Amsterdam Business School
The short-term and long-term effect of
accounting conservatism on firm value
Name: Rob Heemskerk Student number: 10409548
Thesis supervisor: ir. drs. A.C. M. de Bakker Second reader: dr. ir. S.P. van Triest
Date: June 20, 2016 Word count: 13236
MSc Accountancy & Control, specialization Accountancy Faculty of Economics and Business, University of Amsterdam
Statement of Originality
This document is written by student Rob Heemskerk who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Abstract
The purpose of this study is to investigate the effect of accounting conservatism on firm value. Financial information needs to be neutral according to the International Accounting Standards Board (IASB). That means that financial information should not include conservatism (IASCF, 2008). Prior studies conclude that conservatism has both negative and positive effects (Francis, Hasan & Wu, 2013; Kim and Pevzner, 2010; Penman and Zhang, 2002). Building on the different prior studies it is expected that accounting conservatism has a negative effect on firm value in the short-term and a positive effect on firm value in the long-term. A sample is used of active North-America listed firms from 2010 up to and including 2014 to perform a quantitative research. The level of firm-specific conservatism is measured with the C_Score of Khan and Watts (2009) and firm value is measured with the market value of equity. The results of this study are not all in line with the predictions. When a firm performs conservative, the firm value is higher in that year compared to firms which are not conservative. In the long-term the firm value is also higher for conservative firms compared to non-conservative firms. Accounting conservatism and firm value are significantly positively correlated. This is in line with the findings of Park & Chen (2006), who conclude that conservative firms create hidden reserves in the short-term and these firms use hidden reserves to generate future firm value. The results of the study imply that although the critique of the IASB accounting conservatism also has positive outcome. This study can contribute to existing literature as a starting point to investigate more long-term effects of accounting conservatism.
Table of Contents
1. Introduction 5
2. Literature Review 7
2.1. Conservatism 7
2.1.1. Definition of conservatism 7
2.1.2. Explanations for conservatism 8
2.1.3. Khan and Watts’ (2009) motivation for conservatism 9
2.2. Firm value 10
2.2.1. What is firm value? 10
2.2.2. The positive effects of accounting conservatism on firm value 11 2.2.3. The negative effects of accounting conservatism on firm value 12
2.3. Hypotheses development 13
3. Research Methodology 16
3.1. Measure of conservatism 16
3.2. Measure firm value 18
3.3. The models 18 3.3.1. Hypothesis 1 19 3.3.2. Hypothesis 2 19 3.4. Control variables 19 4. Data 21 4.1. Data selection 21 4.2. Sample selection 21 4.3. C_Score determination 22
4.4. Descriptive statistics of hypothesis 1 23
4.5. Descriptive statistics of hypothesis 2 24
5. Results 26
5.1. The short-term effect of accounting conservatism on firm value 26 5.2. The long-term effect of accounting conservatism on firm value 27
5.3. Robustness checks 29
5.4. Additional tests 31
6. Conclusion 34
1.
Introduction
Most firms have to make a financial report of their financial performance during the year. The way of making a financial report is different of each firm. Firms have different possibilities for their financial reporting policies. One financial reporting policy is accounting conservatism. Accounting conservatism is reporting bad news more quickly than good news in the financial statements. This implies systematic differences in reporting between bad news and good news in the timeliness of earnings (Basu, 1997). In a draft about the improved conceptual framework of financial reporting speaks the International Accounting Standards Board (IASB) negative about accounting conservatism. According to them results conservatism in uncertainty in business and economic activities (IASCF, 2008). LaFond and Watts (2008) argue that accounting conservatism reduces the manager’s ability to manipulate the financial statements. This results in less uncertainty about the business and economic activities of the firm. Watts (2003a) argues that conservatism reduces the information asymmetry between the firms and their stakeholders. Both parties know more information about each other when information asymmetry is decreasing. That is a positive aspect of conservatism.
In the literature, two different forms of accounting conservatism are known: conditional and unconditional conservatism (Ball and Shivakumar, 2005). Conditional conservatism is news dependent, while unconditional conservatism is news independent. This implies that unconditional conservatism more relates on systems. Systems determine when earnings and expenses are recognized. So this has no effect on the decisions of managers of firms. Conditional conservatism, which is news dependent relates on when the managers recognize the earnings and expenses. This paper involves only conditional conservatism because unconditional conservatism is news independent, managers have influence on conditional conservatism and users of financial statements can eliminate unconditional conservatism out of the financial reports (Ball and Shivakumar, 2005; Kim & Zhang, 2015).
There is a positive relation between accounting conservatism and firm stock performance. Conservatism is beneficial to debt holders and other outside shareholders of a firm because conservative firms have a better governance relative to non-conservative firms (Francis, Hasan & Wu, 2013). Conservatism affects also investors. It affects the disagreement of investors around annual earnings announcement dates of firms. Conservatism leads to lower proxies of disagreement (D'Augusta, Bar-Yosef & Prencipe, 2015). So conservatism is also beneficial for investors. Besides the positive relation of Francis et al. (2013) and D’Augusta et al. (2015) have Penman and Zhang (2002) found that conservatism has a negative effect on the firm value because conservatism gives more uncertainty to the users of financial reporting. So in the literature are some different conclusions about the effects of performing accounting conservatism. To learn more about the effects of accounting conservatism on
firm value focuses this paper on the long-term effects of accounting conservatism on firm value. This study analyzes the following research question:
What is the short-term and long-term effect of conservatism on firm value?
Current literature about conservatism and firm value misses a link between the degree of conservatism this year and the effects on firm value in the long-term. As D'Augusta et al. (2015) suggest that further research could focus on whether the information effect of conservatism extends to periods other than earnings announcement days. The long-term effect of conservatism is beneficial to know for stakeholders. Current literature discusses also the effects of accounting conservatism of a firm during a financial crisis but current literature lacks a link between a global financial crisis and the effects of conservatism after a financial crisis (Francis et al., 2013). The data of this study originates from the period after the recent financial crisis of 2007 and 2008 (Vyas, 2011).
The research question is answered through a quantitative research. The sample of this paper consists of data of listed North-America companies. The data originates from the years 2010 up to and including 2014. With help of the annual cross-sectional regression model of Khan and Watts (2009) is the firm-specific conservatism for 2010, 2011 and 2014 estimated. These firm-specific conservatism is used in a model to estimate the effect of accounting conservatism on short- and long-term firm value.
A negative short-term relation between accounting conservatism and firm value is expected and in the long-term is a positive relation expected between accounting conservatism of a firm and the firm value.
This study contributes to the ongoing debate about the effect of accounting conservatism on firm value (LaFond & Watts, 2008; Watts, 2003a; IASCF, 2008; Penman & Zhang, 2002). It can contribute because the most recent available data is used. The relation between firm value and accounting conservatism will be more clear and the effects on the long-term firm value are find out.
This paper proceeds as follows. Chapter two reviews the key prior literature on conservatism and the relation between firm value and conservatism. The hypotheses are as well developed in chapter two. In chapter three the research method is described and chapter four includes the data, the sample selection and the descriptive statistics. Chapter five presents the results of the regressions. Chapter six is the conclusion.
2.
Literature Review
In this chapter, an overview of theory on conservatism is provided. The reasons and effects of conservatism are given. Also the impact of conservatism on firm value is discussed. The first paragraph discusses literature about conservatism and the explanations for conservatism. The second paragraph is about firm value and the relation with conservatism.
2.1. Conservatism
This paragraph is about conservatism. The definition and the explanations of conservatism are discussed. Also the motivations of Khan and Watts (2009) are presented.
2.1.1. Definition of conservatism
In the accounting literature is Basu a well-known researcher in the field of conservatism. His general definition of conservatism is “earnings reflecting ‘bad news’ more quickly than ‘good news’ (Basu, 1997, p.3)”. Auditors are conservative when they have a higher degree of verification for recognizing good news than for bad news. So they accept bad news sooner than good news. Unrealized losses are recognized earlier than unrealized gains for example (Basu, 1997). Conservatism is about the differences in the recognition of profits versus losses. Difference in recognition is positively related to the degree of conservatism. It is not the case that a firm has to recognize all his losses immediately and recognize all his profits on the moment that the cash is collected to be conservative. Via two ways can conservatism affect the financial statements. Conservative accounting can be done via recognizing bad news more quickly than good news and firms can also choose to delay the recognition of good news. A result of being conservative as a firm, is that the net asset values are understated (Watts, 2003a). This is one of the critics of investors and standard setters on accounting conservatism. Understatement of the assets brings uncertainty on the reliability of the financial statements. The representational faithfulness, neutrality and comparability of financial statements conflicts with the use of accounting conservatism according to some investors and regulators (IASCF, 2008). Despite these criticism, conservatism exists for about 500 years in the accounting practice. It is a qualitative characteristic in financial reporting. Also is conservatism viewed as the most influential principle of valuation in accounting and in the last decades is accounting becoming more conservative (Watts, 2003a). Basu (1997) concludes that reporting bad news is more timely than good news because a higher sensitivity of earnings was seen by negative unexpected returns than by positive unexpected returns.
This paper only includes condition conservatism. Unconditional conservatism is not related to good or bad news. Unconditional conservatism has to do with the creation of assets and liabilities. It is
related to the choices that are made in the accounting process during the creation. Conditional conservatism has to do with revaluation of existing assets and liabilities and the writing down of the book values when there are unfavorable events and the writing up of the book values when there are favorable events. Conservatism implies that writing down happens sooner than writing up of the books (Beaver and Ryan, 2005).
2.1.2. Explanations for conservatism
One of the four explanations for conservatism is the contracting explanation. Conservatism addresses the problem of moral hazard with the contracting explanation. These problems can be explained with the agency theory. The stakeholders of the firm and the firm itself have asymmetric information about the firm. The firm itself knows a lot more about the firm than the stakeholders. The stakeholders of the firm are the principal and the firm itself is the agent. The principal cannot observe all the actions of the agent. This makes it difficult for the principal to know if the agent works in the best interest for himself or for the principal (Eisenhardt, 1989). Firms inform their stakeholders via their financial statements. The firms try to present their financial statements as good as possible. This motivates the manager of the firm to bias their results. They want to show positive and good results. This bias increases the information asymmetry between the stakeholders and the firm. Conservatism lowers the chance that managers act opportunistic and offsets managerial biases because losses are recognized sooner than profits. Managers are more limited in their actions to act opportunistic when a firm is conservative. So conservatism decreases the information asymmetry between the stakeholders and the firm (Watts, 2003a). The stakeholders and the users of the financial reporting of a firm benefit from conservatism. Getting more reliable information is a benefit for the stakeholders and the users of financial reporting. Skinner (1994) finds that one of the two main reasons to report negative instead of positive is the relationship with the investment community. Negative reporting is practicing earnings management, but in a way that earnings are reported on a conservative way. A good relationship with the investment community is an example of lowering the information asymmetry between the firm and the investors. Another explanation for conservatism is the litigation explanation. Performing conservatism lowers the chance of litigation (Watts, 2003a). Overstatement or understatement of the earnings or net assets of the firm can lead to litigation. A firm tries not to become part of a litigation. A way to avoid litigation is conservatism. The expected litigation costs of overstatement of the earnings or net assets are higher than the expected litigation costs of understatement of the net assets or earnings (Kellogg, 1984). To avoid litigation costs, managers have incentives to be conservative. Managers have also the chance to be personally sued. The danger for possible penalties leads to more conservative and less
conservative is the relation with lawsuits and reputation. Firms disclose negative information timely because they know that when they move up the negative numbers that this will lead to a big negative announcement about their results. Such a big negative announcement will increase the likelihood of lawsuits and some reputation damage for the firm. Practicing conservatism lowers the chance of lawsuits and reputation damage because the firm is honest about the information of the firm and the firm do not withheld any information. The firm increases the likelihood of lawsuits when the firm withheld negative information till they are required to disclose that information. The firm is conservative to prevent themselves from those lawsuits.
The third explanation is the income tax explanation. Tax is dependent of the earnings of the firm. Lower earnings of the firm results in lower tax payments (Watts, 2003a). Conservatism decreases the total earnings and thus the tax payments. Conservatism postpones the payment of tax because the firm understates the earnings or net assets this year, but this will result in higher net assets or earnings in the future. So there is a delay in the payment of taxes.
The fourth explanation for conservatism is the regulatory explanation. It is also known as the political cost explanation. This explanation involves regulators and standard setters of financial reporting. The standard setters want neutrality in financial reporting. This implies that firms cannot overstate or understate their financial statements. But regulators and standard setters have more critique when firms overstate their net assets than when firms understate their net assets (Watts, 2003a). So it is better to be conservative than to be opportunistic.
2.1.3. Khan and Watts’ (2009) motivation for conservatism
The size of a firm has an effect on the information asymmetry between the firms and investors. Larger firms have a better information environment but larger firms have also more complex operations. The better information environment decreases the information asymmetry but the complex operations can increase it. The information asymmetry problem is linked with the first explanation of Watts (2003a). The size of the firm has a relation with the contracting explanation. The overall assumption in the literature is that large firms have lower information asymmetry (Khan and Watts, 2009). So the large firms are expected to be more conservative than small firms because the stakeholders of the firm have much information about the firm. This results in less opportunistic behavior of the firm and more conservative accounting. The taxation and litigation explanations have also a relation with the size of the firm. Large firms have more possibilities to reduce their present tax liability and large firms have more chance to be sued (Khan and Watts, 2009). It is for bigger firms more difficult to recover from a litigation and it is more expensive for the large firms to have a litigation.
Conservatism is directly linked to the market-to-book ratio. The verification for good news and bad news is related to the market values. The assets are understated when a firm has a high market-to-book ratio (Khan and Watts, 2009). This implies conservatism. So the market-to-market-to-book ratio is positive related to conservatism. The market-to-book ratio can also tell much about the growth options and stock returns of the firm. A high market-to-book ratio indicates more growth options and stock returns volatility. This volatility creates uncertainty. It is not known what the firms do with their high market-to-book ratio (Khan and Watts, 2009). Maybe they invest in risky projects and those projects will fail or maybe they invest in good projects and the firm will grow the coming years. The risky projects can lead to litigation cost but the high market-to-book ratio lowers the interest of regulators. So the regulation explanation of conservatism is lower for firms with a high market-to-book ratio.
The third motivation of Khan and Watts (2009) is linked to the process between lenders, shareholders and the firm. The researchers find that highly levered firms have agency conflicts. Leverage is the amount of debt of a firm to finance their assets. A firm has a high leverage when their debt is significant higher than their equity. Conservatism constrains the possibility to act opportunistic. Lenders prefer conservative behavior of firms so lenders react positive on conservative reporting. This has a positive impact on the contracting process of firms and lenders and is related to the contracting explanation. Highly leveraged firms face more uncertainty and risks to be sued. So the litigation explanation is also involved in the third motivation (Khan and Watts, 2009).
2.2. Firm value
This paragraph is about firm value. Different kinds of firm value are discussed and the effect of conservatism on firm value is presented.
2.2.1. What is firm value?
The value of a firm is determined by the underlying economics. The underlying economics are a combination of different factors. The different factors give the opportunity to measure the firm value via different methods (Park & Chen, 2006). Different methods are used in prior literature to determine firm value. Stock prices are a tool to measure firm value (Basu, 1997). Different models can with help of the stock prices determine the value of a firm: book-to-price ratio (B/P), price-to-earnings ratio (P/E), residual income valuation (RIV) and the discounted free cash flows (DFCF) (Lee & Swaminanthan, 1999). Stock prices lead accounting earnings. So stock prices are a better indicator to determine the firm value (Basu, 1997).
2.2.2. The positive effects of accounting conservatism on firm value
Conservatism reduces the likelihood of dividend decreases and earnings decreases (Kim & Pevzner, 2010). This implies that conservative firms can pay higher dividends than firms without conservatism. This is a result of higher earnings. Dividends are often related to the height of the earnings of the firm. The dividends are paid out to the investors and shareholders of the firm. Conservatism leads also to less disagreement of investors on the accounting numbers (D'Augusta et al., 2015). Investors agree more on the financial statements when a firm is conservative. This means that agency-problems decrease with the use of conservatism and that the information asymmetry between the stakeholders and the firm decreases by using conservatism. Less disagreement of the investors results in more investments of the investors in the firm. This can increase the firm value on the short-term and on the long-term. The effect of conservatism on firm value in the short-term is the effect in the same year as the conservatism is measured. Also Watts (2003a) argues that the information asymmetry lowers with the use of conservatism. Conservatism lowers the ability of a manager to manipulate the financial reports. This lowers the information asymmetry between the shareholders of the firm and the manager. This could have a positive effect on the firm value. It would be easier for the firm to borrow money and to invest this money in projects. This can increase the firm value in the long-term. Sohn (2012) found that financial analysts incorporate accounting conservatism into their forecasts. Financial analysts find it easier to forecast the earnings of a conservative firm. This implies that the information asymmetry is lower when a firm performs conservative accounting. Lower information asymmetry makes the market more efficient. The prediction of the future earnings of firms is one of the most important criteria of financial analysts for their performance. So financial analysts search for every available information of the firm. That they also rely on conservatism of the firm means that conservatism is important for the financial analysts (Sohn, 2012).
Firm value is affected via the financing, investments, and other real activities. Firms can get more financing when they perform conservatism. This comes due to a lower cost of debt. The interest rate to borrow money is lower for conservative firms. A lower cost of debt makes external financing easier (Francis et al., 2013). External financing can in the future result in a higher firm value. The firms need some money to invest in projects. These projects can create firm value in the future when they are successful.
Conservative reporting leads also to lowering cost of capital. The cost of capital is the cost of a firm’s debt and equity. Investors see conservative reporting as a low financial reporting strategy risk. Investors can determine the risk better when a firm reports conservative and this results in better investment decisions of the investor. So an investor can make a better investment decision in whether
to invest or not to invest in a conservative firm related to other firms which perform not conservative (Lara, Osma & Penalva, 2011). Conservatism can have an impact on the investors and on the firm value when the risk to invest decreases.
Conservative firms are likely to create more hidden reserves. Hidden reserves are the resources of the firm that are not listed on a balance sheet. These hidden reserves can lead to higher future firm values in the long-term. Investors react positive on the creation of hidden reserves (Park & Chen, 2006). A investor knows that when a firm practice accounting conservatism that the firm creates hidden reserves. The investors are more willing to invest in firms with high hidden reserves than in firms with low hidden reserves because the investors value the firms higher which report conservative and have high hidden reserves.
Accounting conservatism is related to the internal controls of the firm. The internal controls of a firm detect and prevent the firm from weaknesses. These weaknesses result in a lower firm value. Goh and Li (2011) conclude that firms with low weaknesses and a high internal control are more conservative. It seems that strong internal controls automatically leads to conservatism. The internal controls prevent the managers from acting optimistic. The strong internal control environment is related to the contracting explanation and the agency theory. The internal controls provide also reliable accounting information. The reliable accounting information lowers the information asymmetry and a strong internal control environment helps firms in better understanding of their own firm and processes.
Francis et al. (2013) conclude that performing accounting conservatism is important for investors. The investors see conservative reporting as consistent reporting. They find conservative firms also more reliable. The investors find that the firm has reliable and high-quality accounting information when the firm performs consistently conservative reporting. The consistent reporting builds trust and this results in more investments and a higher share price of the firm. The firm value can increase when the firm consistently performs conservative.
2.2.3. The negative effects of accounting conservatism on firm value
On the contradictory, accounting conservatism can also lead to lower firm value. Penman & Zhang (2002) conclude that conservatism can lead to lower quality of earnings and accounting rates of return. Poor-quality earnings exist when changes in investment are coupled with accounting conservatism. When firms practice conservatism via investments, they lower their earnings on the financial statements via writing down the investments. If the change in investments is permanent, than the lower earnings and accounting rates of return are also permanent. This creates lower quality of earnings and rates of
return. Lower rates of return are not an incentive for investors to invest more in a firm. So it would be more difficult for firms to get money and that can lead to a lower firm value.
Accounting conservatism leads to lower price multiples. Price multiples are ratios that are used to compare the share price of the firm with a financial measure which is related with the share price. The price-to-earnings ratio is an example of a price multiple. The effect of conservatism is economically significant on earnings (Chen, Folsom, Paek and Sami, 2013). The higher the price-to-earnings ratio is, the higher is the investor willing to pay for a share. Lower price multiples suggest that investors are willing to only buy shares for a lower price. The interest to invest in a firm is lower and that has some costs for the firm. This can result in a lower firm value in the short-term.
The time period of this study is after a global financial crisis. The financial crisis had an effect on the firm value. Managers of firms manipulate their earnings more often during a crisis. The shareholders are imposed with a greater risk as a result of this manipulation (Francis et al., 2013). This can lead to more negative returns for the investors and firm value losses.
2.3. Hypotheses development
This paper researches the short-term and long-term effect of accounting conservatism on firm value. The research question can be answered with help of a number of hypotheses. Starting with the effect of conservatism on the short-term firm value, the second hypothesis analyzes the long-term firm value and accounting conservatism.
Conservative firms give their investors more uncertainty through the changes in investments. Practice conservative accounting via the investments creates uncertainty for the investors (Penman & Zhang, 2002). This uncertainty increases the information asymmetry between the shareholders of the firm and the firm itself. They do not know whether investments are temporary or permanent. Conservatism lowers the possibility of the manager of the firm to manipulate the books. Manipulating the books can give a higher firm value in the short-term according to the financial reporting (Watts, 2003a). But when a firm is conservative, the manager cannot create these higher firm value. Conservative firms have more losses than profits. It seems that firms which report losses reflect greater conservatism than firms which report profits. Losses can be explained by conservatism. Conservatism increases in the years that a firm has losses relative to years in which a firm has profits (Balkrishna, Coulton & Taylor, 2007). Based on the previous research and the literature review the following hypothesis is formulated:
Lara et al. (2011) conclude that conditional conservatism has influence on different factors that all have positive economic outcomes for the firm. Conservatism leads to increased information precision and reduces the uncertainty about the amount of future cash flows. Information precision is related to information asymmetry. When information is precise, than there is no information asymmetry. So increasing information precision lowers the information asymmetry. Investors use the precision of information to forecast the future cash flows. Conservative reporting gives investors uncertainty about the investments but it also give more certainty about the future cash flows because the information is more precise (Watts, 2003a; Lara et al., 2011).
Lenders benefit after the loan is issued through the timely signaling of default risk and borrowers benefit upfront through lower initial interest rates when a firm is conservative. The default risk is the risk that the firm cannot pay the required payments on their liability. The borrowers benefit upfront because the interest of the loans that applies at the beginning of the loan is lower. This enhances the debt contracting process. The debt contracting process is the process between the borrower and the lender to issue a debt contract. Debt markets are an important source of external financing for firms. The firm is the borrower and the external financer is the lender (Zhang, 2008). A better debt contracting process gives a firm sooner the opportunity to borrow some money. It takes time until a firm creates firm value with their borrowed money. So in the long-term, it will be easier for a firm to generate firm value with this borrowed money than in the short-term. Also Francis et al. (2013) found that external financing is easier for firms which perform accounting conservatism. This comes due a lower cost of debt. Firms have lower costs because the interest rate is lower of the debt. So the costs of the firm to borrow money reduce and their net income can be higher. The findings of Zhang (2008) and Francis et al. (2013) suggest that firms which are conservative also have a good relationship with their external financers. This creates some options for the firm to get easier money from these financers. The firm can use that money to invest in new projects and create future firm value.
Hidden reserves play a significant role for conservative firms. Park and Chen (2006) conclude that conservative firms have higher hidden reserves. These hidden reserves are created in the years that the financial reporting is conservative and the hidden reserves are released when the firm makes investments. Investors are aware of the fact that conservative reporting maybe implies that a firm creates hidden reserves. Investors can value the firm higher when they know that the firm is conservative. This makes it easier to receive money from investors. This money can be used to create future firm value.
Firms with conservative reporting have an improved investment efficiency. They have a higher return on their investment in the future. This implies that firm value also increases in the future. Debt and equity markets have a demand for conservative reporting because the information asymmetry
between the firms and the markets is lower for conservative firms. Firms need debt markets to borrow some money and they need the equity market to get money from investors. Both markets have a positive notion on conservative reporting (Ball, Robin & Sadka, 2008). This makes it easier for the conservative firms to borrow money and to receive money from investors. The firms can use this money for investments to create firm value in the future. Based on the previous studies and the literature review the following hypothesis is formulated:
H2: Accounting conservatism has a positive relation with firm value in the long-term
3.
Research Methodology
In this chapter the research method is explained. First is described how conservatism is measured. After that, firm value is discussed. A model is presented to compare the conservatism of firms and their firm value. The model is built with the help of the models of Basu (1997) and Khan and Watts (2009). Firm-specific conservatism is determined with the model of Khan and Watts (2009). These firm-Firm-specific conservatism is used to create a model that analyzes the relation between conservatism and firm value.
3.1. Measure of conservatism
In this paper the degree of conditional conservatism is measured with the models of Basu and Khan and Watts. The measure of Basu (1997) is the most cited measure of conservatism in accounting literature and is called the asymmetric timeliness measurement. The model of Khan and Watts (2009) is used to determine the firm specific conservatism. Gains are recognized more conservative than losses according to Basu (1997). The regression model is based on the differential reaction of good and bad news in earnings. Stock prices are used as a proxy for good and bad news. Negative stock returns indicate conservatism because losses are recognized earlier than gains and negative stock returns are more related with earnings than positive stock returns. Stock prices should reflect all available information.
To measure the conditional conservatism, the following model of Basu (1997) is used:
𝑋𝑖𝑡 = 𝛽0+ 𝛽1𝐷𝑖𝑡+ 𝛽2𝑅𝑖𝑡+ 𝛽3𝑅𝑖𝑡𝐷𝑖𝑡+ ℇ𝑖
Where X is the earnings per share reported in year t for the firm i, R is the return on a year-time basis,
D is a dummy variable that equals 1 if the annual stock return is less than 0, and 0 otherwise, ℇ is the
residual. β3 is the coefficient that shows when a firm has conservative reporting. β3 is higher for firms
which are more conservative. β3 measures the incremental timeliness for bad news over good news. β2
measures the sensitivity of good news. This is the good news timeliness measure.
To measure the firm-specific conservatism, the model of Khan and Watts (2009) is also included in this study. Khan and Watts (2009) introduced the C_Score. The C_Score is based on Basu’s model and used it to estimate the firm-year measure of financial reporting conservatism. Conservatism is subject to change for each firm. For that reason, it is necessary to include firm-specific characteristics into the measure of conservatism. The size, leverage, market value of equity and book value of equity are firm-specific characteristics and these factors have an effect on the conservatism. These factors have also a link with the explanations of Watts (2003a). The size of the firm can have a relation with the information asymmetry because the size of the firm is positive related with the amount of information
that is published about the firm. Leverage can indicate firms with more debt, those firms are maybe faced with more litigation. The market and book value forecast the growth options of the firm. A higher market to book ratio indicate more growth options (Khan and Watts, 2009; Watts, 2003a). To estimate the timeliness of good and bad news at firm-year level, the following formulas are prepared:
𝐺_𝑆𝑐𝑜𝑟𝑒 = 𝛽2 = µ1+ µ2𝑆𝑖𝑧𝑒𝑖 + µ3𝑀𝑖/𝐵𝑖+ µ4𝐿𝑒𝑣𝑖 𝐶_𝑆𝑐𝑜𝑟𝑒 = 𝛽3 = λ1+ λ2𝑆𝑖𝑧𝑒𝑖+ λ3𝑀𝑖/𝐵𝑖+ λ4𝐿𝑒𝑣𝑖
The G_Score is the timeliness of good news and the C_Score is the timeliness of bad news. The timeliness of bad news is conservatism. The C_Score captures the degree of conservatism and is the incremental bad news timeliness measure. Conservatism is increasing in the C_Score. The G_Score represents the 𝛽2 and the C_Score represents the 𝛽3 in the Basu model. Filling in the G_Score and C_Score creates a firm-specific estimation of conservatism. Summing up both scores is the total bad news timeliness measure. In the model of Khan and Watts (2009) is the natural logarithm of the market value of equity used to measure size. M/B is the market-to-book ratio and Lev is the leverage, this is the long-term plus short-term debt deflated by the market value of equity.
Khan and Watts (2009) made an annual cross-sectional regression model. This model can estimate the values of µ and the values of λ. After estimating the lambdas, the firm-year specific conservatism can be calculated. The C_Score is this specific conservatism. The annual cross-sectional regression model:
𝑋𝑖𝑡 = 𝛽0+ 𝛽1𝐷𝑖𝑡+ 𝑅𝑖𝑡(µ1+ µ2𝑆𝑖𝑧𝑒𝑖+ µ3𝑀𝑖/𝐵𝑖+ µ4𝐿𝑒𝑣𝑖)
+ 𝑅𝑖𝑡𝐷𝑖𝑡(λ1+ λ2𝑆𝑖𝑧𝑒𝑖 + λ3𝑀𝑖/𝐵𝑖 + λ4𝐿𝑒𝑣𝑖) +δ1𝑆𝑖𝑧𝑒𝑖 + δ2𝑀𝑖/𝐵𝑖 + δ3𝐿𝑒𝑣𝑖 + δ4𝐷𝑖𝑆𝑖𝑧𝑒𝑖+ δ5𝐷𝑖𝑀𝑖/𝐵𝑖+ δ6𝐷𝑖𝐿𝑒𝑣𝑖 + ℇ𝑖𝑡
The lambdas are calculated to quantify the effect of conservatism on firm value. Filling the firm-specific characteristics in the C_Score equation gives the opportunity to see if conservatism has an effect on the firm value.
Table 1 Variables definitions of the model of Khan and Watts
Variables Definitions
X Earnings per share
D A dummy variable that equals 1 if return is less than 0, and 0 otherwise
R The return on a year-time basis
Size The natural logarithm of the market value of equity
M/B Market value of equity deflated by the book value of equity
Lev Long-term plus short-term debt deflated by the market value of equity
ℇ Residual
3.2. Measure firm value
Stock prices lead accounting earnings. Stock prices are a tool to measure firm value (Basu, 1997). This is helpful for shareholders and decreases the information asymmetry between shareholders and the firm. Basu (1997) suggests that stock prices are a better measure for firm value than accounting earnings. So accounting earnings are not used to measure firm value. The firm value can also better be predicted through stock price changes than through accounting earnings changes. To estimate the firm value, the market value of equity is used. The market value of equity is useful because it expresses firm value and it can predict the future stock returns. The future stock returns can be predicted because the stock prices should reflect all the available information of the stock of the firm (Penman, 1996).
To measure the firm value the market value of equity is used. The market value of equity is calculated with the total shares of the company and the stock price at year-end.
𝑀𝑎𝑟𝑘𝑒𝑡 𝑣𝑎𝑙𝑢𝑒𝑖𝑡(𝑀𝑉𝑖𝑡) = 𝑇𝑜𝑡𝑎𝑙 𝑜𝑢𝑡𝑠𝑡𝑎𝑛𝑑𝑖𝑔 𝑠ℎ𝑎𝑟𝑒𝑠𝑖𝑡 ∗ 𝑆ℎ𝑎𝑟𝑒 𝑝𝑟𝑖𝑐𝑒𝑖𝑡
The total outstanding shares are all the shares of the firm that are outstanding at the end of the year t for the firm i, the share price is also determined at the end of the year t, Multiplying these two gives the market value of equity in the year t for the firm i. (Berk & DeMarzo, 2013).
3.3. The models
The following models are used to test the hypotheses. To test the hypotheses are the market value of equity and the C_Score in different years analyzed. Firm value is determined with the market value of equity and conservatism is determined with the C_Score of Khan and Watts’ model. The following model is used to test the first and the second hypothesis:
𝑀𝑉𝑖𝑡 = 𝛽0+ 𝛽1𝐶_𝑆𝑐𝑜𝑟𝑒𝑖𝑡+ 𝛽2𝑇𝐴𝑖𝑡 + 𝛽3𝑀𝑖𝑡/𝐵𝑖𝑡+ 𝛽4𝐿𝑒𝑣𝑖𝑡+ ℇ𝑖𝑡
C_Score is the β3 of Basu’s model. The C_Score is firm-specific and the independent variable.
3.3.1. Hypothesis 1
The first hypothesis focuses on the short-term effect of accounting conservatism on firm value. The firm value and the conservatism of the same year are used to test the hypothesis. The model is the same as shown in paragraph 3.3. t is in the model to estimate the relation between conservatism and short-term firm value 2010. So the short-term effect of accounting conservatism on firm value is determined in 2010.
3.3.2. Hypothesis 2
The second hypothesis focuses on the link between accounting conservatism and the long-term firm value. The same value of conservatism is used as in hypothesis 1. The conservatism of 2010 is combined with the firm value of 2014. So 2010 is seen as an base year for conservatism. To see if there is an effect on the long-term firm value, the market value of 2014 is used.
𝑀𝑉𝑖2014= 𝛽0+ 𝛽1𝐶_𝑆𝑐𝑜𝑟𝑒𝑖2010+ 𝛽2𝑇𝐴𝑖2014+ 𝛽3𝑀𝑖2014/𝐵𝑖2014+ 𝛽4𝐿𝑒𝑣𝑖2014 + ℇ𝑖𝑡
3.4. Control variables
It is important to control the risk factors that could impact the effect of conservatism on firm value. The most important risk factors are controlled in this study. There are three control variables. Khan and Watts (2009) argue that when the C_Score is used in a regression as an independent variable that there is a need to control for size, M/B and leverage. These variables are also inputs of the C_Score itself, so the variables are used twice in this study. The total assets are used to control for size because otherwise is the dependent variable also one of the independent variables. The risk factors are
necessary otherwise is there a possibility that a relation is found between conservatism and firm value where there is no relation.
The first control variable is TA. This variable controls the effect of firm size on accounting conservatism. TA are the total assets of the firm. LaFond and Watts (2008) conclude that large firms have less information asymmetry problems. Large firms present more often financial information which reduces the information asymmetry. This give the firms less reasons to report conservative. But large firms have also higher political costs. Watts (2003a) described that high political costs are
related to more accounting conservatism. The firm size is measured with the natural logarithm of the total assets of the firm.
The second control variable is the market-to-book (M/B) ratio. Firms with high market-to-book ratios are expected to perform better than firms with low market-to-book ratios because these firms have more growth and investing options. These firms have more possibilities to increase their firm value (Francis et al., 2013).
The third control variable is Lev. This is the leverage of the firm. Watts (2003a) argues also that firms with high leverage are reporting more conservative. This comes due the conflicts between the debtholder and the shareholders. The leverage is determined as long-term plus short-term debt deflated by the market value of equity.
Table 2 Variables definitions of the control variables
Variables Definitions
TA The natural logarithm of the total assets
M/B Market value of equity deflated by the book value of equity
Lev Long-term plus short-term debt (total debt) deflated by the market value of equity
4.
Data
In this chapter the data and the sample selection is explained. This study uses data from multiple years, which makes it difficult to have a single sample for the multiple years. The hypotheses are tested with the same sample, the additional tests have other samples. The determination of the C_Score is presented. Also the descriptive statistics of hypothesis one and two are presented.
4.1. Data selection
The research question is answered through a quantitative research. The sample of this paper consists of data of listed North-America companies retrieved from COMPUSTAT, which is used as the starting point for the sample formation. North-America is chosen because in this region are most of the firms from the United States of America. In the United States of America is conservatism seen as one of most influential accounting principles (Hellman, 2008). To find if there is a relation between conservatism and firm value a combination of different datasets is used. The financial services industry is excluded from the research. This is a special industry and the financial industry can have an impact on the outcome of the data and research because they have a different regulation
environment. The financial industry has unique accounting practices, which makes it hard to compare this industry with other industries (Ahmed & Duellman, 2007). The financial industry is also not comparable to other industries which makes it better to exclude them from the sample.
The sample originates from the years from 2010 till the latest complete and available data. To contribute to the ongoing debate, it is important that the most recent and complete data is used. The last complete available data is 2014. There is still some data available of 2015 but this data is not complete. Not all the data from all the firms is available of the year 2015. So it is better to use the complete date of 2014 because using data from 2015 results in a small sample. Data on conservatism is from 2010, 2011 and 2014. The period to measure the long-term effect of conservatism on firm value is from 2010 up to and including the year 2014. This is the most recent complete available data of market value of equity. The short-term effect on firm value is measured for the year 2010. 2014 is chosen as an additional test for the short-term effect. For the long-term period is chosen to analyze the conservatism of 2010 and the market value of equity of 2014. The period 2011 up to and including 2014 is chosen for additional testing for the long-term effect.
4.2. Sample selection
The following table shows which data forms the sample for testing the hypotheses. The sample selection started with excluding the financial services firms. Also only active firms in the years from 2010 up to
and including 2014 are selected. The currency for all the selected firms is the U.S. dollar. Only the U.S. dollar is chosen because any fluctuations of other currencies are ruled out of the sample. Starting from this selection the final sample is determined.
Table 3 Sample Selection
Sample procedure Number of observations (N)
Sample 2010 - 2014 7241
After excluding missing data that is required for the regression
3747
After excluding unreliable values and dropping the top and bottom 1% of variables required for the regression
2917
Final sample of firm observations for hypothesis one
2917
Final sample of firm observations from 2010 up to and including 2014 for hypothesis two
2916
For hypothesis two is the same sample used as for hypothesis one, only one missing value is excluded. The samples for the additional tests are not included in table three. The samples for the additional tests are 4484 (analyze conservatism and firm value of 2014) and 3023 (analyze conservatism of 2011 and firm value of 2014). These samples are different because there is more data available for the additional tests.
4.3. C_Score determination
Firstly is the firm-specific C_Score determined. The C_Score is needed because this is firm-specific for conservatism. The C_Score is used to see if conservatism has an effect on the short-term and long-term firm value. In chapter three is explained how this C_Score is determined. The C_Score is determined for the year 2010. The year 2010 is used to test the hypotheses. The following table provides an oversight of the lambdas and the formula to calculate the firm-specific C_Score.
Table 4 Lambdas Lambdas Year: 2010 λ1 6.27 λ2 -0.71 λ3 -0.01 λ4 -4.58 𝐶_𝑆𝑐𝑜𝑟𝑒𝑖 6.27 - 0.71*𝑆𝑖𝑧𝑒𝑖 – 0.01*𝑀𝑖/𝐵𝑖 – 4.58*𝐿𝑒𝑣𝑖
Note: Size is the natural logarithm of the market value of equity, M/B is the market value of equity deflated by the book value of equity and Lev is the long-term plus short-term debt deflated by the market value of equity.
The formula for the firm-specific C_Score is sample-specific. The formula can only be used to the sample they belong. C_Score is positive related with conservatism, so when C_Score is increasing, firm-specific conservatism is also increasing.
4.4. Descriptive statistics of hypothesis 1
Table five reports the descriptive statistics of hypothesis one. The data originates from listed North-America companies. Almost all industries are included in the sample, only the financial industry is excluded. Firm value has a big range because no distinction is made between small and big firms. The mean of the dependent variable MV is 4165.54 and for the independent variable C_Score 0.22.
Table 5 Descriptive statistics for hypothesis 1 (N = 2917)
Variables Mean Standard
deviation (Std.) Minimum Maximum MV 4165.54 16954.82 0.22 364064.50 C_Score 0.22 4.35 -37.75 7.34 TA 5.86 2.46 0.0020 14.16 M/B 3.34 12.74 -94.37 179.87 Lev 0.40 0.91 0 9.05
Note: This table shows the descriptive statistics for hypothesis one. MV is the market value of equity for the fiscal year 2010. C_Score is determined with the model of Khan and Watts (2009) for the year 2010 with the earnings per share including extraordinary items, return on year-time basis, the natural logarithm of the market value of equity, the book value of equity and the total debt. TA is the natural logarithm of the total assets. M/B is the market value of equity deflated by the book value of equity. Lev is the total debt deflated by the market value of equity. All the control variables are for the year 2010.
The following table presents the correlation of the variables. C_Score and MV are negatively correlated (-0.143). The Pearson’s correlation is used.
Table 6 Correlation matrix for hypothesis 1
MV C_Score Size M/B Lev
MV 1.0000 C_Score -0.143 (0.00***) 1.0000 TA 0.418 (0.00***) -0.488 (0.00***) 1.0000 M/B 0.0003 (0.99) 0.0050 (0.79) -0.0766 (0.00***) 1.0000 Lev -0.0309 (0.10*) -0.923 (0.00*) 0.147 (0.00***) -0.0485 (0.01***) 1.0000 *** p < 0.01, ** p < 0.05, * p < 0.1.
Note: This table shows the Pearson’s correlation matrix for hypothesis one. MV is the market value of equity for the fiscal year 2010. C_Score is determined with the model of Khan and Watts (2009) for the year 2010 with the earnings per share including extraordinary items, return on year-time basis, the natural logarithm of the market value of equity, the book value of equity and the total debt. TA is the natural logarithm of the total assets. M/B is the market value of equity deflated by the book value of equity. Lev is the total debt deflated by the market value of equity. All the control variables are for the year 2010.
C_Score and Lev have a correlation of -0.923. This is an indication for multicollinearity (Grewal, Cote & Baumgartner, 2004). Lev is dropped out of the regression of hypothesis one because of the multicollinearity.
4.5. Descriptive statistics of hypothesis 2
Table seven reports the descriptive statistics for hypothesis two. The market value of equity, total assets, market-to-book ratio and leverage are from 2014, so two different tables are presented to show the descriptive statistics and correlation matrix of hypothesis one and two. The sample is the same sample as hypothesis one, only one firm is dropped out of the sample and data from 2014 is used instead of 2010. The mean of the dependent variable MV is 6240.53 and for the independent variable C_Score 0.22.
Table 7 Descriptive Statistics for hypothesis 2 (N = 2916)
Variables Mean Standard
deviation (Std.) Minimum Maximum MV 6240.53 24989.38 0.0095 591015.70 C_Score 0.22 4.35 -37.75 7.34 TA 6.09 2.61 -6.21 13.88 M/B 5.00 48.99 -584.34 1729.12 Lev 0.70 9.02 0 474.30
Note: This table shows the descriptive statistics for hypothesis two. MV is the market value of equity for the fiscal year 2014. C_Score is determined with the model of Khan and Watts (2009) for the year 2010 with the earnings per share including extraordinary items, return on year-time basis, the natural logarithm of the market value of equity, the book value of equity and the total debt. TA is the natural logarithm of the total assets of the year 2014. M/B is the market value of equity deflated by the book value of equity of the year 2014. Lev is the total debt of 2014 deflated by the market value of equity 2014.
In the Pearson’s correlation matrix of hypothesis two is C_Score also negatively correlated with MV. The relation between the control variables and the dependent variable MV is the same as for hypothesis one. Total assets and the market-to-book ratio are positively related and leverage has a negative relation with MV.
Table 8 Correlation matrix for hypothesis 2
MV C_Score Size M/B Lev
MV 1.0000 C_Score -0.148 (0.00***) 1.0000 TA 0.399 (0.00***) -0.450 (0.00***) 1.0000 M/B 0.0116 (0.53) -0.0176 (0.34) -0.0130 (0.48) 1.0000 Lev -0.0123 (0.51) -0.0872 (0.00***) 0.0325 (0.08*) -0.0043 (0.82) 1.0000 *** p < 0.01, ** p < 0.05, * p < 0.1.
Note: This table shows the Pearson’s correlation matrix for hypothesis two. MV is the market value of equity for the fiscal year 2014. C_Score is determined with the model of Khan and Watts (2009) for the year 2010 with the earnings per share including extraordinary items, return on year-time basis, the natural logarithm of the market value of equity, the book value of equity and the total debt. TA is the natural logarithm of total assets of 2014. M/B is the market value of equity deflated by the book value of equity of 2014. Lev is the total debt of 2014 deflated by the market value of equity 2014.
5.
Results
This chapter presents the results of hypothesis one and two. The results are obtained with help of the statistical program STATA. Paragraph one discusses the relation between accounting conservatism and firm value on the short-term and paragraph two reviews the long-term relation of accounting conservatism and firm value. Two robustness checks and additional tests are performed to make the results more reliable. These are explained and presented in paragraph three and four.
5.1. The short-term effect of accounting conservatism on firm value
The short-term effect of accounting conservatism on firm value is measured with the C_Score for accounting conservatism and the market value of equity for firm value. These measures are used because the C_Score is a year- and firm-specific measure for accounting conservatism and the market value of equity represents the firm value and it can predict future stock returns (Khan and Watts, 2009; Penman, 1996). Conservatism and C_Score are substitutes in the results. Based on previous literature is a negative short-term relation between accounting conservatism and firm value expected.
The short-term effect of accounting conservatism on firm value is measured in 2010. One regression is performed to investigate the short-term relation between accounting conservatism and firm value and the results can be found in the following table.
Table 9 Results for hypothesis 1
N = 2917 F-value = 214.81 R-squared = 0.18 Adjusted R-squared = 0.18 Dependent variable: MV
Variables Coefficient Std. T-value P > | t | ViF
C_Score 319.41 57.11 5.59 0.00*** 1.31 TA 3181.33 133.04 23.91 0.00*** 1.32 M/B 46.83 22.39 2.09 0.04** 1.01 Lev 0 (omitted) Constant -14693.7 846.29 -17.36 0.00*** *** p < 0.01, ** p < 0.05, * p < 0.1.
Note: This table shows the results of hypothesis one. MV is the market value of equity for the fiscal year 2010. C_Score is determined with the model of Khan and Watts (2009) for the year 2010 with the earnings per share including extraordinary items, return on year-time basis, the natural logarithm of the market value of equity, the book value of equity and the total debt. TA is the natural logarithm of total assets. M/B is the market value of equity deflated by the book value of equity. Lev is the total debt deflated by the market value of equity. All the control variables are from the year 2010.
The coefficients are relatively high compared to other studies. A explanation for the high coefficient is the relation of the variables with the market value. The real market value is used in this study. The
market value of the firms has relative high values compared to the other variables. The means and standard deviations of the variables can be find in table five. That will clarify the high coefficients.
The R-squared is rounded 0.18. Around 18 percent of the variance of MV is explained by the independent variables. The coefficient of C_Score is 319.41 with a significance lower than 0.001. The positive coefficient implies that an increase of conservatism will result in an increase of the firm value. If the firm specific C_Score an increase has of one, than the market value of equity of the firm will increase with 319.41 in the term. Conservatism has a positive relation with firm value on the short-term, they are positively correlated. The variance inflation factor (ViF) is 1.31 for C_Score. The ViF is close to one, which makes the result valuable.
In untabled results, a correction is made for firms who grew or declined substantially. The coefficient for C_Score is also negative in untabled results for hypothesis one. These results are not used as the main results because some subjective judgements had to be made to come to results.
Summarizing the results, a significantly positive coefficient is found for C_Score. This confirms not the findings of prior studies. Park and Chen (2006) conclude that conservative firms create hidden reserves. The creation of hidden reserves is related to lower net income and a lower market value. This results in a lower firm value. The finding of a positive coefficient, which implies a positive short-term relation between accounting conservatism and firm value is not in line with the study of Park and Chen (2006). Despite the creation of hidden reserves, can accounting conservatism still have a positive effect on the firm value. Hypothesis one is rejected.
5.2. The long-term effect of accounting conservatism on firm value
The long-term effect of accounting conservatism on firm value is measured with the C_Score of 2010 and the market value of equity of 2014. Based on previous literature a positive relation is expected between the C_Score and market value of equity of a firm. One regression is performed to investigate the long-term relation between accounting conservatism and firm value and the results can be found in the following table.
Table 10 Results for hypothesis 2
N = 2916 F-value = 140.11 R-squared = 0.16 Adjusted R-squared = 0.16 Dependent variable: MV
Variables Coefficient Std. T-value P > | t | ViF
C_Score 223.78 109.59 2.04 0.04** 1.26 TA 3999.71 182.10 21.96 0.00*** 1.26 M/B 8.90 8.67 1.03 0.30 1.00 Lev -62.07 47.22 -1.31 0.19 1.01 Constant -18164.93 1199.94 -15.14 0.00*** *** p < 0.01, ** p < 0.05, * p < 0.1.
Note: This table shows the results of hypothesis two. MV is the market value of equity for the fiscal year 2014. C_Score is determined with the model of Khan and Watts (2009) for the year 2010 with the earnings per share including extraordinary items, return on year-time basis, the natural logarithm of the market value of equity, the book value of equity and the total debt. TA is the natural logarithm of the total assets. M/B is the market value of equity deflated by the book value of equity. Lev is the total debt deflated by the market value of equity. All the control variables are from the year 2014.
Also for hypothesis two are the coefficients relatively high compared to other studies. For the market value of 2014 is the real market value used. The relative high values of market value compared to the independent variables are the reason for the high coefficients. The means and standard deviations of the variables can be found in table seven. That will clarify the high coefficients. Almost the same sample is used as for hypothesis one.
The R-squared is rounded 0.16. More than sixteen percent of the variance of MV is explained by the independent variables. The coefficient C_Score is 223.78 with a significance lower than five percent. The positive coefficient implies that an increase in the accounting conservatism of a firm will lead to a higher firm value in the long-term. If the C_score has an increase of one, than the market value of equity of the firm will increase with 223.78 in the long-term. So accounting conservatism and firm value are positively correlated in the long-term. The variance inflation factor (ViF) is 1.26 for C_Score. The ViF is close to one, which makes the result valuable. This is in line with the expectation of hypothesis two, which expects a positive relation. Lev and the market-to-book ratio (M/B) are not significant for a p-value lower than 0.1. The negative relation between MV and Lev is expected from the literature. Khan and Watts (2009) conclude that highly leveraged firms have more uncertainty and risks to be sued. Investors incorporate this information and that leads to a negative effect on the market value of highly leveraged firms.
The control variable TA is positively related to the firm value, both for the short-term and the long-term relation. This is in line with prior literature. Large firms have on average higher total assets
and those large firms have more external shareholders and stakeholders (Watts, 2003a). This implies in most cases also a higher market value of equity. So a positive relation between the market value of equity and total assets is in line with prior literature. The significance is different for all the control variables when the short-term and the long-term are compared.
In untabled results, a correction is made for firms who grew or declined substantially between 2010 and 2014 and is the market value corrected for economic growth. The coefficient C_Score is also positive in the untabled results for hypothesis two. These results are not used as the main results because some subjective judgements had to be made to come to results.
Comparing the results with prior literature confirms the positive C_Score coefficient the findings of other studies. According to Zhang (2008) results conservatism in a better relation with the debt market, which makes it easier to borrow money and generate firm value in the future. Goh and Li (2011) found that firms with high internal controls have low weaknesses and are more conservative. The high internal controls imply better control of the firm, which can lead to a higher future firm value. So the result confirms the investigation of the prior studies. Penman and Zhang (2002) conclude that accounting conservatism leads to lower quality of earnings and accounting rates of return. The positive coefficient implies that despite the lower quality of earnings, there still is a growth in firm value. Hypothesis two is not rejected.
5.3. Robustness checks
The results of the robustness checks are compared with the original tests and if the results are the same, this leads to a more robust and reliable conclusion. The first robustness check is the original test corrected for heteroscedasticity. Heteroscedasticity describes the behaviour of the variances and the standard deviations of the sample. Problems with heteroscedasticity exist when the variances and standard deviations are not constant. The following tables show the same regression as in paragraph 5.1 and 5.2, but they are corrected for heteroscedasticity.
Table 11 Heteroscedasticity for hypothesis 1
N = 2917 F-value = 39.12 R-squared = 0.18 Hypothesis 1 Dependent variable: MV
Variables Coefficient Std. T-value P > | t |
C_Score 319.41 57.11 5.59 0.00*** TA 3181.33 295.27 10.77 0.00*** M/B 46.83 16.19 2.89 0.00*** Lev 0 (omitted) Constant -14693.70 1504.56 -9.77 0.00*** *** p < 0.01, * p < 0.05, * p < 0.1.
Note: See table nine for description of the variables.
Table 12 Heteroscedasticity for hypothesis 2
N = 2916 F-value = 30.76 R-squared = 0.16 Hypothesis 2 Dependent variable: MV
Variables Coefficient Std. T-value P > | t |
C_Score 233.78 68.38 3.27 0.00*** TA 3999.71 379.72 10.53 0.00*** M/B 8.90 2.81 3.16 0.00*** Lev -62.07 28.27 -2.20 0.03** Constant -18164.93 1952.43 -9.30 0.00*** *** p < 0.01, ** p < 0.05, * p < 0.1.
Note: See table ten for description of the variables.
Table eleven has lower p-values than table nine. The results of the controlling for heteroscedasticity are still that accounting conservatism and firm value are significantly positively related on the short-term.
Table twelve has also lower p-values than table ten. The results for the test for heteroscedasticity are still in line with hypothesis two. Accounting conservatism and firm value are positively related in the long-term. The control variables M/B and Lev are significant for a p-value lower than five percent. For hypotheses two is another robustness check performed because the p-value for the effect of C_Score on the long-term market value of equity is not lower than one percent. Another model is used to test the hypothesis. In the current model is the C_Score used. The model of Khan and Watts (2009) describes also the G_Score and according to them is the sum of the G_Score and the C_Score the total bad news timeliness. This is also a way to measure conservatism. For the G_score are the mu’s
estimated: µ1 = -6.42, µ2= 0.77, µ3= 0.017 and µ4= 4.78. The following table shows the model (Panel A) and the results (Panel B) of the model with the sum of G_Score and C_Score.
Table 13 Robustness check for hypothesis 2 Panel A: The model to test hypothesis 2
𝑀𝑉2014= 𝛽0+ 𝛽1(𝐺𝑆𝑐𝑜𝑟𝑒𝑖2010+ 𝐶𝑆𝑐𝑜𝑟𝑒𝑖2010) + 𝛽2𝑇𝐴𝑖2014+ 𝛽3𝑀𝑖2014/𝐵𝑖2014+ 𝛽4𝐿𝑒𝑣𝑖2014 + ℇ𝑖𝑡
Panel B: The results of the robustness check
N = 2916 F-value = 139.04 R-squared = 0.16 Adjusted R-squared = 0.16 Dependent variable: MV
Variables Coefficient Std. T-value P > | t | ViF
G_Score + C_Score 1780.77 2349.74 0.76 0.45 1.69 TA 3731.14 210.59 17.72 0.00*** 1.68 M/B 8.12 8.68 0.94 0.35 1.00 Lev -72.53 47.21 -1.54 0.13 1.01 Constant -17036.71 1081.32 -15.76 0.00*** *** p < 0.01, ** p < 0.05, * p < 0.1.
Note: This table shows the robustness check of hypothesis two. MV is the market value of equity for the fiscal year 2014. The C_Score + G_Score is determined with the model of Khan and Watts (2009) for the year 2010 with the earnings per share including extraordinary items, return on year-time basis, the natural logarithm of the market value of equity, the book value of equity and the total debt. TA is the natural logarithm of total assets. M/B is the market value of equity deflated by the book value of equity. Lev is the total debt deflated by the market value of equity. All the control variables are from the year 2014.
Note: The formula to estimate the G_Score: G_Score = -6.42 + 0.77*Size + 0.017*M/B + 4.78*Lev. Note: The formula to estimate the C_Score: C_Score = 6.27 – 0.71*Size – 0.01*M/B – 4.58*Lev. Note: See table four for description of the variables of G_Score and C_Score.
Based on this robustness check exists still a positive relation between accounting conservatism and firm value on the long-term but there is no significance. Between the original test and the robustness check are not big differences, which suggest that the G_Score has not a great influence on the results of hypothesis two, however on the significance has the G_Score a negative effect.
The results of both robustness checks give homogeneous results and lead in a certain way to a more robust conclusion.
5.4. Additional tests
Two additional tests are performed to make the results more reliable. The additional tests are the same tests as the original tests but for the additional tests are other years chosen. One additional test is
