The Effect of Readability of Annual Reports and Value Relevance of Financial Information on Agency Costs with Analyst Coverage as Moderating Variable

This study aims to determine the effect of readability based on the length of the annual report and the value relevance of the financial information on agency costs. The sample used in this study were 263 firm-year from Kompas100 index. Data were analyzed using multiple linear regression method. The result of this study indicate that the higher the number of pages, words and characters, which reflects the poor readability of the annual report, has a negative effect on the asset turnover ratio, which is an inverse proxy for agency costs. Furthermore, the presence of the analyst coverage variable is able to moderate the positive effect between the number of pages, words and characters in the annual report on the asset turnover ratio. However, no significant effect was found during both test between value relevance to asset turnover ratio and that are moderated by analyst coverage.


INTRODUCTION
All public companies listed on the Indonesia Stock Exchange (IDX) are obliged to disclose company information through annual reports. The information from the annual report allows investors to understand the company's financial condition, performance and cash flow. Through the available information, investors can evaluate the growth and competence of the company's management [11]. Currently, annual reports are starting to develop and contain more information in the form of narrative texts. Annual reports with narrative texts are able to represent an average of 80% of mandatory disclosures from the company's annual report than quantitative data [19]. The increasing use of narrative information that is too much has become a concern for investors and capital holders.
Christopher Cox, who is a former chairman of the board of the SEC stated that investors are turning away from reading the annual report because of the increasing use of verbosity and jargons [20]. This problem was also mentioned by Warren Buffet who revealed that he often had difficulty understanding what was said in the annual report. He also speculated that the writers and authors of the report did not understand what they were writing or even deliberately confused the readers of the report [20].
The shift in the annual report which is increasingly dominated by narrative information makes the report longer. The annual report also includes more professional terms, special notes and nonfinancial information that makes it difficult for investors to understand its content [11]. This condition causes information from within the company is not conveyed properly and causes the emergence of information asymmetry between management and stakeholders. As a result, investors will find it difficult to effectively control the actions taken by management through the information presented in the annual report, resulting in agency costs due to management behavior that prioritizes or maximizes the personal interests of managers rather than companies and investors.
Based on the description above, it is very important if the information in the annual report is easy to read, understandable and the level of readability is measurable. A letter form Otoritas Jasa Keuangan (OJK) of the Republic of Indonesia number 16/SEOJK.04/2021 part 2 point 3 states that the annual report must be made in such a way that it is easy to read. The importance of the role of report readability is also mentioned in PSAK No. 1 (2019) regarding the presentation of financial statements point 17 b which states that presenting information, including accounting policies, by presenting information that is relevant, reliable, comparable and easy to understand. According to Chall readability is a combination of various factors that involve interest, clarity and is easily understood by the reader [11].
There are many measurements of readability that apply in the world and are often used in several studies. However, Indonesian and English are two different languages so that commonly used measurements such as the Fog Index, Flesch Reading Ease and Flesch-Kincaid Indices are not appropriate in measuring the readability of Indonesian report. In this study, the researcher used the most basic measurement based on the length of the annual report in the form of the number of pages, words and characters in measuring the readability level of the annual report [11].
Apart from various measures of readability, there are several studies that have revealed the consequences of readability on agency costs. One research states that companies with a higher level of readability of annual reports will have lower agency costs [11]. There is another study which shows that readability based on the length of the annual report does not have a significant relationship with agency costs [3]. Therefore, this research is important in testing and analyzing this relationship.
In addition to the readability of the annual report, the quality of accounting information also has its own role in the agency relationship and agency costs. Good-quality financial reporting helps investors to better assess firm value, performance and to make good investment decisions [1]. The relevant accounting information will allows investor to measure firm value properly so that investors able to make their own predictions rationally about the future return on security [15]. Therefore, the function of accounting information presented in financial reporting depends on the quality of financial reporting itself, which in this study uses value relevance.
The value relevance of an accounting information will demonstrate the effectiveness and usefulness of the information in making optimal economic decisions for investor. The results of research in 2016 state that there is a negative relationship between the relevance of information and agency costs [14]. These results prove that if the value relevance of information, which is one part of the quality of information, is better, it will reduce agency costs and reduce costs in decision making. Although previously stated research found a negative relationship between relevance and agency costs, the researcher considers it important to test both relationships because the relevance model used in the current study uses the Price Model [17]. This model is able to investigating the association of accounting information such as earning per share, book value and operating cash flow on company's stock prices [9].
Readability and value relevance are very important so that the purpose of the annual report, which is the delivery of company information, can be achieved and is useful for report users in making decisions. According to research in 2018, companies with more readable annual reports have better quality of information disclosure, can reduce the level of information asymmetry faced by shareholders and help them better at monitoring management activities [11]. Furthermore, relevant accounting information is information that is useful in decisions making for investors. The point is that when investors use existing accounting information as a basis for making decisions, it means that investors believe in the truth of the information [12].
Based on the results from previous research [11] and [14], and various other studies, it can be concluded that it is very important to examine and analyze the relationship between readability of annual reports and value relevance to agency costs. This study used asset turnover ratio (ATR) to proxy the level and existence of agency costs within a company. Although asset turnover ratio considered an inverse proxy of agency costs, asset turnover ratio can provide a relative and quantitative measure of the effectiveness of firm investment decisions and the ability of the firm's management to direct asset to their most productive use [6].
In addition to the readability of annual reports and value relevance, corporate governance mechanisms can also reduce management's opportunistic behavior and reduce agency costs. The governance mechanism referred to in this study is analyst coverage. There is an alternative view that analyst is able to deter misreporting and control misbehavior within a company by servings as monitors alongside traditional corporate governance mechanisms [8].
Analyst have the means and motive to be monitors of management behavior. Unlike investors and most stakeholders, analyst are trained to analyze the financial information produced by companies and they have privileged access to management in the company. Furthermore, analyst could look reckless and their reputation suffer if their reports and recommendations were based on manipulated number [4]. A research in 2006 argue that there is a negative relationship between the number of analysts in a company and the level of asymmetry faced by the company, it will also affect the company's funding decisions [2]. Furthermore, the test results are also able to show that companies with more analyst coverage tend to prefer to issue equity compared to debt [2].
Based on the description above, this study aims to examine and analyze the effect of readability of annual reports and value relevance with the analyst coverage mechanism as a moderating variable on agency costs in Indonesia. Most of readability related studies are based on the context of the English speaking countries, and only few have explored the economic consequences of annual report readability in Indonesia. This studies also mainly focus on the role of analyst coverage as a financial and information intermediaries between investor as a principal and management. Due to the presence of analyst, investor will have a better understanding of management performance and competence through financial and non-financial information in the annual report, reducing asymmetric information between the two parties and consequently reducing the agency cost that may happen because of the agency problem.
The study took a sample of companies included in the Kompas100 index on the IDX during the period 2016 to 2019. The sample selection was carried out considering that the companies listed on the Kompas100 index already cover several types of sub-sectors from various types of industries. According to Kompas.id, companies that are members of Kompas100 are companies with stocks that have good fundamentals and performance. Not only that, the company has a market capitalization value of around 70-80% of the total capitalization value of all shares traded on the IDX. The use of the Kom-pas100 index is expected to be able to represent the role of readability and value relevance in the annual reports of all companies listed on the IDX during 2016 to 2019.

The effect of readability of annual reports on agency costs
In agency theory, there is a separation between ownership (capital holders) and control of a company that is run by management. But there is an assumption in agency theory which states that all individuals have a tendency to act in their own interests, even though there is a responsibility to the principal. One of the reasons is because agents have better and more complete information about the company's values and business processes than external parties such as investors, capital holders, financial and non-financial institutions, and so on. If investors do not have access to the same information that management has, investors will find it difficult to value new and innovative investment, as well as difficulties in assessing the inherent risk in company's operations and the company's current position [18].
The existence of financial reports is expected to convey information to shareholders who do not have the same information as the company's management. Therefore, the clarity of financial statements is important so that investors can understand the information contained in these financial statements and can be used for decision making [20]. [11] argues that easier-to-read annual reports have better disclosure quality, which can reduce the level of information asymmetry faced by shareholders and help shareholders to monitor management better. It can be said that an easy-to-read annual report can be useful in reducing the asymmetry of existing information, thereby increasing oversight of the company's management performance and reducing agency costs.
The results of research in 2018 states that there is a negative and significant relationship between readability and agency costs. Companies with a higher level of readability of annual reports have lower agency costs [11]. From the explanation above, the first hypothesis to be tested is as follows: H1: companies with higher readability on annual report will have lower agency costs.

The effect of value relevance on agency costs
In agency theory, there is a separation between capital holders and the control of a company that is run by management. But there is an assumption in agency theory which states that all individuals have a tendency to act in their own interests, even though there is a responsibility to the principal. These actions and uncertainty factors cannot be observed directly by investors so that an accounting mechanism is needed to report and monitor the performance of these agents [14]. Through research [13] related to agency theory, it is revealed that information disclosure can help reduce agency conflict.
The capital market relies on credible financial information [1]. A good quality annual reports and financial reports can help investors measure the value and performance of the company so that better investment decisions can be made. However, the existence of agency problems can result in the content of accounting information in financial statements being useless, so that this can weaken the relationship between accounting information and the value of company shares. Therefore, the function of accounting information presented in financial statements depends on the quality of financial reporting itself. This is also supported by [1], who argues that there are three important elements of earnings quality, namely relevance, earnings management and audit quality. Relevance is one of two basic qualities that can make accounting information useful for decision making [10]. The relationship between relevance to agency problems has also been explored in several literatures such as: Namazi [14]. It can be said that relevance, which is one of the basic qualities of financial reporting, can be an important factor in decision making and has a negative relationship with information asymmetry, thereby reducing agency costs.
The results of previous research in 2016 state that there is a negative and significant relationship between earnings quality criteria, one of which is relevance to agency costs [14]. From the explanation above, the second hypothesis to be tested: H2 : Companies with a higher level of value relevance will have lower agency costs.

Analyst coverage strengthens the relationship between readability of annual reports and agency costs
Disclosure of information in the form of annual reports is an important communication bridge between management and capital holders due to the separation between ownership and control of the company [11]. Furthermore, [11] argues that easierto-read annual reports have better disclosure quality, which can reduce the level of information asymmetry faced by shareholders and help shareholders to monitor management better. In addition to readability, analyst coverage is also negatively related to information asymmetry faced by companies [2]. According to [2] financial analysts can play an important role in information asymmetry because analysts are able to collect complex information and convert it into a form that is more easily understood by investors.
The relationship between analyst coverage and information asymmetry has also been investigated by [2] which states that companies that have a lot of analyst coverage prefer to issue equity rather than debt and companies are not too dependent on favorable market conditions (Market timing behavior) to issue equity. If there is less analyst coverage in the company, which causes information asymmetry within the company, it may open incentives and opportunities for management to take market conditions into consideration to issue equity, especially when there is an overvaluation [2]. Furthermore, [11] states that the negative relationship between readability and agency costs is stronger in companies that have a moderating variable in the form of corporate governance, one of which is analyst coverage. From the explanation above, the third hypothesis to be tested: H 3 : Analyst coverage strengthens the negative effect between readability and agency cost.

Analyst coverage strengthens the relationship between value relevance and agency costs
Financial reporting can help users who do not have the ability to obtain all the financial information they need from an entity, so that these users rely on the information provided in financial statements [10]. The providers of capital (both investors and creditors) of course want the information contained in the annual reports to be useful in decision making, so that basic qualities are needed in the annual report, one of which is the value relevance. It can be said that relevance, which is one of the basic qualities of financial reporting, can be an important factor in decision making and has a negative effect on agency problems in the form of information asymmetry, thereby reducing agency costs.
In addition to value relevance, analyst coverage is also negatively related to information asymmetry faced by the company [2]. According to [2] financial analysts can play an important role in information asymmetry because analyst coverage is able to provide information that is not widely known by market participants, such as the results of discussions and meetings with management or the results of visits to the company's operational sites. This role is very important for investors and market participants in obtaining information and monitoring management performance. It can be concluded that the existence of analyst coverage is able to strengthen the negative relationship between value relevance and agency costs because of the services offered from the analyst coverage. From the explanation above, the fourth hypothesis to be tested:  The sample in this study contained 288 firmyear within the 4 year observation period. In order to fulfill the normality test so that a significance value ≥ 0.05 was obtained there were some outlier data the were excluded, as a result, the total remaining sample used in this study was 263 firmyear samples.
The independent variables in this study were the readability of the annual report (K1, K2 and K3) and value relevance (R). The dependent variable in this study is agency costs (ATR). The moderating variable in this study is analyst coverage (A). The control variables in this study are size, leverage and profitability.
The following is the measurement of each variable in this study: • Size, Leverage dan Profitability Natural logarithm of company's total assets at the year t Company's total debt divided by total assets at the year t Company's net income divided by total assets at the year t To test the four hypotheses in this study, the researcher used a multivariate dependency analysis technique with multiple linear regression analysis. Multivariate analysis with multiple linear regression method was used because this study used numerical data and in this study involved 4 independent variables and 1 dependent variable, where there is correlation between the independent variable and the dependent variable. So the researchers made an equation model to measure whether the 4 independent variables that existed had an effect on the dependent variable. The following is a model of multiple linear regression: = + 1 + 2 + 3 + 4 ( X ) + 5 ( X ) + 6 + 7 + 8 + = Agency cost = Readability of the annual report (K1; K2; K3) = Value relevance = Analyst Coverage = Company's total assets = Company's total debt divided by total assets Profitability= Company's net income divided by total assets = error term

RESULTS AND DISCUSSION
The results of descriptive statistics in this study can be seen in table 2.  The proxy of the readability of the annual report in this study uses the natural logarithm of the number of pages (K1), the natural logarithm of the number of words (K2) and the natural logarithm of the number of characters (K3), where these three numbers can reflect the length and amount of information contained in the annual report.

Value Relevance (R)
The proxy of value relevance (R) in this study uses the regression equation from the Price model [17], where this figure can reflect the effect of information on the current annual report (confirmatory value) to predict the value of the company in the future (predictive value). The

Size
The proxy of the control variable Size in this study uses the natural logarithm of the company's total assets, where this number can reflect the size of Kompas100 companies. The results of the coefficient of determination can be seen in the table below: To test the four hypotheses in this study, the researcher used a multivariate dependency analysis technique with multiple linear regression analysis. The results of the t-test of the regression model (1) using the HAC (Newey-West) covariance method can be seen in the table below: The results of the t-test of the regression model (2) using the HAC (Newey-West) covariance method can be seen in the table below : The results of the t-test of the regression model (3) using the HAC (Newey-West) covariance method can be seen in the table below : The test results from the regression models 1, 2 and 3 consistently state that the more pages, words and characters have a negative effect on the agency costs variable. The higher number of pages, words and characters in the annual report or in other words the readability of the annual report is low, it will make it more difficult for users to obtain valuable information (requires more time and costly), the content of the information becomes tedious, and difficult to understand, and eventually reduce the reader's interest. As a result, it can increase information asymmetry towards stakeholders, especially investors and reduce corporate information transparency.
Shareholders will have little information about management competence and performance, which results in weak oversight of management [11]. Instead of maximizing the value of the company and shareholders by making the company's assets more productive, management may bring up their opportunistic behavior and maximize their own interests by way of inefficiency in investing, controlling operational inputs and outputs, and other excessive consumptive behavior, hence company's asset turnover ratio will be negatively affected. All of these management actions will cause agency costs. The amount of these costs is equal to the value lost by management who maximize their own interests rather than the value of the company [21].
The results of this study are also in line with the reference research conducted by [11] argue that the more information is disclosed in the annual report is directly proportional to the length of the report, resulting in users of the annual report having a hard time to extract valuable information. Furthermore, a long annual report will reduce the interest of its readers, especially with the availability of mass information that is easier to understand about the company. Also stated in [11] that readability is a combination of various factors that involve interest, legibility and ease of understanding by the reader The test results from the regression models 1, 2 and 3 consistently state that there is no significant relationship between value relevance and agency costs variable. Although the calculation of value relevance in this study uses a Price model that involves financial information such as book value per share, earnings per share and cash flow per share which can influence decision making on its use, but when carrying out the supervisory function on management, financial information alone it is not enough. Secondly, financial information such as book value, earnings and cash flow is closely related to and contain management's discretionary accrual policy. Furthermore there is a lot of non-financial information that is closely related to management performance that cannot be ignored nor be underestimated, such as: information contained in management reports (reports from the board of directors), information contained in company profiles (vision, mission, and corporate values/culture; company milestones; company strategy; management profile; awards and certificates; and event highlights), information contained in MDA, other non-financial information related to intellectual capital, and so on.
Our explanation above is supported by the results of 3 regression models where value relevance does not have a significant relationship with the asset turnover ratio which reflect the management performance and an inverse proxy of agency costs. Second, the readability of the annual report which has a negative and significant relationship in this study also shows the importance of non-financial information in relation to management performance and agency costs. Third, the mean of value relevance in descriptive statistics is only 37% which shows the ability of financial information to verify the relationship between the financial information (balance sheet, income statement and cash flow) with the company's performance and market value is only 37%, so that the effect of the financial information is insufficient. The results of the current study of course contradict the results of the reference research [14], where the results of this study found a negative relationship between relevance and agency costs.
The test results of regression models 1, 2 and 3 consistently state that the number of pages, words and characters moderated by analyst coverage has a positive and significant moderating effect on agency costs variable.
One particular form of information intermediaries is analyst coverage. Analyst coverage specifically has knowledge at the company level and industry level, so they can assess the company's performance presented in the annual report and make adjustments. Furthermore, analyst coverage is able to collect complex information and transform it into a form that is more easily understood by investors in the form of analyst reports and recommendations from analysts. So that even though the annual report (number of pages, words and characters) is long, analyst coverage can summarize the information so that it is easier to read and understand with a good level of readability. As a result, the time, effort and costs required by investors are smaller and more efficient, speeding up the decisionmaking process, shareholders will have a better understanding of the performance and competence of management and supervision of management through the information presented in the annual report runs better.
The results of this study are also consistent with [11], where analyst coverage is a governance mechanism that can help strengthen the effect of readability of annual reports in reducing agency costs. The annual report is the main communication bridge between management and investors. So that a mechanism is needed in the form of financial and information intermediaries to increase the credibility of the information contained in the annual report. This mechanism is able to provide added value, especially to investors' knowledge and understanding regarding the company's current and future performance [18].
The test results of regression models 1, 2 and 3 consistently state that the value relevance moderated by analyst coverage does not significantly moderate agency costs variable. Similar to the discussion in the second hypothesis, investors in the decisionmaking process do not depend solely on financial figures, but non-financial information is also taken into consideration. The importance of non-financial information allows users of annual reports and analysts to know the strategy and performance of the company's management and examine the company's economic capabilities at a qualitative level. Furthermore, by analyzing non-financial information such as strategy analysis, it is possible to identify the company's profit drivers and key risks, so that it can help analysts to assess the sustainability of the company's performance and make realistic forecasts of future performance [18]. Those various benefits of non-financial information can not necessarily be found in the financial figures that occur in the financial statements. Furthermore, the role of analyst coverage also does not necessarily make the information more relevant than it already is. This is because analysts through their reports only summarize what is relevant and presented in the annual report into a more concise and easier-tounderstand form.

CONCLUSION
The issue of readability of annual reports has been raised and mentioned several times by experts and scholars in various countries, but few have explored the impact of readability of annual reports and the value relevance of financial information on management performance and agency costs. The purpose of this study is to examine and analyze the effect of readability of annual reports and value relevance with the analyst coverage mechanism as a moderating variable on agency costs in Indonesia. This research sample uses 263 firm-year samples of Kompas 100 companies during the 2016-2019 period.
The number of pages, the number of words and the number of characters in the annual report has a negative effect on the asset turnover ratio. The high number of pages, words and characters in this study reflects the low level of readability of the annual report. An annual report with a low level of readability can increase information asymmetry for stakeholders (especially investors and shareholders), reduce company information transparency, reduce the power and efficiency of supervision on the management opportunistic behavior. These problem will have an impact on asset turnover ratio, where the asset turnover ratio is low reflects the high agency costs. This study did not find any significant relationship between the value relevance of financial information and the asset turnover ratio.
This study also found that the presence of analyst coverage was able to moderate the positive influence between the number of pages, words and characters in the annual report on the asset turnover ratio. Regardless of the number of pages, words and characters in the annual report, analyst coverage can summarize the information and make necessary adjustments so that it is easier for investors to read and understand. As a result, it has an impact on the low information asymmetry between management and investors, supervision of management through the information presented in the annual report is stronger and more efficient so that it affects the high asset turnover ratio which reflects low agency costs. Furthermore, this study failed to find any significant effect between value relevance moderated by analyst coverage and asset turnover ratio.
This study has several theoretical and practical implications. First, the result of this study are expected to become empirical evidences and can be a reference for further research on the same topic related to agency cost and the main factors that influence them such as readability of annual reports, value relevance and analyst coverage. Second, our result are expected to be useful for financial intermediaries, information analyzers and advisors, investors and potential investors to find out how readability of annual reports, value relevance and analyst coverage can affect agency costs for companies in Indonesia.
This study has some limitations, namely first, this study only uses company that included in Kompas100 index with only 4 years observation. As a result, only 263 firm-year sample has been obtained. Second, the proxy of readability on this study used the most basic measurement based on the length of the annual report in the form of the number of pages, words and characters. There is still a lack of limits or standards related to this readability measurement. Third, the measurement of agency costs in this study only uses the asset turnover ratio which is an inverse proxy of agency costs, there are many other measurements that directly reflect agency costs.
With the limitations in this study, further study is highly recommended to use other research object with longer observation period that better at represent the entire capital market in Indonesia. Further study is also recommended to try other measures of readability so that the impact of readability of annual report in Indonesia can be well explored. Further study is recommended to try other measurement of agency cost that is related to management performance so that studies of agency cost in Indonesia can also be well explored.