In chapter one, by exploiting the staggered interstate banking deregulation as exogenous shocks to bank geographic expansion, we examine the causal effect of geographic diversification on systemic risk using the gravity-deregulation approach developed in Goetz, Laeven, and Levine (2013, 2016). We find that bank geographic diversification leads to higher systemic risk measured by the change in conditional value at risk (ΔCoVaR) and financial integration (Logistic(R2)). Furthermore, we document asset similarity and bank inter-connectedness as two channels to explain the documented results. The impact of geographic diversification on systemic risk is more pronounced in BHCs located in states comoving less with the U.S. aggregate economy. In chapter two, by integrating the staggered interstate bank deregulation into a gravity model, we construct a time-varying bank-specific instrument for geographic diversification, and investigate how geographic expansion affects borrowing firms’ innovation. Our approach disentangles the effects of bank deregulation on geographic expansion from competition and isolates its direct impact on innovation via the lending channel. Bank geographic diversification boosts borrowing firms’ innovation input and output, enables firms to expand innovation scope beyond core business, and enhances the economic value of innovation. We find that relaxing debt covenants and alleviating borrowers’ financial constraints are two channels through which bank geographic diversification spurs innovation. In chapter three, we construct a novel bank-specific and time-varying measure of deregulation-induced bank competition following Jiang, Levine, and Lin (2016) and Goetz (2017), and investigate the causal effect of bank competition on borrowing firm’s accounting conservatism. We find that bank competition leads to an increase in firm accounting conservatism. Moreover, we find that bank competition intensifies lenders’ monitoring in that banks impose more strict and intensive covenants on bank loans, and bank monitoring reduces the probability of default of borrowing firms, and thereby result in more conservative reporting of borrowing firms. Our findings are robust to alternative accounting conservatism measure C-Score and potential multicollinearity issue.

This thesis contributes to the discussion on the importance of accounting information for stock market efficiency. As any analysis of market efficiency depends on the use of adequate risk proxies, the thesis first investigates the ability of commonly used risk factors to explain the cross-sectional variation of Swedish stock returns. The findings suggest that capturing risk is indeed very complicated, as neither CAPM beta nor size and book-to-market equity ratio are significantly associated with realized monthly returns. The relative bid-ask spread is found to be the strongest of all the analyzed factors; nevertheless it does not seem to be related to momentum in the manner predicted in the conceptual argument presented earlier in the paper. Second, the thesis documents that simple accounting summary measures, such as earnings and book value of equity, can be used as benchmarks for the identification of mispriced stocks. Contrarian investment strategies (CISs) based on these simple measures earn a substantial value premium of 11-14%. In addition, we find that their effectiveness is compromised by noise that is caused by transitory earnings and accounting conservatism. Controlling for these factors increases the magnitude of the value premium and improves the consistency of the CIS. This suggests that the information reflected in key accounting measures is not fully reflected in stock prices. Finally, this thesis shows that the structure of accounting information matters too. It documents that the transition from the Swedish GAAP to IFRS in 2005 not only changed the average goodwill charges reported by companies, but also affected valuation of goodwill-intensive companies. This suggests that investors were not able to see through the conservative treatment of goodwill prior to the adoption of IFRS and that they recognized its higher persistence only after being provided with accounting information directly highlighting it. It is sometimes suggested that the structure of financial reporting “per se” should not be relevant to the valuation of companies, because the presentation itself does not the affect expected operating performance and because investors and analysts can “see through” and properly discount for various reporting formats. This thesis concludes that, contrary to this proposition, the structure of accounting does matter for equity valuation and that changes in representation do impact on stock prices.

Standard empirical models of operating efficiency (OE) and managerial ability (MA) assume a symmetric linear relation of OE and MA with firm performance. However, OE and MA metrics are likely to respond faster to a demand decrease than a demand increase due to cost stickiness and respond faster to negative returns than to positive returns due to accounting conservatism. As predicted, I find large asymmetries in the behavior of OE and MA measures. OE and MA in levels (changes) are 2.4 (1.4) times and 1.5 (1.6) times more sensitive to demand decreases than demand increases. Similarly, OE and MA levels (changes) are 1.7 (1.5) times and 3.6 (2.3) times more sensitive to negative returns than to positive returns. The incremental explanatory power of modeling asymmetry in OE and MA levels (changes) is 20.9% (39.5%) and 263.3% (27.6%), measured as incremental adjusted R2. Cross-sectionally, asymmetry in OE and MA varies with the determinants of cost stickiness: (1) asset intensity and (2) employee intensity. Moreover, the degree of asymmetry also varies with the determinants of accounting conservatism: (1) book-to-market value of equity, (2) leverage, and (3) market capitalization. In addition, demand decline (i.e., cost stickiness) and bad news (i.e., accounting conservatism) during prior and successive periods have an incremental impact on the asymmetry in OE and MA. The standard models of OE and MA do not control for these correlated-omitted variables or incorporate the cost stickiness and accounting conservatism asymmetries, which yield biased measurements and render incorrect regression estimates and inferences. / Business Administration/Accounting

Försiktighetsprincipen är och har länge varit en av de viktigaste principerna inom svensk redovisning. Sedan år 2005 måste dock samtliga börsnoterade företag i Sverige upprätta sin redovisning i enlighet med det internationella regelverket International Financial Reporting Standards (IFRS). Tidigare forskning har anmärkt att IFRS ger relativt mycket utrymme för subjektiva bedömningar och har ett komplicerat förhållande till försiktighet som är ett omdebatterat ämne bland redovisare.Att upprätta redovisning är en subjektiv process där personliga egenskaper hos redovisaren och redovisarens omgivning har en inverkan på sättet den upprättas. Uppsatsens syfte är därför att förklara vilka faktorer som påverkar tillämpningen av försiktighet i redovisning i svenska börsnoterade företag. Studien fyller ett gap kring styrelsesammansättningens påverkan på mängden försiktighet i redovisning som inte är ett väl utforskat område, speciellt angåendesvenska förhållanden. Den undersöker även hur ägarstruktur och företagets storlek påverkar redovisningen i fråga om tillämpningen av försiktighet. För att genomföra studien har nödvändig data inhämtats från databaser och årsredovisningar. Market-to-Book valuehar använtssom en metod för att mäta försiktighet i redovisning. Studienhar kommit fram till några intressanta resultat som bidragit till forskningen kring försiktighetsprincipens framtid. Det har nämligen visat sig att företag med en större andel kvinnori styrelsen tenderar att vara mer försiktiga i sin redovisning. Större företag har även påvisats visa mindre tendenser till en försiktig redovisning vilket inte var förenligt med studiens ena hypotes. Försiktighet i redovisning är ett omdiskuterat ämne och det finns fortfarande mycket kvar att undersöka. Framtida forskning skulle kunna undersöka hur andra faktorer påverkar redovisningens försiktighet exempelvis redovisares personlighetsdrag. Andra metoder borde även användas för att mäta graden av tillämpning av försiktighetsprincipen.

本文從債務契約的角度，調查穩健會計對財務報表使用者的經濟效益，以及編製者提供穩健會計的經濟誘因。穩健會計的最主要受益者為債權人，選擇債務契約係基於，首先，本研究利用當期公司穩健特性預期次期該公司契約是否違約，研究結果發現穩健會計得以使債權人及時偵知債務人倒帳風險，其次，本研究調查編製者的經濟誘因，實證結果未能發現，其中經濟誘因的變數包括銀行貸款利率、異常應計與盈餘反應係數，本文建議未來研究可繼續探求財報提供者報導穩健之誘因，並進行跨國分析比較。 / Perceived from the perspective of debt contract, this paper investigates the benefit of accounting conservatism for users of financial statements and the motivation of financial preparers to provide conservative reporting. I choose this topic because main beneficiaries of accounting conservatism are the lenders. First, I use the measure of conservatism in current period to predict the default risk of borrowers in next period. The evidences show that conservatism benefits lenders through the timely signaling of default risk. Second, I try to discover the motivation of borrowers to provide conservatism accounting through the proxies of cost of capital, abnormal accruals and earning response coefficient but fail. One potential avenue for future research is to find the economic motivation of financial reporting preparers to report conservative accountings and do the cross-country comparisons.

Syfte: Större revisionsbolag har länge inom redovisningslitteraturen associerats med bättre revisionskvalitet jämfört med mindre revisionsbolag. På senare tid har frågan lyfts om större revisionsbolags höga revisionsarvoden beror på högre revisionskvalitet eller marknadsmakt. Denna studie bidrar med nya empiriska underlag inom jämförelsestudier mellan större och mindre revisionsbolag, där revisionsarvodet används som proxy för revisionskvalitet. Vidare tar denna studie, till skillnad från tidigare studier, även hänsyn till revisionskvalitet inom olika riskmiljöer.   Metod: Studien har använt sig av en kvantitativ metod med ett positivistiskt förhållningssätt genom arbetet. En deduktiv forskningsansats har tillämpats där tidigare forskning har legat som grund till studiens framförda hypoteser. Insamling av finansiella sekundärdata för totalt 2518 företag har utförts via databasen Thomson Reuters Datastream.   Resultat & slutsats: Studiens resultat påvisar signifikanta skillnader i revisionskvalitet mellan större och mindre revisionsbolag i studiens olika riskmiljöer. För studiens europeiska länder är förhållandet mellan större och mindre revisionsbolag likvärdiga medan de amerikanska revisionsbolagen skiljer sig signifikant revisionskvalitetsmässigt.   Förslag till fortsatt forskning: Studien har genomförts utan hänsyn tagen till kvalitativa faktorer som kan komma att påverka revisionskvalitet, vilket öppnar ett utrymme för komparativa studier med en kvalitativ inriktning. Det finns även möjlighet att utöka antalet börsmarknader för respektive land eller utvidga antalet länder i olika riskmiljöer.   Uppsatsens bidrag: Studien lämnar två bidrag till redovisningslitteraturen i form av nya empiriska bevis inom revisionskvalitet mellan större och mindre revisionsbolag samt unik forskning kring revisionskvalitetsstudier mellan riskmiljöer. Vidare lämnar studiens resultat incitament åt praktiker att granska revisionsmarknaden för eget vinstintresse samt svarar på normgivares funderingar kring marknadsbalansen mellan större och mindre revisionsbolag. / Aim: Big audit firms have long been associated with higher audit quality, compared to smaller audit firms. Recent studies suggest that the higher audit fees from bigger audit firm is affected by market misuse rather than better audit quality. This study provides new empirical evidence between the comparison of big vs small audit firm, where audit fees are used as proxy for audit quality. Furthermore, this study investigates different litigation environment that could affect audit quality.   Method: This study uses an quantitative based method with an positivist, deductive approach, were earlier studies have had an impact on our hypotheses. Financial information from 2518 companies has been collected from Thomson Reuters Datastream.   Result & Conclusions: This study's result provides significant differences of audit quality between big and small audit firms in different risk environments. For this study, the audit quality relationship between big and small audit firms are equivalent for the European countries while audit quality between big and small audit firms in the US differ significantly.   Contribution of the thesis: This study leaves two contributions to the extent audit literature, in terms of empirical evidence of audit quality between big and small audit firms and unique research results of audit quality in different litigation environments. Furthermore, the results of this study creates incentives for practitioners to review the audit market for self interests and answer legal setters concerns about unbalanced audit markets.   Suggestions for future research: The study has been carried out without consideration of qualitative factors that may affect audit quality. It’s opening a space for comparative studies with an qualitative approach. It is also possible to expand the number of stock exchanges for a country or expand the number of countries in different risk environments.

Muito em função da dificuldade de se observar o dia-a-dia dos conselheiros nas empresas, acadêmicos e profissionais se questionam sobre a importância dessas estruturas de governança corporativa (GC) quando o assunto é propriedade dos números contábeis. Nesse sentido, esta tese, além de discutir extensivamente e traçar um panorama geral acerca da configuração desses conselhos no Brasil ao longo dos últimos cinco anos, estuda de que forma essas e outras estruturas de GC se associam com medidas de qualidade da informação contábil das firmas brasileiras. Diferentemente do que se encontra na literatura existente, a pesquisa em tela buscou compreender o ambiente dos conselhos de administração (CAs) e Fiscais (CFs) de modo mais amplo, avaliando de forma agregada 35 características estruturais, de remuneração e de composição desses órgãos, considerando suas inter-relações e implicações quando da associação destas com variáveis que medem as propriedades informacionais contidas no lucro. Para tanto, analisou-se, em média, cerca de 325 empresas por ano, compreendendo análise de 19.487 currículos de conselheiros de administração e fiscais efetivos no país, além da análise das outras variáveis de governança, tais como avaliação formal de conselheiros, frequência de reuniões, tipos de contrato de remuneração e outras. Posteriormente, utilizou-se a técnica de Análise Fatorial para estabelecer constructos subjacentes e não correlacionados entre si, de modo que os coeficientes estimados nas regressões posteriores fossem reportados de forma mais consistente do que se construídos com um número limitado de variáveis individuais. Cinco fatores representam mais de 80% da variância dos dados, sinalizando que características como idade, formação básica em administração, contabilidade e/ou economia, remuneração, board interlocking e tamanho do conselho formam um agregado parcimonioso de características que representam a maior parte da diversidade subjacente dos conselhos brasileiros. Destarte, após análise de simultaneidade dos modelos, testou-se as associações existentes entre as propriedades dos números contábeis e os fatores de GC por intermédio de modelos de regressão múltipla estimados pelo método OLS, e de modelos de regressão logística, controlados por setor, tamanho das empresas, adesão aos níveis diferenciados de governança, emissão de ADR, presença de CF permanente instalado, estrutura de propriedade e outros. Os resultados indicam uma associação inversa e significante entre a proporção dos membros que participam de outros CAs e os níveis de persistência, sinalizando que empresas com CAs formados por membros dedicados tendem a reportar, na média, lucros mais persistentes. A remuneração variável é positivamente relacionada com a persistência dos lucros reportados sugerindo que a remuneração dos conselheiros é influenciada pela substância econômica dos resultados divulgados e com o conservadorismo contido no lucro sinalizando para redução do conservadorismo contábil na medida em que a remuneração variável dos conselheiros aumenta. A idade também mostrou ser fator relevante para explicar os níveis de conservadorismo e de gerenciamento de resultados das firmas, sinalizando tendência de maior grau de conservadorismo e menor grau de gerenciamento de resultados por parte dos conselheiros mais velhos. Adicionalmente, não foi possível afirmar que o conselho fiscal afeta de forma positiva o monitoramento da qualidade dos lucros reportados no mercado de capitais brasileiro / In view of the difficulty of observing the day-to-day activities of members of oversight board and board of directors, academics and professionals wonder about the importance of corporate governance (CG) structures when the subject is the propiety of accounting numbers. In this sense, besides discussing extensively and tracing an overview of the configuration of these boards in Brazil over the past five years, I also study how these and other CG structures are associated with proxies of accounting information of Brazilian firms. Contrary to what has been found in the existing literature, I seek to understand the environment of the board of directors (BoD) and oversight board (OB) more broadly, evaluating in an aggregate way 35 variables of structural, remuneration and composition characteristics of these boards, considering their interrelations and implications when associated with variables that measure the informational properties contained in earnings. For that, I analyze, on average, 325 companies per year and 19,487 résumés of the members of these boards, in addition to analyzing other variables such as their attendance at meetings, compensation contracts, board interlocking and others. Later, I use factor analysis to establish underlying constructs not correlated with each other, so that the estimated coefficients in subsequent regressions are reported more consistently than if built with a limited number of individual variables. Five factors represent more than 80% of the variance of the data, signaling that characteristics such as age, basic training in business administration, accounting and/or economics, board compensation, board size and interlocking form a thrifty aggregate of characteristics that represent most of the underlying diversity of Brazilian boards. Thus, after analysis of simultaneity, I test the existing associations between the earnings quality and factors of CG through multiple regression models estimated by OLS and logistic regression models, controlled by sector, firm size, adherence to the different levels of corporate governance, issuance of ADRs, presence of a permanent OB, ownership structure and others. The results indicate a significant inverse association between the proportion of members who participate in other BoDs and persistence levels of earnings, signaling that companies with BoDs formed by dedicated members tend to report, on average, more persistent earnings. The variable pay for board members is positively related to earnings persistense suggesting that the board members compensation is influenced by the economic substance of the disclosed results and with accounting conservatism signaling a decrease in accounting conservatism when the remuneration of these members increases. The age of members also proved to be a relevant factor to explain the levels of conservatism and earnings management of firms, signaling a trend toward greater levels of conservatism and a lesser levels of earnings management of when these boards are formed of older members. Additionally, it was not possible to affirm that the oversight board has a positive effect on the quality of monitoring accounting information reported in the Brazilian capital market.

會計保守性為盈餘品質特性之一，其已存在好幾個世紀，且有證據顯示在過去三十年間有增加的趨勢，對於會計實務影響深遠。會計保守性會造成資產的帳面價值被低估，但卻提高會計盈餘品質。 本研究探討會計保守程度與財務困難機率之關係、機構投資人對於公司之持股與會計保守性之關係以及財務困難機率與機構投資人持股之關係。實證結果發現：會計保守程度較高之公司其財務危機機率較低，顯示會計保守程度較高之公司其財務體質較佳；公司財務困難機率愈低，則機構投資人持股愈多；而會計保守程度對於機構投資人之持股行為亦有顯著之影響。 / Conservatism is an attribute of accounting earnings, which has been present in accounting practice for centuries and appears to be more popular in the last 30 years. It persistently understates the book value of net assets while increases earnings quality. The influence of conservatism on accounting is significant. This study examines the relations among accounting conservatism, financial distress, and institutional holding. The findings indicate that the higher the accounting conservatism the lower the financial distress probability, the lower the financial distress probability the more the institutional holding, and the accounting conservatism has also significant effect on institutional holding.

Muito em função da dificuldade de se observar o dia-a-dia dos conselheiros nas empresas, acadêmicos e profissionais se questionam sobre a importância dessas estruturas de governança corporativa (GC) quando o assunto é propriedade dos números contábeis. Nesse sentido, esta tese, além de discutir extensivamente e traçar um panorama geral acerca da configuração desses conselhos no Brasil ao longo dos últimos cinco anos, estuda de que forma essas e outras estruturas de GC se associam com medidas de qualidade da informação contábil das firmas brasileiras. Diferentemente do que se encontra na literatura existente, a pesquisa em tela buscou compreender o ambiente dos conselhos de administração (CAs) e Fiscais (CFs) de modo mais amplo, avaliando de forma agregada 35 características estruturais, de remuneração e de composição desses órgãos, considerando suas inter-relações e implicações quando da associação destas com variáveis que medem as propriedades informacionais contidas no lucro. Para tanto, analisou-se, em média, cerca de 325 empresas por ano, compreendendo análise de 19.487 currículos de conselheiros de administração e fiscais efetivos no país, além da análise das outras variáveis de governança, tais como avaliação formal de conselheiros, frequência de reuniões, tipos de contrato de remuneração e outras. Posteriormente, utilizou-se a técnica de Análise Fatorial para estabelecer constructos subjacentes e não correlacionados entre si, de modo que os coeficientes estimados nas regressões posteriores fossem reportados de forma mais consistente do que se construídos com um número limitado de variáveis individuais. Cinco fatores representam mais de 80% da variância dos dados, sinalizando que características como idade, formação básica em administração, contabilidade e/ou economia, remuneração, board interlocking e tamanho do conselho formam um agregado parcimonioso de características que representam a maior parte da diversidade subjacente dos conselhos brasileiros. Destarte, após análise de simultaneidade dos modelos, testou-se as associações existentes entre as propriedades dos números contábeis e os fatores de GC por intermédio de modelos de regressão múltipla estimados pelo método OLS, e de modelos de regressão logística, controlados por setor, tamanho das empresas, adesão aos níveis diferenciados de governança, emissão de ADR, presença de CF permanente instalado, estrutura de propriedade e outros. Os resultados indicam uma associação inversa e significante entre a proporção dos membros que participam de outros CAs e os níveis de persistência, sinalizando que empresas com CAs formados por membros dedicados tendem a reportar, na média, lucros mais persistentes. A remuneração variável é positivamente relacionada com a persistência dos lucros reportados sugerindo que a remuneração dos conselheiros é influenciada pela substância econômica dos resultados divulgados e com o conservadorismo contido no lucro sinalizando para redução do conservadorismo contábil na medida em que a remuneração variável dos conselheiros aumenta. A idade também mostrou ser fator relevante para explicar os níveis de conservadorismo e de gerenciamento de resultados das firmas, sinalizando tendência de maior grau de conservadorismo e menor grau de gerenciamento de resultados por parte dos conselheiros mais velhos. Adicionalmente, não foi possível afirmar que o conselho fiscal afeta de forma positiva o monitoramento da qualidade dos lucros reportados no mercado de capitais brasileiro / In view of the difficulty of observing the day-to-day activities of members of oversight board and board of directors, academics and professionals wonder about the importance of corporate governance (CG) structures when the subject is the propiety of accounting numbers. In this sense, besides discussing extensively and tracing an overview of the configuration of these boards in Brazil over the past five years, I also study how these and other CG structures are associated with proxies of accounting information of Brazilian firms. Contrary to what has been found in the existing literature, I seek to understand the environment of the board of directors (BoD) and oversight board (OB) more broadly, evaluating in an aggregate way 35 variables of structural, remuneration and composition characteristics of these boards, considering their interrelations and implications when associated with variables that measure the informational properties contained in earnings. For that, I analyze, on average, 325 companies per year and 19,487 résumés of the members of these boards, in addition to analyzing other variables such as their attendance at meetings, compensation contracts, board interlocking and others. Later, I use factor analysis to establish underlying constructs not correlated with each other, so that the estimated coefficients in subsequent regressions are reported more consistently than if built with a limited number of individual variables. Five factors represent more than 80% of the variance of the data, signaling that characteristics such as age, basic training in business administration, accounting and/or economics, board compensation, board size and interlocking form a thrifty aggregate of characteristics that represent most of the underlying diversity of Brazilian boards. Thus, after analysis of simultaneity, I test the existing associations between the earnings quality and factors of CG through multiple regression models estimated by OLS and logistic regression models, controlled by sector, firm size, adherence to the different levels of corporate governance, issuance of ADRs, presence of a permanent OB, ownership structure and others. The results indicate a significant inverse association between the proportion of members who participate in other BoDs and persistence levels of earnings, signaling that companies with BoDs formed by dedicated members tend to report, on average, more persistent earnings. The variable pay for board members is positively related to earnings persistense suggesting that the board members compensation is influenced by the economic substance of the disclosed results and with accounting conservatism signaling a decrease in accounting conservatism when the remuneration of these members increases. The age of members also proved to be a relevant factor to explain the levels of conservatism and earnings management of firms, signaling a trend toward greater levels of conservatism and a lesser levels of earnings management of when these boards are formed of older members. Additionally, it was not possible to affirm that the oversight board has a positive effect on the quality of monitoring accounting information reported in the Brazilian capital market.

本研究以美國證券市場為樣本，探討銀行業之法規開放是否對非金融業公司之會計穩健性造成影響。銀行法規開放可使銀行透過併購於州內擴張，借款公司可能面對之銀行議價能力上升，使得銀行要求融資顧客出具高品質之財務報表。另一方面，州際銀行法規開放可能使他州銀行進入本地金融市場，進而提高金融市場之競爭度。由於會計穩健性可降低管理階層與投資人的資訊不對稱，被視為高品質的財務報表，本文推測銀行業法規鬆綁會影響公司會計穩健性。實證結果顯示，州內銀行法規開放與會計穩健性間具有顯著正向關係；然而，州際銀行法規開放與會計穩健性間具有顯著負向關係。額外分析顯示，大型企業更傾向於對銀行法規開放作出反應，而較依賴發行公司債融資方式之公司在面對銀行法規開放時則減少財務報導中認列損失之即時性。 / This study uses US banking deregulation as a research setting to examine whether banking deregulation has an impact on the degree of accounting conservatism of non-financial firms. Since banking deregulation would enable large banks expand within states through merge and acquisition, bank’s bargaining power to lenders may increase. Banks may require their clients to provide higher quality of financial statements. On the other hand, banking deregulation may increase competition in local financial markets by enabling large banks to enter into local markets. As accounting conservatism could alleviate information asymmetry between management and investors, and serves as an indicator of high quality of financial statements, I expect that banking deregulations would have an impact on firms’ reporting conservatism. I find that there is a significantly positive relationship between intrastate banking deregulation and firms’ timely loss recognition. In contrast, interstate banking deregulation has a significantly negative influence on firms’ timely loss recognition. Additional analyses reveal that large firms are more likely to respond to the banking deregulation and that firms that rely more on public debt financing respond to the banking deregulation by decreasing their degree of timely loss recognition in financial reporting.