Amsterdam Business School
What is the effect of fair value adjustments of investment properties on a firms dividend policy and does a strong corporate governance structure moderates this
effect?
Name: Martin Bond
Student number: 10004474 Date 17-08-2015
Word Count: 9675
MSC Accountancy & Control, specialization [Accountancy] Faculty of Economics and Business, University of Amsterdam
2 Statement of Originality
This document is written by student Martin Bond 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.
3 Abstract
In this paper the impact of adjustments in fair value of investment properties on dividend policy is investigated. To investigate this relationship, the research of Goncharov and Van Triest (2011) is replicated and some additional models are included for the contribution of this paper. Furthermore, the moderating effect of the corporate governance structure of firms on the relationship between adjustments in fair value and dividend policies is researched in this paper. When fair value adjustments have a transitory nature and managers of firms have the ability to identify their implications for earnings in the future, it is expected that including adjustments of fair value in net income will have no distribution consequences. However, when relevant stakeholders are not able to assess the information in the earnings components correctly, incorporating unrealized gains in net income potentially may introduce noise in the process of decision making. When dividend distribution is based on unrealized income with a transitory nature, this will increase the risk for companies. In this paper a sample is used with 462 observations of five different countries within the European Union. The countries
included in the sample are Germany, Greece, Poland, Portugal and the United Kingdom. No empirical evidence is found regarding the increase in dividends in relation with adjustments in fair value. In this paper a significant negative relationship between adjustments in fair value and distributed dividends of firms is found. Also, an significant evidence is found for the effect of the corporate governance structure of firms on the relationship between adjustments in fair value for investment properties and distributed dividends. Based on the result of this research can be suggested that it is more likely that companies with a high attendance average of the board will decline their dividends.
4 Contents
1.Introduction... 5
1.1.Background...5
1.2 Research question... 7
1.3 Motivation and contribution... 8
1.4 Structure…...9
2. Literature review and hypotheses………...10
2.1. Financial reporting………...10
2.1.1. Agency theory……… ..10
2.1.2. The role of accounting regulation...12
2.1.3. Persistence of earnings components……….12
2.2. Dividend policies……….13
2.3. Fair value adjustments and dividends………..14
2.4. Corporate governance structure………...………15
2.4. The role of CG in moderating the relation between fair value adjustments and dividends…...16
3. Method.………...……….. 17
3.1. Sample selection……….…….…..17
3.2.1. Empirical models hypothesis 1……… 17
3.2.2. Empirical models hypothesis 2………..19
3.2.3. Measurement of control variables……….20
4. Results………..21
4.1. Descriptive statistics………..21
4.2. Main analyses……….22
4.2.1. Results empirical models hypothesis 1……….. 22
4.2.2. Results empirical models hypothesis 2……… ..28
5. Conclusion and limitations………30
6. References……… 32
5 1. INTRODUCTION
1.1.BACKGROUND
Fair value accounting is becoming more and more important in accounting standards. The implementation of the International Financial Reporting Standards (IFRS), particularly in the European Union, has led to a discussion that these standards are fair value based standards. The implementation of IFRS has led to frequent comments from regulators that the
International Accounting Standards Board (IASB) moves towards ‘full fair value accounting’ (Ernst & Young, 2005). They suggest that historical cost accounting is a more reliable
measure to use in financial reporting compared to fair value accounting. Other regulators, like the IASB, state that fair value accounting leads to improved financial reporting, because it is expected to deliver more timely and value relevant information (Financial Accounting
Standards Board, 2000). Fair value accounting should benefit for investors, because fair value measurement allows them to make more informed and better decisions. (Barth, 2007; Barth and Landsman, 1995). As a consequence, fair value accounting is promoted by the IASB in the International Financial Reporting Standards (Barth, 2007). However, fair value accounting can lead to an increase of unrealized, transitory gains in net income. Consequently this will increase the managerial discretion (Penman, 2007; Plantin et al., 2008). When the relevant stakeholders are not able to assess the information in the earnings components correctly, incorporating unrealized gains in net income potentially may introduce noise in the process of decision making (Hung and Subramanyam, 2007).
Prior literature examined the relationship between dividend policies and fair value adjustments. The framework of Lintner (1956) predicts that only the unrealized earnings that have a persistent nature should have an impact on the distribution of dividends of a firm, assuming that transitory fair value adjustments in unrealized earnings are identified in a correct manner by firm management and shareholders. So fair value adjustments and earnings consist of transitory and persistent components. Transitory components of earnings are items that will not persist in earnings on the long run while persistent components will not occur one time, but will also return in the future. So according to the framework of Lintner (1956) a firm’s dividend policy will not be influenced by the transitory components of earnings. However, there are some issues with Lintner’s framework in practice. According to Lintner transitory fair value adjustments in unrealized earnings are not distributed assuming that these transitory fair value adjustments are identified correctly by the management of the firm. This
6 means that no relationship is expected between transitory fair value adjustments and
distributed dividends according to Lintner. However, other researchers concluded that Lintner’s assumption that the persistence of fair value adjustment can correctly be identified by the management of a firm does not hold in practice (DeAngelo et al., 1996; Enria et al.,2004). They show that there is a lacking ability of firm management to understand the persistence of fair value adjustments and current earnings related to earnings in the future. DeAngelo (1996) and Enria et al. (2004) show that firm management is often too optimistic in assessing the persistence of fair value adjustments and earnings components. So they state that firm management often get the wrong impression about the persistence of earnings. This might cause noise in the dividend policies of companies. In addition, Sloan (1996)
investigates in his research if investors are processing the earnings persistence in a correct manner. This research concludes that investors are not processing the earnings persistence correctly, so according to this research investors cannot perfectly assess the difference between transitory and persistent unrealized gains.
When dealing with earnings components with a transitory nature, this might lead to problems. For example, when unrealized transitory gains are recognized as fair value
adjustments with a persistent nature, this might affect a firm’s dividend policy choices. When firms choose to pay out these transitory unrealized gains in fair value, without knowing that these fair value gains are not persistent, this could be at the expense of the firm’s long-term value (Enria, et al., 2004). They show in their research that regulators are worried that companies will increase their dividend pay-outs, because upward fair value adjustments with a transitory nature are recognized as persistent. Due to the increase in dividend pay-outs based on transitory unrealized gains in fair value, regulators have concerns about the growth in leverage. This increase in leverage may go at the expense of both equity holders and debt holders (Enria et al., 2004). Therefore, the question whether and in what way dividend policies of firms are influenced by fair value accounting is important to investigate.
In prior literature, where the relationship between dividend policies and fair value adjustments was examined, researchers came to contradicting conclusions. In the research of Goncharov and Van Triest (2011) was examined what effect positive adjustments in fair value will have on dividend policies of firms in Russia. Their main focus was on changes in
financial securities. As stated before, the framework of Lintner (1956) predicted that only the unrealized earnings that have a persistent nature should have a positive impact on the
distribution of dividends. In contrast, Goncharov and Van Triest (2011) show that firm do change their dividend pay-outs as a result of upward fair value adjustments in net income, but
7 they instead paid out less dividend. They gave two possible explanations for their findings that dividends decline when positive fair value adjustments in income are present. Their first possible explanation is that managers possibly use upward adjustments in fair value
opportunistically to justify relatively low levels of dividend, so they use it as an excuse for cutting dividends. Goncharov and Van Triest (2011) state that dividends are cut by firms and they defend it with the excuse that upward changes in fair value are transitory. According to them this will mainly be the case in companies that have weak corporate governance
structures. Consistent with this, Gugler and Yurtoglu (2003) show in their research that companies that have a weak corporate governance structure cut dividends significantly more often, compared to firms with a strong corporate governance structure.
The second possible explanation of the findings by Goncharov and Van Triest (2011) is that upward adjustments in fair value can be associated with an unobservable reaction by the management to high growth. They explain that maintaining the level of dividend pay-out, by looking at the level of income before fair value adjustments, could increase the riskiness of a firm. This possibly explains the relatively lower payout of dividends. Based on the research of Brav et al. (2005), which shows that management makes decisions related to dividends after the consideration of investment needs, it can be stated that the management of a firm possibly pays lower dividends relatively to finance investments. Consequently, it can be concluded that there is a negative relationship between unrealised fair value gains and dividends of a firm.
1.2. RESEARCH QUESTION
As stated in the background section, fair value accounting is becoming increasingly important in IFRS. Fair value accounting should make the information of the financial statements more timely, relevant and transparent (Financial Accounting Standards Board, 2000). Although, other regulators state that fair value accounting has also some downsides. They suggest that historical cost accounting is a more reliable measure to use in financial reporting compared to fair value accounting (Ernst & Young, 2005). These regulators show that there is a lacking ability of firm management to understand the persistence of fair value adjustments and current earnings related to earnings in the future (DeAngelo, 1996: Enria et al., 2004).
Goncharov and Van Triest (2011) show that firms paid out less dividend, when upward
8 that have a weak corporate governance structure cut dividends significantly more often, compared to firms with a strong corporate governance structure.
This led to the following research question: What is the effect of fair value
adjustments of investment properties on a firms dividend policy and does a strong corporate governance structure moderates this effect?
This paper examines the research question by replicating the paper of Goncharov and Van Triest (2011) in a different setting and for a longer time period between 2005 and 2012. The relationship between dividend policies and adjustments in fair value from investment properties will be examined. The countries that are used to answer the research question are firms of Germany, Greece, Poland, Portugal and the United Kingdom. This paper contributes to their results by also examining the moderating effect of corporate governance on the relationship between dividend policies and fair value adjustments. This paper answers to the suggestion of Goncharov and Van Triest (2011) to examine if the results are varying for weak or strong corporate governance structures. So this paper examines the first possible
explanation of Goncharov and Van Triest (2011) that is that managers may use positive fair value adjustments opportunistically to justify a relatively lower level of dividends. So in this paper is examined if managers use fair value adjustment as an excuse for declining their dividends. This paper also contributes to Gugler and Yurtoglu (2003) by examining the moderating effect of corporate governance on the effect of upward fair value adjustments on dividend policy choices.
1.3.MOTIVATION AND CONTRIBUTION
In this section the contribution of this paper is described. The contribution of this paper addresses scientific and societal aspects.
First, I contribute to the literature on fair value accounting. As explained in the background section, literature on fair value accounting does not only focuses on benefits, but also on the potential costs of including unrealized gains in financial reporting. I contribute to the literature by documenting whether the discretion inherent in fair value adjustments can be exploited to increase dividend payouts, as well as investigate how corporate governance quality moderated this relationship.
Second, I contribute to the literature on dividend policies. Despite that regulators are worried about the distribution of transitory adjustments in fair value, Linter’s (1956)
9 dividend policy choices. I examine this relationship for a long window and for a broad sample of European firms.
The contribution to practitioners is also an important reason for addressing the aforementioned research question. This paper is important for corporate outsiders. For example, it is important to know for debtholders and small shareholders how fair value adjustments affect dividend policies of firms. This paper enables them to anticipate and correct for potential opportunistic actions, such as inclusion or tightening of dividend
provision in debt covenants. There is at the moment much debate about fair value accounting, so it is important to know the benefits and costs associated with fair value accounting. This paper will provide a better understanding of the impact that fair value adjustment may have on dividend policy and if corporate governance quality moderates this relationship.
1.4.STRUCTURE
The remainder of this paper will be structured as follows. In section two I discuss the relevant literature and develop the hypotheses. Section three reviews the sample, regression models and measurement of variables. In section four an overview of the empirical results is
provided. Finally, in section five the conclusion of this research is presented and the research limitations are discussed.
10 2. LITERATURE REVIEW AND HYPOTHESES
2.1.Financial reporting
2.1.1. Agency theory
In prior literature there are several findings related to cutting dividends when firms deal with fair value adjustments. Goncharov and Van Triest (2011) show that firms cut dividends when upward adjustments in fair value are present. A possible explanation they provide is that a firms management can use positive fair value revaluations as an excuse for declining their dividends. The excuse in this case could be, that those adjustments in fair value are of a transitory nature. This possible explanation could be clarified by the agency theory.
Most firms are characterized by separation of ownership and control. According to Jensen and Meckling (1976) this has some benefits such as the potential to diversify risk for shareholders, but also creates agency problems, because of the asymmetric information between owners and shareholders. Separation of ownership and control leads to two specific problems: moral hazard and diverse selection. Financial reporting is used to overcome these problems. On the basis of financial reporting, shareholders can assess to what extent managers acted in their best interest and it provides measures of performance that they can use in the incentive contract of managers. In this way financial reporting can address the moral hazard problem. Furthermore, it provides outside investors with information that they can use to assess the future prospect of the firm (e.g., the persistence of current performance as well as the exploitation of new opportunities) such that they can make informed decisions about their ownership. In this way the problem of adverse selection can be addressed (Shivakumar, 2012).
By Jensen and Meckling (1976) the agency relationship is defined as a contract between agent (managers) and principals (the owners of the firm). So the agency theory is about the separation of ownership and control. The principal has to control the behaviour of the agent, by aligning the interest of the agent with the concern of the firm by providing incentives and monitoring their behaviour.
It is important to note that agency conflicts are not exclusively for the relationship
between the management and the shareholders. Jensen and Meckling (1976) state that there is also an agency cost of debt. They show that managers can take actions on behalf of
11 holders to overcome agency problems. Bharath, Sunder and Sunder (2008) state in their research that borrowers that have more opaque financial statements are facing higher interest rates and have more strict borrowing terms. Consistent with the importance of financial statements for setting the interest rates, Minnis (2011) concluded that borrowers will receive more attractive interest rates from lenders if they get their financial statements voluntarily audited. Debtholders can protect themselves by means of debt covenants. These covenants are intended at mitigating debtholders needs to monitor managers closely by aligning the interests of debtholders with the interests of shareholders (Christensen and Nikolaev, 2012).
Dividend policies could also be used to reduce agency problems in firms (Easterbrook, 1984; Gugler and Yurtoglu, 2003; Jensen, 1986), by distributing dividends and attract capital from the market, a firm is monitored by outside parties like debt holders and investment banks more severe (Easterbrook, 1984). Or by eliminating the free cash flow problem, so that management can’t waste cash on bad investments (Jensen, 1986). Dividend distribution could also be used to protect small shareholders against rent extraction, as Gugler and Yurtoglu (2003) argue that large shareholders have the discretion and incentives to extract rents of the firm in their own interest. While dividends are paid on a pro-rata payout ratio, dividends are equally distributed to all shareholders, both small and large. So dividend pay outs are a good way to protect small shareholders from rent extraction by large shareholders. So dividends could be used to signal to small shareholders that firms do not allow large shareholders to extract rents from them. The rent extraction hypothesis is about that the large shareholders are cutting dividends because then there’s more cash in the firm from which they can extract rents in their own self-interest (Gugler and Yurtoglu, 2003). They examined if those agency
problems like the rent extraction hypothesis are more likely to occur when companies have a weak corporate governance structure in place, and their findings support this hypothesis. So weak corporate governance structure could explain why firms are cutting dividends when they’re dealing with upward adjustments in fair value.
12 2.1.2. The role of accounting regulation
According to International Accounting Standards Board (IASB) fair value accounting leads to improved financial reporting, because it is expected to deliver more and value relevant
information (Financial Accounting Standards Board, 2000). Consequently, fair value accounting is promoted by the IASB in the International Financial Reporting Standards (Barth, 2007). Fair value accounting should benefit for investors, because fair value
measurement allows them to make more informed and better decisions (Barth, 2007; Barth and Landsman, 1995). They state that fair value valuations are more transparent compared to historical cost valuations and they better reflect the current values of items on the balance sheet. So this will make the process of making decision for investors more accurate.
Fair value accounting also has negative effects. Laux and Leuz (2009) show in their research that the International Accounting Standards board (IASB) and the Financial Accounting Standards Board (FASB) are dealing with a trade-off between reliability and relevance. As stated before, using fair value measurement increases the relevance of
information. On the other hand fair value measurement could be causing noise in the financial statements, because market prices possibly contain noise (Hung and Subramanyam, 2007). With fair value accounting assets are valued based on the current market value of assets that are identical. Although, sometimes there are no identical assets or there is no liquid market for these individual assets. Less liquid markets lead to less informative prices as individual
transactions exert a significant influence on the market prices for which assets are traded. So consequently, fair value accounting could reduce the reliability of the measure of assets.
2.1.3. Persistence of earnings components
Sloan (1996) identifies that earnings consists of both accrual and cash components, where accruals represent those adjustments to cash flows (e.g., deprecation) to come up with, amongst others, a more informative measure of performance (earnings) compared to the balance of cash flows. When current earnings components will return in periods in the future, those components are of a persistent nature, while one-time items in earnings are transitory (Sloan, 1996). Earnings consist of different components. Accounting regulation to some point distinguishes persistent from transitory components in the financial statements to facilitate investor decision making (e.g., extra-ordinary items are presented separately in the income statement). Goncharov and Van Triest (2011) identify another important component of
13 earnings which is of increasing importance under IFRS accounting standards, namely fair value components. Adjustments in fair value are most of the time not realised directly before the period ends.
Sloan (1996) found in his research that the persistence of accrual and cash components in earnings are not always equal. He also found that investors can’t fully process the
persistence of those individual earnings components effectively. So firms with high accruals are overvalued by investors, while firms with low amount of accruals are undervalued. Firm management also isn’t effective when processing the persistence of different earnings components (DeAngelo et. al., 1996). So it’s difficult for investors and managers to assess which components in income will persist in future periods. This could lead to the distribution of transitory accrual earnings and transitory fair value gains in dividends, because the
persistence is not successfully assessed.
2.2. Dividend policies
The framework of Lintner (1956) predicts that there is a smooth relationship between a firms distributed dividends and the earnings of these companies. So firms define their distributed dividends on dividend payments of the year before and the earnings of the previous year and current year. Between those variables companies try to make a smooth relationship.
DeAngelo and DeAngelo (1990) concluded that firms cut dividends if their earnings are negative. When these companies ignore dividend pay outs, investors could think that they have troubles with their liquidity. Because companies try to generate a smooth relationship between dividend distribution and earnings and the fact that both management (DeAngelo et al., 1996) and investors (Sloan, 1996) cannot perfectly assess the persistence of earnings components, transitory components of earnings could be distributed.
There are a couple of theories on dividend policy choices. Modigliani and Miller (1961) state that dividend policies do not matter for investors. The distribution of dividends could also be used to solve agency problems. A firm could distribute dividends and attract money from capital markets, the debt holders and/or banks are monitoring the firm closely which protects the shareholders of the firm as well (Easterbrook, 1984). Dividend policies could be used for different purposes, cash flow signalling explanation, free cash flow
hypothesis and rent extraction hypothesis. The cash flow signalling explanation states that the management of the firm has more knowledge on the future cash flows of the company than outsiders and the firm has incentives to signal information about this to investors (Gugler and
14 Yurtoglu, 2003). According to the free cash flow hypothesis companies that have high cash flows could better distribute an amount of this free cash. In this way there is less chance for the firm to invest this cash in project that have a negative net present value (Gugler and Yurtoglu, 2003; Jensen, 1986). The free cash flow hypothesis is mainly useful for companies that have problems with overinvestment.
The last way to use dividends is given by the rent extraction hypothesis. This
hypothesis states that dividend policies could be used to protect small shareholders from rent-extracting behaviour of the largest shareholders (Gugler and Yurtoglu, 2003). When
distributing dividends on a pro-rata basis there’s less free cash to extract from the company in the largest shareholders self-interest. By distributing dividends a firm could signal that they do not allow rent extraction by large shareholders from the firm. When a firm is cutting dividends this could be a symptom that the largest shareholders are extracting rents from the firm. The research of Gugler and Yurtoglu (2003) concludes that firms with weak corporate governance structures are more likely to decline dividends.
2.3. Fair value adjustments and dividends
In this paper the effect of upward fair value adjustments on dividend policy choices is examined. In prior literature this effect is examined before by Goncharov and Van Triest (2011). They found opposing results from what they expected. From Lintner’s framework (1956) it could be expected that dividends are in a smooth relationship with earnings. Further, following Lintner (1956), they assume that investors and managers can fully process the persistence of earnings components like fair value adjustments. So, they expect that firms do not distribute transitory upward fair value adjustments and that only earnings components that have a persistent nature are included in the distribution of dividends. Enria et. al. (2004) described the concerns that regulators had on this issue, as regulators expected the distribution of transitory components of upward fair value adjustments. This concern comes from prior research that indicates that investors (Sloan, 1996) and management (DeAngelo et. al., 1996; Jensen, 1993) can’t fully process the persistence of earnings components effectively. More specifically, DeAngelo et. al. (1996) and Jensen (1993) show that managers can’t interpret the persistence of earnings. Furthermore they show that managers are often too optimistic in evaluating the earnings components persistence. In addition Sloan (1996) explains that investors can fully process the difference between transitory and persistent unrealized gains. So, dependent whether the fair value adjustments are transitory and whether investors and
15 management can perfectly assess the persistence of fair value adjustments, no effect or a positive effect on distributed dividends could be expected when positive fair value
adjustments exists. On the contrary, Goncharov and Van Triest (2011) show that a firm with upward fair value adjustments declined their distribution of dividends. On the basis of this mixed evidence, I formulate my hypothesis in a non-directional manner.
Hypothesis 1: Changes in fair values have consequences on the distribution of transitory
components of fair value adjustments on investment properties
2.4. Corporate governance structure
Goncharov and Van Triest (2011) propose the corporate governance structure of a firm as an alternative moderating variable on the effect of fair value adjustments on dividend policy choices. Corporate governance can be described as a set of mechanisms and processes that help ensure that companies are directed and managed to create value for their owners, while fulfilling responsibilities to other stakeholders (Merchant, 2007). Corporate governance deals with the ways in which suppliers of finance to corporations assure themselves of getting a return on their investment (Schleifer and Vishny, 1997). Corporate governance is a multi-faceted construct that, amongst others, looks at indications of good board governance such as the independence of directors and committees, the size of the board, whether the roles of chairman and CEO are not combined in one single officer, whether the supervisory board members are active and not too busy on other boards, the quality and independence of the auditor, the activity of the audit committee and the capability of the audit committee members (Khanchel, 2007). Most research about corporate governance is focused on observable
attributes of board governance, such as an independent chairman, the size of the board and board interlocks. Many studies examined the effect of higher quality corporate governance on the manifestation of agency problems within firms (e.g., Core et al., 1999). Gugler and Yurtoglu (2003) examined the relationship between dividend policies and corporate
governance structures. They concluded that there is a direct positive effect between dividend policy choices and corporate governance structures. The dividends are higher with a stronger corporate governance structure. Gugler and Yutoglu (2003) show that corporate governance structures are essential for protecting minority shareholders when dividends are distributed and in other cases
16 2.4.1. The role of CG in moderating the relation between fair value adjustments and
dividends
Goncharov and Van Triest (2011) show that firms with upward fair value adjustments are declining their dividends. This while they expected no effect when: i) fair value adjustments are transitory or ii) when fair value adjustments are persistent but investors and managers failed to assess the persistence accurately, and expected a positive effect when the fair value adjustments are persistent and this is accurately identified by investors and management. A potential explanation for this is firms with upward fair value adjustments are using those adjustments in fair value as an excuse to decline dividends. They argue that those gains are unrealized and thus not available for distribution. The rent extraction hypothesis could possibly explain this, because management may have a preference to more free cash flows which they can spend of negative NPV projects that only increase their own benefits and/or because of the conflict of interest between the large and smaller shareholders, where large shareholders can expropriate rents at the expense of small shareholders. Corporate governance may act to mitigate these agency problems. For example, improved corporate governance limits the opportunities for managers to overinvest by allocating corporate resources to investment projects that do not benefit the firm (e.g., pet projects). Furthermore,
large shareholders have more discretion to extract rents from the firm when the corporate governance structure is weaker (Gugler and Yurtoglu, 2003). So, consistent with Goncharov and Van Triest (2011), firms with weak corporate governance structures are expected to be more inclined to use adjustments in fair value as an excuse to decline dividends. This leads to the second hypothesis.
Hypothesis 2: Firms with weak corporate governance structure are more likely to decline
17 3. Method
3.1. Sample Selection
This paper’s sample will be based on five different countries within the European Union from 2005 to 2012. The reason that this timeframe is chosen is that it is required since 2005 to provide the financial statements following IFRS. The data is collected from the Datastream database. The sample consist of 462 observations. The five countries that are chosen for this research are the Germany, Greece, Poland, Portugal and the United Kingdom.
The five countries stated above are chosen, because within the European Union the different countries have different guidelines for the distribution of dividends of firms. Some countries in the EU have legislation or guidelines, which limits the possibility to distribute unrealized income on investment properties (KPMG, 2008). For this research, two countries are chosen which allow the distribution of unrealized income on investment properties and two countries are chosen where this is not allowed. In this way the effect of this legislation on the dividend distribution can be tested. The countries where distribution is allowed are Greece and Portugal. In Poland and Great Britain distribution is not allowed. Finally, Germany is included, because of the large data which is available for this country regarding corporate governance information. In Germany distribution is also not allowed.
3.2.1. Empirical models hypothesis 1
For the first hypothesis it is conditional to measure the persistence of adjustments in fair value on investment properties. To assess the persistence of adjustments in fair value of investment properties, an earnings persistence regression model is used similar to the research of Sloan (1996) and Goncharov and Van Triest (2011). The persistence of adjustments in fair value of investment properties is assessed through the following model:
NIit = γ0 + γ1NIBREVit-1 + γ2REVit-1 + CONTROLS + εit (1)
In this model net income is regressed on lagged net income before revaluations and lagged revaluation income. NIit is net income in year t, NIBREVit-1 is net income before
adjustments in fair value of investment property and REVit-1 is the revaluation of investment property in year t-1. The magnitude and the sign of γ2 is dependent on the persistence of
18 adjustments in fair value of investment property. If the adjustments in fair value are fully transitory this coefficient is predicted to be zero. Assuming that the fair value adjustments are persistent, the coefficient on REV γ2 should be greater than zero for good news and this coefficient should be negative if the adjustments indicate a negative performance for the future. As in Sloan (1996) and Goncharov and Van Triest (2011) all variables are scaled by average total assets.
The second model of this research evaluates what impact the revaluation and historical cost components will have on dividends. For this model the partial adjustment model of Lintner (1956) is used, which also is included in the research of Goncharov and Van Triest (2011) and Goergen et al. (2005). The set of independent variables consists of current
earnings and lagged earnings and dividends. The empirical model described by equitation (2) assesses the relationship between the dividend policy and earnings components and looks as follows:
ΔDIVit = α0 + α1NIit + α2NIit−1 + α3DIVit−1+ CONTROLS + Ɛit (2)
ΔDIVit is the change in dividend from year t-1 to year t and DIVit-1 are the lagged dividends. NIit is net income in year t and NIit-1 is net income in year t-1. For this model all variables are also scaled by average total assets.
Based on the research of Goncharov and Van Triest (2011), for the third model I decompose net income into net income before revaluation of fair value adjustments of investment property. The model looks as follows:
ΔDIVit = β0 + β1NIBREVit + β2NIBREVit−1 + β3REVit + β4DIVit−1+ CONTROLS + Ɛit (3)
All variables are defined as in the first two models. The first hypothesis of this research is tested by looking at the magnitude and the sign of the coefficient on adjustments in fair value of investment properties(REVit). As stated before, if adjustments in fair value are fully transitory this coefficient is expected to be zero. When investors are able to measure the persistence for these future earnings correctly, hypothesis 1 predicts that adjustments in fair value are not distributed and β3 will be zero in this model. In contrast, if investors are not able to correctly measure fair value adjustments persistence, it is expected that β3 will be greater than zero and that β3 will be equal to β1. This is because companies will pay out a similar proportion of adjustments in fair value of investment property.
19
For the fourth empirical model an additional test is created for countries where the distribution of unrealised gains in fair value is (not) allowed. For Greece and Portugal its allowed to distribute these earnings, so the dummy variable equals one for these countries. For Germany, Poland and the United Kingdom distributing these earnings is not allowed, so the dummy variable for the countries equals zero. I expect that the dummy variable will have a moderating effect on the relationship between unrealised gains in fair value of investment properties and the change in dividends scaled over total assets.
ΔDIVit = β0 + β1NIBREVit + β2NIBREVit-1 + β3REVit + β4REV*ddistr_allowed +
β5ddistr.allowed + β6DIVit-1 + CONTROLS + Ɛit (4)
All variables are defined as in the first two models.
3.2.2. Empirical models hypothesis 2
This paper examines the moderating effect of a firms corporate governance structure on the effect of positive fair value adjustments on the dividend policy choices of a firm. The corporate governance, is measured by an equation composed of a couple determinants explained in the paper of Khanchel (2007). A corporate governance dummy is used to measure the moderating effect on the first hypothesis. Proxies are used to test the corporate governance structure. The proxies included in this research are (High CG Score), (High boardsize) and (Highmeetingaverage). The dummy variable equals one when the variable is above the median split and otherwise the variable is zero and will have no effect on the results.When the dummy is 1, the corporate governance structure on this variable is strong, when it’s 0 the corporate governance structure is weak. A moderated regression is used to test the effect of corporate governance structure on the effect of positive fair value adjustments on dividend policy choices (hypothesis 1). The following models are used:
ΔDIVit = β0 + β1NIBREVit + β2NIBREVit-1 + β3REVit + β4REV*High CG Score + β5High CG Score + β6DIVit−1 + CONTROLS + Ɛit
20
ΔDIVit = β0 + β1NIBREVit + β2NIBREVit-1 + β3REVit + β4REV*High boardsize +
β5high boardsize + β6DIVit−1 + CONTROLS + Ɛit (6)
ΔDIVit = β0 + β1NIBREVit + β2NIBREVit-1 + β3REVit + β4REV*Highmeetingaverage +
β5Highmeetingaverage + β6DIVit−1 + CONTROLS + Ɛit (7)
ΔDIVit is the change in dividends from year t-1 to year t. NIBREVit and NIBREVit-1 are net income before adjustments in fair value of investment property in year t and year t-1. REVit is the revaluation of investment property in year t. Finally, DIVt-1 is the distributed dividend in year t-1.
In these three models β3 assesses the relationship between adjustments in fair value and changes in dividends for firms with a weak corporate governance structure. Additionally, β3 + β4 shows the relationship between fair value adjustments and dividend changes for companies with a strong corporate governance structure. So β4 shows the difference between firms with a strong and weak corporate governance structure, with regard to the relationship between adjustments in fair value and dividend changes. On the basis of my second
hypothesis, it could be expected that β4 will be negative.
3.2.3. Measurement of control variables
Several variables are included in the empirical model to control for aspects that influence the adjustments in fair value accounting. First, the proxy for leverage, which is the total debt over total assets (LEVit) is included to the empirical model. Second the size of the firm is included (SIZEit). This is the natural logarithm of total assets (TAit). Finally cash over total assets (CASHit) and the growth of sales over total assets (GROWTHit) are included to the empirical model. The variables (SIZEit), (LEVit), and (GROWTHit) are chosen, because according to the research of Bhattacharya (1979) and Rees (1997) agency costs and signaling cost have influence on a firm dividend payout. Furthermore, the variable (CASHit) is included to control for financial constraints. By including this additional variables of control, the negative effect of adjustments in fair value is better pronounced in statistical and economic terms.
21 4. Results
4.1. Descriptive statistics
Table 1 reports descriptive statistics on variables used in the empirical models of this research. The final sample for adjustments in fair value for investment properties consist of 462 observations. The median of the variable (REVit), which is scaled over total assets, is (0) and the mean is (-0.00022). This suggests that the majority of the firms do not make any revaluations for investment property. Net income before the adjustments of unrealized fair value on investment property has a median of (0.031) and a mean of (-0.0015). Regarding the control variables that are used in this research, firmsize (SIZEit) has a median of (12.325) and a mean of (12.419). The median of debt scaled over total assets, the indicator for leverage, has a median of (0.219) and a mean of (0.234). The median of the growth in sales over total assets (GROWTHit) is (0.035) and a mean of (0.028). Finally, scaled cash (CASHit) has a median of (0.078) and a mean of (0.138). The median and mean of all the control variables, do not show a big difference between each other. This suggests that the mean of the different control variables is not affected by outliers. All the variables are explained in Appendix A.
Table 1: Descriptive statistics
Variable N Mean Std. dev Min Median Max
NIit 1532 0.01343 0.14507 -0.6912 0.038 0.335 NIit-1 1365 0.015 0.1445 -0.6912 0.0395 0.335 REVit 462 -0.00022 0.02657 -0.15298 0 0.14006 REV_Dit 462 .64285 0.47968 0 1 1 ∆DIVit 1393 0.00107 0.01514 -0.06366 0 0.076945 DIVit-1 1413 0.01195 0.02175 0 0.00305 0.12874 SIZEit 1685 12.41863 2.10055 1.60944 12.32499 17.96449 LEVit 1678 0.23398 0.19952 0 0.21897 0.98020 GROWTHit 1464 0.02858 0.28714 -1.29140 0.03518 0.83159 CASHit 1682 0.13790 0.16237 0.00002 0.07777 0.78162
22 Table two gives an overview of the Pearson correlations. The revaluation decision (REV_Dit) for adjustments in fair value of investment property is correlated with most of the other variables with a highly significant coefficient. Regarding the control variables the revaluation decision is significantly correlated with all the four variables. This means that a firms decision to revaluate is dependent on the size of the firm (0.138), the leverage position of a company (-0.149), the financial growth of sales (0.109) and total cash that a company owns (0.137).
Table 2: Spearman correlation matrix
Variable NIit NIit-1 REVit REV_Dit ∆DIVit DIVit-1 SIZEit LEVit GROWTHit CASHit
NIit 1 NIit-1 1.529*** 1 REVit 0.273*** 0.112** 1 REV_Dit -0.107** 0.004 0.106** 1 ∆DIVit 0.287*** 0.404*** 0.086 -0.010 1 DIVit-1 0.234*** 0.254*** 0.018 0.083 0.030 1 SIZEit 0.170*** 0.191*** 0.050 0.095* 0.138*** 0.539*** 1 LEVit -0.392*** -0.318*** 0.042 0.175*** -0.149*** -0.021 0.187*** 1 GROWTHit 0.395*** 0.198*** 0.086 -0.118** 0.109** 0.019 0.131** -0.187*** 1 CASHit 0.226*** 0.003*** -0.005 -0.232*** 0.137** -0.090* -0.070 -0.416*** 0.187*** 1
***, **, and * indicate statistical significance at the 1, 5, and 10 percent levels, respectively.
4.2. Main analyses
4.2.1. Results empirical models hypotheses 1
Table 3 reports regression estimates for the empirical model described by equitation (1). There is no evidence found to suggest that adjustments in fair value are persistent. The coefficient on (REVit-1) is unequal to zero, which indicates that the fair value
adjustments are of a persistent nature. However, the coefficient on γ2 (REVit-1) is reported in column (a) and this coefficient is not significant (p=0.281). In column (b) the coefficient on γ2 (REVit-1) is replaced by (REV_Dit-1). This coefficient on γ2 equals one if a revaluation has occurred and equals zero if a company did not revalued their financial assets. This method is used to mitigate the influence of outliers on the results. The coefficient on this variable is not statistically significant (p=0.976), so based on this regression model no evidence is found to suggest that adjustment in fair value are persistent.
23 Table 3: Earnings persistence analysis
Variables (a) (b) Constant -0.112 (-2.63)*** -0.101 (-1.91)** NIBREVit-1 0.474 (9.63)*** 0.396 (5.84)*** REVit-1 -0.248 (-1.08) REV_Dit-1 0.0006 (0.03) SIZEit 0.009 (2.75)*** 0.013 (2.84)*** LEVit -0.060 (-1.76)* -0.179 (-3.94)*** GROWTHit 0.132 (5.46)*** -0.133 (-1.68)* CASHit 0.067 (1.38) 0.156 (4.41)*** N 365 211 Adj. R2 0.410 0.437 F score (P>F) 41.46 (0.000)*** 28.15 (0.000)***
24 Table 4 reports regression estimates for the empirical models described by equitation (2) and (3). The results of empirical model described by equitation (2) are shown in column (a) in table 4. The coefficient α3 (DIVit-1) in column (a) reports a negative result and is highly significant. The coefficients α1 (NIit) and α3 (NIit-1) in column (a), which are current and lagged earnings, show a positive result and are both highly significant. These results are consistent with the research of Goncharov and Van Triest (2011) and the framework of Lintner (1956).
In column (b) till (e) the results are shown of the empirical model described by
equitation (3). The coefficient on β3 (REVit) is replaced by (REV_Dit) in the columns (c) and (e). In the columns (d) and (e) also control variables are included to control for aspects that influence the adjustments in fair value accounting. The effect of net income before fair value adjustments (NIBREVit) reported in column (b) is positive for both current and lagged
earnings, but is only significant for lagged earnings (NIBREVit-1). The coefficient (REVit) is negative and not significant. The variable (REVit) is replaced by (REV_Dit) in column (c) and this has no effect on the coefficients. In columns (d) and (e) the model is extended with the control variables (SIZEit), (LEVit), (GROWTHit), and (CASHit). The control variables do not have a big impact relating on the coefficients tabled in column (a) till (c). Regarding the control variables, only (SIZEit) has a significant impact on the distributed dividends. Based on the results reported in table 4 can be suggested that (SIZEit) has a significant positive impact for a firms dividend distribution.
25 Table 4: Dividend policy analyses
Variables (a) (b) (c) (d) (e)
Constant 0.002 (4.92)*** 0.003 (2.02)** 0.002 (0.89) -0.019 (-2.08)** -0.019 (-2.15)** NIit 0.018 (4.74)*** NIit-1 0.015 (3.68)*** NIBREVit 0.010 (0.94) 0.011 (1.03) 0.008 (0.65) 0.008 (0.67) NIBREVit-1 0.030 (2.37)** 0.029 (2.31)** 0.020 (1.48) 0.019 (1.48) REVit 0.034 (0.59) 0.011 (0.19) REV_Dit 0.001 (0.25) 0.001 (-5.32) DIVit-1 -0.202 (-10.15)*** -0.287 (5.37)*** -0.287 (5.37)*** -0.294 (-5.35)*** -0.292 (-5.32)*** SIZEit 0.002 (2.25)** 0.002 (2.15)** LEVit -0.003 (-0.40) -0.003 (-0.39) GROWTHit -0.001 (-0.25) -0.001 (-0.22) CASHit 0.021 (1.84)** 0.023 (1.89)** N 1127 178 178 178 178 Adj. R2 0.108 0.134 0.133 0.155 0.155 F score (P>F) 46.27 (0.000)*** 7.87 (0.000)*** 7.78 (0.000)*** 5.04 (0.000)*** 5.06 (0.000)***
26 Table 5 is an additional test and reports regression estimates for the empirical models
described by equitation (4). In this table the effect is tested whether the allowance of
distributing unrealized gains in fair value has an effect on the relationship between unrealized gains in fair value of investment properties and the change in dividends scaled over total assets. I expect that the dummy variable will have a moderating effect on the relationship between unrealised gains in fair value of investment properties and the change in dividends scaled over total assets. In column (a) the dummy is multiplied by the variable (REVit). As in the previous tables, in column (b) this variable is replaced by (REV_D). This variable equals one if revaluation has occurred and equals zero if a company did not revalued their financial assets.
Looking at the results, in column (a) the coefficient on (REV*distr. Allowed) is (0.044) and the coefficient on (Distr. Allowed) is (-0.002). Both the coefficients are insignificant so no suggestions can be made. In column (b), the coefficients on
(REV_Ddistr.allowed) and (Distr. allowed) are -0.002 and 0.0001. Both the coefficients are insignificant so based on this empirical model no evidence is found to suggest that the
distribution allowance has an effect on the relationship between unrealized gains in fair value and change in dividends.
27 Table 5: Effect of allowance dividend distribution
Variables (a) (b) Constant -0.012 (-2.09)** -0.021 (-2.20)** NIBREVit 0.008 (0.70) 0.008 (0.72) NIBREVit-1 0.0195 (1.47)** 0.0190 (1.46) REVit 0.000 (0.00) REV_Dit 0.002 (0.47) Rev*distr.allowed 0.044 (0.33) Distr. allowed -0.002 (-0.56) REV_Ddistr.allowed -0.002 (-0.27) Distr. allowed 0.0001 (0.01) DIVit-1 -0.298 (-5.34)*** -0.296 (-5.28)*** SIZEit 0.002 (2.33)** 0.002 (2.21)** LEVit -0.002 (-0.26) -0.002 (-0.21) GROWTHit -0.002 (-0.31) -0.001 (-0.21) CASHit -0.021 (1.79)* 0.023 (1.87)* N 178 178 Adj. R2 0.147 0.147 F score (P>F) 0.0001*** 0.0001***
28 4.2.2. Results empirical models hypothesis 2
Table 6 reports regression estimates for the empirical models described by equitation (5) till (7). In this table the moderating effect of the corporate governance structure of firms is tested on the relation between unrealized adjustments in fair value for investment properties and dividend distribution. Three proxies are used to test the corporate governance structure. The proxies included in this research are (High CG score), (High boardsize) and
(Highmeetingaverage). The results of empirical model described by equitation (5) are shown in column (a). In this column the proxy (High CG Score) is tested. In column (b) the results of model 6 are used and the proxy (High boardsize) is tested. Finally in column (c) the proxy (Highmeetingaverage) is tested and the results of model 7 are shown in this model.
The first two proxies that are tested in column (a) and (b) results in insignificant coefficients. The coefficients are (4.059) and -0.010 for column (a), (5.288) (-0.011) for column (b). For the proxy (Highmeetingaverage) reported in column (c) the coefficients are (4.122) for the variable (REV*Highmeetingaverage) and (-0.017) for (Highmeetingaverage). For (REV*Highmeetingaverage) the coefficient is significant at 10% level and for the
coefficient (Highmeetingaverage) at a 1% level. Based on the results in column (c) can be suggested that it is more likely that companies which have a high attendance average on board will change their dividends payments. These results are not expected, because based
on the research of Gugler and Yurtoglu (2003), the opposite results would be expected.
Gugler and Yurtoglu (2003) concluded that companies that have a weak corporate governance structure cut dividends significantly more often, compared to firms with a strong corporate governance structure. A possible explanation for the results showed in table 6 column (c) is based on the free cash flow hypothesis. With a strong corporate governance, firms will pay out more dividend, because in this way there is less change for the firm to invest this cash in projects that have a negative net present value (Gugler an Yurtoglu, 2003). When these companies base this decision to distribute earnings on unrealised earnings with a transitory nature, this can explain that the coefficient (Highmeetingaverage) is positive. Another
possible explanation for the different result, compared to the research of Gugler and Yurtoglu (2003), is that the significance of this coefficient is influenced by the low number of
observations for this variable (20). Based on the results in table 6 column (c) hypothesis 2 is rejected.
29 Table 6: The moderating effect of corporate governance
Variables (a) (b) (c) Constant 0.025 (0.33) 0.025 (0.34) 0.105 (3.35)*** NIBREVit 0.097 (0.72) 0.010 (0.68) 0.088 (2.02)* NIBREVit-1 0.107 (0.88) 0.130 (1.07) 0.104 (2.13)* REVit -4.310 (-0.96) -5.540 (-0.93) -4.344 (-2.20)* Rev*High CG score 4.059 (0.89) High CG score -0.010 (-0.94) REV*High boardsize 5.288 (0.87) High boardsize -0.011 (-0.65) REV*Highmeetingaverage 4.122 (2.09)* High board meeting average -0.017
(-3.50)*** Rev*distr.allowed Distr. allowed DIVit-1 -0.578 (-2.31)** -0.590 (-2.35)** -0.666 (-4.17)*** SIZEit -0.002 (-0.41) -0.002 (-0.31) -0.008 (-3.98)*** LEVit 0.040 (0.99) 0.030 (0.74) 0.085 (5.39)*** GROWTHit -0.013 (-0.35) -0.004 (-0.11) -0.004 (-0.18) CASHit 0.049 (0.47) 0.035 (0.32) 0.096 (2.38)** N 34 34 20 Adj. R2 0.201 0.175 0.843 F score (P>F) 0.112 0.141 0.0006***
30 5. Conclusion and limitations
Due to the implementation of the international accounting reporting standard s in 2005 within the European Union, fair value accounting became much more important for European listed firms. However, when relevant stakeholders are not able to assess the information in the earnings components correctly, incorporating unrealized gains in net income potentially may introduce noise in the process of decision making. When the distribution of dividends is based on unrealized earnings that are transitory, this will lead to an increase on the risk for
companies. Therefore, in this paper the relationship is investigated between unrealised fair value changes of investments properties and dividends of a firm. In this paper the research of Goncharov and Van Triest (2011) is replicated. Some additional models are included for the contribution of this research. To test the relationship between adjustments in fair value of investments properties and dividends, the framework of Lintner (1956) is used. With this framework it is possible to control the level current earnings plus current and past policy of dividends. The first hypothesis of this research is that if adjustments in fair value are transitory, and for investors it is possible to process this information efficiently, fair value adjustments should not have distribution consequences.
This paper’s sample is based on five different countries within the European Union from 2005 to 2012. The reason that this timeframe is chosen is that it is required since 2005 to provide the financial statements following IFRS. The five countries that are chosen for this research are Germany, Greece, Poland, Portugal and the United Kingdom. The data is collected from the Datastream database. The sample consist of 462 observations.
The results of earnings persistence in model 3 do not show any significant results, so based on this regression model no evidence is found to suggest that adjustment in fair value are persistent. The results of empirical model described by equitation (2) show that
adjustments in fair value are negatively related with changes in dividends. I came to these results, after controlling for lagged dividends, earnings before revaluation and additional control variables.
I contribute to the literature by documenting whether the discretion inherent in fair value adjustments can be exploited to increase dividend payouts, as well as investigate how corporate governance quality moderated this relationship. The results of empirical model described by equitation (7) show that companies which have a high attendance average on board will change their dividend payments. Based on this result, hypothesis 2 can be rejected.
31 To this study there are several limitations. First of all, there are no industry, country and year dummies included. Second, for several variables there is not much data available, which has a negative impact on the significance of the coefficients. Finally, regarding the variables of corporate governance structures of firms, there is also not much data available. Therefore, for future research larger samples can be used by including not only European countries, but also countries outside the European Union. For future research, regarding the corporate governance structures of firms, it is important that countries are chosen which have relatively much data available for the proxies for the corporate governance structure, like the corporate governance score, the size of the board and board attendance average.
32 6. References
Bharath, S., Sunder, J. and Sunder, S., 2008. Accounting quality and debt contracting, The Accounting Review 83(1), 1–28.
Barth, M.E. (2007). Standard-setting measurement issues and the relevance of research.
Accounting and Business Research, 37(1), 7-15.
Barth, M.E. and Landsman, W.R. (1995). 'Fundamental issues related to using fair value accounting for financial reporting'. Accounting Horizons, 9(4): 97-107.
Brav, A., Graham, J.R., Harvey, C.R., and Michaely, R. (2005). Payout policy in the 21st century. Journal of financial economics, 77(3), 483–527.
Christensen, H.B. and Nikolaev, V.V., (2012). Capital versus performance covenants in debt contracts, Journal of Accounting Research 50(1), 75–116.
Core, J.E., Holthausen, R.W., Larcker, D.F. (1999). Corporate governance, chief executive officer compensation, and firm performance Journal of Financial Economics, (51), 317-490.
DeAngelo, H., & DeAngelo, L. (1990). An Empirical Investigation of Troubled NYSE Firms.
The Journal of Finance, 45(5), 1415-1431.
DeAngelo, H., DeAngelo, L., and Skinner, D.J. (1996). Reversal of future dividend signaling and the disappearance of sustained earnings growth. Journal of financial
economics, 40(3), 341–371.
Easterbrook, F.H. (1984). Two Agency-Cost Explanations of Dividends. American
Economic Review, 74, 650–59.
Enria, A., Cappiello, L., Dierick, F., Grittini, S., Haralambous, A., Maddaloni, A., Molitor, P., Pires, F. and Poloni, P. (2004). 'Fair value accounting and financial stability'. European Central Bank Occasional Paper Series, 13.
Ernst & Young (2005) How Fair is Fair Value? (London: Ernst & Young). International Accounting Standards Board (2006) International Financial Reporting Standards 2006 (London: IASCF Publications).
Farinha, J. (2003). Dividend policy, corporate governance and the managerial entrenchment hypothesis: an empirical analysis. Journal of Business Finance and Accounting, 30 (9-10), 1173-1209.
Financial Accounting Standards Board (2000). 'Statement of Financial Accounting Concepts No. 7: Using cash flow information and present value in accounting measurements'. Norwalk, CT.
33 Goncharov, I., & Van Triest, S. (2011). Do fair value adjustments influence dividend
policy? Accounting and Business Research, 41(1), 51-68.
Gugler, K., &Yurtoglu, B.B. (2003). Corporate governance and dividend pay-out policy in Germany. European Economic Review, 47, 731 – 758.
Hung, M. and Subramanyam, K.R. (2007). 'Financial statement effects of adopting international accounting standards: the case of Germany'. Review of Accounting Studies, 12(4): 623-657.
Jensen, M. (1986). Agency Costs of Free Cash-flow, Corporate Finance and Takeovers.
American Economic Review, 76, 323–29.
Jensen, M.C., & Meckling, W.H. (1976). Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of financial economics, 3, 305-360.
Khanchel, I. (2007). Corporate governance measurement and determinant analysis.
Managerial auditing journal, 22(8), 740-760.
KPMG (2008). Feasibility study on alternative to capital maintenance regime established by the Second Company Law Directive 77/91/EEC of 13 December 1976 and an examination of the impact on profit distribution of the new EU accounting regime. Larcker & Tayan, 2011, Seven myths about corporate governance, Working paper, Stanford
University)
Laux, C., & Leuz, C. (2009). The crisis of fair-value accounting: making sense of the recent debate. Accounting, Organizations and Society, 34(6-7), 826-834.
Lintner, J. (1956). 'Distribution of incomes of corporations among dividends, retained earnings, and taxes'. American Economic Review, 46(2): 97-113.
Minnis, M., 2011. The Value of Financial Statement Verification in Debt Financing:
Evidence from Private U.S. Firms, Journal of Accounting Research 49(2), 457-506. Penman, S. (2007). 'Financial reporting quality: is fair value a plus or a minus?' Accounting
and Business Research, 37(Special issue): 33-44.
Plantin, G., Sapra, H. and Shin, H.S. (2008). 'Marking-to-market: panacea or Pandora’s box?' Journal of Accounting Research, 46(2): 435-460.
Shleifer, A., & Vishny, R.W. (1997). The limits of arbitrage. Journal of finance, 52(1), 35–55. Shivakumar, L. (2012). ‘The role of financial reporting in contracting’. London Business
School
Sloan, R.G. (1996). 'Do stock prices fully reflect information in accruals and cash flows about future earnings?' The Accounting Review, 71(3): 289-315.
34 Appendix A
Within this appendix variables are explained
NIit Net income current year
NIBREV Net income before fair value adjustments
REVit Adjustments in fair value on investment
property scales over total assets
REV_Dit Dummy variable which equals 1 when
adjustments in fair value on investment properties are revalued
DIVit Current dividend scaled by total assets
DIVit-1 Lagged dividends scaled by total assets
ΔDIVit Change in dividends
Ddistr_allowed Dummy variable which equals 1 when
distribution is allowed
SIZEit Natural logarithm total assets
LEVit Ratio of debt over assets
GROWTHit % change in sales
