o ensure the quality of information, quality standards are required by regulators in the Brazilian market, primarily via the disclosure of companies' financial statements. To ensure that the information reported by financial statements is truthful, external auditor examinations are conducted. The independent audit of the financial statements provides users with an assessment of the information disclosed by the company, informs their investment decisions, and encourages transparency among the company management (Jensen;Meckling 1976;Chow 1982). In this context, the audit report becomes a source of information that can be used by stakeholders in their decision-making processes. The audit report is also a formal, legal, and unbiased report that validates and ensures the regularity and reliability of the financial statements produced and disclosed by the companies (Araújo 2003).

Therefore, the present study seeks to answer the following question: is a company's variance of the average share price as listed in the Brazilian capital market related to the level of information disclosed in audit reports? The motivation for this research originates from the lack of studies on Brazil's capital market related to disclosure using an independent audit report as a source for analyses.

Since the 1990s and due to the economic globalization process, there has been an increase in foreign investors in Brazil's capital market. Additionally, several Brazilian companies have entered foreign markets, listing their shares on foreign stock exchanges, such as the New York Stock Exchange, as American Depositary Receipts (ADRs). These changes have forced Brazilian companies to improve their management practices, including the development of corporate governance.

To encourage better corporate governance practices in the Brazilian market and boost investor confidence, some institutional and governmental initiatives have also been implemented. Notably, corporate law No. 6.404/76, which instituted modern rules to govern publicly listed companies, and law No. 6.385/76, which created the Securities Commission (Comissão de Valores Mobiliários -CVM), were implemented. The creation of the New Market and of Levels 1 and 2 of Corporate Governance by the São Paulo Stock Exchange (Bovespa) represents more milestones in this process.

The CVM operates on matters related to Brazilian corporate law (6.404/76) and is empowered to discipline, regulate, and monitor the activities of the several members of Brazil's securities market: the publicly listed companies, the financial intermediaries, and the investors. Law 6.404/76 (amended by Law 11.638/07) stipulates that at the end of each fiscal year, each company's board will draft financial statements that should clearly express the company's net worth and the changes that occurred in that fiscal year. These statements should be complemented with explanatory notes and the other analytical tools necessary for the clarification of assets and earnings for that period. In addition, "the financial statements of publicly listed companies will follow the regulations implemented by the CVM and will be audited by independent auditors" (art. 177 § 3rd -amendment introduced by Law 11.941/09).

Through the preparation of financial statements, it is possible to obtain information regarding an entity's financial position, assets, earnings, and financial flow, which is useful for a wide range of users to make decisions. These statements should show the management of resources entrusted to the administration, according to CVM resolution No. 488/2005. Disclosure of corporate information in the capital market, whether voluntary or compulsory, enables the reduction of informational asymmetry between the parties and the economic agents involved.

Contractual relationships may be considered the essence of a company (Jensen and Meckling 1976). These relationships are made not only with investors, but also with the employees, customers, and suppliers, among other economic agents. The conditions established in these contracts form the basis of the organization (Alchian; Demsetz 1972). Not all of the economic agents possess the same amount of information regarding the company. The uneven access to information gives rise to a situation of information asymmetry between the economic agents (Jensen and Meckling 1976). Considering the capital market environment, the problem lies in the fact that the agent (management) has insights that are hardly observable by the principal (shareholder/investor).

The different levels of information that exist between the economic agents create opportunities and conditions that would limit the market's efficiency (Arrow, 1963). In this context, the informational asymmetry establishes implications for adverse selection (ex ante opportunism) and moral hazard (ex post opportunism) (Macagnan, 2005). The problem of adverse selection occurs before the signing of a contract, in which one of the parties holds more information on the transaction than the other party (Akerlof 1970;Rothschild;Stiglitz 1976;Mishkin 1991;Kreps 1994;Cutler and Zeckhauser 1997). The problem of moral hazard would be an opportunist post-contract consequence related to the difficulty in observing whether the actions of a particular party are aligned with the agreement (Milgron and Roberts 1992). This type of situation could lead to economic inefficiency in the capital market (Akerlof 1970).

Beginning from the assumption that all economic activity can be reduced to a series of bilateral contracts, which can be made or broken at any time by either party (Jensen; Meckling 1976), the problems of moral hazard and adverse selection could generate costs. These costs include the following: the principal monitoring of the agent; the agent demonstrating that their behavior is line with the behavior desired by the principal; and insurance and residual losses (Jensen;Meckling 1976). To minimize these problems, the agency theory proposes the disclosure of information, as well as the hiring of external auditors, which could improve market transactions (Verrecchia 1999; Healy and Palepu 2001). The agency theory stresses the significance of information disclosure as a way to mitigate both the informational asymmetry between the principal and the agent, as well as the problems with the high costs of monitoring the agents' actions to ensure that they comply with institutional standards. Corporate disclosure becomes essential in ameliorating the problems related to moral hazard and adverse selection derived from informational asymmetry.

Information Disclosure to the Market

The Financial Accounting Standards Board (FASB, 2008) considers relevant information to be information that can "make a difference" in a future decision. In addition, relevant information should allow users to make predictions concerning the outcome of observed events and confirm or correct a previous expectation, which is called the predictive value and feedback value of information. Accordingly, the disclosure of financial, economic, and management information through statements could serve as a decision-making aid, helping to set goals for attracting new resources and risk management, thereby increasing investor confidence. This confidence would raise the likelihood of investors to purchase company shares. Risk reduction also would generate better stock pricing, which in turn encourages new capital openings and strengthens the stock market (Levitt 1998). In particular, the disclosure of information through audited financial statements (external and independent audit) could be regarded as a means of reducing agency costs (Jensen;Meckling 1976;Chow 1982).

The studies of Ball and Brown (1968), Chow and Wong-Boren (1987), Cooke (1989aCooke ( , 1989b)), Raffournier (1995), Patton and Zelenka (1997), Owusu-Ansah (1997), Leventis and Weetman (2000), Piacentini (2004), and Zhang (2008) indicated the significance of disclosure in the market. Furthermore, Firth (1979) and Wallace, Naser, and Mora (1994) noted that audit firms have an influence on the content shown in companies' annual reports. The larger and more recognized the audit firm, the greater this influence. Studies by DeAngelo (1981) and Inchausti (1997) reported that the large audit firms have incentives to provide a higher level of auditing quality through information disclosure in their reports and also that they risk losing their prestige if they are linked to clients with poor disclosure practices. Craswell and Taylor (1992) suggested that the audit firm choice is likely to be associated with the decision to disclose more or less information.

The examination of financial statements by external auditors promotes the disclosure of information that is sufficient to guide users' decisions (Jensen and Meckling 1976;Chow 1982 andHealy andPalepu 2001). The hiring of external auditors is a way to signal

Volume XIV Issue V Version I Year ( ) D the good practices of the company's management to foreign investors (Jensen and Meckling 1976). The hiring of these services originates from a market demand to increase the credibility of financial reports (Healy and Palepu 2001).

Regarding the Brazilian market, the goal of an accounting audit is to increase the users' degree of confidence in the financial statements. This process is achieved by an audit report stating that statements meet the regulatory requirements (NBC TA 200 CFC Resolution N. 1.203/09). The audit report would be the end product of the auditor, a formal means of communication that provides information to stakeholders.

The Brazilian Accounting Standard NBC T 11/05 that was promulgated by the Federal Accounting Council (Conselho Federal de Contabilidade -CFC, 2005) states the form and content guidelines for reports issued on the financial statements of organizations by independent auditors and clarifies the main components that must be included. When issuing an opinion, the auditor assumes technical and professional responsibility for the report that will then be directed to shareholders, investors or partners, and the board of directors, or the company equivalent. In certain circumstances, the audit report will be distributed to contracted service providers.

Depending on the nature of the opinions expressed in the audit report, the report is classified as one of the following: a (1) report without caveats; (2) report with caveats; (3) adverse report; and (4) report omitting opinion. In any of the referred reports, the auditor should identify the following information: (a) the financial statements on which the opinion is based; (b) the name of the entity and the dates and periods reported; and (c) if the work is performed by an audit firm, the auditor should indicate the name and the number under which the firm is registered with the Regional Council of Accountancy. The independent audit report essentially consists of 3 parts: (1) identifying the financial statements and defining the responsibilities of the directors and auditors; (2) describing the extent of the work; and (3) expressing the auditor's opinion on the financial statements. Finally, the audit report must be dated and signed by the accountant responsible for the work (noting their registration number with the Regional Council of Accountancy), and the date should correspond to the date of the audit report closure.

The present study is based on the premise that the disclosure of information to the market through audit reports is considered to be a form of informing stakeholders on the legal, financial, and economic situation of the organization, thereby enabling stakeholders to support a company's future plans by the buying or selling stocks. By examining the variation of prices before or after the report publication, it is possible to determine if the values reflect the type and the degree of information disclosed in the market or if the report does not reflect relevant information. Therefore, the research hypothesis presented in this study is the following: H1: the information disclosed in the audit report is related to the variation in the average value of company shares.

Therefore, to model estimates, we employed a multiple linear regression model using the ordinary least squares (OLS) method and panel data. The model proposed and tested in the present study was exploratory because it has not been used in similar studies. The econometric model is specified below:

The dependent variable (MVVit) is the variance of the mean value of the companies' share price. For this variable, the daily share prices of the companies were obtained the official website of the Brazilian stock exchange 60 days before and 60 days after the publication of the audit report. With these daily stock prices, it was possible to obtain the average value for the periods before and after the report release. Preferred and ordinary shares were considered together because their variations were quite similar.

The choice of the period for the event study was based on the audit report publication date for each company's earnings statement. The magnitude of the event was 120 days, beginning 60 days (-60) before the event (0) and ending 60 days (+60) after the event. The days considered when calculating the arithmetic mean were only those days that trading occurred in the Bovespa.

The formula used to calculate the arithmetic mean of the values is as follows:

where Xa is the mean share price prior to the event, Xp represents the mean share price after the event, ?Sp( -60) is the sum of the values of the daily share prices in the period of 60 days immediately before publication of the audit report, ?Sp(+60) represents the sum of the values of the daily share prices in the period of 60 days immediately after publication of the audit report, and N is the number of days in which the shares were traded within the period of 60 days.

Once the means for companies were calculated for the years analyzed, the next step was to verify the variance in the mean share prices (increase or decrease) after the publication of audit reports. To do so, we used the following formula: MVV=?(Xp(ON)+Xp(PN)) -? (Xa(ON)+Xa(PN)) , (4) where MVV is the mean value variance, ?(Xp(ON)+Xp(PN)) represents the sum of ON and PN share mean value after the event, and ?(Xa(ON)+Xa(PN)) is the sum of ON and PN share mean value before the event.

The explanatory variables, ?1EDit (discontinuity indicators), ?2EAPit (accounting practices indicators), and ?3EDEit (indicators of other events), were developed as described below. First, analyses of the audit report contents were performed to identify the disclosed information and create the respective indicators. To construct these indicators, the primary aspects described by the auditors in their caveats were extracted, as well as all of the content described in the report, including paragraphs on relevant information presented in the series of caveats. The decision to include these paragraphs was made because the audit report should be analyzed in its entirety and in view of the (impartial) informational content presented because of the significance given to that document by the legislation (Law 6.404/76 and subsequent amendments), regulators (CVM and CFC), and the capital market, which requires it in their own regulations. The starting point for the construction of indicators was the NBC T 11.10 -CFC Resolution N. 1.037/05 (CFC 2005). This resolution dictates the main aspects that are to be noted by the auditors regarding evidence of the audited company's operating discontinuity.

From the content analyses of the audit reports, 37 new indicators were created. In conducting the research, the 19 NBC T 11.10 operational discontinuity indicators and the 37 indicators created were considered together for a total of 55 indicators. After defining the indicators, categories of indicators were created that allowed for the level of information disclosure to be measured. These indexes are illustrated in Figure 1. In the category "indicators of accounting practices", we aimed to address the main information disclosed in accounting practices, such as evaluations, bookkeeping, provisions, adjustments, reclassifications, divergences, etc. This category covers information on the assets and liabilities of the companies, as well as on the financial statements and the limitations of audit procedures. There are 16 indicators of information disclosure on the current or non-current assets and 3 indicators of disclosure related to the company's current or non-current liabilities. In addition, a sub-category, "other financial information", was created with 5 indicators. Another category was created to disclose a company's strategic and administrative information and relevant facts that could somehow impact the current or future earning; this category was "disclosure indicators -other events" and included 12 indicators. After the construction of the indicators, disclosure indices were calculated for each of the 33 companies. The 55 analytical indicators are presented in Appendices I and II.

A set of categories representing the indicators was created to measure the level of information disclosure. These categories are represented with the indexes calculated for each company using the following formula: where nj is the number of indicators expected for each company, j is the sub-index of each company, and i is the number of indicators. If the indicator Xij is disclosed, it assumes a value of 1; otherwise, the value is 0.

Control variables were inserted in the model and were considered to be indicators capable of aiding investment decisions. Therefore, these variables can influence the stock value (increase or decrease). The control variables are the total assets (?4TA it ), net income (?5NI it ), net profit (?6NP it ), equity (?7Eq it ), indebtedness (?8Ind it ), beta (?9Beta it ), general liquidity (?10GL it ), and cash-yield per share (?13CashYield it ). These values were obtained from the Economática database. The economic control variables were the exchange rate (?11EX_R it ) and the gross domestic product-GDP growth rate (?12GDP it ), obtained from the Brazilian Institute of Economics (Instituto Brasileiro de Economia (IBRE)) of the Getúlio Vargas Foundation website. Established in 1951, the IBRE focuses on the production and dissemination of macroeconomic statistics and applied economic research and pioneered calculating the Brazilian GDP. The company's sector of activity (?14SECTOR it ) was also included as a control variable.

The study population was composed of 255 publicly traded companies with shares traded on Bovespa's traditional market. Of this population, 44 financial entities were excluded because they operate in regulatory and competitive environments that are different from the other companies. Of the remaining non-financial companies, 211 were selected because they disclosed at least one report with caveats over a period of 5 years (2005 to 2009), which resulted in a final sample of 33 companies and a total of 165 observations. VII.

The research results are presented below using two methods of analyses: (1) descriptive and correlation analyses and (2) linear regression.

Table 1 reports the descriptive analyses results of the explanatory and control variables in the econometric model. The means of indicators are based on the content analyses of the company's audit reports, XD (indicators of operational discontinuity), XDE (indicators of other events), and XPC (indicators of accounting practices), and also of the control variables, TA, beta, cash-yield, Ind, GL, NP, GDP, Eq, NI, and EXR. Among the explanatory variables, the disclosure of accounting practices (XAP) has the highest incidence (0.9757), followed by the disclosure of operational discontinuity indicators (0.8727).

The correlation analysis allowed for the opportunity to identify the multicollinearity relationships that occur with 2 or more independent variables, as shown in Table 2: net income (NI) and total assets (TA); disclosure of accounting practices (XAP) with disclosure of operational discontinuity (XD); disclosure of accounting practices with disclosure of other events (XOE); equity (Eq), and net profit (NP). The variables excluded from the model were XD (disclosure of operational discontinuity), XOE (disclosure of other events), TA (total assets) and NP (net profit) because they had large P-values (probability). The variables TA and NP, although they were proposed to measure the influence of size, were not included in the regression because they did not contribute to the model.

To achieve a model with significant explanatory variables, a significance level of up to 10% was considered. Under this condition, the following variables were excluded: equity (Eq), indebtedness (Ind), and general liquidity (GL) because they were not significant. However, the only relevant variable continued to be disclosure of indicators of accounting practices (XAP), and this variable became more significant (near 0.10) when controlling for the sector of activity.

Table 3 shows the results obtained with the regression model using the ordinary least squares method to identify the relationship between the explanatory variables, the level of information disclosure in the audit reports, and the dependent variable, variation in the mean company share price. The indicator of disclosure of accounting practices has a coefficient of 0.7454, which indicates that for each unit of the XAP variation, the average value of shares varies 0.7454 (i.e., it explains variance sensitivity between the variable (Y) and the variable (XAP) by 0.7454 or 74.54%). This positive relationship indicates that as companies disclose more information for that indicator, the change in the average value of its shares on the market will be positive, thus confirming the agency theory discussed in this study.

The control variables that remained in the model, beta, cash-yield per share, GDP growth rate and exchange rate, had P-Values of 0.002, 0.0757, 0.0010, and 0.00, respectively. Beta had a negative relationship with the dependent variable, and the other control variables had a positive relationship, as expected from Equation 4.

The negative relationship demonstrated by beta indicates that if this variable increases 1 unit, the variation of the average share price will be -1.19 during the period examined. An increase in the GDP growth rate of 1% per year will be reflected in an increase of 0.2675 units in the calculated mean share price variation. An increase in the exchange rate of 1% per year results in an increase of 7.67 units in the calculated the mean share price variation. An increase of one unit in cash-yield per share yields an increase of 0.9735 units in the calculated mean share price variation.

The present study aimed to identify if the disclosure of information through independent audit reports is correlated with variation in the mean prices of stocks traded after the disclosure and was based on a sample of traditional companies listed on Bovespa (from 2005 to 2009).

The results suggest that greater disclosure via an audit report and when considering a specific indicator provides investors with a better assessment of the company, which can reduce informational asymmetry and raise the level of confidence in the audited company.

The indicators of operational continuity and other events were not significant to the model and were therefore excluded. Thus, it is not possible to know whether this information has any relevance to users because it did not explain the mean share price of the companies studied.

The research reveals that statistically, the only indicator relevant to the model is the disclosure of accounting practices, and therefore, disclosing such information in the audit reports increases the variation in the mean share price of the companies (listed in the sample) after the audit report is disclosed.

This research does not claim to be a comprehensive review of this subject given Brazil's current regulatory scenario, in which information is disclosed through financial statements and audit reports to meet the standards that were based on the international market.