption of international accounting standards. Journal of Accounting Research. 45(1): Danbolt, J., and W. Rees An experiment in fair value accounting: UK investment vehicles. European Accounting Review. 17(2): Daniel, K., D. Hirshleifer, and A. Subrahmanyam Investor psychology and security market under-and overreactions. The Journal of Finance, 53(6), Daske, H., L. Hail, C. Leuz, and R. Verdi Mandatory IFRS reporting around the world: Early evidence on the economic consequences. Journal of Accounting Research. 46: DeFond, M., M. Hung, and R. Trezevant Investor protection and the information content of annual earnings announcements: International evidence. Journal of Accounting and Economics. 43(1): DeFond, M., X. Hu, M. Hung, and S. Li The impact of mandatory IFRS adoption on foreign mutual fund ownership: The role of comparability. Journal of Accounting and Economics, 51(3),

46 Doukakis, L The persistence of earnings and earnings components after the adoption of IFRS. Managerial Finance. 36: Dye, R. and S. Sunder Why not allow FASB and IASB standards to compete in the US? Accounting Horizons. 15(3): Fields, T., T. Lys, and L. Vincent Empirical research on accounting choice. Journal of Accounting and Economics. 31(1): Francis, J., S. Huang, I. Khurana, and R. Pereira Does corporate transparency contribute to efficient resource allocation? Journal of Accounting Research. 47(4): Hayn, C The information content of losses. Journal of accounting and economics, 20(2), Healy, P., and K. Palepu Information asymmetry, corporate disclosure, and the capital markets: A review of the empirical disclosure literature. Journal of accounting and economics. 31(1): Healy, P., A. Hutton, and K. Palepu Stock performance and intermediation changes surrounding sustained increases in disclosure. Contemporary Accounting Research, 16(3), Hope, O., W. Thomas, and G. Winterbotham, G Geographic earnings disclosure and trading volume. Journal of Accounting and Public Policy, 28(3), Horton, J., G. Serafeim, and I. Serafeim "Does Mandatory IFRS Adoption Improve the Information Environment?" Contemporary Accounting Research. 30(1): Jeanjean, T., and H. Stolowy Do accounting standards matter? An exploratory analysis of earnings management before and after IFRS adoption. Journal of Accounting and Public Policy. 27(6): Kabir, M., F. Laswad, and M. Islam Impact of IFRS in New Zealand on accounts and earnings quality. Australian Accounting Review. 20(4): Kim, O. and R. Verrecchia Market reaction to anticipated announcements. Journal of Financial Economics. 30(2): Kim, O. and R. Verrecchia Trading volume and price reactions to public announcements. Journal of Accounting Research. 29(2): Kim, O. and R. Verrecchia Pre-announcement and event-period private information. Journal of Accounting and Economics. 24(3):

47 Kothari, S Capital markets research in accounting. Journal of Accounting and Economics. 31(1): Lambert, R., C. Leuz, and R. Verrecchia Accounting information, disclosure, and the cost of capital. Journal of Accounting Research. 45(2): Landsman, W., E. Maydew, and J. Thornock The information content of annual earnings announcements and mandatory adoption of IFRS. Journal of Accounting and Economics. 53(1): Leuz, C, and P. Wysocki Economic consequences of financial reporting and disclosure regulation: A review and suggestions for future research. Working paper. Available at: abstract= Leuz, C., and R. Verrecchia The Economic Consequences of Increased Disclosure (Digest Summary). Journal of Accounting Research, 38, Lev, B On the usefulness of earnings and earnings research: lessons and directions from two decades of empirical research. Journal of Accounting Research. 40: Li, S Does mandatory adoption of International Financial Reporting Standards in the European Union reduce the cost of equity capital? The Accounting Review. 85(2): O'Brien, P Changing the Concepts to Justify the Standards. Accounting Perspectives. 8(4): O'Brien, P., and W. Scott Contract theory: The effectiveness of a conceptual framework for financial reporting has been debated for years. CA Magazine, April 1. PriceWaterhouseCoopers Putting IFRS in motion: Key accounting issues for pharmaceuticals and life sciences companies. Scott, W. R Financial accounting theory. Pearson Education Canada. Sharpe, I. and R. Walker Asset revaluation and stock market prices. Journal of Accounting Research. 13: Standish, P, and S.-I. Ung Corporate signaling, asset revaluation and stock prices of British companies. The Accounting Review. 57(4): Tan, H., Wang, S., and M. Welker Analyst following and forecast accuracy after mandated IFRS adoptions. Journal of Accounting Research, 49(5), Verrecchia, R Essays on disclosure. Journal of Accounting and Economics. 32(1-3):

48 Watts, R Conservatism in accounting part I: Explanations and implications. Accounting Horizons. 17(3): Watts, R Conservatism in accounting part II: Evidence and research opportunities. Accounting Horizons 17(4): Welker, M Disclosure policy, information asymmetry and liquidity in equity markets. Contemporary Accounting Research. 11: Yu, G Accounting standards and international portfolio holdings: Analysis of cross-border holdings following mandatory adoption of IFRS. Dissertation. University of Michigan. 40

49 Table 2-2 Descriptive Statistics for the combined TSX and TSXV stock markets prior to and post IFRS adoption Prior to IFRS Adoption Post IFRS Adoption Difference in Mean Difference in Median Variable N Mean Std. Std. Median Mean Dev. Dev. Median AVAR *** *** AVOL INST_OWN * REPORTLAG *** *** SIZE LEVER *** OIL_MIN COVERAGE *** *** UE * EXCHANGE ADR NUMEST *** *** DLOSS RET * PRICE ** ** * if significance <0.10, ** if significance <0.05, *** if significance <0.01 Post IFRS adoption is for fiscal years 2012 and AVAR is the abnormal return volatility, AVOL is abnormal volume, SIZE is the natural logarithm of the market value of equity calculated at fiscal year-end, NUMEST is the number of analyst that follow a firm at any time during the firm s fiscal year, REPORTLAG is the time distance from the firm s fiscal year-end to the earnings announcement date, LEVER is calculated as total liabilities at fiscal year-end standardized by total assets at fiscal year-end, DLOSS is equal to 1 if the reported earnings per share (EPS) at fiscal year-end is negative, and 0 otherwise, UE is the absolute difference between the actual annual EPS and the actual annual EPS in the prior year, standardized by the corresponding closing stock price, ADR is equal to one if the firm is cross-listed on a U.S. exchange and zero otherwise. POST x EXCHANGE is an indicator variable and is 1 if it is TSXV firm and in the POST period. EXCHAGE is 1 if TSXV firm and 0 if TSX firm. The OIL_MIN variable is equal to 1 if the firm belongs to the oil, gas, or mineral industry, and 0 otherwise. INSTITUTIONAL_OWNERSHIP represents the percentage of firm shares held by institutional investors in year t. The COVERAGE variable is equal to 1 when number of analyst is greater than 0, and 0 otherwise. RET is the absolute value of the cumulative return during the 5 days event window. We winsorize AVAR, AVOL, REPORTLAG, LEVER and UE variables at the 1st and 99 th percentiles. We used t-tests for difference in means and Wilcoxon test for difference in medians. 41

50 Table 2-3 AVAR and AVOL for the combined TSX and TSXV stock markets broker into subcategories Oil, Gas and Mineral (prior to and post IFRS adoption) and Non-Oil, Gas and Mineral (prior to and post IFRS adoption) Oil, Gas and Mineral in the combined Markets prior to IFRS adoption (n=450) Oil, Gas and Mineral in the combined Markets post to IFRS adoption (n=450) Difference in above two rows Non-Oil, Gas and Mineral in the combined Markets prior to IFRS adoption (n=362) Non-Oil, Gas and Mineral in the combined Markets post to IFRS adoption (n=362) Difference in above two rows AVAR Mean AVAR Median AVAR Std. Dev. AVOL Mean* AVOL Median* AVOL Std. Dev.* *** *** * ** * if significance <0.10, ** if significance <0.05, *** if significance <0.01 AVAR is the abnormal return volatility and AVOL is the abnormal volume. We used t- tests for difference in means and Wilcoxon test for difference in medians. *The sample size for AVOL in non-oil, Gas and Mineral is 360 both prior to and post- IFRS adoption. 42

51 Table 2-4 Descriptive statistics for firms in TSX stock market Prior to IFRS Adoption Post IFRS Adoption Difference in Mean Difference in Median Variable N Mean Std. Std. Median Mean Dev. Dev. Median AVAR *** *** AVOL INST_OWN * REPORTLAG *** *** SIZE LEVER ** ** OIL_MIN COVERAGE *** *** UE ADR NUMEST *** *** DLOSS RET PRICE ** ** Post IFRS adoption is for fiscal years 2012 and AVAR is the abnormal return volatility, AVOL is abnormal volume, SIZE is the natural logarithm of the market value of equity calculated at fiscal year-end, NUMEST is the number of analyst that follow a firm at any time during the firm s fiscal year, REPORTLAG is the time distance from the firm s fiscal year-end to the earnings announcement date, LEVER is calculated as total liabilities at fiscal year-end standardized by total assets at fiscal year-end, DLOSS is equal to 1 if the reported earnings per share (EPS) at fiscal year-end is negative, and 0 otherwise, UE is the absolute difference between the actual annual EPS and the actual annual EPS in the prior year, standardized by the corresponding closing stock price, ADR is equal to one if the firm is cross-listed on a U.S. exchange and zero otherwise. POST x EXCHANGE is an indicator variable and is 1 if it is TSXV firm and in the POST period. EXCHAGE is 1 if TSXV firm and 0 if TSX firm. The OIL_MIN variable is equal to 1 if the firm belongs to the oil, gas, or mineral industry, and 0 otherwise. INSTITUTIONAL_OWNERSHIP represents the percentage of firm shares held by institutional investors in year t. The COVERAGE variable is equal to 1 when number of analyst is greater than 0, and 0 otherwise. RET is the absolute value of the cumulative return during the 5 days event window. We winsorize AVAR, AVOL, REPORTLAG, LEVER and UE variables at the 1st and 99 th percentiles. We used t-tests for difference in means and Wilcoxon test for difference in medians. 43

52 Table 2-5 Descriptive statistics for firms in TSXV stock market Prior to IFRS Adoption Post IFRS Adoption Difference in Mean Difference in Median Variable N Mean Std. Std. Median Mean Dev. Dev. Median AVAR * * AVOL INST_OWN REPORTLAG *** ** SIZE LEVER ** OIL_MIN COVERAGE UE ** ** ADR NUMEST DLOSS RET * PRICE ** ** Post IFRS adoption is for fiscal years 2012 and AVAR is the abnormal return volatility, AVOL is abnormal volume, SIZE is the natural logarithm of the market value of equity calculated at fiscal year-end, NUMEST is the number of analyst that follow a firm at any time during the firm s fiscal year, REPORTLAG is the time distance from the firm s fiscal year-end to the earnings announcement date, LEVER is calculated as total liabilities at fiscal year-end standardized by total assets at fiscal year-end, DLOSS is equal to 1 if the reported earnings per share (EPS) at fiscal year-end is negative, and 0 otherwise, UE is the absolute difference between the actual annual EPS and the actual annual EPS in the prior year, standardized by the corresponding closing stock price, ADR is equal to one if the firm is cross-listed on a U.S. exchange and zero otherwise. POST x EXCHANGE is an indicator variable and is 1 if it is TSXV firm and in the POST period. EXCHAGE is 1 if TSXV firm and 0 if TSX firm. The OIL_MIN variable is equal to 1 if the firm belongs to the oil, gas, or mineral industry, and 0 otherwise. INSTITUTIONAL_OWNERSHIP represents the percentage of firm shares held by institutional investors in year t. The COVERAGE variable is equal to 1 when number of analyst is greater than 0, and 0 otherwise. RET is the absolute value of the cumulative return during the 5 days event window. We winsorize AVAR, AVOL, REPORTLAG, LEVER and UE variables at the 1st and 99 th percentiles. We used t-tests for difference in means and Wilcoxon test for difference in medians. 44

53 Table 2-6 Firm-level regressions of abnormal return volatility and abnormal trading volume for full sample with 2009 and 2010 as pre-ifrs and 2012 and 2013 as post-ifrs period (standard errors clustered by firms) Abnormal return volatility (AVAR) Abnormal trading volume (AVOL) Variables Pred. Sign Coefficient t-stats Coefficient t-stats Intercept? *** POST *** ** EXCHANGE? ** POST x EXCHANGE ** REPORTLAG * * SIZE? *** * LEVER UE DLOSS *** * NUMEST *** * ADR ** COVERAGE * INSTIT_OWN? OIL_MIN *** ** RET *** *** POST x RET? EXCHANGE x RET POST x EXCHANGE x RET? * *? Mean Adj. R Squared Number of observations * if significance <0.10, ** if significance <0.05, *** if significance <0.01 Post IFRS adoption is for fiscal years 2012 and AVAR is the abnormal return volatility, AVOL is abnormal volume, SIZE is the natural logarithm of the market value of equity calculated at fiscal year-end, NUMEST is the number of analyst that follow a firm at any time during the firm s fiscal year, REPORTLAG is the time distance from the firm s fiscal year-end to the earnings announcement date, LEVER is calculated as total liabilities at fiscal year-end standardized by total assets at fiscal year-end, DLOSS is equal to 1 if the reported earnings per share (EPS) at fiscal year-end is negative, and 0 otherwise, UE is the absolute difference between the actual annual EPS and the actual annual EPS in the prior year, standardized by the corresponding closing stock price, ADR is equal to one if the firm is cross-listed on a U.S. exchange and zero otherwise. POST x EXCHANGE is an indicator variable and is 1 if it is TSXV firm and in the POST period. EXCHAGE is 1 if TSXV firm and 0 if TSX firm. The OIL_MIN variable is equal to 1 if 45

54 the firm belongs to the oil, gas, or mineral industry, and 0 otherwise. INSTITUTIONAL_OWNERSHIP represents the percentage of firm shares held by institutional investors in year t. The COVERAGE variable is equal to 1 when number of analyst is greater than 0, and 0 otherwise. RET is the absolute value of the cumulative return during the 5 days event window. We winsorize AVAR, AVOL, REPORTLAG, LEVER and UE variables at the 1st and 99 th percentiles. 46

55 Table 2-7 Firm-level regressions of abnormal return volatility and abnormal trading volume for full sample with 2009 as pre-ifrs and 2013 as post-ifrs period (standard errors clustered by firms) Abnormal return volatility (AVAR) Abnormal trading volume (AVOL) Variables Pred. Sign Coefficient t-stats Coefficient t-stats Intercept? *** POST *** *** EXCHANGE? ** POST x EXCHANGE * REPORTLAG SIZE? *** LEVER UE DLOSS *** *** NUMEST ** ** ADR * COVERAGE * INSTIT_OWN? * OIL_MIN *** ** RET *** *** POST x RET? *** ** EXCHANGE x RET POST x EXCHANGE x RET? *** ? Mean Adj. R Squared Number of observations * if significance <0.10, ** if significance <0.05, *** if significance <0.01 Post IFRS adoption is for fiscal years 2012 and AVAR is the abnormal return volatility, AVOL is abnormal volume, SIZE is the natural logarithm of the market value of equity calculated at fiscal year-end, NUMEST is the number of analyst that follow a firm at any time during the firm s fiscal year, REPORTLAG is the time distance from the firm s fiscal year-end to the earnings announcement date, LEVER is calculated as total liabilities at fiscal year-end standardized by total assets at fiscal year-end, DLOSS is equal to 1 if the reported earnings per share (EPS) at fiscal year-end is negative, and 0 otherwise, UE is the absolute difference between the actual annual EPS and the actual annual EPS in the prior year, standardized by the corresponding closing stock price, ADR is equal to one if the firm is cross-listed on a U.S. exchange and zero otherwise. POST x EXCHANGE is an indicator variable and is 1 if it is TSXV firm and in the POST period. EXCHAGE is 1 if TSXV firm and 0 if TSX firm. The OIL_MIN variable is equal to 1 if 47

56 the firm belongs to the oil, gas, or mineral industry, and 0 otherwise. INSTITUTIONAL_OWNERSHIP represents the percentage of firm shares held by institutional investors in year t. The COVERAGE variable is equal to 1 when number of analyst is greater than 0, and 0 otherwise. RET is the absolute value of the cumulative return during the 5 days event window. We winsorize AVAR, AVOL, REPORTLAG, LEVER and UE variables at the 1st and 99 th percentiles. 48

57 Table 2-8 Firm-level regressions of abnormal return volatility and abnormal trading volume for full sample with 2010 as pre-ifrs and 2013 as post-ifrs period (standard errors clustered by firms) Abnormal return volatility (AVAR) Abnormal trading volume (AVOL) Variables Pred. Sign Coefficient t-stats Coefficient t-stats Intercept? *** POST *** *** EXCHANGE? POST x EXCHANGE REPORTLAG ** SIZE? *** LEVER UE DLOSS *** *** NUMEST *** *** ADR ** COVERAGE INSTIT_OWN? ** OIL_MIN *** ** RET *** *** POST x RET? *** * EXCHANGE x RET POST x EXCHANGE x RET Mean Adj. R Squared Number of observations? *** ? * if significance <0.10, ** if significance <0.05, *** if significance <0.01 Post IFRS adoption is for fiscal years 2012 and AVAR is the abnormal return volatility, AVOL is abnormal volume, SIZE is the natural logarithm of the market value of equity calculated at fiscal year-end, NUMEST is the number of analyst that follow a firm at any time during the firm s fiscal year, REPORTLAG is the time distance from the firm s fiscal year-end to the earnings announcement date, LEVER is calculated as total liabilities at fiscal year-end standardized by total assets at fiscal year-end, DLOSS is equal to 1 if the reported earnings per share (EPS) at fiscal year-end is negative, and 0 otherwise, UE is the absolute difference between the actual annual EPS and the actual annual EPS in the prior year, standardized by the corresponding closing stock price, ADR is equal to one if the firm is cross-listed on a U.S. exchange and zero otherwise. POST x EXCHANGE is an indicator variable and is 1 if it is TSXV firm and in the POST period. EXCHAGE is 1 if TSXV firm and 0 if TSX firm. The OIL_MIN variable is equal to 1 if 49

58 the firm belongs to the oil, gas, or mineral industry, and 0 otherwise. INSTITUTIONAL_OWNERSHIP represents the percentage of firm shares held by institutional investors in year t. The COVERAGE variable is equal to 1 when number of analyst is greater than 0, and 0 otherwise. RET is the absolute value of the cumulative return during the 5 days event window. We winsorize AVAR, AVOL, REPORTLAG, LEVER and UE variables at the 1st and 99 th percentiles. 50

59 Chapter 3: Did Mandatory Adoption of IFRS Increase Liquidity in the Canadian Stock Markets? Abstract: This study investigates whether average liquidity for three groups of firms - Canadian, non-us (international) and U.S. firms - traded on Canadian stock exchanges increased or decreased after mandatory adoption of IFRS in Canada. It consider two competing forces affecting liquidity from IFRS adoption: enhanced comparability of firms within industries that span international boundaries and less tailoring of financial reporting to satisfy local investor needs. This study finds that liquidity decreased for Canadian and U.S. firms, suggesting that the benefits of global comparability were not sufficient to offset the loss of local investor accommodation. However, there is some evidence that liquidity improved for non-u.s. international firms suggesting that enhanced comparability of firms within industries did span international boundaries. To provide perspective on these results for Canadian exchanges, this study compares liquidity before and after IFRS adoption for all three groups of firms traded on the U.K. and Australian exchanges. These three countries share the British influence on financial reporting historically and have similar legal and institutional settings. This study expects that the benefits of global comparability would be relatively greater for the U.K. exchanges relative to the Canadian and Australian exchanges given the U.K. s proximity and commerce with other European countries. It finds that liquidity increased for the U.K. exchanges but decreased for the Australian exchange, supporting the interpretation of the Canadian findings. 51

60 3.1 Introduction Accounting is shaped by economic and political forces (Ball, 2006, p. 8) In his discussion of IFRS, Ball (2006) points out that there is little theory or evidence on the advantages and disadvantages of uniform accounting rules within a country. Ball suggested that local forces and politics were too strong for the idea of international convergence of financial reporting standards to work as intended and the benefits of comparability were doubtful. However, the attractive notion that IFRS would enhance the global movement of capital persuaded more than 100 countries to adopt IFRS. This is evident in the calls for IFRS made by Sir David Tweedie, International Accounting Standards Board (IASB) Chairman, when he was advocating the case for international uniformity in accounting standards (Ball, 2006): As the world s capital markets integrate, the logic of a single set of accounting standards is evident. A single set of international standards will enhance comparability of financial information and should make the allocation of capital across borders more efficient. (p. 15) This underlying motivation for globalization of capital markets based on the notion of enhanced comparability under IFRS is also highly visible in the objectives of International Financial Reporting Standards (IFRS) as laid out by the International Accounting Standards (IAS) body: 1. develop high quality, understandable and enforceable global accounting standards that require high quality, transparent and comparable information to help participants in the world s capital markets and other users 52

61 2. promote the use and rigorous application of those standards 3. bring about convergence The international community rejoiced in the fact that Canada decided to adopt IFRS with the backdrop of U.S. uncertainty. Some thought that Canada s decision might help persuade the U.S. to join the IFRS bandwagon. The decision was also significant considering that Canada and the U.S. are connected in many ways. U.S. is Canada s largest trading partner and U.S. firms, still reporting under U.S. GAAP, have a significant presence (50% of international firms) in both Canadian stock exchanges TSX and TSXV (see table 3-1). Therefore, Canada offers a unique experimental setting for investigating the impact of IFRS adoption on liquidity of Canadian, non-u.s. international and U.S. firms. This study investigates the impact of mandatory adoption of IFRS on the liquidity of domestic, non-u.s. international, and U.S. international firms in the Canadian stock markets. It defines an international firm as one whose headquarter address in not in the country of listing for the stock exchange. However, this study takes out the effect of U.S. firms by controlling for them in the regression analysis. The primary research question addresses the core objective of Canada adopting IFRS that Canada wanted to integrate its capital markets with the rest of the world (non-u.s.) even though it was already well integrated with the U.S. markets. Evidence of increases in liquidity, measured by zero returns and bid-ask spread, for both international (non-u.s. and U.S) and domestic firms between the pre- and post-ifrs Canadian stock markets would support a hypothesis that IFRS adoption has fostered the globalization of Canadian stock markets. However, this study finds that the liquidity, measured by zero-returns and bid-ask spreads, did not 53

62 increase for Canadian and U.S. firms in the post-ifrs adoption period. Moreover, it finds evidence that non-u.s. international firms traded on the Canadian exchanges gained in liquidity, measured by zero-returns and bid-ask spreads, in the post-ifrs adoption period. If we can t export the scenery, we we ll import the tourists. W.C. Van Horne, the Man in charge of building the Royal Canadian Pacific Railways, 1886 Greater globalization of stock markets is appealing because more participation by international investors and firms in a country s stock markets enhances capital allocation. The information environments of the stock markets improve as a result of increased globalization resulting in less information asymmetry (Healy and Palepu, 2001). This helps investors to diversify risks and reduces volatility in those stock markets. In turn, the prices of investments in those stock markets get closer to their true underlying value. These benefits increase the overall economic prosperity of a country as the capital allocation process improves (Scott, 2011). Healy and Palepu (2001) conclude that the trends of capital markets becoming increasingly global justify the increased need for global accounting standards. They discuss three trends that explain why we need global accounting standards; the increasing demand for information as institutional investors look to diversify by investing around the globe; corporations seeking capital wherever the market conditions are most attractive; and individual investors trading more easily using internet-based technologies. Healy and Palepu s (2001) discussion motivates this study s research question relating International accounting standards to their impact on globalization of stock markets and liquidity of the exchanges in a country. 54

63 However, the globalization of capital markets also means that local conditions or industry needs may be ignored. Domestic industry is where domestic jobs are created. The local communities and their cultures need local economic development to foster the sustainability of those local environments. U.S. and Canada have very close economic ties with Canada being the largest trading partner. The stakes were inherently high in Canada s decision to adopt IFRS. Ball et al. (2003) specifically points to this dilemma of underlying economic and political institutions influencing the effectiveness of mandatory IFRS adoption by a country. O Brien (2009) stated the following: As 2011 approaches, many of us try to reassure ourselves that international financial reporting standards (IFRS) are not so different from our familiar Canadian generally accepted accounting principles (GAAP). At the same time worldwide convergence in standard setting is accelerating fundamental changes in the accounting landscape. These changes include emphasizing the balance sheet over the income statement, valuation over performance, and the hypothetical over the concrete. This explains why it is important to study Canada s decision to adopt IFRS. IFRS adoption in Canada offers a unique opportunity to study the trade-off that IFRS may enhance global capital allocation at the cost of local information and its effect on the domestic economy. Canada is a country where strong IFRS implementation (enforcement) is a reality due to her strong legal and governance related institutions. There is no issue of concurrent changes in governance and regulatory institutions in this 55

64 study that were potential problems in previous European and Asian IFRS studies (Daske et al., 2008; Landsman et al., 2012). The domestic Canadian economy is heavily dependent on two strong industries the financial and extractive industries. Canada has two major stock exchanges the senior Toronto Stock Exchange or TSX and the junior TSX Venture Exchange or TSXV. The TSXV, where domestic retail investors are heavily involved in raising capital for the extractive industries uses many innovative financial solutions (Pandes, A. and Robinson, M., 2013). This helps local entrepreneurs invest capital in research and development of technologies geared towards solving issues facing domestic extractive industries (e.g. reducing carbon emissions in Oil Sands projects). Canadian extractive industries are carefully developed and managed by the domestic economy. Similarly, a reason that Canada was saved from the wrath of the financial crisis while being so integrated to the U.S. economy is that her strong banking environment is managed by the domestic Canadian banking regulatory institution, the Office of Superintendent of Financial Institutions (OSFI), with stringent financial reporting and capital requirements. This is interesting as Canadian banks are still somewhat isolated from the rest of the world outside North America. Local GAAP standards were developed based on domestic conditions and domestic demand in each country for properties of accounting information. This localization quality in accounting is reflected in the high demand for conservatism in accounting in Canada (Ball, 2006). Conservatism in accounting helps prevent overstated performance in the accounting statements. High conservatism in accounting in Canada may be desirable due to the high concentration of resource-based firms in Canada. The 56

65 fair values of these resources are volatile in nature, making fair value information less useful for holding managers accountable (O Brien, 2009). Thus, one might argue that the concept of globalization of accounting standards is justified by the benefits to the world investor community and the concept of localization of accounting standards can be explained in terms of the benefits to the domestic investors and domestic industries. For example, the analytical report on the U.S. IFRS website talks about one of the benefits of globalized markets that the U.S. is missing in this trade-off: SEC s continued indecision on IFRS leaves the U.S. isolated and limits U.S. companies access to international capital. This study investigates the impact of countrywide adoption of IFRS on the liquidity of domestic, non-u.s. international, U.S. international firms listed on the Canadian stock markets. Increases in liquidity for international and domestic firms listed in the Canadian stock markets in the post-ifrs period would indicate that Canadian stock markets benefitted in terms of more comparability of financial reports with international companies from the decision to mandatorily adopt IFRS as its dominant set of accounting standards. Instead, this study finds that liquidity, measured by bid-ask spreads and zeroreturns, decreased for both domestic and U.S. firms listed in the Canadian capital markets. These results are interesting in the context that Canada and U.S. are strong regional trading partners. However, liquidity increased for non-u.s. international firms, indicating that non-u.s. international firms benefitted from IFRS adoption in Canada. 57

66 For comparative purposes, this study applies the liquidity tests to two other countries Australia and the U.K. Both of these countries are similar to Canada in many ways because of their Commonwealth heritage. Their legal, political and economic institutions share common grounds. Australia also has a strong presence of domestic extractive industries like Canada while the U.K. has a strong financial industry sector. However, they differ in terms of their stock markets along the lines of domestic vs. global forces. The U.K. s stock markets are strongly global in the context of its close proximity to the European Union. Most countries in the European Union adopted IFRS during the same time frame as when U.K. adopted IFRS. There are more global and more diversified firms (table 3-5 and table 3-19) in the U.K. stock exchanges as there are relatively larger international firms listed in the U.K. compared to new and small international firms mostly focused in the extractive industries listed in Canada and Australia (table 3-5). U.K. banks are more global while Canadian and Australian banks are mostly isolated from the rest of the world. Australia and Canada are naturally rich in extractive resources and hence the domestic forces are stronger in these countries. Canada is also unique among these countries because of its close proximity and trading relationships with U.S., world s largest economic power, and it has a large set of U.S. firms listings (50% of international firms) in the Canadian stock markets (see table 3-1). Just like Canada s results, this study finds that liquidity for domestic firms, measured using bid-ask spreads and zero-returns, decreased after IFRS adoption in the more domestic Australian stock market. However, in the more globally oriented stock market of the U.K., liquidity generally increased for domestic and non-u.s. international 58

67 firms. The study s comparative results during the post-ifrs adoption period also show that liquidity, measured by bid-ask spreads and zero-returns, was mostly indifferent for U.S. firms in the U.K. stock markets but decreased for U.S. firms in the Canadian stock markets. Finally, the same liquidity regression results for domestic firms also hold in Canada, U.K, and Australia, even when this study takes out all firms from extractive industries and financial industries. For robustness, this study conducted tests by firm-size quintiles for each country under the assumption that larger firms are more connected and integrated with the rest of the world increasing the likelihood that they have peer companies that investors can compare them with. The results indicate that smaller firms in both Canada and Australia experienced lower liquidity while larger firms experienced higher liquidity in both Canada and U.K. stock exchanges in the post-ifrs adoption period. Finally, this study conducted a number of geographic segments tests for those size quintiles. More geographic segments in a firm compared to another firm indicate that the firm has operations in more countries that the other firm. The results indicate that smaller firms in Canada and Australia have lower numbers of geographic segments while larger firms in both the Canada and the U.K. stock exchanges have higher numbers of geographic segments. This supports the assumption that larger firms are more international. A surprising part is that smaller size firms in U.K. have higher numbers of geographic segments, indicating that U.K. firms are, in general, truly international irrespective of size. This result is consistent with this study s main conjecture that IFRS resulted in higher liquidity for more internationally diversified exchanges where global forces were more active. 59

68 3.2 Literature Review Several recent studies have documented the capital market advantages of IFRS. Some studies showed voluntary IFRS adoption resulted in lower cost of equity (Barth et al., 2008, Leuz and Verrechia, 2000). Some studies argued that one set of uniform financial accounting standards is likely to make financial reports more comparable across firms in many countries, which ultimately is expected to reduce cost of equity for these firms (Armstrong et al., 2010). Barth et al. (2008) found evidence that firms that adopted IFRS generally showed higher accounting quality as compared to firms that were still following local GAAP. Another study found that, once firms adopted IFRS, it results in enhanced comparability of IFRS accounting amounts with US GAAP amounts assuming that US GAAP amounts are of high quality (Barth et al., 2012). Armstrong et al. (2010) tested how the market reacted to the adoption of IFRS in Europe. They found positive market reaction to the news of adoption of IFRS by those firms who had low information quality or higher information asymmetry before the adoption of IFRS. That is reasonable with the argument that markets expected net information quality markets benefits in the post-ifrs adoption. However, they found negative market reaction to the IFRS adoption news of those firms especially from countries where the enforcement was doubtful in the first place. In another study investigating IFRS adoption by European countries by Beuselinck et al. (2010), it was evident that IFRS-based disclosures were firm-specific information revealing in the post IFRS adoption period. They also linked their result to 60

69 reduced future surprise news. Most of these early IFRS adopting countries featured in these studies switched from low quality accounting standards to higher quality IFRS (Ashbaugh and Pincus, 2001; Botosan, 1997; Easley and O Hara, 2004; Lambert et al., 2007). Thus, these studies document improvements in information quality from the substitution of IFRS for the previous local accounting. There are other studies that argue that IFRS may not benefit developing countries due to two main reasons; lack of institutions that provide proper implementation of IFRS standards; due to lack of homogeneity in the developing world (Cairns, 1990; Briston, 1978; Samuels and Oliga, 1982). These countries are from different geographic locations and their economies are quite different from each other (Africa, Asia, Latin America, the Middle East, Oceania, and Western Europe). Most of these developing countries were under the rule of different colonial powers before their independence (Chamisa, 2000). This study is specifically focused on those countries that already had high quality accounting and had strong legal and governance institutions before the adoption of IFRS. A good example of such study is from Brochet et al. (2013). Their study finds that abnormal returns to insider purchases were reduced in the post-ifrs adoption period in U.K. markets. They attributed this capital market benefit to improved comparability than higher quality accounting under IFRS adoption. They argued that U.K. domestic standards were very similar to IFRS, and hence the capital market benefits are likely due to improved comparability better precision of across-firm information. In this IFRS study, Canada is moving from a high quality local Canadian GAAP which was already comparable with the U.S. GAAP to another high quality IFRS. Hence, in the Canadian 61

70 mandatory IFRS adoption experience, this study expects net liquidity benefits of IFRS to come from the comparability advantage of IFRS. This study is similar to Daske et al. (2008) as it investigates into the impact of IFRS on liquidity of capital markets in IFRS adopting countries. Daske et al. (2008) includes sample from many IFRS adopting countries and their control group was non- IFRS adopting countries. However, this study is different as it is particularly interested in the impact of IFRS on liquidity for local and international firms in a single country, Canada. Capital market benefits for international firms who list in foreign countries is one of the expected benefit of adoption of IFRS (Healy and Palepu, 2001). This particular IFRS benefit from the aspect of international listers has not been studied in the previous literature. Canada has very interesting experimental setting for studying IFRS where many U.S. and international firms actively list their stocks unlike any other country. Liquidity of non-u.s. international firms compared to Canadian (domestic) and U.S. firms in the post-ifrs adoption period compared to pre-ifrs adoption period has not been studied in earlier studies in the context of these two competing forces; global comparability advantages and less tailoring of information needs of local investors. Next, this study discusses the important studies that provide the logics why the study expects benefits for liquidity for domestic and international firms listed in the Canadian stock markets due to the adoption of IFRS. Bae, Tan and Welker (2008) found that there is negative relationship between the differences in accounting standards across countries and foreign analyst following. They also find a negative relationship between differences in accounting standards and analyst forecast accuracy. They concluded that differences in accounting standards impose economic costs for financial analysts and their 62

71 work in interpreting financial information in accounting reports. That means fewer analysts would follow firms from countries with accounting standards different from the standards in their home country. DeFond and Hung (2003) argued that heterogeneity in accounting choice (differences in LIFO versus FIFO inventory methods) makes it difficult to compare earnings across firms. Bae et al. (2008) found evidence that local analysts have an information advantage over foreign analysts, i.e., local analysts forecast more accurately than their foreign peers. Hope (2003a) found that financial analysts forecasts are more accurate for firms that disclose their accounting policies. Such disclosures help analysts make financial statements more comparable. Ashbaugh and Pincus (2001) found that analysts earnings forecast accuracy was negatively related to the difference between IFRS and local GAAP in the pre-ifrs adoption period. Moreover, they noticed that analysts forecasts become more accurate after firms adopted IFRS. Several other studies found consistent evidence of higher levels of analyst following and forecast accuracy, and reduced analyst s dispersion in earnings forecasts in the post-ifrs adoption period (Byard et al. 2011; Horton et al., 2012; De Franco et al, 2011). Chen et al (2015), in a recent study, investigated the link between greater comparability and better acquisition decisions. Their study found that acquiring firms make more profitable acquisition decisions higher merger announcement returns, higher acquisition synergies, and better future operating performance - when target firms financial statements were more comparable. An interesting part of their study is that the acquirers benefitted more from the comparability quality of accounting when acquirers were exposed to higher information asymmetry before the acquisition decision, operated in volatile operating environments, and knew relatively little about the target. This is 63

72 consistent with the argument made by Daske et al. (2008) that even if the accounting quality did not improve under mandatory IFRS in some countries, it is possible that the financial reporting was more useful with IFRS adoption. Daske et al. (2008) point to the comparability of financial statements that would allow investors to differentiate between lower and high quality firms, thus reducing information asymmetry and/or estimation risk. Hence the switch to IFRS offers positive externalities for other firms (Coffee, 1984; Dye, 1990; Lambert et al., 2007). The flag argument that convinced countries to adopt IFRS and reject local accounting standards was that IFRS is a single set of high quality standards that would make financial statements more comparable across countries. The relationship between liquidity (bid ask and zero return) and analyst coverage has been widely discussed in the finance literature. Foucault et al. (2013) argues that the probability of informed trading (PIN) in the financial sector during the financial crisis explains the widened bid-ask spreads on financials. Their study also clarifies the role of information asymmetry in explaining variations in market liquidity (bid-ask spreads). The role of financial analyst is to collect, produce, interpret, and disseminate information about the fundamental value of the firm. Their role reduces the PIN as company insiders lose advantages in trading compared to uninformed traders. This is why, the authors explain, when the last analyst ceases covering a firm, the expected result is a widened bid-ask spread compared to those of comparable stocks (Ellul and Panaydes, 2011). The reason is that uninformed traders start thinking that some traders (informed) may have material non-public information. Easley et al. (1998) provide evidence that the probability of informed trading is higher when a firm does not have analyst coverage, which is consistent with the findings 64

73 of Ellul and Panaydes (2011). Hence, this paper argues that more comparable financial statements under IFRS will result in reducing the PIN as more market players e.g. analyst, media, investor groups, etc. will play greater roles in reducing adverse selection costs related to information asymmetry. This in turn will decrease bid-ask spreads in the market, meaning higher liquidity. Several studies have shown that larger firms have lower probability of informed trading (PIN). Easley et al. (2002) investigated the PIN of all NYSE stocks from 1983 to 1988, and found that their PIN median estimate (0.19) did not vary much across stocks. However, informed trading is negatively correlated with the firm size and positive correlated with the volatility and the bid-ask spread. The firm size results are consistent with Hasbrouck (1988) and many other studies in the finance literature. Bhushan (1989) found that the number of analysts is positively related to firms size, and subsequent studies also confirmed this relationship (Brennan and Hughes, 1991). This is important as this study expects the impact of IFRS comparability characteristics to play a bigger role in larger firms due to the greater role that analysts and other accounting users play in the larger firms. Smaller firms have less analyst and media coverage, and hence the comparability benefit may not be material. This is also consistent with stock market regulators (like Larry Harris) explanation that that new and growth firms are often harder to value than older firms in the same industry. This uncertainty leads to higher bid-ask spreads. Several researchers investigated the relationship between firms choice of accounting standards and international investment allocation decisions. Bradshaw et al. (2004) argued that many international firms adopted U.S. GAAP to attract U.S. 65

74 institutional investors since these investors face reduced information processing costs with more familiar U.S. GAAP and perceive U.S. GAAP as of high quality. Covrig et al. (2007) found that foreign mutual funds holdings from around the world increased for firms that adopted IFRS. They argued that more familiar or high quality IFRS attracts foreign investors. Hence, it can be argued that the adoption of IFRS in Canada could be followed by increased scrutiny by non-north American international analysts and non-north American international institutional investors. Both local and international firms in Canada may experience a richer information environment, i.e. more comparable information in the case of Canadian and international investors, in the post-ifrs adoption period, meaning that that the stock will be traded more widely and become more liquid. More comparable information under IFRS helps investors price firms better. Hence this study can relate the period of adoption of IFRS with the higher levels of comparable information about the firms. This should result in adjusting prices of these firms to be based on more available public information in the post-ifrs adoption period compared to that of the pre-ifrs adoption period. Several studies provide evidence that changes in information availability can lead to price changes (Beneish and Gardner, 1995; Arbel and Strebel, 1982). They showed that wider information availability impacts the costs that investors incur to collect, analyze, and disseminate information about a stock. It is expected that investors would find it less costly to collect comparable information under IFRS regime, which would reduce information asymmetry and/or estimation risk, leading to greater liquidity in the market (Ball, 2006). Horton et al. (2012) found that consensus forecast errors for firms decreased 66

75 in the mandatory IFRS adoption period relative to forecast errors of the control group. Yip and Young (2012) examined the comparability characteristics among the mandatory IFRS adoption firms in the EU, and found that mandatory IFRS adoption improved crosscountry information comparability. Beneish and Gardner (1995) showed evidence consistent with an information cost/liquidity argument that investors demand higher returns for higher trading costs (higher bid-ask spreads) and for holding securities with higher information collection costs. Barth et al. (1999) argued that the idea of international accounting harmonization is likely to help reduce information acquisition cost for international investors as they gain expertise in interpreting financial statement that follow local accounting standards. De Franco et al. (2011) suggested, based on the results of their study, that the comparability quality in financial accounting helps lower the cost of acquiring information, and further helps analysts increase the overall quantity and quality of information about the firm. Bradshaw et al. (2009) investigated another empirical proxy in their study to measure comparability by using the commonality of accounting choices and examined the impact of comparability on analyst forecasts. They found that the changes of accounting procedures impose processing costs on financial statement users in interpreting that information - larger analyst forecast errors and increased forecast dispersion. Moreover, there is this spill over effect from global advantage of IFRS that information obtained from firms in one country following IFRS can be compared in valuing firms in other countries that follow IFRS. This advantage will further reduce estimation risk and the cost of equity (Dye, 1990). These advantages become more magnified as more countries adopt IFRS. Covrig et al. (2007) find evidence consistent with these arguments that foreign mutual fund ownership, on average, is higher among voluntary IFRS adopters as information under IFRS is familiar under IFRS. 67

76 DeFond et al. (2011) linked mandatory IFRS adoption to higher foreign mutual fund ownership investment due to higher comparability under IFRS regime. 3.3 Hypothesis Development It may be argued that all domestic, U.S. and non-u.s. international firms would gain capital market benefits due to provision of more comparable financial information in IFRS based financial statements with firms in the rest of the world after the mandatory adoption of IFRS in Canada. It would be less costly for investors to compare financial reports across countries and even within countries with regards to international firms (Daske et al., 2008; Armstrong et al., 2007; Covrig, DeFond, and Hung, 2007). The comparability of financial information for international firms would help the capital markets in pricing both domestic and international firms when both domestic and international firms report under IFRS. Further, investors would be able to differentiate low quality firms from high quality firms when comparing financial performance, leading to lower information asymmetries and/or lower estimation risk. In fact, reporting incentives of firms in different markets may change as a result of mandatory adoption of IFRS (Coffee, 1984; Dye, 1990; Lambert et al., 2007; Daske et al., 2008). This would lead investors to more informed pricing of both domestic and international firms, and estimation risk would be reduced overall in the market. Also, it can be argued that IFRS may result in greater cross-border investment, as it would make it easier for foreign investors to invest in firms in Canadian markets using more comparable financial reports under IFRS. This would 68

77 lead to higher liquidity benefits in the Canadian stock markets as firms investor bases increase, eventually improving risk sharing and lowering cost of capital (Merton, 1987). EC Regulation No. 1606/2002 indicates that financial reports under IFRS are intended to be more transparent and of higher quality especially with respect to disclosures. However, as Canadian GAAP was already of high quality, the capital market benefits may be small with regard to transparency or higher quality of financial reporting under IFRS. Instead in the Canadian case, it is primarily the comparability advantage that domestic and international firms would benefit from in the post-ifrs adoption period (Brochet et al., 2013). Mandatory adoption of IFRS in Canada also means that domestic investors would become more skilled in analyzing IFRS-based financial statements. This would lead to a reduction in potential bias by domestic investors to focus on domestic companies against international firms reporting under IFRS in the pre-ifrs adoption period. Many international firms were reporting under IFRS in the pre-ifrs adoption period in Canada. Hence this study hypothesizes that liquidity, measured by bid-ask spreads and zero-returns, would be higher in the post-ifrs adoption period for all domestic, U.S. and non-u.s. international firms in Canada (hypothesis 1). On the other hand, considering that Canada s decision to adopt IFRS meant moving away from its largest trading partner, the U.S., and also that the Canadian economy is mostly focused in the extractive industries and the majority of its international firms are from the extractive industries, adoption of IFRS may not lead to higher liquidity for domestic, U.S. and even non-u.s. international firms. In fact, Canadian securities regulators allowed the use of U.S. GAAP for U.S. firms listed in Canada even after the implementation of IFRS. Ball (2006) argued that local forces and 69

78 politics were too strong for the idea of international convergence of financial reporting standards to work as intended (Ball, Kothari, and Robin, 2000; Ball Robin, and Wu, 2003; Leuz, 2003, Ball and Shivakumar, 2005; Burgstahler, Hail, and Leuz, 2006; Daske et al., 2007; Daske et al., 2008) because there is considerable use of judgment and private information when accounting standards are applied. The inherent discretion for firms in applying IFRS standards cannot be counted on blindly. This discretion is used in relation to the prevailing reporting incentives in a country including the country s legal institutions, various market forces, and firms operating characteristics (Ball, 2006). In the Canadian experiment, IFRS may not be as well suited to the domestic Canadian economic environment as the former domestic standards. Previous literature argues that analysts may have problems forecasting earnings under the IFRS regime and this problem may be exacerbated in Canada due to its greater oil, gas and mineral based industry (oil, gas, mineral problem of new discoveries with the earnings announcement). Therefore, the question of the IFRS liquidity impact in Canada is an empirical question. Hypothesis 1: Liquidity would be higher for domestic, U.S. and non-u.s. international firms listed in the Canadian stock markets after mandatory IFRS adoption than in the pre-ifrs period. 3.4 Empirical Model Following previous literature, liquidity is measured using the daily bid-ask spread and zero-returns for each firm s stock (Daske et al., 2008). Bid-ask spread is measured at the end of each trading day as the difference between the bid and ask price divided by the 70

79 midpoint of the difference. Zero-returns is calculated as the proportion of days with zero daily stock returns out of all-potential trading days in a given month. This study divides IFRS adopters into six categories: mandatory domestic, voluntary domestic, mandatory international, voluntary international, mandatory U.S. and voluntary U.S. This study believes that each group may benefit from mandatory IFRS adoption for different reasons. Mandatory international, domestic, and U.S. adopters represent switchers that changed to IFRS from local GAAP based financial statements when the whole country mandatorily adopted IFRS. These three groups of firms would possibly benefit due to IFRS adoption in the country due to the comparability advantages of IFRS. The voluntary domestic, international, and U.S. adopters include those firms that adopted IFRS before the country adopted IFRS. These three set of firms adopted IFRS for other incentives, e.g. to raise capital. This study expects the voluntary domestic, international, and U.S. IFRS adopters to benefit from the mandatory adoption of IFRS in the country as the local investors become more familiar with using the IFRS based financial information. International voluntary IFRS adopters would also benefit from the less biased local investor community especially when both local and international firms are reporting financial results under IFRS after the country adopted IFRS. The study s first indicator variable is IFRS, which takes the value of one when a firm reports under IFRS irrespective of whether the country had adopted IFRS. The IFRS variable can capture the liquidity impact of voluntary IFRS adoption. Another binary varidable is Country_Adopt that takes a value of one for periods ending after the calendar year in which the country adopts IFRS. This variable will capture possible liquidity benefits of mandatory adoption of IFRS in a country. This study picked the calendar year 71

80 in the post-ifrs adoption period in which the majority of firms in that country adopted IFRS. This study is interested to know if the liquidity improved, on average, after the country s decision to mandatorily adopt IFRS for both local, international and U.S. firms. This study takes out the year of adoption because the transition period is not fully reflective of the IFRS mandatory adoption liquidity effects. The main constructs of interest are the local, international (non-u.s.) and U.S. switchers, meaning those domestic, international and U.S. firms that switched from local GAAP to IFRS after the mandatory adoption of IFRS reflecting the forced change. These firms reflect the main average results for the comparability advantage of mandatory adoption of IFRS. This study is interested to know if the liquidity improved, on average, after the country s decision to mandatorily adopt IFRS for both local, international (non-u.s.) and U.S. firms. Based on the six groups discussed earlier, this study introduces several variables and their interaction terms: INTER_FIRM is a dummy variable and stands for international firms (headquarter address not in the country of listing). IFRS is a dummy variable and stands for the use of IFRS before country adoption of IFRS. COUNTRY_ADOPT is an indicator variable that takes the value of 1 in calendar years after mandatory adoption of IFRS (when the majority of firms adopted IFRS). COUNTRY_ADOPT x INTER_FIRM stands for the incremental impact of country adoption on liquidity for international firms. IFRS x COUNTRY_ADOPT stands for the incremental impact of voluntary IFRS reporting on liquidity after the country adopts IFRS. IFRS x INTER_FIRM stands for incremental impact on liquidity of voluntary IFRS reporting by international firms. IFRS x COUNTRY_ADOPT x INTER_FIRM is a 72

81 three way interaction dummy variable and captures the incremental effect on liquidity for international voluntary IFRS adopters when the country adopts IFRS. INTER_FIRM_US_FIRM is a two-way interaction dummy variable and captures the incremental effect on liquidity for U.S. firms. COUNTRY_ADOPT x INTER_FIRM x US_FIRM is a three way interaction dummy variable and captures the incremental effect for U.S. firms after the country adopts IFRS. IFRS x INTER_FIRM x US_FIRM is a three way interaction dummy variable and captures the incremental effect for U.S. firms after the U.S. firms voluntarily adopt IFRS. COUNTRY_ADOPT x IFRS x INTER_FIRM x US_FIRM is a four way interaction dummy variable and captures the incremental effect on liquidity when the country adopts IFRS for U.S. firms that were already IFRS voluntary adopters. CROSSLIST is an indicator variable equal to 1 if a firm is cross-listed on a major U.S. stock exchange. LN_SD_RETURN is log of standard deviation of the returns in the previous month. LN_MEDIAN_ST_OVER is the log of the median of the share turnover in the previous month. LN_MKT_VALUE is the log of the market value of the firm s stock in the previous month. This study introduces the control variables based on previous studies (Daske et al., 2008) - firm size, share turnover and return variability. Finally, this study controls for industry fixed effects and firm fixed effects. (MEDIAN_SPREAD)it = β0 + β1(ifrs)it + β2(country_adopt)it + β3(ifrs x COUNTRY_ADOPT)it + β4(international_firm)it + β5(ifrs x INTERNATIONAL_FIRM)it + β6(country_adopt x INTERNATION_FIRM)it + β7(ifrs x COUNTRY_ADOPT x INTERNATIONAL_FIRM)it + β8(us_firm x INTERNATIONAL_FIRM)it + β9(us_firm x IFRS x INTERNATIONAL_FIRM)it + 73

82 β10(us_firm x INTERNATIONAL_FIRM x COUNTRY_ADOPT)it + β11(us_firm x IFRS x INTERNATIONAL_FIRM x COUNTRY_ADOPT)it + β12crosslistit + β13ln_sd_returnit + β14ln_median_st_overit + β15ln_mkt_valueit + εit (Equation 1) (ZERO_RETURN)it = β0 + β1(ifrs)it + β2(country_adopt)it + β3(ifrs x COUNTRY_ADOPT)it + β4(international_firm)it + β5(ifrs x INTERNATIONAL_FIRM)it + β6(country_adopt x INTERNATION_FIRM)it + β7(ifrs x COUNTRY_ADOPT x INTERNATIONAL_FIRM)it + β8(us_firm x INTERNATIONAL_FIRM)it + β9(us_firm x IFRS x INTERNATIONAL_FIRM)it + β10(us_firm x INTERNATIONAL_FIRM x COUNTRY_ADOPT)it + β11(us_firm x IFRS x INTERNATIONAL_FIRM x COUNTRY_ADOPT)it + β12crosslistit + β13ln_sd_returnit + β14ln_median_st_overit + β15ln_mkt_valueit + εit (Equation 2) This study obtained all the financial data including the bid-ask and price data from Datastream and other data identifying the IFRS adoption firms from Capital IQ. Note: MEDIAN_SPREAD is monthly median bid-ask spread and is measured using prices at the end of each trading day where spread is the difference between the bid and ask price divided by the midpoint of the difference. INTER_FIRM is a dummy variable and stands for international firms (headquarter address not in the country of listing). IFRS is a dummy variable and stands for the use of IFRS either before or after country adoption of IFRS. COUNTRY_ADOPT is an indicator variable that takes the value of 1 in periods after mandatory adoption of IFRS (when the majority of firms 74

83 adopted IFRS). COUNTRY_ADOPT x INTER_FIRM stands for the incremental impact of country adoption on liquidity for international firms. IFRS x COUNTRY_ADOPT stands for the incremental impact of IFRS reporting on liquidity after the country adopts IFRS. IFRS x INTER_FIRM stands for incremental impact on liquidity of IFRS reporting by international firms. IFRS x COUNTRY_ADOPT x INTER_FIRM is a three way interaction dummy variable and captures the incremental effect on liquidity for international voluntary IFRS adopters when the country adopts IFRS. INTER_FIRM_US_FIRM is a two-way interaction dummy variable and captures the incremental effect on liquidity for U.S. firms. COUNTRY_ADOPT x INTER_FIRM x US_FIRM is a three way interaction dummy variable and captures the incremental effect for U.S. firms when the country adopts IFRS. IFRS x INTER_FIRM x US_FIRM is a three way interaction dummy variable and captures the incremental effect for U.S. firms when the U.S. firms voluntarily adopt IFRS. COUNTRY_ADOPT x IFRS x INTER_FIRM x US_FIRM is a four way interaction dummy variable and captures the incremental effect on liquidity when the country adopts IFRS for U.S. firms that were voluntary adopters. CROSSLIST is an indicator variable equal to 1 if a firm is cross-listed on a U.S. stock exchange. LN_SD_RETURN is log of standard deviation of the returns in the previous month. LN_MEDIAN_ST_OVER is the log of the median of the share turnover in the previous month. LN_MKT_VALUE is the log of the market value of the firm s stock in the previous month. 3.5 Sample Description This study strictly followed the Capital IQ database when selecting the years when that country adopted IFRS. It selected the post-ifrs and pre-ifrs years based on the 75

84 criteria that the majority of firms in that country declared IFRS as their official adopted accounting standards. Canada adopted IFRS in But this study excluded calendar year 2011, as many non-december 31 firms did not adopt IFRS till It included calendar years 2012 and 2013 for the post-ifrs period for the Canadian sample as the majority of the firms in Canada were reporting under IFRS in those years. For the pre- IFRS period, it included calendar years 2009 and Similarly, for Australia, this study included calendar years 2006 and 2007 as the pre-ifrs adoption period and 2009 and 2010 as the post-ifrs adoption period was the IFRS transition year for Australian firms and this study excluded that year. For the U.K. s sample, we included calendar years 2003 and 2004 as the pre-ifrs adoption period and calendar years 2006 and 2007 as the post-ifrs adoption period. 3.6 Descriptive Statistics Table 3-1 provides the overall percentage of international firms in the Canadian, Australian and U.K s stock markets. The reader can notice from table 3-2 that, if you include the U.S. firms, Canada has the highest percentage of international firms listed in her stock markets. However, the picture changes significantly when U.S. firms are taken out. These two tables also show that roughly 50% of the international firms in Canada are U.S. firms. This situation is unique to Canada and shows that U.S. firms have significant and close relations in the Canadian capital markets. Canada decided to adopt IFRS to obtain capital market benefits from the non-u.s. international world while Canada had important ongoing significant economic relations with the U.S. We also notice from table 3-2 that the U.K. has about 9% international firms listed in her capital markets. It is also 76

85 noticeable from TSX and TSXV documents in the How to List pamphlet that many new and growth oriented international firms raise capital in Canadian stocks markets. Table 3-5 shows that Canada s stock exchanges have about 50% of their listed firms in oil, gas, and mineral sectors and about 16% in the financial sectors. It is also clear from table 3-5 that Australia has about 49% firms in her stock exchanges from the oil, gas, and mineral sector. Hence, both Canada and Australia are naturally rich in these extractive resources. It is noticeable that U.K. industry composition is quite diverse except for the strong presence of financial sector firms in its stock exchanges. It is well known that U.K. banks have a strong presence all over the world e.g. HSBC, Standard Chartered etc. It also shows in that within Canada, the oil, gas, and mineral industry has stronger presence (65%) in the TSXV exchange as compared to (41%) the TSX exchange. Most of the oil, gas, and mineral firms in the TSXV exchange are growth oriented and smaller in size while there are more established firms from this sector in the TSX exchange. Similarly, the financial sector is more prevalent in the TSX exchange. In table 3-3, panel A, it is clear that mean bid ask spreads for firms in the Canadian stock exchanges significantly increased (0.040) in the post-ifrs adoption period (p-value of less than 1% level) compared to the pre-ifrs adoption value (0.032). This means that the liquidity decreased, on average, for firms in the Canadian stock exchanges in the post IFRS adoption period. The median bid ask spreads, however, decreased for firms in the Canadian stock exchanges meaning that liquidity increased. We have conflicting result from the perspective of mean and median bid ask spreads for all firms listed on the Canadian stock exchange. Similarly, in table 3-3 panel A, it is clear that mean and median bid ask spreads for firms in the Australian stock exchange 77

86 increased significantly (p-value of less than 1% level) in the post-ifrs adoption period (0.036 and 0.023) compared to the pre-ifrs adoption values (0.022 and 0.016). This means that the liquidity decreased for firms in the Australian stock exchanges in the post IFRS adoption period. In table 3-3 panel A, it is clear that mean and median bid ask spreads for firms in the U.K. s stock exchange significantly (p-value of less than 1% level) decreased, on average, in the post-ifrs adoption period (0.022 and 0.013) compared to the pre-ifrs adoption period (0.031 and 0.018). This means that the liquidity increased for firms in the U.K. s stock exchanges in the post IFRS adoption period. In table 3-3, panel A, it is clear that mean and median zero returns for firms in the Canadian stock exchanges significantly (p-value of less than 1% level) increased (0.216 and 0.105) in the post-ifrs adoption period compared to the pre-ifrs adoption values (0.178 and 0.100). This means that liquidity decreased, on average, for firms in the Canadian stock exchanges in the post IFRS adoption period. Similarly, in table 3-3, panel A, it is clear that mean and median zero returns for firms in the Australian stock exchange significantly (p-value of less than 1% level) increased in the post-ifrs adoption period (0.224 and 0.190) compared to the pre-ifrs adoption values (0.196 and 0.167). This means that the liquidity decreased for firms in the Australian stock exchanges in the post IFRS adoption period. In table 3-3, panel A, it is clear that mean and median zero returns for firms in the U.K. s stock exchange significantly (p-value of less than 1% level) decreased in the post-ifrs adoption period (0.148 and 0.050) compared to the pre-ifrs adoption values (0.304 and 0.250). This means that liquidity increased for firms in the U.K. s stock exchanges in the post IFRS adoption period. 78

87 In table 3-3, panel B, mean bid ask spreads for international firms in the Canadian stock exchange increased (0.024) but not significantly in the post-ifrs adoption period compared to the pre-ifrs adoption value (0.023). Similarly, in table 3-3, panel B, mean and median bid ask spreads for international firms in the Australian stock exchange significantly increased (p-value of less than 1% level for mean and p-value of less than 5% level for median) in the post-ifrs adoption period (0.037 and 0.026) compared to the pre-ifrs adoption values (0.029 and 0.025). This means that the liquidity decreased, on average, for international firms in the Australian stock exchanges in the post IFRS adoption period. In table 3-3, panel B, it is clear that mean and median bid ask spreads for international firms in the U.K. s stock exchange significantly (p-value of less than 1% level) decreased, on average, in the post-ifrs adoption period (0.024 and 0.017) compared to the pre-ifrs adoption period (0.039 and 0.025). This means that liquidity increased for international firms in the U.K. s stock exchanges in the post IFRS adoption period. In table 3-3, panel B, the zero returns tables for international firms show consistent results for firms in all three countries, Canada, Australia and U.K. It shows that the international firms liquidity results are consistent for both bid ask spreads and zero returns. Finally, we calculate descriptive statistics for U.S. firms in table 3-3, panel C. These tables largely show that liquidity decreased, for both bid ask spreads and zero returns, for U.S. firms listed in the Canadian stock exchanges. This similarity of results for U.S. firms listed in Canada with all firms listed in Canada may be explained by the North American regional or local factors; close proximity and strong economic 79

88 relationship between Canada and U.S. However, this is not true for U.S. firms listed in the U.K. stock exchanges. Pearson correlations are provided in table 3-4. There is no strong correlations that may affect the regression results. Finally table 3-6 provides how the study got to the final number of observations. 3.7 Multivariate Statistics Tables 3-7, 3-8 and 3-9 provide a template how the Analysis of differences are calculated in the next regression tables. First in table 3-7, the study defined the intercepts and stopes in the regressions. Next in table 3-8, it groups the intercepts and slopes according to each category of domestic, international and U.S. firms. Then in table 3-9, the study provides the explanation for the F-tests by differentiating two groups in both pre- and post-ifrs periods. The main regression results for Canada (combined tables 3-10 & Table 3-11 last six rows under Analysis of differences heading) show that all three switchers groups domestic firms, international firms, and U.S. firms who switched from domestic local GAAP to IFRS lost liquidity (higher bid ask spreads and higher zero returns) in the post IFRS adoption period. The results are opposite to the study s hypothesis 1. The individual slopes for International firms ( for bid ask spreads and for zero returns) and U.S.*International ( for bid ask spreads and for zero returns) show that international (U.S. and non-u.s.) firms listed in Canada, in general, enjoyed better liquidity in the pre-ifrs adoption period. This may be due to international firms that afford dual-listing abroad having better reputations. On the other hand, U.S. firms listed 80

89 in Canada enjoyed better liquidity than local firms indicating preferred treatment of U.S. firms in the Canadian stock markets. The TMX group that represents both TSX and TSXV exchanges aggressively attracts U.S. firms to list in the Canadian stock markets. Further results also indicate (both at the slope level and the Analysis of differences at the bottom of tables 3-10 & 3-11) that it is the international firms (non-u.s.), who had already adopted IFRS before Canada adopted it, that gained liquidity in the post-ifrs adoption period (IFRS*International*Country adoption has coefficients for bid ask spreads and for zero returns). This is not true for U.S. firms that adopted IFRS before Canada did. These U.S. firms, on average, did not benefit from the mandatory IFRS adoption in Canada. The advantage of higher liquidity for U.S. firms in the pre- IFRS adoption period was lost in the post IFRS period. Hence, both the Canadian and U.S. firms lost liquidity but there is some evidence that the non-u.s. international firms gained liquidity in the post-ifrs adoption period. This may be good news for policy makers in Canada as Canada wanted to adopt IFRS to be more attractive place for non- U.S. international firms. In table 3-12, the study took out the oil, gas and mineral firms from Canada s full sample and reran the same regression. Again this study found generally consistent results with the main regression that the liquidity (both higher zero returns and higher bid ask spreads) significantly decreased for switchers in the post regression period for domestic, non-us international firms, and U.S. firms. Most non-u.s. international firms in Canada are from the oil, gas, and minerals industries (extractive industries) and thus the benefits of comparability were gained by the non-u.s. international firms in these industries only due to a large presence of these industries in Canada (DeFond et al, 2011). In table 3-14, 81

90 the study took out the financial firms from the Canada s full sample and reran the regression. Again it found consistent results with the main regression that the liquidity (both higher zero returns and higher bid ask spreads) significantly decreased in the post IFRS adoption period for switchers among both domestic, non-u.s. international, and U.S. firms. Again we do not find better liquidity results for non-u.s. international firms while most non-u.s. international firms are from the oil, gas, and mineral industries. In the final regression analysis (table 3-16) for Canada, the study took out both the dominating sectors oil, gas, and mineral firms and financial firms and again found that switchers from the domestic, non-u.s. international and U.S. firms significantly lost liquidity (both higher zero returns and higher bid ask spreads) in the post IFRS period. In general, these results are surprising because Canada switched from Canadian GAAP to IFRS considering that IFRS would be more comparable for firms in the international marketplace and hence that should mean better liquidity benefits for all groups including domestic firms and U.S. firms. Again, there is some evidence that liquidity increased for non-u.s. firms but only in the oil, gas and mineral industries. Previous studies have shown that the comparability advantage works better under IFRS when there is a large number of firms from a specific industry that join IFRS as a result of adoption of IFRS. 3.8 Australia and United Kingdom Liquidity tests Finally this study does the same tests for all domestic, non-u.s. international, U.S. firms in Australia and U.K. s stock markets. Canada, Australia and U.K. share the British Commonwealth background and similar legal, political, and governance institutions. Further, this study picked Australian stock markets because it is clear from table 3-5 that 82

91 Australia has similar industry characteristics both Canada and Australia have many (close to 50%) firms listed in their stock exchanges from the oil, gas, and mineral sectors. Hence the local factors are quite similar between Canada and Australia. However there is a comparatively smaller number of international firms listed in the Australian stock market compared to the Canadian stock markets. The international firms that list in these two countries are mostly from the oil, gas, and mineral sectors. The U.K. stock exchange is also interesting because, like Canada, the U.K. has a strong financial sector. However, there is a significantly higher number of non-u.s. international firms present in the U.K. from diverse industries (table 3-5). The regression results (table 3-10 and 3-11) for the Australian full sample show lower liquidity, for both zero returns and bid ask spreads, for local and non-u.s. international switchers groups in Australia. There are very few U.S. firms that list in the Australian stock exchange. Hence we omit the coefficients for U.S. firms and their interactions in the table for Australia s regression results. In further analysis (table 3-13) the study took out the dominating sectors from the Australian sample oil, gas, and mineral firms and again found that switchers from the domestic and international firms largely lost liquidity (higher zero returns and higher bid ask spreads) in the post IFRS period. The regression results (table 3-10 and 3-11) for the U.K. full sample show significantly lower zero returns and bid ask spreads for both non-u.s. international and domestic firms. U.S. firms listed in U.K. stock exchanges were either indifferent or benefitted in terms of liquidity in these exchanges. In further analysis (table 3-15), the study took out the dominating sector financial firms from U.K s full sample and again 83

92 found that switchers both international, U.S. and domestic firms largely gained liquidity (lower zero returns and lower bid ask spreads) in the post IFRS period. This is interesting as U.K. gained liquidity for its non-u.s. international and domestic firms while Canada and Australia lost liquidity for domestic firms who switched to IFRS after the adoption of IFRS in Canada. The U.K. s stock markets have greater international firms presence even before the adoption of IFRS and its industry profile is more diversified compared to Canada s and Australia s stock markets. 3.9 Size tests This study tests liquidity for firm size under the assumption that larger firms are more international than smaller firms. It divides the firms into quintiles separately for each country and runs regressions for each quintile for both zero-returns and bid-ask spreads. The results support the overall story that larger firms in quintiles 4 and 5 (table 3-17 and table 3-18) have higher liquidity than smaller firms (quintiles 1 and 2) in the post IFRS adoption period. The results also support the international nature of the U.K. stock markets where liquidity increased for both smaller and larger firms. The study then calculated the number of geographic segments for each quintile for each country (table 3-19), and again finds support for the overall results that even smaller U.K. firms have higher number of geographic segments (truly international) while the study finds lower numbers of geographic segments for smaller Australian and Canadian firms confirming the local nature of these firms. 84

93 3.10 Summary & Future Research Direction This study investigates the impact of IFRS on liquidity in a unique setting in Canada. The study documents that both local and U.S. international firms who switched to IFRS from local GAAP in Canada did not gain liquidity. However, there is some evidence that non-u.s. international firms gained liquidity in the Canadian stock markets. This is especially true for those non-u.s. international firms that had adopted IFRS in the pre-ifrs adoption period. The study then compares these three groups of firms to similar groups of firms in the Australian and U.K. stock exchanges when IFRS was adopted in those two countries. And interesting finding of this study is that the group of U.S. firms that adopted IFRS in Canada lost liquidity advantage in the post-ifrs period but that is not true for U.S. firms in the U.K. exchange in the post-ifrs adoption period. The study concludes that the reason that we see that liquidity is lost for both Canadian domestic and U.S. firms listed in the Canadian stock markets is the regional closeness between U.S. and Canada. Local Canadian and U.S. GAAPs were more suitable to convey information about the local business conditions to the local investors and IFRS, at least in the first two years, did not benefitted these two groups of firms. Based on the results of IFRS adoption in the U.K. stock markets, the study concludes that diversified industry presence and more international firms in the stock markets are important factors to capture the advantages of IFRS. The study s Canadian liquidity results may also be explained by the general shift in the prices of commodities or other macro-economic factors. It is clear from the study s analysis that Australian economy is very similar to the Canadian economy especially because roughly 50% of the economies in both countries are from extractive industries 85

94 i.e. oil, gas, and mineral firms. Future research work can do more to control for these macro-level factors. One way may be to test liquidity in Australian firms during the same time frame as when Canada adopted IFRS ( as pre-ifrs adoption period and as post-ifrs adoption period) and similarly test liquidity in Canadian firms during the same time frame as when Australia adopted IFRS ( as pre-ifrs adoption period and as post-ifrs adoption period). Future research can investigate if the lower liquidity for Australian and Canadian domestic firms in the post-ifrs adoption period were at harmful levels for local economies. Similarly, future research can investigate if the gain in liquidity for non-u.s. international firms was significant enough advantage for Canadian economy that Canada opted to adopt IFRS at the cost of rejecting well-functioning local Canadian GAAPs. 86

95 Bibliography Aboody, D., M. E. Barth, and R. Kasznik Revaluations of fixed assets and future firm performance: Evidence from the U. K. Journal of Accounting and Economics. 26: Ahmed, A. S., M. Neel, and D. Wang Does mandatory adoption of IFRS improve accounting quality? Preliminary evidence. Contemporary Accounting Research. 30(4): Ahmed, A. S., R. A. Schneible, and D. E. Stevens An empirical analysis of the effects of online trading on stock price and trading volume reactions to earnings announcements. Contemporary Accounting Research 20(3): Amiram, D Financial information globalization and foreign investment decisions. Journal of International Accounting Research. 11(2): Arbel, A., and P. Strebel The neglected and small firm effects. The Financial Review. 17(4): Armstrong, C., M. Barth, A. Jagolinzer, and E. Riedl Market reaction to the adoption of IFRS in Europe. The Accounting Review. 85(1): Ashbaugh, H., and M. Pincus Domestic accounting standards, international accounting standards, and the predictability of earnings. Journal of Accounting Research. 39(3): Ausseneg, W., P. Inwinkl, and G. Schneider Earnings management and local vs International accounting standards of European public firms. Working Paper, Vienna University of Technology. Bae, K. H., Tan, H., & Welker, M. (2008). International GAAP differences: The impact on foreign analysts. The Accounting Review, 83(3), Ball, R. and P. Brown An empirical evaluation of accounting income numbers. Journal of Accounting Research Ball, R., A. Robin, and J. S. Wu Incentives versus standards: properties of accounting income in four East Asian countries. Journal of accounting and economics. 36(1): Ball, R., S. Kothari, and A. Robin The effect of international institutional factors on properties of accounting earnings. Journal of Accounting and Economics. 29(1):

96 Ball, R International Financial Reporting Standards (IFRS): pros and cons for investors. Accounting and Business Research. 36(sup1): Ball, R. and L. Shivakumar Earnings quality in UK private firms: comparative loss recognition timeliness. Journal of Accounting and Economics. 39(1): Bamber, L., O. Barron, and D. Stevens Trading volume around earnings announcements and other financial reports: Theory, research design, empirical evidence, and directions for future research. Contemporary Accounting Research. 28(2): Barron, O., R. Schneible, and D. Stevens What do we really Know about the Firm size Effect on Trading Volume Reactions to Earnings Announcement? A Re-Examination and Extension. Working paper. Barth, M., W. Landsman, and M. Lang International accounting standards and accounting quality. Journal of Accounting Research, 46(3), Barth, M., W. Landsman, M. Lang, and C. Williams Are IFRS-based and US GAAP-based accounting amounts comparable? Journal of Accounting and Economics. 54(1): Basu, S The conservatism principle and the asymmetric timeliness of earnings. Journal of Accounting and Economics. 24(1): Beaver, W Alternative accounting measures as predictors of failure. The Accounting Review. 43(1): Beaver, W. and S. Ryan Conditional and unconditional conservatism: concepts and modeling. Review of Accounting Studies. 10(2-3): Beneish, M. D., and J. C. Gardner Information costs and liquidity effects from changes in the Dow Jones Industrial Average list. Journal of Financial and Quantitative Analysis. 30(01): Beuselinck, C., P. Joos, I. Khurana, and S. van der Meulen Mandatory IFRS Reporting and Stock Price Informativeness. CentER Discussion Paper; Vol Tilburg University. Bhushan, R. Firm characteristics and analyst following. Journal of Accounting and Economics. 11(2): Biddle, G, G. Hilary, and R. Verdi How does financial reporting quality relate to investment efficiency? Journal of Accounting and Economics. 48(2): Blanchette, M., F.-E. Racicot, and J.-Y. Girard The effects of IFRS on financial ratios: Early evidence in Canada. Certified General Accountants Association of Canada. 88

97 Botosan, C. A Disclosure Level and the Cost of Equity Capital. The Accounting Review. 72(3): Brennan, M. and P. Hughes Stock prices and the supply of information. The Journal of Finance. 46(5): Briston, R.J The Evolution of Accounting in Developing Countries. International Journal of Accounting: Education and Research. Fall: Brochet, F., A. D. Jagolinzer and E. J. Riedl, Mandatory IFRS adoption and financial statement comparability. Contemporary Accounting Research. 30(4): Byard, D., Y. Li, and Y. Yu The effect of mandatory IFRS adoption on financial analysts information environment. Journal of Accounting Research. 49(1): Cairns, D Aid for the Developing World. Accountancy. March: Chamisa, E.E The Relevance and Observance of the IASC Standards in Developing Countries: Case Studies of Botswana and Zimbabwe. Unpublished Doctoral Thesis, University of Birmingham, England. Chen, C. W., D. W. Collins, T. D. Kravet, and R. Mergenthaler Financial statement comparability and the efficiency of acquisition decisions. Available at SSRN Coffee Jr, J. C Market failure and the economic case for a mandatory disclosure system. Virginia Law Review Covrig, V., M. Defond, and M. Hung Home bias, foreign mutual fund holdings, and the voluntary adoption of international accounting standards. Journal of Accounting Research. 45(1): Danbolt, J., and W. Rees An experiment in fair value accounting: UK investment vehicles. European Accounting Review. 17(2): Daniel, K., D. Hirshleifer, and A. Subrahmanyam Investor psychology and security market under-and overreactions. The Journal of Finance, 53(6), Daske, H., L. Hail, C. Leuz, and R. Verdi Mandatory IFRS reporting around the world: Early evidence on the economic consequences. Journal of Accounting Research. 46: DeFond, M. L., and M. Hung An empirical analysis of analysts cash flow forecasts. Journal of Accounting and Economics. 35(1):

98 DeFond, M., M. Hung, and R. Trezevant Investor protection and the information content of annual earnings announcements: International evidence. Journal of Accounting and Economics. 43(1): DeFond, M., X. Hu, M. Hung, and S. Li The impact of mandatory IFRS adoption on foreign mutual fund ownership: The role of comparability. Journal of Accounting and Economics, 51(3), De Franco, G., et al The benefits of financial statement comparability. Journal of accounting research. 49(4): Doukakis, L The persistence of earnings and earnings components after the adoption of IFRS. Managerial Finance. 36: Dye, R. A Mandatory versus voluntary disclosures: The cases of financial and real externalities. Accounting Review. 1990: Dye, R. and S. Sunder Why not allow FASB and IASB standards to compete in the US? Accounting Horizons. 15(3): Easley, D., S. Hvidkjaer, and M. O hara Is information risk a determinant of asset returns? The journal of finance. 57(5): Easley, D., and M. O hara Information and the cost of capital. The journal of finance. 59(4): Easley, D., M. O'hara, and P. S. Srinivas Option volume and stock prices: Evidence on where informed traders trade. The Journal of Finance. 53,(2): Fields, T., T. Lys, and L. Vincent Empirical research on accounting choice. Journal of Accounting and Economics. 31(1): Foucault, T., O. Kadan, and E. Kandel Liquidity cycles and make/take fees in electronic markets. The Journal of Finance. 68(1): Francis, J., S. Huang, I. Khurana, and R. Pereira Does corporate transparency contribute to efficient resource allocation? Journal of Accounting Research. 47(4): Harris, Larry Trading and exchanges: Market microstructure for practitioners. Oxford University Press Hasbrouck, J Trades, quotes, inventories, and information. Journal of Financial Economics. 22(2): Hayn, C The information content of losses. Journal of accounting and economics, 20(2),

99 Healy, P., and K. Palepu Information asymmetry, corporate disclosure, and the capital markets: A review of the empirical disclosure literature. Journal of accounting and economics. 31(1): Healy, P., A. Hutton, and K. Palepu Stock performance and intermediation changes surrounding sustained increases in disclosure. Contemporary Accounting Research, 16(3), Hope, O. K Analyst following and the influence of disclosure components, IPOs and ownership concentration. Asia-Pacific Journal of Accounting & Economics. 10(2): Hope, O., W. Thomas, and G. Winterbotham, G Geographic earnings disclosure and trading volume. Journal of Accounting and Public Policy, 28(3), Horton, J., G. Serafeim, and I. Serafeim Does Mandatory IFRS Adoption Improve the Information Environment? Contemporary Accounting Research. 30(1): Jeanjean, T., and H. Stolowy Do accounting standards matter? An exploratory analysis of earnings management before and after IFRS adoption. Journal of Accounting and Public Policy. 27(6): Kabir, M., F. Laswad, and M. Islam Impact of IFRS in New Zealand on accounts and earnings quality. Australian Accounting Review. 20(4): Kim, O. and R. Verrecchia Market reaction to anticipated announcements. Journal of Financial Economics. 30(2): Kim, O. and R. Verrecchia Trading volume and price reactions to public announcements. Journal of Accounting Research. 29(2): Kim, O. and R. Verrecchia Pre-announcement and event-period private information. Journal of Accounting and Economics. 24(3): Kothari, S Capital markets research in accounting. Journal of Accounting and Economics. 31(1): Lambert, R., C. Leuz, and R. Verrecchia Accounting information, disclosure, and the cost of capital. Journal of Accounting Research. 45(2): Landsman, W., E. Maydew, and J. Thornock The information content of annual earnings announcements and mandatory adoption of IFRS. Journal of Accounting and Economics. 53(1): Leuz, C, and P. Wysocki Economic consequences of financial reporting and disclosure regulation: A review and suggestions for future research. Working paper. Available at: abstract=

100 Leuz, C., and R. Verrecchia The Economic Consequences of Increased Disclosure (Digest Summary). Journal of Accounting Research, 38, Leuz, C IAS versus US GAAP: information asymmetry based evidence from Germany's new market. Journal of Accounting Research. 41(3): Lev, B On the usefulness of earnings and earnings research: lessons and directions from two decades of empirical research. Journal of Accounting Research. 40: Li, S Does mandatory adoption of International Financial Reporting Standards in the European Union reduce the cost of equity capital? The Accounting Review. 85(2): Merton, R. C A simple model of capital market equilibrium with incomplete information. The journal of Finance. 42(3): O'Brien, P Changing the Concepts to Justify the Standards. Accounting Perspectives. 8(4): O'Brien, P., and W. Scott Contract theory: The effectiveness of a conceptual framework for financial reporting has been debated for years. CA Magazine, April 1. Pandes, J. A. and M. J. Robinson The Canadian junior IPO market and the capital pool company program. Handbook of Research on IPOs: Panayides, M. A., and A. Ellul Do Financial Analysts Restrain Insiders' Informational Advantage? Working Paper: Available at PriceWaterhouseCoopers Putting IFRS in motion: Key accounting issues for pharmaceuticals and life sciences companies. Samuels, J.M. and J.C. Oliga Accounting Standards in Developing Countries. International Journal of Accounting Education and Research. Fall: Scott, W. R Financial accounting theory. Pearson Education Canada. Sharpe, I. and R. Walker Asset revaluation and stock market prices. Journal of Accounting Research. 13: Standish, P, and S.-I. Ung Corporate signaling, asset revaluation and stock prices of British companies. The Accounting Review. 57(4): Tan, H., Wang, S., and M. Welker Analyst following and forecast accuracy after mandated IFRS adoptions. Journal of Accounting Research, 49(5),

101 Verrecchia, R Essays on disclosure. Journal of Accounting and Economics. 32(1-3): Watts, R Conservatism in accounting part I: Explanations and implications. Accounting Horizons. 17(3): Watts, R Conservatism in accounting part II: Evidence and research opportunities. Accounting Horizons 17(4): Welker, M Disclosure policy, information asymmetry and liquidity in equity markets. Contemporary Accounting Research. 11: Yip, R. W. and D. Young Does mandatory IFRS adoption improve information comparability? The Accounting Review. 87(5): Yu, G Accounting standards and international portfolio holdings: Analysis of cross-border holdings following mandatory adoption of IFRS. Dissertation. University of Michigan. 93

102 Table 3-1 International listings in Canada by country/region Country/Region All Canada TSX TSXV U.S. 51% 47% 51% Asia 13% 8% 25% Australia/NZ/PNG 11% 18% 5% U.K./Europe 9% 12% 8% Latin America 8% 5% 6% Africa 3% 5% 1% Other 5% 5% 4% Total 100% 100% 100% Table 3-2 Percentage of U.S. listings by country Number of Number of Percentage of international firms international firms international firms including U.S. excl. U.S. firms excl. U.S. firms firms Percentage of international firms including U.S. firms Total number of firms in each country Canada 180 5% % 3486 Australia 40 2% 40 2% 2000 U.K % 243 9%

103 Table 3-3 Descriptive Statistics for bid ask spreads, zero returns and size Prior to IFRS Adoption Post IFRS Adoption variables n Mean St. Dev. Median n Mean St. Dev. Median Dif. in mean Panel A All firms Bid Ask spreads Canada *** *** Australia *** *** U.K *** *** Zero Returns Canada *** *** Australia *** *** U.K *** *** Size Canada *** *** Australia *** *** U.K *** *** Panel B International firms Bid Ask spreads Canada *** Australia *** ** U.K *** *** Zero Returns Canada *** *** Australia U.K *** *** Size Canada Australia U.K *** *** Panel C U.S. firms Bid Ask spreads Canada ** *** Australia *** *** U.K *** *** Zero Returns Canada *** *** Australia *** *** U.K *** *** Size Canada Australia *** *** U.K *** *** Dif. in median 95

104 Table 3-4 Pearson Correlation tables Bid Ask Bid Ask Zero IFRS Inter U.S. Country Crosslist SD Share Size Zero Return IFRS Inter firm U.S. firm Country Adopt Crosslist SD Share turnover Size Table 3-5 Industry Composition by Country Domestic International Canad Australi United Canad Australi United Oil, Gas and Minerals 49% 49% 5% 61% 52% 14% Non-Durables 2% 3% 5% 3% 14% 2% Durables 0% 1% 1% 0% 0% 0% Manufacturing 4% 5% 9% 4% 0% 2% Chemicals 8% 1% 1% 7% 0% 0% Industry 6% 6% 9% 7% 10% 7% Telecom 1% 2% 2% 6% 0% 2% Utilities 1% 1% 2% 0% 0% 2% Retail 3% 4% 7% 0% 0% 3% Health 3% 8% 3% 6% 14% 3% Finance 16% 13% 40% 3% 0% 44% Other 7% 7% 16% 3% 10% 21% Total 100% 100% 100% 100% 100% 100% 96

105 Table 3-6 Sample Selection Sample selection Firms N Total number of observations for 5 years Drop Australia transition year Drop Canada transition year Drop U.K transition year firms Drop if missing values of share turnover, SD return Drop if SIC missing or zero Drop firms that had less than 46 months of observation Final sample

106 Table 3-7 Defining the intercept and slopes Intercept = β0 IFRS = β1 Country adoption = β2 IFRS*Country adoption = β3 International = β4 IFRS*International = β5 International*Country adoption = β6 IFRS*International*Country adoption = β7 U.S.*International = β8 U.S.*IFRS*International = β9 U.S.*International*Country adoption = β10 U.S.*IFRS*International*Country adoption = β11 Table 3-8 Grouping of interpretation Domestic firms not using IFRS before adoption = β0 Domestic firms using IFRS before adoption = β0+β1 Domestic firms not using IFRS after adoption = β0+β2 Domestic firms using IFRS after adoption = β0+β1+β2+β3 International firms not using IFRS before adoption = β0+β4 International firms using IFRS before adoption = β0+β4+β1+β5 International firms not using IFRS after adoption = β0+β4+β2+β6 International firms using IFRS after adoption = β0+β4+β1+β5+β2+β6+β7 U.S. firms not using IFRS before adoption = β0+β4+β8 U.S. firms using IFRS before adoption = β0+β4+β8+β1+β5+β9 U.S. firms not using IFRS after adoption = β0+β4+β8+β2+β6+β10 U.S. firms using IFRS after adoption = β0+β4+β8+β1+β5+β9+β2+β6+β10+β11 98

107 Table 3-9 Analysis of differences tests Domestic using IFRS after adoption vs. Domestic using IFRS before adoption = β0+β1+β2+β3-β0-β1 =β2+β3=0 (Test 1) Domestic using IFRS after adoption vs. Domestic not using IFRS before adoption = β0+β1+β2+β3-β0 =β1+β2+β3=0 (Test 2) International using IFRS after adoption vs. International using IFRS before adoption = β0+β4+β1+β5+β2+β6+β7- β0-β4-β1-β5 =β2+β6+β7=0 (Test 3) International using IFRS after adoption vs. International not using IFRS before adoption = β0+β4+β1+β5+β2+β6+β7-β0-β4=β1+β5+β2+β6+β7=0 (Test 4) U.S. using IFRS after adoption vs. U.S. using IFRS before adoption = β0+β4+β8+β1+β5+β9+β2+β6+β10+β11-β0- β4-β8-β1-β5-β9 =β2+β6+β10+β11=0 (Test 5) U.S. using IFRS after adoption vs. U.S. not using IFRS before adoption = β0+β4+β8+β1+β5+β9+β2+β6+β10+β11-β0-β4-β8=β1+β5+β9+β2+β6+β10+β11=0 (Test 6) 99

108 Table 3-10 Firm level regressions of bid-ask spreads for full sample with industry fixed effects and firm fixed effects Canada Australia U.K. Intercept β *** 0.098*** 0.108*** IFRS β *** ** *** Country adoption β *** *** IFRS * Country adoption β *** International β *** IFRS * International β ** 0.005** International * Country adoption β ** *** IFRS * International* Country adoption β *** U.S. * International β *** 0.032*** U.S. * IFRS * International β * U.S. * International * Country adoption β *** U.S. * IFRS * International * Country adoption β Crosslist *** 0.039*** LN_SD_Return 0.019*** 0.010*** 0.010*** LN_MEDIAN_ST_OVER *** *** *** LN_MKT_VALUE *** *** *** Fixed effects Industry Firm Industry Firm Industry Firm Observations Mean adjusted R Analysis of differences F-tests Domestic using IFRS after adoption vs. domestic using IFRS before adoption Domestic using IFRS after adoption vs. domestic not using IFRS before adoption International using IFRS after adoption vs. international using IFRS before adoption International using IFRS after adoption vs. international not using IFRS before adoption U.S. using IFRS after adoption vs. U.S. using IFRS before adoption U.S. using IFRS after adoption vs. U.S. not using IFRS before adoption Test *** 0.010*** 0.002*** Test *** 0.008*** *** Test ** 0.007*** *** Test *** *** Test Test **

109 TABLE 3-11 Firm level regression of zero returns for full sample with industry fixed effects and firm fixed effects Canada Australia U.K. Intercept β *** 0.506*** 1.049*** IFRS β *** 0.015*** * Country adoption β *** *** IFRS * Country adoption β ** *** International β *** 0.153*** * IFRS * International β ** International * Country adoption β ** IFRS * International* Country adoption β *** *** U.S. * International β ** 0.029*** U.S. * IFRS * International β *** 0.078** U.S. * International * Country adoption β *** U.S. * IFRS * International * Country adoption β * 0.279*** Crosslist *** 0.103*** LN_SD_Return 0.009*** *** *** LN_MEDIAN_ST_OVER *** *** *** LN_MKT_VALUE *** *** *** Fixed effects Industry Firm Industry Firm Industry Firm Observations Mean adjusted R Analysis of differences F-tests Domestic using IFRS after adoption vs. domestic using IFRS before adoption Domestic using IFRS after adoption vs. domestic not using IFRS before adoption International using IFRS after adoption vs. international using IFRS before adoption International using IFRS after adoption vs. international not using IFRS before adoption U.S. using IFRS after adoption vs. U.S. using IFRS before adoption U.S. using IFRS after adoption vs. U.S. not using IFRS before adoption Test *** *** Test *** 0.019*** *** Test ** *** Test *** *** Test ** Test

110 TABLE 3-12 Firm level regression of bid-ask spreads and zero returns for Canadian full sample without oil, gas, and minerals industries with industry fixed effects and firm fixed effects Bid Ask Zero Returns Intercept β *** 0.559*** IFRS β *** 0.008** Country adoption β *** IFRS * Country adoption β *** International β *** *** IFRS * International β ** 0.118*** International * Country adoption β *** IFRS * International* Country adoption β7 U.S. * International β *** 0.076*** U.S. * IFRS * International β ** *** U.S. * International * Country adoption β *** * U.S. * IFRS * International * Country adoption β Crosslist *** LN_SD_Return 0.012*** LN_MEDIAN_ST_OVER *** *** LN_MKT_VALUE *** *** Fixed effects Industry Firm Industry Firm Observations Mean adjusted R Analysis of differences F-tests Domestic using IFRS after adoption vs. domestic using IFRS before adoption Domestic using IFRS after adoption vs. domestic not using IFRS before adoption International using IFRS after adoption vs. international using IFRS before adoption International using IFRS after adoption vs. international not using IFRS before adoption U.S. using IFRS after adoption vs. U.S. using IFRS before adoption U.S. using IFRS after adoption vs. U.S. not using IFRS before adoption Test * Test *** 0.014*** Test *** 0.030* Test *** 0.156*** Test Test *** 102

111 TABLE 3-13 Firm level regression of bid-ask spreads and zero returns for Australian full sample without oil, gas, and mineral industries with industry fixed effects and firm fixed effects Bid Ask Zero Returns Intercept β *** 0.519*** IFRS β ** Country adoption β *** IFRS * Country adoption β ** International β * ** IFRS * International β International * Country adoption β IFRS * International* Country adoption β7 Crosslist *** ** LN_SD_Return 0.008*** *** LN_MEDIAN_ST_OVER *** *** LN_MKT_VALUE *** *** Fixed effects Industry Firm Industry Firm Observations Mean adjusted R Analysis of differences F-tests Domestic using IFRS after adoption vs. domestic using IFRS before adoption Domestic using IFRS after adoption vs. domestic not using IFRS before adoption International using IFRS after adoption vs. international using IFRS before adoption International using IFRS after adoption vs. international not using IFRS before adoption Test *** Test *** 0.011*** Test Test *** U.S. using IFRS after adoption vs. U.S. using IFRS before adoption Test 5 U.S. using IFRS after adoption vs. U.S. not using IFRS before adoption Test 6 103

112 TABLE 3-14 Firm level regression of bid-ask spreads and zero returns for Canadian full sample without financial industry with industry fixed effects and firm fixed effects Bid Ask Zero Returns Intercept β *** 0.763*** IFRS β *** 0.028*** Country adoption β * 0.019* IFRS * Country adoption β International β *** 0.155*** IFRS * International β *** IFRS * International* Country adoption β U.S. * International β *** 0.307*** U.S. * IFRS * International β * *** U.S. * International * Country adoption β *** 0.042*** U.S. * IFRS * International * Country adoption β11 Crosslist *** *** LN_SD_Return 0.021*** 0.014*** LN_MEDIAN_ST_OVER *** *** LN_MKT_VALUE *** *** Fixed effects Industry Firm Industry Firm Observations Mean adjusted R Analysis of differences F-tests Domestic using IFRS after adoption vs. domestic using IFRS before adoption Domestic using IFRS after adoption vs. domestic not using IFRS before adoption International using IFRS after adoption vs. international using IFRS before adoption International using IFRS after adoption vs. international not using IFRS before adoption Test *** 0.011*** Test *** 0.039*** Test * Test *** U.S. using IFRS after adoption vs. U.S. using IFRS before adoption Test *** 0.029*** U.S. using IFRS after adoption vs. U.S. not using IFRS before adoption Test ***

113 TABLE 3-15 Firm level regression of bid-ask spreads and zero returns for U.K. full sample without financial industry with industry fixed effects and firm fixed effects Bid Ask Zero Returns Intercept β *** 1.297*** IFRS β ** Country adoption β *** *** IFRS * Country adoption β *** *** International β *** *** IFRS * International β *** International * Country adoption β *** IFRS * International* Country adoption β *** U.S. * International β *** U.S. * IFRS * International β ** U.S. * International * Country adoption β U.S. * IFRS * International * Country adoption β11 Crosslist 0.100*** LN_SD_Return 0.009*** *** LN_MEDIAN_ST_OVER *** *** LN_MKT_VALUE *** *** Fixed effects Industry Firm Industry Firm Observations Mean adjusted R Analysis of differences F-tests Domestic using IFRS after adoption vs. domestic using IFRS before adoption Domestic using IFRS after adoption vs. domestic not using IFRS before adoption International using IFRS after adoption vs. international using IFRS before adoption International using IFRS after adoption vs. international not using IFRS before adoption Test *** Test *** Test Test *** *** U.S. using IFRS after adoption vs. U.S. using IFRS before adoption Test U.S. using IFRS after adoption vs. U.S. not using IFRS before adoption Test * 0.208*** 105

114 TABLE 3-16 Firm level regression of bid-ask spreads and zero returns for Canadian full sample without financial, oil, gas, and mineral industries with industry fixed effects and firm fixed effects Bid Ask Zero Returns Intercept β *** 0.632*** IFRS β * 0.014*** Country adoption β IFRS * Country adoption β International β *** *** IFRS * International β ** 0.109*** International * Country adoption β * 0.031* IFRS * International* Country adoption β7 U.S. * International β *** 0.120*** U.S. * IFRS * International β *** *** U.S. * International * Country adoption β U.S. * IFRS * International * Country adoption β11 Crosslist 0.057*** *** LN_SD_Return 0.014*** 0.008*** LN_MEDIAN_ST_OVER *** *** LN_MKT_VALUE *** *** Fixed effects Industry Firm Industry Firm Observations Mean adjusted R Analysis of differences F-tests Domestic using IFRS after adoption vs. domestic using IFRS before adoption Domestic using IFRS after adoption vs. domestic not using IFRS before adoption International using IFRS after adoption vs. international using IFRS before adoption Test Test *** 0.012*** Test ** International using IFRS after adoption vs. international not using IFRS before adoption Test *** 0.152*** U.S. using IFRS after adoption vs. U.S. using IFRS before adoption Test *** 0.054*** U.S. using IFRS after adoption vs. U.S. not using IFRS before adoption Test *** 106

115 Table 3-17 Bid Ask spreads firm level quintile size tests for all three countries. 5 is largest quintile and 1 is smallest Coefficients Coefficients Coefficients Coefficients Coefficients Canada Australia U.K. Domestic using IFRS after adoption vs. Domestic using IFRS before adoption Domestic using IFRS after adoption vs. Domestic not using IFRS after adoption Domestic using IFRS after adoption vs. Domestic using IFRS before adoption Domestic using IFRS after adoption vs. Domestic not using IFRS after adoption Domestic using IFRS after adoption vs. Domestic using IFRS before adoption Domestic using IFRS after adoption vs. Domestic not using IFRS after adoption *** *** *** * ** *** 0.025*** 0.002** 0.010*** 0.007*** *** 0.005*** 0.007*** 0.006*** 0.006*** 0.011*** *** *** *** *** *** *** 107

116 Table 3-18: Zero returns firm level quintile size tests for all three countries. 5 is largest quintile and 1 is smallest Coefficients Coefficients Coefficients Coefficients Coefficients Canada Australia U.K. Domestic using IFRS after adoption vs. Domestic using IFRS before adoption Domestic using IFRS after adoption vs. Domestic not using IFRS after adoption Domestic using IFRS after adoption vs. Domestic using IFRS before adoption Domestic using IFRS after adoption vs. Domestic not using IFRS after adoption Domestic using IFRS after adoption vs. Domestic using IFRS before adoption Domestic using IFRS after adoption vs. Domestic not using IFRS after adoption *** 0.035*** 0.090*** 0.136*** *** *** 0.012*** 0.048*** 0.099*** ** 0.024*** *** 0.020*** *** *** *** 0.034*** *** *** *** *** 108

117 Table 3-19: Mean number of geographic segments for each firm by quintile for all three countries. 5 is largest quintile and 1 is smallest Coefficients Coefficients Coefficients Coefficients Coefficients Canada Australia U.K

118 Chapter 4: Has Mandatory Adoption of IFRS Increased Non-U.S. International Institutional Investment in the Canadian Stock Markets? Abstract: This exploratory study investigates whether non-u.s. international institutional investment increased in the Canadian stock markets in the post-ifrs adoption period. A priori, it can be argued that that IFRS would bring more institutional investment in Canada from international countries due to the existence of more comparable IFRS-based financial statements in the post-ifrs adoption period. Canada may have decided to adopt IFRS in part to attract more international investment from the non-u.s. international world. This study, however, does not find supporting evidence that non-u.s. international institutional investors increased their investment in Canada in terms of higher percentage of shares owned or greater number of non-u.s. international institutional investors in the post-ifrs adoption period. It is likely that non-u.s. international institutional investors, on average, did not find comparability under IFRS a significant enough advantage in their decisions whether to invest in Canadian firms. In international circles among non-u.s. institutional investors, Canada already enjoyed a high reputation of high quality local accounting standards and established and reputable economic and legal institutions. 4.1 Introduction Canada decided to adopt IFRS in It adopted IFRS in the context of the U.S. being its largest trading partner. The decision to adopt IFRS may be interpreted as a decision to join the international community at the potential cost of Canada s economic closeness to the U.S. markets and investors. U.S. institutional investors constitute the largest international investment group in the Canadian stock markets. Hence, there is an inherent 110

119 trade-off when it comes to attracting international institutional investment in Canada s move to adopt IFRS. This exploratory study investigates whether non-u.s. international institutional investors found Canadian stock markets more attractive after mandatory IFRS adoption. The control group for the study consists of U.S. firms listed in U.S. stock markets. Most U.S. firms continued to follow U.S. GAAP. This comparative approach using Canadian and U.S. firms was not used in previous studies that investigate changes in investment related to IFRS adoption in Europe and Asia (Florou and Pope, 2012; Defond et al., 2012; Defond et al., 2011). The fact that Canada moved from a high quality local reporting regime to global accounting permits direct analysis of the question whether adoption of global accounting standards increased the attractiveness of its stock markets to non-u.s. international institutional investors. From here on the study uses the term international (institutional) investors to refer to non-u.s. international institutional investors who invest in the Canadian stock markets (U.S. stock markets). Previous studies that examine the incremental benefits of IFRS standards in many countries that have adopted IFRS cannot disentangle the effects of an increase in the quality of the accounting system itself and the influence of specific characteristics of IFRS, e.g. comparability, as some of those countries had lower-quality accounting before the adoption of IFRS (Landsman et al., 2012, Barth et al., 2008). Moreover, it is difficult to control for many local intra-country factors that can affect the results of these types of studies e.g. better auditing rules in the post-ifrs adoption period (Landsman et al., 2012; Armstrong et al., 2010; Barth et al., 2008). However, for the purposes of this study, both Canada and the U.S. already possessed the necessary underlying institutional framework, with developed accounting 111

120 and auditing professions and evolved and reputable capital markets (Ball, 2006). Hence, this study exploits a more controlled setting than previous related studies. Canada s economy has close links to the U.S. economy and does roughly 73 percent of its international trading with the U.S. The decision to accept IFRS as the dominant set of accounting standards in Canada provides an experimental setting for investigating whether Canadian adoption of IFRS resulted in increased participation of international investors in the Canadian capital markets. Proponents have used increased financial statement comparability as a key motivating factor to convince countries to adopt IFRS. The idea is that IFRS would eliminate many international differences in accounting standards and standardize accounting reporting formats (Ball, 2006). This would result in analysts makiag fewer adjustments in order to make companies financial statements more comparable across the globe. Hence, institutional investors would benefit, as they create large, standardized-format financial databases (Ball, 2006). It may also be argued that IFRS were created to satisfy the information needs of investors in terms of global markets. In other words, IFRS provide financial information at the international level. This is achieved as accounting information from different countries becomes more comparable (DeFond et al., 2012). The principles-based approach of IFRS helps local and international analysts cover more industries and firms across national boundaries, as there is less specific learning required once they understand the broader accounting principles (Simon, 1959). IFRS avoid detailed accounting rules, as ubiquitous rules exist under local accounting standards. IFRS rely more on auditors professional judgment in following general economic principles in order to ensure that the application of accounting standards is not misleading for the investors (Scott, 2011). 112

121 Hence, IFRS can benefit international investors, as there is increasing evidence in the finance and accounting literature that investors face limits on their ability to collect and process information and that these limits are likely to have stronger effects on international investors who are removed from the local scene and lack local connections, especially if the financial statements are prepared under local GAAPs (Scott, 2011). However, there are opposing arguments. Despite increased adoption of one set of global IFRS across the world, financial reporting practice is still influenced by local economic and political forces (Ball, 2006). Canadian firms are generally younger and growth oriented and mostly from oil, gas and mineral related industries. Ball (2006) argues that the possibility that international investors may have to deal with uneven IFRS implementation may result in increasing rather than decreasing information processing costs. The accounting inconsistencies under IFRS would be buried at a deeper and less transparent level for investors when comparing financial numbers from many countries than was the case under local accounting standards, when international investors could specialize in understanding those local differences in the local context in a more focused and transparent way. This exploratory study s main multivariate regression results show that the percentage ownership of international (non-u.s.) institutional investors in the Canadian firms has not changed significantly in the Canadian stock markets as compared to that in the U.S. markets in the post-ifrs period. The results may simply reflect Canada s established reputation as one of the best markets for international oil, gas and mineral firms to raise capital in the world. Ball (2006) argued that countries with a reputation of high quality accounting and strong governance institutions may not see much advantage 113

122 under IFRS. For the second dependent variable, the number of international (non-u.s.) investors who own stock in Canadian firms, the regression results also indicate insignificant changes, indicating that there are no significant differences in the number of international investors that owned stocks in Canadian firms in the Canadian stock markets as compared to the U.S. markets in the post-ifrs period. It may be observed that on average U.S. firms are larger and different on many dimensions as the U.S. is a much larger economy. For this reason, this study also ran as an additional robustness test the regression using a propensity score matched sample that included U.S. firms that were comparable to Canadian firms based on the given characteristics of the Canadian firms (DeFond et al., 2012). Based on the propensity score matched sample, the results for the second dependent variable, the number of international investors who own stocks in the Canadian firms, indicate that Canadian firms actually lost ground in terms of the number of international investors who own stocks in the Canadian firms compared to comparable U.S. firms in the post-ifrs adoption period. This paper adds to the stream of research that investigates the real effect of accounting still considered rare in the literature (Beneish et al., 2010; Florou and Pope, 2009; Yu, 2010; DeFond et al., 2011). The paper investigates the real economic impact international investment - when Canada adopted global accounting standards as a national policy instrument for reducing the economic distance between countries. However, the results show that Canada did not benefit from more investment by international (institutional) investors as a result of adopting IFRS. These results do not mean that Canada did not benefit in other ways from IFRS adoption. 114

123 Section 2 provides a literature review. In section 3, the paper develops the hypothesis. In Section 4 and 5, the paper presents the research design and the sample description, respectively. Section 6, and 7 provide the descriptive statistics, the multivariate tests and the results respectively. Finally section 8 concludes the study. 4.2 Literature Review Canada joined the international community in adopting IFRS in 2011 with the rationale that they are the gold standards for financial reporting in the global capital markets. On the other hand, the U.S., Canada s largest trading partner, did not. Ball (2006) argues that mandatory uniform standards (IFRS in this case) are a solution to the problem of negative information externalities. Externalities refer to the costs that the use of different accounting methods impose on others, as investors have to convert across different accounting standards to make sense of financial statements. International investors also lack local connections. Limiting information externalities in turn helps overcome the limited attention and information-gathering capabilities of international investors. These arguments for a principles-based approach and mandatory uniform standards provide support for the proposition that IFRS would enhance the integration of capital markets around the world. However, Ball (2006) provides a counterargument to this proposition by suggesting that local economic and political forces may hinder the benefits that countries can achieve from mandatory adoption of uniform accounting standards such as IFRS. Ball argues: Powerful local economic and political forces therefore determine how managers, auditors, court regulators and other parties influence the implementation of rules. These forces have exerted a substantial influence on financial reporting practice 115

124 historically, and are unlikely to suddenly cease doing so, IFRS or no IFRS. Achieving uniformity in accounting standards seems easy in comparison with achieving uniformity in actual reporting behavior. The latter would require radical change in the underlying economic and political forces that determine actual behavior. (p. 15) Several studies have investigated the economic consequences of increased comparability. De Franco et al. (2011) constructed a measure of financial statement comparability based on earnings-based outputs. This comparability measure captures the degree to which firms accounting numbers are similar to those of their industry peers. The researchers in this study used stock returns to capture the similarity in accounting numbers. They showed that financial statement comparability reduced the costs of acquiring information, and increased the quantity and quality of information available to analysts. Barth et al. (2012), in a different context, investigated the extent to which the accounting amounts from non-u.s. firms that follow IFRS are comparable to the accounting amounts from U.S. firms that follow U.S. GAAPs. They assessed accounting system comparability and value relevance comparability. They found that the amounts from IFRS firms are more comparable to amounts from U.S. GAAP firms. These studies support the argument that global investors may find relevant information advantages in global IFRS. Bradshaw et al. (2009) investigated another empirical proxy to measure comparability: the commonality of accounting choices, and examined the impact of comparability on analyst forecasts. They found that change in accounting procedures imposed processing costs on financial statement users in interpreting that information: larger analyst forecast errors and increased forecast dispersion. Yip and Young (2012) 116

125 examined comparability characteristics among mandatory-ifrs-adoption firms in the EU using an earnings-based comparability measure, and found that mandatory IFRS adoption improved cross-country information comparability. Improved financial statement comparability from IFRS may result in international investors taking advantage of reduced information acquisition costs and, as a result, investing more in international firms (Kang and Stulz, 1997). This is the market-wideadvantages argument that IFRS would provide positive externalities: investors would be able to differentiate between lower- and higher-quality firms (Daske et al., 2008). To capture such market-wide benefits, mandatory adoption of IFRS is beneficial (Coffee, 1984; Dye, 1990; Lambert, Leuz, and Verrecchia, 2007). Many empirical studies support the argument that IFRS facilitate integration of capital markets in terms of higher crossborder investment as a result of harmonization around IFRS, increasing comparability of firms across markets and countries (Armstrong et al., 2010; Hail et al., 2010; Li 2010; DeFond et al., 2011). However, others argue that we might not see an increase in cross-border investments as a result of adoption of IFRS. Ball (2006) blames enforcement institutions in each country as one of the key reasons that we might not see the benefits of IFRS. He argues that uneven enforcement of IFRS curtails the ability of uniform standards to reduce information costs and information risk. Since the comparability advantages of IFRS are connected to how IFRS users in other countries implement IFRS, this may hinder the expected benefits of IFRS adoption in Canada. 117

126 Moreover, Ball argues that information processing costs after the implementation of IFRS could increase for international investors, as the international differences in reporting standards would be buried at a much deeper and less transparent level. Ball (2006) further argues that this in turn would take away the signalling capacity of highquality reporting regimes that follow better standards than low-quality regimes. Canada already had high-quality accounting standards before its implementation of IFRS. Hence, it is not clear whether Canada would achieve its objective of better integration of its stock markets with the global markets, especially when it might lose some of its high-quality reporting regime s signalling power in the post-ifrs adoption period. Substantial international differences in financial reporting quality are inevitable, and my major concerns are that investors will be misled into believing that there is more uniformity in practice than actually is the case and that, even to sophisticated investors, international differences in reporting quality now will be hidden under the rug of seemingly uniform standards. (Ball, 2006, p. 22) A few recent studies have investigated the change in cross-border investments caused by IFRS adoption. This literature looks into the potential real effects of IFRS from a globalization advantage perspective (Beneish et al., 2010; Florou and Pope, 2009; Yu, 2010, DeFond et al., 2011). Beneish et al. (2010) investigated the impact on crossborder investments in both the equity and debt markets resulting from mandatory IFRS adoption. They found that it is foreign debt rather than foreign equity investment that flows more freely as a result of IFRS adoption. They found improved financial statement quality more important than comparability of the benefits of IFRS adoption. Florou and Pope (2009) found greater institutional investment in the post-ifrs adoption period. 118

127 4.3 Hypothesis Development Improved comparability under IFRS is the key driving force that may lead to increased investment by international (non-u.s.) investors in the Canadian stock markets in the post-ifrs adoption period. This is evident in the calls by many proponents of IFRS including the European Commissioner for the Internal Market, Frits Bolkestein: investors and other stakeholders will be able to compare like with like (GAAP Convergence, 2001). Both investment professionals and the academic literature support the argument that improved comparability under IFRS should lead to increased international investors investment in the IFRS adopting country. Investment professionals consider crosscountries accounting reconciliations that clarify the differences in accounting standards as very time consuming and costly (Morgan Stanley Dean Witter, 1998). Other academic studies have pointed to the high costs of acquiring and processing information about international firms as a key reason why international investors are reluctant to invest in international countries (Kang and Stulz, 1997; Bradshaw et al, 2004; Chan et al., 2005; Covrig et al., 2007; DeFond et al., 2011). Hence this paper hypothesizes that there will be relatively more international institutional investment in the Canadian stock markets as compared to that in the U.S. stock markets after the adoption of IFRS in Canada. Despite increased adoption of one set of global IFRS across the world, financial reporting practice is still influenced by local economic and political forces (Ball, 2006). Canadian firms are generally younger and growth oriented, and are mostly from oil, gas and mineral related industries. Many international oil, gas and mineral related firms 119

128 already list in the Canadian stock exchanges. It can be argued that Canadian GAAP were more suitable for local Canadian industries and local investors in order to make local economic decisions. Besides, Canada already had a reputation among international circles as an important developed country with high quality local accounting standards and with reputable economic and legal institutions. Ball (2006) argues that high reputation countries with high quality local accounting and strong local institutions may not gain benefit from IFRS adoption. Hence, this paper argues that gaining a comparability advantage under IFRS may not be a sufficiently strong enticement for non-u.s. international institutional investors when it comes to investing in Canada. Ball (2006) also argues that international investors may have to deal with uneven IFRS implementation, which may result in increasing rather than decreasing information processing costs. Barth et al. s (2008) study argues that a more principles-based approach under IFRS may increase managerial flexibility in reporting and thus increase opportunities for earnings management. Barth et al. s (2008) study and others also point out that reducing alternative accounting methods under IFRS may result in management providing a less true and faithful representation of the firm s underlying economic situation (Ashbaugh and Pincus, 2001). Since counterarguments for less investment by international investors in the Canadian stock markets are also valid, only empirical investigation can settle this study s broader research question Hypothesis 1: There is relatively more international institutional investment greater share of percentage ownership by international (non-u.s.) institutional investors and greater number of international institutional investors who own stocks - in the 120

129 Canadian stock markets as compared to that in the U.S. stock markets after the adoption of IFRS in Canada. 4.4 Research Design and empirical model This study s main tests compare the international (non-u.s.) institutional investment in Canadian firms against the benchmark international (non-canadian) institutional investment in U.S. firms in the post-ifrs adoption period. If Canada s decision to adopt IFRS has impacted the way international investors make investment decisions in Canada e.g. high comparability advantage, then the study may see a relative change in the international (non-u.s.) investment greater percentage share of ownership by international (non-u.s.) institutional investors and greater number of international (non-u.s.) institutional investors - in the Canadian firms listed in the Canadian exchanges compared to that in the U.S. firms listed in the U.S. exchanges. However, it is also possible that Canada did not benefit from its decision to adopt IFRS. Canada is predominantly an oil, gas and mineral based economy with close economic links to U.S. economy. Canada already enjoyed a high reputation in the international investor community as a country with high quality accounting and high quality regulations and institutions. For testing purposes, the study uses three dummy variables: POST, which is 1 in the post-ifrs period and 0 in the pre-ifrs periods, CANADA, which is 1 for Canadian firms listed in the Canadian stock exchanges and 0 for U.S. firms listed in the U.S. stock exchanges. The main variable of interest is the interaction term, POST x CANADA, which captures the effect of what happens to Canadian firms in terms of international 121

130 institutional investment in the post-ifrs period relative to that of U.S firms listed in the U.S. stock exchanges. If international (non-u.s.) institutional investment increased for Canadian firms in the post-ifrs as compared to the international (non-canadian) institutional investment in the benchmark U.S. sample, the study would observe a positive sign for coefficient of the interaction term, POST x CANADA. The study uses standard errors clustered by firms in all its main regression analysis. The study uses numerous control variables in the regression model that have been used in the previous literature. In order to alleviate the concern that the U.S. and Canadian firms are not comparable, the study also uses as additional robustness test propensity score matching methodology to identity a matched set of U.S. firms based on known characteristics of Canadian firms (Dehejia and Wahba, 2002; DeFond et al., 2012). U.S. firms are generally bigger in size, greater in numbers than Canadian firms and majority of Canadian firms belong to extractive industries. Having a much larger U.S. data that belong to larger firms from different set of industries may create a bias in the overall comparison between Canadian and U.S. firms in our investigation. The propensity score matching procedure uses logit regression to model the probability of being a Canadian firm using the Canadian and the U.S. sample. It uses firm level control variables as predictors in the logit regression. This will give the propensity score for each firm using the predicted probabilities obtained from the logit model. Next, it matches each Canadian firm (treatment firm) to the U.S. firm (control firm) using size, big 4 auditor, turnover, cash ratio, sales growth, book to market, dividend yield, leverage, return on equity, EP-ratio, returns and returns variation. Finally the study estimates the main regression equation using the matched sample just obtained from the previous steps. 122

131 Equations 1 and 2 are for the main inter-country tests. The main regression equations for the two dependent variables, international (non-u.s.) institutional percentage ownership and number of international (non-u.s.) institutional investors, are as follows: INT_INS_OWNit = β0 + β1postit + β2canadait + β3(post x CANADA)it + β4sizeit + β5big4_auditorit + β6turnoverit + β7adrit + β8cashit + β9sales_growthit + β10book_to_marketit + β11div_yieldit + β12leverageit + β13roeit + β14e-p_ratioit + β15returnsit + β16returns_variation+ εit (1) NUM_INT_INS_INVit = β0 + β1postit + β2canadait + β3(post x CANADA)it + β4sizeit + β5big4_auditorit + β6turnoverit + β7adrit + β8cashit + β9sales_growthit + β10book_to_marketit + β11div_yieldit + β12leverageit + β13roeit + β14e-p_ratioit + β15returnsit + β16returns_variation+ εit (2) Note: INT_INS_OWN stands for the international institutional ownership percentage. NUM_INT_INS_OWN stands for log of 1 plus the number of international institutional investors who own stocks in the period. POST is a dummy variable, 1 for 2013 in the post-ifrs period and 0 for 2009 in the pre-ifrs period. CANADA is another dummy variable equal to 1 for Canadian firms and 0 for U.S. firms. POST x CANADA is an interaction term used to capture the incremental impact on the dependent variable for Canadian firms in the post-ifrs period. SIZE is the log of the market size of the firm. BIG4_AUDITOR is 1 if the firm is audited by one of the big 4 auditors and 0 otherwise. TURNOVER is the annual share volume divided by the shares outstanding. ADR is 1 if the firm is cross-listed in the U.S. stock exchange. CASH is the ratio of cash and short-term investment to total assets at year-end. SALES_GROWTH is the annual growth rate in the net sales. BOOK_TO_MARKET is the book value of equity divided by market value of equity at year-end. DIV_YIELD is the total dividends divided by the market value of equity at year-end. LEVERAGE is the total liabilities divided by total assets. ROE is net income before extraordinary items divided by book value of equity at year-end. E- P_RATIO is net income divided by market value of equity at year-end. RETURNS variable is the stock returns over the calendar year. RETURN_VARIATION is the standard deviation of monthly stock returns over the calendar year. 123

132 4.5 Sample Description This study collected the data from the S&P Capital IQ database for percentage of ownership and number of international owners. It collected most of the accounting data from S&P Capital IQ and Compustat. It collected the cross-listings data from S&P Capital IQ database and collected the number of analyst data from IBES database. The study collected the returns and standard deviation of returns data from Datastream database. The study uses the TSX and TSX Venture exchanges firms in the Canadian sample while it uses the NYSE, NASDAQ and NYSE LLC market exchanges firms in the sample for the U.S. For the pre-ifrs period, it collected fourth quarter observations for percentage ownership for each international owner for 2009 calendar year. The study excluded the year 2011 to avoid transition effects as it is the first year of adoption of IFRS for many firms, and also many firms in Canada did not adopt IFRS until For the post-ifrs period the study collected fourth quarter observations for percentage ownership for each international owner for the 2013 calendar year. The study uses the U.S. sample as a benchmark against which to compare if the Canadian firms experienced greater international investment. In the first set of multivariate tests in tables 4-5 and 4-6, it has 8917 firm-year observations for the both the U.S. and Canadian firms. In the second set of tests using propensity score model in tables 4-7 and 4-8, it has 4128 firm-year observations for the U.S. and the Canadian firms. 124

133 Following the previous literature (DeFond et al., 2012), the study winsorizes the top and bottom 1% of all of the scaled control variables (i.e. ROE, Returns, Return Variation, Leverage, Dividend Yield, Book to Market, E-P ratio, Sales Growth, Stock Turnover, Cash and Closely Held). The study uses two dependent variables at the firm level: percentage ownership of international (non-u.s.) investors and number of international owners that own a stock in each firm. Both of these dependent variables have been used in the previous literature (DeFond et al., 2012; Bradshaw et al., 2004). The first dependent variable, percentage ownership of international institutional investors in each firm, indicates the size of international investment in terms of percentage ownership share. If the study found that there is a higher percentage ownership share, on average, by international (non-u.s.) institutional investors in the Canadian firms in the post-ifrs period compared to that in the U.S. firms, then the study can attribute the greater percentage ownership in Canadian stocks by international (non-u.s.) institutional investors to IFRS adoption. In the multivariate regression analysis, the study takes the cumulative percentage ownership for international (non-u.s.) institutional investment for each firm. The second dependent variable, number of international investors that own stocks in each firm, indicates how extensively Canadian stocks are held among the international investors. If the study found that there are a higher number of international investors in Canadian firms in the post-ifrs period, it would indicate that international investors are increasing their investment in Canadian stocks in terms of greater number of international investors that own stocks in Canadian firms. For the multivariate regression analysis, the 125

134 study measures this variable as log of 1 plus the number of international investors that own stock in each sample firm. 4.6 Descriptive Statistics Tables 4-1 and 4-2 show the descriptive statistics for all variables in the pre-ifrs and post-ifrs period for the Canadian firms respectively. For the number of institutional (non-u.s.) owners variable both the mean and the median increased in the post-ifrs period (7.53 mean and 2 median) compared to the mean and median in the pre-ifrs period (4.42 mean and 1 median). For the international (non-u.s.) institutional ownership percentage variable both the mean and the median increased in the post-ifrs period (0.04 mean and 0.01 median) compared to the mean and median in the pre-ifrs period (0.02 mean and 0.00 median). For the size variable in table 4-1 and 4-2, both the mean and the median decreased in the post-ifrs period (4.27 mean and 4.12 median) compared to the mean and median in the pre-ifrs period (4.56 mean and 4.39 median). Table 4-3 and 4-4 show the descriptive statistics for all variables in the pre-ifrs and post-ifrs period for the U.S. firms respectively. For the number of institutional owners variable both the mean and the median increased in the post-ifrs period (10.77 mean and 8 median) compared to the mean and median in the pre-ifrs period (6.81 mean and 4 median). For the international institutional ownership percentage variable both the mean and the median increased in the post-ifrs period (0.03 mean and 0.02 median) compared to the mean and median in the pre-ifrs period (0.02 mean and 0.01 median). For the size variable in table 4-3 and 4-4, both the mean and the median 126

135 decreased in the post-ifrs period (6.56 mean and 6.52 median) compared to the mean and median in the pre-ifrs period (6.06 mean and 5.98 median). It is evident that U.S. firms are generally larger firms compared to Canadian firms (4.27 for Canadian firms and 6.56 for U.S. firms). Share turnover is, on average, higher for U.S. firms than for Canadian firms (0.35 for Canadian firms and 1.58 for U.S. firms). This is because U.S. stock markets are generally more liquid than Canadian stock markets. Propensity score matching sample testing may take care of some of the inherent differences between the Canadian and U.S. firms. 4.7 Multivariate Statistics The study used industry fixed effects and robust standard errors clustered by firms in all main regression analysis. In table 4-5, the first regression for international (non- U.S.) institutional percentage ownership shows that there was no relative change in the post-ifrs period in Canada compared to that in the U.S. firms as POST x CANADA is negative and not significant. This means that IFRS adoption did not benefit Canadian firms from the perspective of more international ownership as compared to that in the U.S. firms. But the variable CANADA is positive and significant meaning Canadian firms generally enjoyed higher international (non-u.s.) percentage ownership than that in the U.S. firms even in the pre-ifrs period. The POST variable is positive and significant, meaning that U.S. firms, on average, experienced higher international (non-canadian) percentage ownership in the post-ifrs period. The results are informative as they indicate that Canada generally enjoyed a high reputation in the international world as a destination to invest in and the likely policy 127

136 change of Canada to move towards IFRS instead of staying with Canadian GAAP maintained the status quo of its high reputation among the international (non-u.s.) investors. Ball (2006) argued that countries that had high quality accounting and high reputation with high quality regulatory institutions might not benefit from IFRS. The Size control variable is positive and significant, meaning that larger firms attract greater international (non-u.s.) percentage ownership. In table 4-6, the second regression for number of international institutional investors shows that there was no significant change in the number of international owners in the post-ifrs period in Canada compared to that in the U.S. firms as POST x CANADA is not significant. This means that IFRS did not benefit Canadian firms from the perspective of more international institutional investors investing in Canadian firms as compared to that in the U.S. firms. Again, the results are informative. The policy choice of Canada moving towards IFRS instead of staying with Canadian GAAP did not made much difference to the decision making of investing in Canada by international institutional investors. Again the reason may be that Canada already had a high reputation among international institutional investors as one of the high quality destinations to invest in oil, gas and mineral related stocks. This confirms again what Ball (2006) was arguing that countries that already have high quality accounting and reputation might not benefit from IFRS adoption. The size control variable is positive and significant, meaning that larger firms attract greater international (non-u.s.) percentage ownership. Other control variables incluing turnover, leverage, returns, number of analyst, and returns variation are significant and generally have similar signs as in previous studies. 128

137 This study then analyses the propensity score matched sample regressions. In table 4-7, the propensity score matched sample regression for international institutional percentage ownership shows that there was no change in the percentage ownership in the post-ifrs period in Canada as POST x CANADA is negative and not significant. This result is same as the main results in table 4-5. In table 4-8, the propensity score matching sample regression for number of international institutional investors shows that the number of international owners decreased in the post-ifrs period in Canada as POST x CANADA is negative and significant at the 5% level (p<0.05). This means that IFRS may have had a negative impact on overall Canadian firms from the perspective of the number of international institutional owners investing in Canadian firms as compared to that in the matched U.S. firms. Based on the empirical tests for two years, 2009 and 2013 in this study, it can be concluded that there is no evidence that Canada benefitted from the adoption of IFRS from the perspective of international (non-u.s.) institutional investment compared to that in the U.S. In fact, there is some evidence in the propensity score matched sample regression results that Canada may have become a little disadvantaged as there are fewer international institutional investors who own stocks in Canadian firms in the post-ifrs adoption period compared to that in each U.S. firm. 4.8 Summary & Future Research Direction This study investigates Canada s decision to adopt IFRS, as Canadian adoption experience is very interesting in the backdrop that Canada and U.S. are similar and economically interconnected in many ways. Canada decided to adopt IFRS but U.S. did 129

138 not. Based on the evidence this study collected, it can be concluded that Canada did not benefitted from its decision to adopt IFRS from the perspective of international (non- U.S.) institutional investment compared to that in the U.S. in the post-ifrs adoption period. It is probable that international institutional investors found a high reputation with local high quality accounting and reputable economic and legal institutions advantages for Canada as more important signals than gaining comparability advantage under IFRS (Ball, 2006). Future research can look into doing a comparison between non-u.s. international institutional investment for Canadian and Australian firms during the respective IFRS adoption periods in both countries. The reason is that this study s Canadian results may also be explained by the general shift in the prices of commodities or other macroeconomic factors. It is possible that non-u.s. international institutional investment for Canadian firms may be lower in the post-ifrs adoption period due to lower commodity prices. It is also clear from my second study s analysis that Australian economy is very similar to the Canadian economy especially because roughly 50% of the economies in both countries are from extractive industries i.e. oil, gas, and mineral firms. These new comparative tests for non-u.s. international institutional investment for Canadian and Australian firms may control for these macro-level factors. 130

139 Bibliography Aboody, D., M. E. Barth, and R. Kasznik Revaluations of fixed assets and future firm performance: Evidence from the U. K. Journal of Accounting and Economics. 26: Ahmed, A. S., M. Neel, and D. Wang Does mandatory adoption of IFRS improve accounting quality? Preliminary evidence. Contemporary Accounting Research. 30(4): Ahmed, A. S., R. A. Schneible, and D. E. Stevens An empirical analysis of the effects of online trading on stock price and trading volume reactions to earnings announcements. Contemporary Accounting Research 20(3): Amiram, D Financial information globalization and foreign investment decisions. Journal of International Accounting Research. 11(2): Arbel, A., and P. Strebel The neglected and small firm effects. The Financial Review. 17(4): Armstrong, C., M. Barth, A. Jagolinzer, and E. Riedl Market reaction to the adoption of IFRS in Europe. The Accounting Review. 85(1): Ashbaugh, H., and M. Pincus Domestic accounting standards, international accounting standards, and the predictability of earnings. Journal of Accounting Research. 39(3): Ausseneg, W., P. Inwinkl, and G. Schneider Earnings management and local vs International accounting standards of European public firms. Working Paper, Vienna University of Technology. Bae, K. H., Tan, H., & Welker, M. (2008). International GAAP differences: The impact on foreign analysts. The Accounting Review, 83(3), Ball, R. and P. Brown An empirical evaluation of accounting income numbers. Journal of Accounting Research Ball, R., A. Robin, and J. S. Wu Incentives versus standards: properties of accounting income in four East Asian countries. Journal of accounting and economics. 36(1): Ball, R., S. Kothari, and A. Robin The effect of international institutional factors on properties of accounting earnings. Journal of Accounting and Economics. 29(1):

140 Ball, R International Financial Reporting Standards (IFRS): pros and cons for investors. Accounting and Business Research. 36(sup1): Ball, R. and L. Shivakumar Earnings quality in UK private firms: comparative loss recognition timeliness. Journal of Accounting and Economics. 39(1): Bamber, L., O. Barron, and D. Stevens Trading volume around earnings announcements and other financial reports: Theory, research design, empirical evidence, and directions for future research. Contemporary Accounting Research. 28(2): Barron, O., R. Schneible, and D. Stevens What do we really Know about the Firm size Effect on Trading Volume Reactions to Earnings Announcement? A Re-Examination and Extension. Working paper. Barth, M., W. Landsman, and M. Lang International accounting standards and accounting quality. Journal of Accounting Research, 46(3), Barth, M., W. Landsman, M. Lang, and C. Williams Are IFRS-based and US GAAP-based accounting amounts comparable? Journal of Accounting and Economics. 54(1): Basu, S The conservatism principle and the asymmetric timeliness of earnings. Journal of Accounting and Economics. 24(1): Beaver, W Alternative accounting measures as predictors of failure. The Accounting Review. 43(1): Beaver, W. and S. Ryan Conditional and unconditional conservatism: concepts and modeling. Review of Accounting Studies. 10(2-3): Beneish, M. D., and J. C. Gardner Information costs and liquidity effects from changes in the Dow Jones Industrial Average list. Journal of Financial and Quantitative Analysis. 30(01): Beneish, M. D., et al The effect of IFRS adoption on cross-border investment in equity and debt markets. Working paper. Available at: com/abstract Beuselinck, C., P. Joos, I. Khurana, and S. van der Meulen Mandatory IFRS Reporting and Stock Price Informativeness. CentER Discussion Paper; Vol Tilburg University. Bhushan, R. Firm characteristics and analyst following. Journal of Accounting and Economics. 11(2): Biddle, G, G. Hilary, and R. Verdi How does financial reporting quality relate to investment efficiency? Journal of Accounting and Economics. 48(2):

141 Blanchette, M., F.-E. Racicot, and J.-Y. Girard The effects of IFRS on financial ratios: Early evidence in Canada. Certified General Accountants Association of Canada. Botosan, C. A Disclosure Level and the Cost of Equity Capital. The Accounting Review. 72(3): Bradshaw, M. T., B. J. Bushee, and G. S. Miller Accounting Choice, Home Bias, and US Investment in Non US Firms. Journal of Accounting Research. 42(5): Bradshaw, M. T., G. S. Miller, and G. Serafeim Accounting method heterogeneity and analysts forecasts. Unpublished paper, University of Chicago, University of Michigan, and Harvard University. Brennan, M. and P. Hughes Stock prices and the supply of information. The Journal of Finance. 46(5): Briston, R.J The Evolution of Accounting in Developing Countries. International Journal of Accounting: Education and Research. Fall: Brochet, F., A. D. Jagolinzer and E. J. Riedl, Mandatory IFRS adoption and financial statement comparability. Contemporary Accounting Research. 30(4): Byard, D., Y. Li, and Y. Yu The effect of mandatory IFRS adoption on financial analysts information environment. Journal of Accounting Research. 49(1): Cairns, D Aid for the Developing World. Accountancy. March: Chamisa, E.E The Relevance and Observance of the IASC Standards in Developing Countries: Case Studies of Botswana and Zimbabwe. Unpublished Doctoral Thesis, University of Birmingham, England. Chan, K., V. Covrig, and L. Ng What determines the domestic bias and foreign bias? Evidence from mutual fund equity allocations worldwide. The Journal of Finance 60(3): Chen, C. W., D. W. Collins, T. D. Kravet, and R. Mergenthaler Financial statement comparability and the efficiency of acquisition decisions. Available at SSRN Coffee Jr, J. C Market failure and the economic case for a mandatory disclosure system. Virginia Law Review Covrig, V., M. Defond, and M. Hung Home bias, foreign mutual fund holdings, and the voluntary adoption of international accounting standards. Journal of Accounting Research. 45(1):

142 Danbolt, J., and W. Rees An experiment in fair value accounting: UK investment vehicles. European Accounting Review. 17(2): Daniel, K., D. Hirshleifer, and A. Subrahmanyam Investor psychology and security market under-and overreactions. The Journal of Finance, 53(6), Daske, H., L. Hail, C. Leuz, and R. Verdi Mandatory IFRS reporting around the world: Early evidence on the economic consequences. Journal of Accounting Research. 46: DeFond, M. L., and M. Hung An empirical analysis of analysts cash flow forecasts. Journal of Accounting and Economics. 35(1): DeFond, M., M. Hung, and R. Trezevant Investor protection and the information content of annual earnings announcements: International evidence. Journal of Accounting and Economics. 43(1): DeFond, M., X. Hu, M. Hung, and S. Li The impact of mandatory IFRS adoption on foreign mutual fund ownership: The role of comparability. Journal of Accounting and Economics, 51(3), DeFond, M. L., et al Has the Widespread Adoption of IFRS Reduced US Firms' Attractiveness to Foreign Investors? Journal of International Accounting Research. 11(2): Dehejia, R. H., and S. Wahba Propensity score-matching methods for nonexperimental causal studies. Review of Economics and statistics. 84(1): De Franco, G., et al The benefits of financial statement comparability. Journal of accounting research. 49(4): Doukakis, L The persistence of earnings and earnings components after the adoption of IFRS. Managerial Finance. 36: Dye, R. A Mandatory versus voluntary disclosures: The cases of financial and real externalities. Accounting Review. 1990: Dye, R. and S. Sunder Why not allow FASB and IASB standards to compete in the US? Accounting Horizons. 15(3): Easley, D., S. Hvidkjaer, and M. O hara Is information risk a determinant of asset returns? The journal of finance. 57(5): Easley, D., and M. O hara Information and the cost of capital. The journal of finance. 59(4):

143 Easley, D., M. O'hara, and P. S. Srinivas Option volume and stock prices: Evidence on where informed traders trade. The Journal of Finance. 53,(2): Fields, T., T. Lys, and L. Vincent Empirical research on accounting choice. Journal of Accounting and Economics. 31(1): Florou, A. and P. Pope Mandatory IFRS adoption and investor asset allocation decisions. Working Paper: Available at Foucault, T., O. Kadan, and E. Kandel Liquidity cycles and make/take fees in electronic markets. The Journal of Finance. 68(1): Francis, J., S. Huang, I. Khurana, and R. Pereira Does corporate transparency contribute to efficient resource allocation? Journal of Accounting Research. 47(4): Hail, L. and C. Leuz Capital market effects of mandatory IFRS reporting in the EU: Empirical evidence. Available at: Hail, L., C. Leuz, and P. Wysocki Global accounting convergence and the potential adoption of IFRS by the US (Part I): Conceptual underpinnings and economic analysis. Accounting Horizons. 24(3): Harris, Larry Trading and exchanges: Market microstructure for practitioners. Oxford University Press Hasbrouck, J Trades, quotes, inventories, and information. Journal of Financial Economics. 22(2): Hayn, C The information content of losses. Journal of accounting and economics, 20(2), Healy, P., and K. Palepu Information asymmetry, corporate disclosure, and the capital markets: A review of the empirical disclosure literature. Journal of accounting and economics. 31(1): Healy, P., A. Hutton, and K. Palepu Stock performance and intermediation changes surrounding sustained increases in disclosure. Contemporary Accounting Research, 16(3), Hope, O. K Analyst following and the influence of disclosure components, IPOs and ownership concentration. Asia-Pacific Journal of Accounting & Economics. 10(2): Hope, O., W. Thomas, and G. Winterbotham, G Geographic earnings disclosure and trading volume. Journal of Accounting and Public Policy, 28(3),

144 Horton, J., G. Serafeim, and I. Serafeim Does Mandatory IFRS Adoption Improve the Information Environment? Contemporary Accounting Research. 30(1): Jeanjean, T., and H. Stolowy Do accounting standards matter? An exploratory analysis of earnings management before and after IFRS adoption. Journal of Accounting and Public Policy. 27(6): Kabir, M., F. Laswad, and M. Islam Impact of IFRS in New Zealand on accounts and earnings quality. Australian Accounting Review. 20(4): Kang, J. K. and R. Stulz Why is there a home bias? An analysis of foreign portfolio equity ownership in Japan. Journal of financial economics. 46(1): Kim, O. and R. Verrecchia Market reaction to anticipated announcements. Journal of Financial Economics. 30(2): Kim, O. and R. Verrecchia Trading volume and price reactions to public announcements. Journal of Accounting Research. 29(2): Kim, O. and R. Verrecchia Pre-announcement and event-period private information. Journal of Accounting and Economics. 24(3): Kothari, S Capital markets research in accounting. Journal of Accounting and Economics. 31(1): Lambert, R., C. Leuz, and R. Verrecchia Accounting information, disclosure, and the cost of capital. Journal of Accounting Research. 45(2): Landsman, W., E. Maydew, and J. Thornock The information content of annual earnings announcements and mandatory adoption of IFRS. Journal of Accounting and Economics. 53(1): Leuz, C, and P. Wysocki Economic consequences of financial reporting and disclosure regulation: A review and suggestions for future research. Working paper. Available at: abstract= Leuz, C., and R. Verrecchia The Economic Consequences of Increased Disclosure (Digest Summary). Journal of Accounting Research, 38, Leuz, C IAS versus US GAAP: information asymmetry based evidence from Germany's new market. Journal of Accounting Research. 41(3): Lev, B On the usefulness of earnings and earnings research: lessons and directions from two decades of empirical research. Journal of Accounting Research. 40:

145 Li, S Does mandatory adoption of International Financial Reporting Standards in the European Union reduce the cost of equity capital? The Accounting Review. 85(2): Merton, R. C A simple model of capital market equilibrium with incomplete information. The journal of Finance. 42(3): O'Brien, P Changing the Concepts to Justify the Standards. Accounting Perspectives. 8(4): O'Brien, P., and W. Scott Contract theory: The effectiveness of a conceptual framework for financial reporting has been debated for years. CA Magazine, April 1. Pandes, J. A. and M. J. Robinson The Canadian junior IPO market and the capital pool company program. Handbook of Research on IPOs: Panayides, M. A., and A. Ellul Do Financial Analysts Restrain Insiders' Informational Advantage? Working Paper: Available at PriceWaterhouseCoopers Putting IFRS in motion: Key accounting issues for pharmaceuticals and life sciences companies. Samuels, J.M. and J.C. Oliga Accounting Standards in Developing Countries. International Journal of Accounting Education and Research. Fall: Scott, W. R Financial accounting theory. Pearson Education Canada. Sharpe, I. and R. Walker Asset revaluation and stock market prices. Journal of Accounting Research. 13: Simon, H. A Theories of decision-making in economics and behavioral science. The American economic review 49(3): Standish, P, and S.-I. Ung Corporate signaling, asset revaluation and stock prices of British companies. The Accounting Review. 57(4): Tan, H., Wang, S., and M. Welker Analyst following and forecast accuracy after mandated IFRS adoptions. Journal of Accounting Research, 49(5), Verrecchia, R Essays on disclosure. Journal of Accounting and Economics. 32(1-3): Watts, R Conservatism in accounting part I: Explanations and implications. Accounting Horizons. 17(3): Watts, R Conservatism in accounting part II: Evidence and research opportunities. Accounting Horizons 17(4):

146 Welker, M Disclosure policy, information asymmetry and liquidity in equity markets. Contemporary Accounting Research. 11: Yip, R. W. and D. Young Does mandatory IFRS adoption improve information comparability? The Accounting Review. 87(5): Yu, G Accounting standards and international portfolio holdings: Analysis of cross-border holdings following mandatory adoption of IFRS. Dissertation. University of Michigan. 138

147 Table 4-1 Descriptive Statistics for all variables in the pre-ifrs period for Canadian Firms Variable n Mean S.D. Min.25 Mdn.75 Max Number of foreign Institutional owners Foreign institutional ownership Size Big4 Auditor Turnover ADR CASH Sales Growth Book to Market Div. Yield LEV ROE EP Ratio Returns Returns Variation

148 Table 4-2 Descriptive Statistics for all variables in the post-ifrs period for Canadian Firms Variable n Mean S.D. Min.25 Mdn.75 Max Number of foreign Institutional owners Foreign institutional ownership Size Big4 Auditor Turnover ADR CASH Sales Growth Book to Market Div. Yield LEV ROE EP Ratio Returns Returns Variation

149 Table 4-3 Descriptive Statistics for all variables in the pre-ifrs period for U.S. Firms Variable n Mean S.D. Min.25 Mdn.75 Max Number of foreign Institutional owners Foreign institutional ownership Size Big4 Auditor Turnover ADR CASH Sales Growth Book to Market Div. Yield LEV ROE EP Ratio Returns Returns Variation

150 Table 4-4 Descriptive Statistics for all variables in the pre-ifrs period for U.S. Firms Variable n Mean S.D. Min.25 Mdn.75 Max Number of foreign Institutional owners Foreign institutional ownership Size Big4 Auditor Turnover ADR CASH Sales Growth Book to Market Div. Yield LEV ROE EP Ratio Returns Returns Variation

151 Table 4-5 Multivariate regression for foreign institutional percentage ownership with 2009 as pre-ifrs period and 2013 as post-ifrs period (with industry fixed effects and standard errors clustered by firms) Variables Coefficients t-values p-values POST CANADA POST x CANADA Size Big4 Auditor Turnover ADR Cash Sales Growth Book-to-Market Div. Yield LEVERAGE ROE wep_ratio Nanalyst Return Return Variation Intercept R-Squared Number of observations

152 Table 4-6 Multivariate regression for number of foreign institutional owners with 2009 as pre-ifrs period and 2013 as post-ifrs period (with industry fixed effects and standard errors clustered by firms) Variables Coefficients t-values p-values POST CANADA POST x CANADA Size Big4 Auditor Turnover ADR Cash Sales Growth Book-to-Market Div. Yield LEVERAGE ROE wep_ratio Nanalyst Return Return Variation Intercept R-Squared Number of observations

153 Table 4-7 Propensity score matching sample - multivariate regression for foreign institutional percentage ownership with 2009 as pre-ifrs period and 2013 as post-ifrs period (with industry fixed effects) Variables Coefficients t-values p-values POST CANADA POST x CANADA Size Big4 Auditor Turnover ADR Cash Sales Growth Book-to-Market Div. Yield LEVERAGE ROE wep_ratio Nanalyst Return Return Variation Intercept R-Squared Number of observations

154 Table 4-8 Propensity score matching sample - multivariate regression for number of foreign institutional owners with 2009 as pre-ifrs period and 2013 as post-ifrs period (with industry fixed effects) Variables Coefficients t-values p-values POST CANADA POST x CANADA Size Big4 Auditor Turnover ADR Cash Sales Growth Book-to-Market Div. Yield LEVERAGE ROE wep_ratio Nanalyst Return Return Variation Intercept R-Squared Number of observations

155 Chapter 5: Summary In my thesis, I examine the costs and benefits of implementing IFRS in Canada from two perspectives: from the perspective of local investors and from the perspective of international investors. This examination involved completing three research papers. Below I provide a summary of my thesis constituting three papers and the limitations of my thesis. In my first paper in Chapter 2, I examine whether IFRS adoption resulted in greater information content in earnings announcements for TSX and TSXV firms in Canada. In other words, I investigate if the quality of earnings information increased under IFRS in two Canadian exchanges TSX and TSXV. Increased value relevance of information provided by IFRS earnings would likely reflect increased measurement of changes in net asset values based on expectations as opposed to realizations. Because values based on expectations are subject to greater divergence of opinion than values based on realizations, greater value relevance is likely to be accompanied by higher abnormal return volatility and abnormal trading volume. Consistent with this argument, this study finds abnormal volatility and abnormal volume during the earnings announcement periods was higher in post-ifrs announcements periods than in pre-ifrs announcement periods for firms traded on the TSX. However, the benefit of higher abnormal returns volatility and abnormal volume due to more relevant information in the earnings announcements was not gained for the TSXV firms. This study discriminates across the two exchanges in terms of information quality based on the mix of institutional and retail investors, analyst following, concentration in 147

156 the oil, gas and mining sectors and size of firm. Institutional investors are more prominent in the TSX information environment while retail investors participate heavily in the TSXV environment. These results indicate that IFRS earnings announcements led to greater divergence of opinion among investors in TSX firms, possibly due to the more forward-looking measurement emphasis under IFRS. I also find some evidence that information content of earnings was higher for TSXV firms relative to TSX firms in the pre-ifrs period. I cautiously suggest that TSXV results of lower abnormal volatility and abnormal volume in the post-ifrs period may reflect a greater reliance of TSXV investors on realized earnings information emphasized more under Canadian GAAP. My first paper suggests that Canada s big exchange, TSX, may have benefitted in terms of more information content in the earnings announcements in the post-ifrs adoption periods but may have lost ground in terms of less information content in the IFRS based earnings announcements for TSXV firms in the post-ifrs adoption periods. TSXV exchange is one of an important component of Canadian economic story. Lots of young, entrepreneurial and innovative Canadian firms obtain initial capital in this exchange. Hence the benefits of IFRS are mix when we consider the information content of IFRS based earnings announcements in the post-ifrs adoption periods. In my second paper in Chapter 3, I examine whether average liquidity for three groups of firms - Canadian, non-us international and U.S. firms - traded on Canadian stock exchanges increased or decreased after mandatory adoption of IFRS in Canada. I consider two competing forces affecting liquidity from IFRS adoption: enhanced comparability of firms within industries that span international boundaries and less tailoring of financial reporting to satisfy local investor needs. I find that liquidity 148

157 decreased for Canadian and U.S. firms, suggesting that the benefits of global comparability were not sufficient to offset the loss of local investor accommodation. However, I find some evidence that liquidity improved for non-u.s. international firms suggesting that enhanced comparability of firms within industries did span international boundaries. To provide perspective on these results for Canadian exchanges, I also compare liquidity before and after IFRS adoption for all three groups of firms traded on the U.K. and Australian exchanges. These three countries share the British influence on financial reporting historically and have similar legal and institutional settings. This study expects that the benefits of global comparability would be relatively greater for the U.K. exchanges relative to the Canadian and Australian exchanges given the U.K. s proximity and commerce with other European countries. I find that liquidity increased for the U.K. exchanges but decreased for the Australian exchange, supporting the interpretation of the Canadian findings. These results are interesting as I expected the comparability characteristics of IFRS to benefit the Canadian listed firms. I infer based on the evidence that diversified industry presence and more international firms are important factors to capture the comparability advantages of IFRS. The initial results show Canadian firms and investors may have been disadvantaged by the adoption of IFRS in Canada. Future research can investigate if the lower liquidity for Canadian and Australian domestic firms in the post- IFRS adoption period were at harmful levels for local economies. Similarly, future research can investigate if the gain in liquidity for non-u.s. international firms in Canada was significant enough advantage for Canadian economy that Canada opted to adopt IFRS at the cost of rejecting well-functioning local Canadian GAAPs. 149

158 In my third and final paper in Chapter 4, I examine whether non-u.s. international institutional investment, measured by percentage ownership by international investors and number of international investors, increased in the Canadian stock markets in the post- IFRS adoption period. I consistently find evidence that Canada did not benefit from higher non-u.s. international institutional investment in the Canadian stock markets in the post-ifrs period. Ball (2006) argued that countries that already have high quality accounting and that already have well-established capital markets and other regulatory institutions may not benefit from IFRS. A developed country like Canada that already has high reputation among the international community in the pre-ifrs adoption period with its high quality accounting and strong local governance institutions may not find a strong advantage by adopting IFRS mandatorily. The results in my third paper show that Canada did not see any noticeable gain from more non-u.s. international institutional investment in the Canadian firms in the post-ifrs adoption period. Future research can look into doing a comparison between non-u.s. international institutional investment for Canadian and Australian firms during the respective IFRS adoption periods in both countries. The reason is that this study s Canadian results may also be explained by the general shift in the prices of commodities or other macro-economic factors. It is possible that non-u.s. international institutional investment for Canadian firms may be lower in the post-ifrs adoption period due to lower commodity prices. It is also clear from my second study s analysis that Australian economy is very similar to the Canadian economy especially because roughly 50% of the economies in both countries are from extractive industries i.e. oil, gas, and mineral firms. 150

159 These new comparative tests for non-u.s. international institutional investment for Canadian and Australian firms may control for these macro-level factors. Based on the three studies results, I can conclude that Canada may not have benefitted from the adoption of IFRS in Canada. Paper 1 s results showed that the information content increased for larger firms in the TSX exchange in the short window earnings announcement test but decreased for firms in the TSXV exchange in the post- IFRS adoption period. Considering that firms in the TSX sample were generally larger in size considering our sample criteria in paper 1, paper 2 s results are congruent with paper 1 s results as size tests (additional robustness tests in table 3-17 and 3-18 in paper 2) showed that larger firms generally experienced higher liquidity while smaller firms experienced lower liquidity in the post-ifrs adoption period. However, the overall results in paper 2 for Canadian domestic firms showed that liquidity decreased for domestic firms in Canada. Similarly, paper 3 s results also showed that Canadian firms listed in the Canadian stock markets did not benefitted from more non-u.s. international institutional investment in the post-ifrs adoption period. However in paper 2, there is limited evidence that that liquidity increased for non-u.s. international firms listed in the Canadian stock markets in the post-ifrs adoption period, meaning that there is some evidence that Canadian stock markets may have become more global in the post-ifrs adoption period. The above results about IFRS implementation in Canada do have limitations. In all three papers, I compare the pre-ifrs adoption period with the post-ifrs adoption period. In my first two studies, I picked 2009 and 2010 as the pre-ifrs adoption period and 2012 and 2013 as the post-ifrs period. In my third study, I picked 2009 as the pre- 151

160 IFRS adoption period and 2013 as the post-ifrs adoption period. It is possible that the time period may be capturing some unrelated effects that impact the results of my study, and I may find different results if I include more years in my sample. Also, the costs and benefits of IFRS studied in my thesis are not comprehensive in nature. There is a possibility that IFRS may have benefitted Canada in other situations that I have completely ignored in my thesis. 152

161 Appendix: Copyright permission documents 153

162 154

163 155