Long term effect of practicing a conservative accounting policy in
the financial crisis
Name: Rolf Bouma
Student number: 10589376 Thesis supervisor: Dr. A. Sikalidis Date: 26 June 2017
Word count: 12006
MSc Accountancy & Control, specialization Accountancy Faculty of Economics and Business, University of Amsterdam
Statement of Originality
This document is written by student Rolf Bouma 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 relationship of accounting conservatism during the financial crisis with a firm’s long-term value. The International Accounting Standards Board (IASB) has removed conservatism from their Conceptual Framework because it would not adequately deal with uncertainty. Prior literature provides both positive and negative effects of conservatism (Penman and Zhang, 2002; Watts, 2003; Lafond and Roychowdhury, 2008; Gigler et al., 2009). This study aims to provide the answer whether accounting conservatism is beneficial for firms in a time of uncertainty (the financial crisis). Penman & Zhang’s (2002) theory is tested about the trust of investors in conservative firms. The empirical research is done with the Basu regression model, complemented by the research of Watts & Zuo (2012). The results contradicted the theory of Penman & Zhang (2002). Conservatism during the financial crisis is significantly positively related to the long-term firm value, but negatively (not significantly) related to the capital expenditures. One of the possible reasons for these results comes from Ramalingegowda and Yu (2012), they claim that the influence and demand from long-term institutional investors is of vital importance. One other possible reason comes from the study of Zhang (2008), which states that it is easier for conservative firms to attract loans and that would eventually lead to an increase in firm value and decreasing capital expenditures. Measurement errors are always possible when measuring conservatism and biases could influence the results. So, more research should be done in this area to get to know more precisely what the underlying theory and the effects of accounting conservatism are.
Contents
1 Introduction ... 1
2 Literature Review ... 4
2.1 Conservatism ... 4
2.2 Discussion ... 5
2.3 Short term results of conservatism in a financial crisis ... 11
2.4 Hypotheses development ... 12 3 Research Methodology ... 15 3.1 Conservatism measurement... 15 3.2 Empirical models ... 16 4 Data ... 18 4.1 Data collection ... 18
4.2 Descriptive statistics of the data... 19
5 Results... 21
5.1 Pearson correlation matrix ... 21
5.2 Regression results... 22
5.3 Analysis of the results ... 24
6 Conclusion... 29
1
1 Introduction
The accounting practice has become more and more conservative, as is shown by the recent empirical studies on this topic (Watts, 2003). This is interesting because many capital market regulators, standard setters and academics do not favor conservatism. Conservatism was also seen as a quality of financial reporting information; therefore, it was also in the Conceptual Framework of the International Accounting Standards Board (IASB). Yet, the IASB changed their mind and removed conservatism from the Conceptual Framework (Hellman, 2008). The IASB mentions that conservatism is not an adequate way of dealing with uncertainty. To partially see if this is a valid notion of the IASB this study focuses on conservatism in time of uncertainty, the financial crisis, and whether that has a positive or negative effect on the future firm value.
Some research has already been done on the effect of a conservative accounting policy during a financial crisis, yet still no research is focused on the long-term effect of a conservative accounting policy during a financial crisis. This research is relevant for two main reasons. Firstly, it has not been investigated before, only some research has been done on the long-term effect of a conservative accounting policy and some research on the influence of a conservative accounting policy in the financial crisis, but never combined. Secondly, for firms and all the stakeholders surrounding the firms it is important to know whether the chosen accounting policy is beneficial for them in the long-term, at the time of a financial crisis. The literature until now on conservatism is not in alignment. In particular, some researchers argue that a conservative accounting policy is beneficial for the firm’s value (Watts, 2003; Lafond and Roychowdhury, 2008). Other researchers point out that conservatism can potentially destroy firm value (Penman and Zhang, 2002; Gigler et al., 2009). The relationship that is going to be tested is whether the firm value of firms that reported conservatively during the financial crisis in 2008 is greater than the firm value of firms not reporting conservatively on the long run. The financial crisis has had a large impact on the global financial market, it is therefore of great importance to know whether conservative accounting is beneficial for firms during a crisis as well as the following years. Therefore, the research question is:
What effect does a conservative accountancy policy, during a financial crisis, have on the firm’s long-term value?
Considering the increase of firm value in the short run for conservative reporting firms, it would seem likely that the firm value also goes up in the subsequent years. Yet, Penman and Zhang investigated an interest aspect of accounting conservatism and the investments of firms (2002). In the financial crisis most firms lowered their investments, so the future earnings would
2 also decrease, but the current earnings would be higher (Campello et al., 2011). The conservative accounting firms state higher current earnings than their competitors since they build up reserves with their conservative accounting policy (Penman and Zhang, 2002). Firms tend to let the current earnings provide a basis for the future earnings and consequently for the firm’s value. For these conservative reporting firms, it gives the false image of their future earnings. So, in the financial crisis conservative reporting firms could show higher earnings, yet due to the decrease of investments in the financial crisis the future earnings also decreased. Non-conservative reporting firms also have lower future earnings because of the decrease of investments, but it is better reflected in their financial reports. Investors do not appreciate how conservatism and changes in investment raise questions about the quality of the earnings (Penman and Zhang, 2002). So, investors would rather invest in non-conservative firms, making it easier for them to increase their firm value more than their conservative competitors. This downside of conservative reporting could lead to the decrease of firm value in the long term. Next to Penman & Zhang (2002), Gigler et al. (2009) also find a possible reason which could lead to a worse firm value from conservative firms in comparison to non-conservative firms. The debt contracting efficiency that is claimed to go hand in hand with conservative reporting firms by several researchers (Watts, 2003; Zhang, 2008), is less efficient than the debt contracting of non-conservative reporting firms (Gigler et al., 2009). This inefficiency in debt contracting can lead to a decreasing firm value in the long term, especially at times when loans are already hard to get. This background provides an interesting reason to investigate the mentioned research question.
The results of this study however contradict the theory of Penman and Zhang (2002). The findings show a significant positive relationship between the increase of firm value in 2015 and accounting conservatism during the financial crisis. On the other hand, the capital expenditures are negatively related to during-crisis conservatism, but this result is not significant. For these results several other explanations are brought forward, one of which is found in the study of Ramalingegowda and Yu (2012). Following their theory (Ramalingegowda and Yu, 2012), the results would occur because of the demand and satisfaction of long-term institutional investors, contradicting the theory that investors would be unsatisfied with accounting conservatism. Next to this theory, the theory of Zhang (2008) could also provide a possible explanation for the results. Yet, more research is necessary to assess more precisely what the reason is for these results. A combination of different theories is also plausible.
In the following chapters, the literature around this research is introduced together with the used hypotheses. After the literature section, the research methodology is explained, followed
3 by the explanation of the data that was used. The results are noted and analyzed in the next section. The last chapter is the conclusion, which also contains opportunities for future research.
4
2 Literature Review
This chapter provides an overview of the theory on conservatism and the introduction of the hypotheses. First conservatism is explained in general. Second a discussion follows on the positive and negative effects of conservatism coming from both the equity market point of view as well as the debt market point of view. Third the short-term effect of accounting conservatism on firm value is discussed and fourth and final is the hypotheses development.
2.1 Conservatism
Accounting conservatism is described by Ruch and Taylor (2015) as accounting policies or tendencies that result in the downward bias of accounting net asset value relative to economic net asset value. In conservatism, there are two distinct types that are used in the literature: conditional conservatism and unconditional conservatism. Although there are two distinct types, they have a lot of shared purposes (Beaver & Ryan, 2005). These purposes include capturing investors and others perceived asymmetric loss functions, minimizing firms’ litigation, tax or regulatory costs and enabling regulators to minimize economic instability and avoid criticism, according to Beaver and Ryan (2005).
According to Basu (1997) conditional conservatism means reflecting economic ‘bad news’ more quickly than ‘good news’. This is also referred to as asymmetric timeliness. Having a conservative accounting policy as a firm also leads to a difference in the persistence of earnings. Common examples of conditional conservatism are goodwill impairment, inventory recorded at the lower of cost or market and asymmetry in gain/loss contingencies (Ruch and Taylor, 2015). The emphasis in literature on conditional conservatism is on improving contracting efficiency, given managers’ incentives to report upward-biased accounting numbers (Beaver & Ryan, 2005).
Unconditional conservatism is consistent and does not depend on economic news events. As Watts (2003) mentions, an important consequence of unconditional conservatism is the persistent understatement of net asset values. This understatement in the current period could lead to an overstatement of earnings in future periods because the future expenses are understated (Watts, 2003). Common examples of unconditional conservatism are accelerated depreciation methods, expensing R&D costs and expensing advertising costs (Ruch and Taylor, 2015). The emphasis in literature on unconditional conservatism is on the difficulty of valuing certain types of economic assets and liabilities and to determine the effect of those values on the future income (Beaver & Ryan, 2005).
5 Even with some of their shared purposes it is important to distinguish these two types of conservatism because they have different effects on the financial statements. Also, the use of unconditional conservatism may preempt the use of conditional conservatism, since the understatement of the assets limits the magnitude of bad news corrections. Beaver and Ryan (2005) also mention that although it is less obvious, conditional conservatism also affects unconditional conservatism through the resetting of the cost bases of net assets. Lastly, the conditions in which the types of conservatism are used differ. In this research, it is important to keep this in mind and to what extent they could influence a change in firm value.
2.2 Discussion
According to the study of Ball, Robin and Sadka (2008) accounting conservatism is more aimed at the debt market than the equity market. They acknowledge that both markets are important for firms, but the debt market is requiring more conservatism of firms (Ball et al., 2008). The reason that is mentioned in the study is that equity markets rely less on information from the financial report alone, they have several other information channels more suited to their needs (Ball et al., 2008). In the end, it is necessary for firms to gain capital from both markets so that they can increase their firm value in the long-term, so for this study both markets will be considered and will be discussed in this literature section.
Watts (2003) outlines in his paper the prior research that has been done and provides his own insights on why conservatism is beneficial for stakeholders of firms. According to him, accounting conservatism is deemed to be necessary as an efficient contracting mechanism (Watts, 2003). The reason Watts gives is that conservatism offsets managerial bias, defers earnings and understates cumulative earnings and net assets (2003). These effects constrain managements’ opportunistic payments to themselves and others, which lead to an increase in firm value (Watts, 2003). Sometimes a lender demands conservative reported financial statements, because they need to know the risk of default of the debtor. That is also the reason that banks prefer firms that report more conservatively. So, by a conservative accounting policy a firm can attract more investments and therefore increase their firm value (Ruch & Taylor, 2015). The constrains on managers to serve the shareholders is an example of the benefits of conservatism on the equity market side, whereas the preference from banks for conservative reporting firms is an example of conservatism on the debt market side.
Hidden reserves occur more often with conservative reporting firms. These so called hidden reserves cannot be found on the balance sheet of a firm as a listed item. It is likely that the hidden reserves will lead to higher firm values in the long-term, since investors react positive
6 on the creation of them, according to Park & Chen (2011). If a firm is reporting conservatively it is a signal towards the investor that it is creating hidden reserves. Conservative reporting firms are valued higher by investors, because they are more willing to invest in a firm with high hidden reserves than a firm with low hidden reserves. The hidden reserves will be used for future profits and that is what specifically attracts the investors (Park & Chen, 2011). Applying conservative accounting and creating hidden reserves with that, enables firms to attract more investors. The investors are lured by the prospect of future profits because of the hidden reserves, allowing conservative reporting firms to use the investments of the investors to increase their firm value. Accounting conservatism also has a positive effect on the cash holdings of firms (Louis, Sun & Urcan, 2012). Accounting conservatism would lead to more efficient use of cash holdings, therefore mitigating the effect of value destruction that is associated with cash holdings (Louis et al., 2012). This result is supportive of the notion of Watts (2003), because managers deal more efficiently with the cash holdings and thus prevents those managers to destroy firm value. By making these more efficient decisions the firm value can only increase over time. Also, making the most efficient decision during a financial crisis is of vital importance for the firm in general.
For conservative firms, it is easier to borrow money (Zhang, 2008). Conservative reporting firms have two advantages over their non-conservative reporting competitors: first, conservative firms get lower interest rates and second, banks are more willing to lend money to firms because of a timelier signal of default (Zhang, 2008). In this study Zhang (2008) appeals to the standard setters, that they should reconsider accounting conservatism as an aspect of their Framework, because it would yield benefits for both the lenders as well as the borrowers. Accounting conservatism bears costs and it is important to get to know the cost-benefit tradeoff according to Zhang (2008). If conservative reporting firms are still able to borrow more money than their competitors in the financial crisis, this will most likely result in a higher firm value in the long run. Although, of course it is necessary to keep in mind that banks were extra careful with providing loans during the financial crisis, so by just providing evidence that you are a conservative reporting firm does not mean you could borrow money easier at that time.
Li (2013) investigated the influence of conservatism on debt contracts via negotiable covenants. Whether and to what extent a covenant is renegotiable is very important to the influence conservatism has on it. If it is not possible to renegotiate, or the cost are very high, then conservatism reduces the efficiency of the debt contract (Li, 2013). If it is possible to renegotiate the covenant for a small or medium price, then it suddenly becomes more efficient to use conservatism from a firm’s point of view (Li, 2013). Nikolaev (2010) looked only at the
7 influence of covenants at the start of a debt contract. He found that firms who use conservatism get more favorable contracts than the firms that do not use conservatism (Nikolaev, 2010). The reason for this is that there will be greater reliance on the covenant in a debt contract if a firm uses conservatism, this reliance makes it harder to get a good covenant if you are a non-conservative firm (Nikolaev, 2010). So, the studies of Li (2013) and Nikolaev (2010) contradict each other, because the non-renegotiable covenant of Li’s study and the covenants of Nikolaev’s study have the same beginning as debt contracts. While Li’s (2013) study says that conservatism would be inefficient for those contracts, Nikolaev’s (2010) study argues that it is better to use accounting conservatism for firms who are trying to get a debt contract with a covenant. Overall, it looks like it is beneficial to be conservative as a firm if they want to borrow money, especially since the study of Li (2013) was a theoretical model with own assumptions and Nikolaev (2013) tested his theory empirically with data from practice. Considering the study of Zhang (2008), it would mean that debt contracts with covenants are easier to get for conservative firms. This would again lead to the conclusion that firms who are able to get more money for investment will get a higher firm value in the long-term. Also, these covenants will take away some of the uncertainties lenders have when providing capital, uncertainty that is already at a high level during the financial crisis.
Multiple benefits of a conservative accountancy policy are provided by Lafond and Watts (2008). They claim that the effect of information asymmetry between managers and the outside equity market make firms to report conservatively (Lafond & Watts, 2008). Because this conservative reporting is generated the firms will have increased firm and equity values (Lafond & Watts, 2008). The standard setters of the Financial Accounting Standards Board (FASB) are specifically targeted in this study of Lafond and Watts (2008). They even state that if the FASB were to be successful in achieving their goal of eliminating conservatism it would increase information asymmetry and not decrease it, which is much more desirable (Lafond & Watts, 2008). Also, the need for voluntary disclosures of bad news is less when a firm is reporting conservatively. Voluntary disclosures and conservative accounting work as substitutes (Lafond & Watts, 2008). If the management of a firm uses conservative accounting then it is less necessary to also disclose a part of the bad news (Hui, Matsunaga and Morse, 2009). By mitigating the information asymmetry effect with the equity market firms can again attract more investments to gain future economic benefits and increase firm value.
Supportive of the study of Lafond & Watts (2008), is the study of Kim and Zhang (2016). They investigated the relationship between stock price crash risks and conservatism (Kim & Zhang, 2016). The results of the study pointed out that the degree of conservatism is
8 significantly and negatively associated with future crash risk (Kim & Zhang, 2016). The reason Kim and Zhang (2016) give comes from the study of Lafond and Watts (2008), because conservatism would play an important role in reducing information asymmetry within the equity market. Given that firms are unlikelier to have the risk of a future stock crash gives the equity market participants more trust to invest in them. With more investments, it is easier to increase the firm value.
Lafond and Roychowdhury (2008) found that there is a demand from the shareholders for accounting conservatism. Long-term institutional investors especially demand accounting conservatism from the firms they invest in, according to the study of Ramalingegowda and Yu (2012). Those long-term institutional investors are specific shareholders that demand conservatism from their firms, which might seem strange since they are likely to have privileged access to inside information of the firm. Yet, it is often costlier and more time consuming to directly monitor managers or gather the information yourself, that is why conservatism is a perfect way in disciplining managers’ investment decisions (Ball, 2001). Individual investors would not be interested in monitoring a firm, especially its financial statements, because on average they are more motivated by liquidity concerns or speculations (Ramalingegowda & Yu, 2012). This lack of individual investor interest results in a lack of demand for conservatism as Ramalingegowda and Yu (2012) argue. Ramalingegowda and Yu (2012) are backed up by Roychowdhury (2010) in the sense that conservatism within firms does not stand alone. There are corporate governance mechanisms by which conservative reporting is enforced (Roychowdhury, 2010). This last argument is also found in other studies (Ahmed & Duellman, 2007; Lara, Osma & Penalva, 2009). If conservatism keeps important shareholders satisfied it is more likely that they will keep investing and that will have a positive effect on the long-term firm value.
On the other side, Penman and Zhang (2002) provide evidence that conservatism can upset investors by misleading the earnings predictability, through changes in investment and the creating of reserves. In good times, conservative firms stack up reserves, which they can use when the bad times are upon them. A perfect example of bad times for most firms was of course the financial crisis. That period could perfectly be used to release the reserves and inflate the earnings. The present earnings are of great importance to the calculation of the earnings forecast (Penman & Zhang, 2002). Investors do not like this ambiguity in earnings predictability and prefer to invest their money in other firms that are not misleading in their earnings predictability. So, the quality of earnings will be negatively affected by a firm when it is practicing conservativism in their accounting (Penman & Zhang, 2002). This view is opposite of the earlier
9 expressed view of Park and Chen (2011). Overall, conservatism adversely affects the quality of earnings, because of the reduction in earnings predictability and persistence (Ruch & Taylor, 2015).
A study from Balkrishna, Coulton and Taylor (2007) in Australia provides evidence that firms that have consecutive years of losses increase their conservative reporting. Through this increasing of conservative reporting reversal of the losses is very hard to do (Balkrishna et al., 2007). It is known that Australian General Accepted Accounting Principles has a pervasive aspect of conservatism, but they claim that conservatism still is more evident with firms that record losses (Balkrishna et al., 2007). These results can have a pervasive influence on the outcome of the sample in this study, because although conservative firms seem to have benefits in both the equity market as well as the debt market they are not making proper use of it and keep recording losses. It is unknown whether this effect is also observable in firms of the United States of America who have their own General Accepted Accounting Principles (US GAAP) that tends to erase the notion of accounting conservatism under the guidance of the FASB.
The two different types of conservatism lead to different costs of equity capital (Chan, Lin & Strong, 2009). Unconditional conservatism would be associated with a lower cost of equity capital and conditional conservatism would be associated with a higher cost of equity capital (Chan et al., 2009). Accounting conservatism is a signal towards the investors regarding the quality of the firm’s current and future earnings. Investors usually require a higher rate of return if conservative firms are more vulnerable to opportunistic managerial decisions (Chan et al., 2009). Also, unconditional conservatism is ex ante and conditional conservatism is ex post, which leads to this difference in cost of equity capital (Chan et al., 2009). Lara, Osma and Penalva (2011) have done study in the United States of America and yield different results than the study of Chan et al. (2009), who did their study in the United Kingdom. According to the study of Lara et al. (2011), conditional conservatism lowers the cost of capital. The reason given is that the information uncertainty is reduced by increasing the bad news reporting precision (Lara et al., 2011). The argument that this would-be ex post and thus too late for a lower cost of equity capital is not mentioned in the study of (Lara et al., 2011). An explanation for these mixed results could be in the study of Beaver and Ryan (2005), since they found that unconditional conservatism and conditional conservatism can preempt one another. So, by saying that unconditional conservatism is ex ante and yields the benefit of a lower cost of equity capital, it is logical that conditional conservatism does not have the same benefit anymore. Whereas if you only look to conditional conservatism you might find the lower cost of equity capital, because the other form of conservatism is not considered. The difference could also be explained by the
10 difference in countries that are investigated, because the United States of America and the United Kingdom have different accounting policies and regulations. Overall, it seems that conservatism in its broadest meaning leads to a lower cost of equity capital, which makes it easier for conservative firms to gain investments. Eventually these investments will lead to a higher firm value.
Accounting conservatism biases in financial numbers lead to inefficient decisions according to Gigler et al. (2009). Their research challenges the beforementioned notion of efficient debt contracting (Watts, 2003; Zhang, 2008). Gigler et al. (2009) claim that there are three components with debt contracting that are not fully explored: the change in information content, shift in debt covenants and the efficiency of debt contracting. In the model they produced they argue that lenders more often do not enter a debt contract, when it would be profitable for them to do so. So, accounting conservatism would not enhance the efficiency of debt contracts. The difference in results can be explained by the difference in definition Gigler et al. (2009) use for efficiency of debt contracts. Most researchers including Watts (2003) and Zhang (2008) use the interest rate as a proxy for an efficient debt contract, but Gigler et al. (2009) use the expected opportunity costs as proxy for an efficient debt contract. Given the research of Gigler et al. (2009) accounting conservatism results in less investment space and finally in less opportunity to increase the firm value in the long run.
This discussion part of the literature will now briefly be summarized, so that the most important pieces of information are known. Ball et al. (2008) discussed the importance of accounting conservatism on the equity market, but most of all the importance on the debt market. Watts (2003) followed with explaining the effect of accounting conservatism on managerial investment decisions. Conservative reporting makes managers to make more efficient investment decisions for the shareholders of the firm. Conservative accounting allows for the creation of hidden reserves, which according to Park and Chen (2011) is what investors prefer. The cash holdings of firms also benefit from accounting conservatism as can be seen from the study of Louis et al. (2012). Another benefit of a conservative accounting policy is that it is easier to attract loans, because of two reasons: the interest rate is lower and there is timelier information of default (Zhang, 2008). Li (2013) and Nikolaev (2010) reflect in their studies on the covenants in debt contracting, overall it seems that conservatism allows firms to have easier access to debt contracts. Accounting conservatism reduces the information asymmetry and therefore will eventually increase the firm value (Lafond & Watts, 2008). Furthermore, this reduction in information asymmetry helps to mitigate stock crash risks (Kim & Zhang, 2016). Shareholders and specifically long-term institutional investors demand conservatism from the
11 firms they invest in (Ramalingegowda & Yu, 2012). Penman & Zhang (2002) provide evidence of unsatisfied investors, because of the use of conservatism and creation of reserves. Accounting conservatism would influence the earnings forecast wrongly (Penman & Zhang, 2002). That investors do not like the making of reserves within a conservative accounting policy is contradicted by the study of Park and Chen (2011). In Australia firms that get losses one year are very likely to continue with losses the following years due to the presence of increasing conservatism (Balkrishna et al., 2007). Chan et al. (2009) and Lara et al. (2011) find contradicting results in regard to the effect of conditional conservatism and cost of equity capital. The cost of equity capital would be higher with conditional conservatism, according to Chan et al. (2009), but according to Lara et al. (2011) the cost of equity capital would actually decrease if a firm where to report conservatively. Possible explanations are the different countries in which the studies are done or the preemption of one conservatism to the other, which is described by Beaver and Ryan (2005). The improving efficiency of the debt market by the use of conservative accounting is contradicted by the research of Gigler et al. (2009). The wrong measurements are used to capture the efficiency of the debt market under conservative accounting, according to Gigler et al. (2009). These studies are used to provide the necessary explanation of the hypotheses and the results of the empirical research.
2.3 Short term results of conservatism in a financial crisis
Watts and Zuo (2012) researched the effect of conservative accounting in the financial crisis of 2008, they concluded that in the short run firm value increases if firms report conservatively. Moreover, the conservative accounting firms invest more than their non-conservative competitors and conservative accounting firms borrow more than their non-conservative competitors, because conservatism enables them to do so (Watts & Zuo, 2012). One of the reasons they provide is that managerial opportunism is limited by accounting conservatism (Watts and Zuo, 2012). The other reason is that a conservative reporting firm’s borrowing capacity outperforms the capacity of non-conservative reporting firms (Watts & Zuo, 2012). The crisis period stock return has a positive association with accounting conservatism. Watts and Zuo (2012) try to give different explanations for this, which they have mostly tested. The one explanation that stands out is the one of long-term institutional shareholders demand for conservatism (see Ramalingegowda and Yu, 2012). The long-term institutional shareholders are significantly and positively correlated with pre-crisis accounting conservatism (Watts & Zuo, 2012). Considering that conservative firms perform better during the financial crisis it is
12 expected that they will continue this in the future, leading to a higher firm value than other non-conservative reporting firms.
There is a positive relationship between the abnormal stock returns during a financial crisis and measures of conditional and unconditional conservatism (Francis, Hasan and Qiang, 2013). So, the firms that used a conservative accounting policy suffered less (in term of losses) than the firms they compete with. Francis et al. (2013) use different time frames in the financial crisis and use different robustness checks to make sure that their findings are correct, since there is always a risk of measurement error in the research of conservatism. The studies of Watts and Zuo (2012) and Francis et al. (2013), thus agree that the firm value of firms that used accounting conservatism outperformed their non-conservative competitors during the financial crisis in 2008. Francis et al.’s (2013) research supports the notion of Lafond and Watts (2008) that conservatism mitigates information asymmetry between the internal managers and the outside stakeholders. Furthermore, Francis et al. (2013) stress that standard setters should fully understand the economic consequences of accounting conservatism and the way it helps to mitigate information risk and helps to control the agency problem. However, they do remind everyone that studies in the field of accounting conservatism are prone to measurement errors and that it cannot be ruled out that it is also present in their study, despite the different proxies and robustness checks (Francis et al., 2013).
2.4 Hypotheses development
Most literature so far seems to favor a conservative accounting policy in a financial crisis. It would increase different opportunities to gain investments from the debt market as well as the equity market (Watts, 2003; Lafond & Watts, 2008; Zhang, 2008). The positive effects of accounting conservatism are mostly deducted from the idea that conservatism increases efficient decision making. The managers make better decisions for the shareholders and the firms can get the most efficient loans from the banks (Watts, 2003). Though an interesting theory around the concept of accounting conservatism is made and tested by Penman and Zhang (2002). Their study is mostly focused on unconditional conservatism, in which they provide evidence that firms make reserves in the years they are performing well. These reserves are used in a year when the firm’s performance is not as good. So, their current earnings are better than the competitors, but by doing this their future earnings will be forecasted incorrectly (Penman & Zhang, 2002). Especially if it goes together with a decline in investments, which is very likely in a financial crisis. Investors do not like it when the future earnings forecasts seem to be misleading (Penman & Zhang, 2002). They withdraw their investment in that firm and their trust is mostly gone. This
13 will have a negative impact on those firms, because investors invest less in firms if their trust declines (Bottazzi, Da Rin & Hellman 2012). This decrease in investment will lead to a lower firm value in the long-term. This way a conservative accounting policy during a financial crisis can harm the firm’s long-term value.
Also, important to mention is the research of Gigler et al. (2009), which states that the debt market does not work efficiently for firms with a conservative accounting policy. Lenders would sometimes be reluctant to enter a debt contract, while it would be beneficial for them to do so (Gigler et al., 2009). Especially in the financial crisis, because banks are even more reluctant to lend money to firms in that period. Although some studies claim that conservatism improves the efficiency of the debt market (Watts, 2003; Zhang, 2008). Still, the debt contracts that will be closed at the time of the financial crisis could have dire consequences for the future firm value. So, in two diverse ways it is expected that the firm value of firms will decrease if they applied a conservative accounting policy in the financial crisis. Both from the equity market point of view as well as from the debt market point view this negative relationship can be expected. To test this, the first hypothesis is as follows:
H1: Does a firm’s long-term value decrease if it applied a conservative accounting policy in the financial crisis?
For my second hypothesis, the focus is more on the investments firms make some time after the financial crisis. Since the effect that investors will have, of not being satisfied with a conservative accounting policy, will only show in the investments after some years (the investors lose trust in the firm). Trusting a firm is very important for the decision whether to invest in that certain firm (Bottazzi et al., 2012). Penman and Zhang’s (2002) theory will be tested by looking at the capital expenditures of firms applying a conservative accounting policy after seven years. The expectation is that the investments of firms applying conservative accounting will be decreased more than those of their non-conservative competitors. Mainly because of the loss of trust of investors in the conservative reporting firms that are showing wrong predictions of future earnings. Another possibility is that since the conservative accounting made inefficient debt contracts, according to Gigler et al. (2009), the conservative reporting must slow down their capital expenditures, because of a lack of funds raised. Again, there is a possible theory from both the equity market as well as the debt market. So, to test this, the second hypothesis is as follows:
H2: Are the investments of firms decreased in the long term if those firms applied a conservative accounting policy in the financial crisis?
15
3 Research Methodology
The research method is explained in this chapter. In three parts, the measures and the models are explained. First the measure for conservatism is described, followed by the measure for the firm value and finally the regression models for the hypotheses are discussed. The measures and models are based on the papers of Basu (1997) and the model of Watts and Zuo (2012).
3.1 Conservatism measurement
As basis for the measurement of conservatism the model of Basu will be used (1997). This model is widely used model among researchers (Francis et al., 2013; Lafond & Watts, 2008; Watts, 2003; Watts & Zuo, 2012). The model of Basu has certain limitations, but nonetheless this model is a better measure for conservatism than other accounting conservatism measures (Ryan, 2006). Some of the limitations that Ryan (2006) sums up are: earnings may not immediately reflect bad news, because of buffers allowed under the General Accepted Accounting Principles (GAAP) in the United States of America; some types of discretionary accounting behavior yield asymmetric timeliness and some types of economic phenomena yield asymmetric timeliness. These limitations are acknowledged by most of the researches that use this model, for example Watts (2003) and Francis et al. (2013). Return is used as proxy for economic news, good news means a positive return and bad news means a negative return in the Basu regression model. This measurement relies on the fact that stock prices reflect unbiased economic news (Dietrich, Muller and Riedl, 2007). The earnings presented by a firm will provide bad news sooner than good news, according to the accounting conservatism principle. The regression model proposes that there is evidence of accounting conservatism if the earnings respond more strongly to bad news than good news (Dietrich et al., 2007). The Basu regression model looks as follow:
𝑁𝐼 = 𝛽1+ 𝛽2𝐷 + 𝛽3𝑅 + 𝛽4𝐷×𝑅 + 𝑒
The NI stands for net income. R is the compounded return and D is a dummy variable coded 1 if R is less than 0, and 0 otherwise. The 𝛽4 is the so called Basu coefficient, it captures the
differential verifiability that is required for the recognition of gains versus losses, in other words it captures the accounting conservatism (Watts & Zuo, 2012).
16 3.2 Empirical models
For the calculation of the firm value the total stock return is used. The stock price reflects the firm value, calculating the return of this stock price means that the firm value has changed in that period. A positive return implies an increase of firm value, whereas a negative return implies a decrease in firm value. This measure is common as indication of the firm value (Watts & Zuo, 2012). So, the Basu model is expanded with the total stock return, which creates this model:
𝑁𝐼 = 𝛽1+ 𝛽2𝐷 + 𝛽3𝑅 + 𝛽4𝐷×𝑅
+𝛽5𝑇𝑜𝑡𝑅𝑒𝑡 + 𝛽6𝑇𝑜𝑡𝑅𝑒𝑡×𝐷 + 𝛽7𝑇𝑜𝑡𝑅𝑒𝑡×𝑅 + 𝛽8𝑇𝑜𝑡𝑅𝑒𝑡×𝐷×𝑅 + 𝑒
In addition to this model there will also be some firm-level controls added, as is done by most previous studies. The firm-level controls that are added are: leverage, size and market-to-book ratio. The research of Ball, Kothari and Nikolaev (2012) shows that these control variables are an effective way to eliminate possible biases in the Basu regression model. The debt contracting demand of conservatism is captured by the leverage and it is expected that leverage is positively related with conservatism. Since on average larger firms have less information asymmetry it is expected that size is negatively related to conservatism. On the future asymmetric timeliness of earnings, market-to-book ratio is expected to be affected negatively, because the market-to-book ratio reflects past asymmetric timeliness and growth options (Roychowdhury & Watts, 2007). The model to test H1 will be as follow:
𝑁𝐼 = 𝛽1+ 𝛽2𝐷 + 𝛽3𝑅 + 𝛽4𝐷×𝑅
+𝛽5𝑇𝑜𝑡𝑅𝑒𝑡 + 𝛽6𝑇𝑜𝑡𝑅𝑒𝑡×𝐷 + 𝛽7𝑇𝑜𝑡𝑅𝑒𝑡×𝑅 + 𝛽8𝑇𝑜𝑡𝑅𝑒𝑡×𝐷×𝑅
+𝛽9𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒 + 𝛽10𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒×𝐷 + 𝛽11𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒×𝑅 + 𝛽12𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒×𝐷×𝑅
+𝛽13𝑆𝑖𝑧𝑒 + 𝛽14𝑆𝑖𝑧𝑒×𝐷 + 𝛽15𝑆𝑖𝑧𝑒×𝑅 + 𝛽16𝑆𝑖𝑧𝑒×𝐷×𝑅
+𝛽17𝑀𝑇𝐵 + 𝛽18𝑀𝑇𝐵×𝐷 + 𝛽19𝑀𝑇𝐵×𝑅 + 𝛽20𝑀𝑇𝐵×𝐷×𝑅 + 𝑒
The coefficients 𝛽4, 𝛽8, 𝛽12, 𝛽16 and 𝛽20 are all Basu coefficients. The coefficient which is the most interesting and important for this research is the 𝛽8, if it is negative it is consistent with H1.
In that case there is a negative relation between the total stock return and during-crisis accounting conservatism.
17 To test H2 total return (TotRet) is replaced by capital expenditures (Capex). The expectation is that the coefficient 𝛽8 will be negative, implying a negative relation between the capital expenditures and during-crisis accounting conservatism. So, the model to test H2 will be as follow: 𝑁𝐼 = 𝛽1+ 𝛽2𝐷 + 𝛽3𝑅 + 𝛽4𝐷×𝑅 +𝛽5𝐶𝑎𝑝𝑒𝑥 + 𝛽6𝐶𝑎𝑝𝑒𝑥×𝐷 + 𝛽7𝐶𝑎𝑝𝑒𝑥×𝑅 + 𝛽8𝐶𝑎𝑝𝑒𝑥×𝐷×𝑅 +𝛽9𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒 + 𝛽10𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒×𝐷 + 𝛽11𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒×𝑅 + 𝛽12𝐿𝑒𝑣𝑒𝑟𝑎𝑔𝑒×𝐷×𝑅 +𝛽13𝑆𝑖𝑧𝑒 + 𝛽14𝑆𝑖𝑧𝑒×𝐷 + 𝛽15𝑆𝑖𝑧𝑒×𝑅 + 𝛽16𝑆𝑖𝑧𝑒×𝐷×𝑅 +𝛽17𝑀𝑇𝐵 + 𝛽18𝑀𝑇𝐵×𝐷 + 𝛽19𝑀𝑇𝐵×𝑅 + 𝛽20𝑀𝑇𝐵×𝐷×𝑅 + 𝑒
18
4 Data
Everything concerning the data that is used for this study is explained in the following section. First the way of how the data was collected is mentioned. Second the descriptive statistics of the collected data is given, together with a table for a clear view of all the descriptive statistics.
4.1 Data collection
The data for this empirical research is taken from two databases, the Compustat database and the Center for Research and Security Prices (CRSP). The stock prices to calculate the compounded return and the total return is collected from the CRSP database, all the other data is collected from the Compustat database. All the data that is used is from firms from the United States of America. There are two reasons to use American firms, because the sample that can be taken for firms under relatively the same conditions is larger and according to Hellman (2008) one of the most important accounting principles in the United States of America is conservatism. Furthermore, most research on accounting conservatism is done in the United States of America (Lafond & Watts, 2008; Hui et al., 2009; Lara et al., 2011; Francis et al., 2013; etc.) Firms in the financial industry are not included in the dataset, because those are very specific type of firms that could influence the dataset. Furthermore, accounting techniques from firms in the financial industry differ from other industries, making the comparison even harder (Ahmed & Duellman, 2007). Also, firms from which the data of certain variables lacks are omitted, including firms that defaulted in the period between 2008 and 2015. After those steps, there are 1091 sample firms left for H1 and 1090 for H2. To come to the final sample the largest outliers are removed from the sample, leaving 1071 sample firms to test H1 and 1070 sample firms to test H2.
The monthly stock prices from the last nine months of 2008 were collected to calculate the compounded return (R) in that period, the choice for the nine months compound return comes from the study of Watts & Zuo (2012). 2008 is chosen as a starting point, because the financial crisis was in full effect that year and when discussing the financial crisis researchers all have at least 2008 in the period they define as the financial crisis period (Francis et al., 2013; Watts & Zuo, 2012). To calculate the total return (TotRet) the stock prices from 2014 to 2015 were taken from the CRSP database. The stock prices from 2014 and 2015 were used because enough time had passed since the financial crisis to get a clear view of the effect of conservatism on the long-term value. Net income is scaled by the market value of the previous year. So, the net income is from 2008 and the market value to calculate the value of NI in the calculation is from 2007. The leverage is calculated by dividing the book value of debt by the book value of
19 total assets. The size is the natural logarithm of the firm’s market value of equity. Market-to-book ratio (MTB) is calculated by dividing the market value by the Market-to-book value of equity. All the data to calculate the leverage, size and MTB are from the annual reports from 2008 that are in the Compustat database. The Capex variable is the capital expenditures from 2015 divided by the lagged net assets, this data was available in the Compustat database. In Table 1 all the variables with their definitions can be found.
Table 1
Variables with their definition
Variables Definitions TotRet NI R D Leverage Size MTB Capex
The total return from 2014 to 2015
Net Income of 2008 scaled by the market value of equity of 2007
Compounded return in 2008
A dummy variable that equals 1 if return is less than 0, and 0 otherwise
Book value of debt divided by the book value of total assets of 2008
The natural logarithm of the market value of equity of 2008 Market value divided by the book value of equity of 2008 The capital expenditures of 2015 divided by the lagged net assets
4.2 Descriptive statistics of the data
In Table 2 the descriptive statistics for the variables of H1 and H2 are presented. The dependent variable NI has a mean of -0.051 and a standard deviation of 0.452. The total return variable (TotRet) has a mean of 6.2%, which means that on average firms gained some market value from 2014 to 2015. The compounded return of 2008 (R) gives a mean of -28.5%. The mean of the dummy variable (D) suggests that 76.3% of the firms experienced negative returns in the year 2008. Considering that it was very hard to gain market value in that year of the financial crisis this number makes sense. The average firm in this sample has a leverage of 0.208. The mean of
20 the natural logarithm of the market value of equity (Size) is 6.373. The MTB is on average 2.055. The Capex variable has a mean of 0.042.
Table 2
Descriptive statistics for hypotheses 1 and 2
Variables N Mean
Standard
Deviation Min Max
NI 1,071 -0.051 0.452 -5.260 9.344 TotRet 1,071 0.062 0.899 -1.810 13.628 R 1,071 -0.285 0.609 -2.390 8.082 D 1,071 0.763 0.426 0 1 Leverage 1,071 0.208 0.248 0 3.676 Size 1,071 6.373 2.320 0.076 12.892 MTB 1,071 2.055 3.177 -9.642 38.743 Capex 1,070 0.042 0.050 0 0.541
21
5 Results
In this section, the results of the empirical research are outlined. Starting with the Pearson correlation matrix between the different variables with their significance level. Then, the regression results of the regressions will be outlined. To be concluded by the analysis of the results, with reflection of the studies discussed in the literature review part of this study.
5.1 Pearson correlation matrix
In Table 3 the Pearson correlation matrix is presented, together with their significance level. The correlations with a significance level at 10% or better are marked with an asterisk. Total return has a negative significant correlation with net income (NI), but a non-significant positive relationship with the compound return (R). Total return is not significantly correlated with leverage, size or MTB. Though it does show a positive relation with MTB which is consistent with prior studies (Lemmon & Lins, 2003; Watts & Zuo, 2012).
The capital expenditures are only significantly correlated with the leverage. The low level of correlation of the total return and the capital expenditures to the other variables is most likely caused by the seven years difference of the data. Total return and the capital expenditures are based on 2015 and all the other variables are based on 2008. It is likely that in the meantime a lot of firms have taken over other firms or merged, which also influences the correlation.
22
Table 3
Pearson correlation matrix
TotRet NI R D Leverage Size MTB
NI -0.0725 (0.0176)** R 0.0147 0.1873 (0.6312) (0)* D -0.0594 -0.0528 -0.5876 (0.0519)* (0.084)* (0)* Leverage 0.0085 -0.2299 -0,0687 0.0592 (0.7815) (0)*** (0,0245)** (0.5058) Size -0.002 0.139 0,1348 -0.0204 0.0168 (0.9489) (0)*** (0)*** (0.5058) (0.5829) MTB 0.0323 0.0648 0,1487 -0.1425 0.0219 0.2069 (0.2903) (0.0339)** (0)*** (0)*** (0.4735) (0)*** Capex -0.0536 0.0157 -0,0342 0.0268 0,1599 0.0293 -0.011 (0.0796)* (0.6079) -0,2638 (0.382) (0)*** (0.3377) (0.7185) * indicate significance at the 10% level.
** indicate significance at the 5% level. *** indicate significance at the 1% level.
5.2 Regression results
As can be seen in Table 4 the coefficient 𝛽4 (D*R) is significantly negative at the 1% level (p-value <0.1%). More importantly the coefficient 𝛽8 (TotRet*D*R) is significantly positive at 1%
(p-value < 0.01). The coefficient 𝛽12 (Lev*D*R) is significantly positive (p-value <0.01), this is
consistent with prior studies (Watts & Zuo, 2012). The coefficients that show the interaction of the negative returns with the firm’s size 𝛽16 (Size*D*R) and the market-to-book 𝛽20
(MTB*D*R) are both positive, but the negative return in interaction to the market-to-book is not significant (p-value = 38.3%). All in all, these results reject hypothesis 1, which means that: A firm’s long-term value does not decrease if it applied a conservative accounting policy in the financial crisis.
23
Table 4
Regression analysis of H1
Variable Coef. P>t [95% Conf. Interval] D -0.034 0.769 -0.264 0.195 R 1.034 0 0.764 1.304 D*R -1.024 0 -1.355 -0.693 TotRet 0.124 0.002 0.045 0.203 TotRet*D -0.264 0 -0.362 -0.165 TotRet*R -0.365 0 -0.542 -0.189 TotRet*D*R 0.250 0.009 0.061 0.439 Leverage -0.425 0.001 -0.683 -0.167 Lev*D 0.838 0 0.518 1.158 Lev*R -0.281 0 -0.386 -0.176 Lev*D*R 1.324 0 1.078 1.570 Size 0.036 0.013 0.008 0.065 Size*D -0.020 0.253 -0.055 0.014 Size*R -0.209 0 -0.278 -0.140 Size*D*R 0.197 0 0.120 0.274 MTB 0.019 0.07 -0.002 0.040 MTB*D -0.023 0.081 -0.050 0.003 MTB*R -0.053 0.056 -0.107 0.001 MTB*D*R 0.029 0.383 -0.036 0.094 Constant -0.133 0.158 -0.317 0.052 N 1,071 Adj. R² 22.95%
In Table 5 coefficient 𝛽4 (D*R) is again significantly negative at the 1% level (p-value <0.01).
Again, the most important coefficient is 𝛽8, now it is negative yet not significantly (p-value =
38.9%). Once again, the coefficient on the interaction of negative returns with leverage 𝛽12 is
significantly positive (p-value < 0.01). Furthermore, the coefficients 𝛽16 and 𝛽20 are positive
again, but again only 𝛽16 is significantly positive (p-value < 0.01). Overall, these results neither support nor rejects hypothesis 2. This would imply that: The investments of firms are not influenced in the long-term if those firms applied a conservative accounting policy in the financial crisis.
24
Table 5
Regression analysis H2
Variable Coef. P>t [95% Conf. Interval]
D -0.036 0.766 -0.272 0.200 R 0.966 0 0.686 1.247 D*R -0.992 0 -1.338 -0.645 Capex -0.133 0.857 -1.584 1.318 Capex*D 0.565 0.51 -1.116 2.246 Capex*R 1.849 0.236 -1.209 4.906 Capex*D*R -1.466 0.389 -4.806 1.874 Leverage -0.474 0 -0.738 -0.209 Lev*D 0.851 0 0.522 1180 Lev*R -0.178 0.001 -0.286 -0.070 Lev*D*R 1.201 0 0.949 1453 Size 0.038 0.011 0.009 0.067 Size*D -0.022 0.229 -0.057 0.014 Size*R -0.203 0 -0.274 -0.133 Size*D*R 0.193 0 0.115 0.271 MTB 0.023 0.036 0.001 0.044 MTB*D -0.028 0.042 -0.055 -0.001 MTB*R -0.061 0.03 -0.116 -0.006 MTB*D*R 0.036 0.281 -0.030 0.103 Constant -0.150 0.12 -0.338 0.039 N 1,07 Adj. R² 20.38%
5.3 Analysis of the results
Given the previous results of the regression these will now be analyzed with referring to previous literature to try to give a logical explanation why these results occurred. For starters, the paper of Penman & Zhang (2002) will be discussed, since the hypotheses were built on the findings in their study. Possible explanations for why the hypotheses were not confirmed are that the paper of Penman and Zhang (2002) does not consider what an exogenous shock as the financial crisis can do the investment decisions of investors, in a financial crisis the overall trust is already low with the investors so the influence of the possibly misleading earnings forecast does not apply. As seen from the study from Bottazzi et al. (2012), investors tend to invest in the firms they trust. If the performance of firms is better than that of the competitors these investors will go for that, since the firms that used a conservative accounting policy had a better performance they
25 will attract more investments (Watts & Zuo, 2012; Francis et al., 2013). The ambiguity of earnings forecast just does not seem to be that relevant anymore to investors or they can make better forecasts themselves in 2015, considering that the research from Penman & Zhang (2002) comes from 2002. Technology and more availability of data about firms allows investors to make better investment decisions. Penman & Zhang’s (2002) theory does still hold for the second hypothesis, but there the model does not show enough significance. The capital expenditures could go down because of lost trust with the investors, but there is no proof yet.
There might still be a possibility that Penman & Zhang (2002) are right since the data could contain a survivorship bias. This is because firms that delisted during the period 2008 and 2015 are not considered in this sample. The firms that delisted might have done so because of poor results, which would partially confirm the theory of Penman & Zhang (2002). Since firms that are not receiving as much investments as their competitors could be out competed in the long-term. Yet, Watts & Zuo (2012) tried to avoid the survivorship bias and included it in another test, but both approaches yielded comparable results. Considering that the time window of Watts & Zuo (2012) is a lot shorter than in this study the survivorship bias could still influence the results. The problem is that it is hard to find out whether and to what extent the effect of survivorship bias has on the results of this study. There should be no influence on the capital expenditures by the remaining firms.
Not only the equity market side theory of Penman & Zhang (2002), but also the debt market theory of Gigler et al. (2009) cannot provide an explanation for these results. So, it does appear that the debt market is more efficient than Gigler et al. (2009) study mentions it is. The conservative reporting firms could attract enough investments to increase their firm value more than the non-conservative reporting firms. Maybe the assumptions Gigler et al. (2009) use are not as representative in practice as the ones used by other studies (Watts, 2003; Zhang, 2008; Watts & Zuo, 2012). Another explanation for the failure of Gigler et al.’s (2009) theory could be that the financial crisis is a very specific setting which is not compatible with theory they have outlined. The theory could still be applicable for the capital expenditures, because the conservative reporting firms would have made inefficient debt contracting decisions which makes that they are unable to still borrow the same amount of money they need for the capital expenditures.
If assumed that Penman & Zhang’s (2002) theory is incorrect as well as the theory of Gigler et al. (2009) than another theory should be found to clarify these results. A plausible reason for this result could be that, since the firms who practiced a conservative accounting
26 policy in the financial crisis outperformed their industry competitors (Watts & Zuo, 2012; Francis et al., 2013), will continue to do so in the future. This would be backed up by the positive significance of the coefficient on the interaction of negative returns with the total returns, 𝛽8. So,
the theory that is discussed in the literature section 2.3 would apply to this case. Yet, this would only be an explanation for hypothesis one and not for two. The coefficient 𝛽8 in hypothesis two
suggests a negative relationship, but it is not significant. This could mean that whether a firm applies a conservative accounting policy in the financial crisis does not influence the capital expenditures in the long-term. So, maybe the positive effect of increased capital expenditures during the financial crisis if accounting conservatism is practiced, which is seen in the study of Watts and Zuo (2012), is no longer in place after seven years. The non-conservative accounting policy practicing competitors have picked up to the same or even better level of capital expenditures.
Another theory could be deducted from the study in the equity market of Ramalingegowda and Yu (2012). Since their research to the long-term institutional investors claims that they demand conservatism of the firms they invest in (Ramalingegowda & Yu, 2012). This will lead to higher satisfaction of those long-term institutional investors, which leads to more investments and eventually higher firm value. Whereas individual investors are not interested in conservatism nor demand it from the firm, they have diverse ways of information to base their investment decision on (Ramalingegowda & Yu, 2012). This could explain why the theory of Penman and Zhang (2002) is incorrect and immediately provides a new one. Since the long-term institutional investors keep investing during the financial crisis their firm value will have increased more than their competitors in the long-term. Moreover, when the shock is over the long-term investors probably will not invest as much as they did, leading to less capital expenditures. So, this would also explain why the capital expenditures are lower or at the same level as the firm’s competitors.
Another possibility that exists is that certain qualified or skilled managers that implement conservatism are better aware of its benefits, by this knowledge they can make better use of it during the financial crisis. So, the difference eventually is just made by the quality of the managers and is not directly a conservatism benefit. But, Watts & Zuo (2012) found that the theory deducted from Ramalingegowda and Yu (2012) better reflects the influence on conservatism than the skills of managers. So, the governance role of the long-term institutional investors has more influence on conservatism and its outcome (Watts & Zuo, 2012). Furthermore, for the capital expenditure the same counts that the governance role is more important.
27 Zhang’s (2008) notion that it is easier for conservative firms to borrow money, could also be an explanation for the significant positive interaction of negative during-crisis returns with total returns. Opposed to the equity market sided theory of Ramalingegowda and Yu (2012), this theory comes from the debt market side. The findings of Zhang (2008) are also backed up with the study of Nikolaev (2010) in which the result was that conservative firms more easily get debt contracts with covenants, which is more reassuring for banks and makes them more willing to agree with a contract. Even if this is in a time of financial uncertainty. Watts and Zuo (2012) also confirmed this theory by showing that conservative firms were better able to attract money from lenders than non-conservative competitors. Since conservative firms can invest more than their non-conservative competitors, which will lead to a higher firm value in the long run. The strange thing is that the capital expenditures does not seem to be influenced, while it would make sense that those conservative firms are still able to borrow more money and invest it into their firm. Especially now the interest rate is a lot lower than during the financial crisis. A possible explanation is that firms first want to repay the debt of the loans they got in the financial crisis, with a higher interest rate before applying for new ones.
To conclude, in this analyzing part it is shown that it is very likely that Penman and Zhang’s (2002) theory does not hold for this empirical research. Only if the survivorship bias made the sample completely irrelevant, which it is not likely to do. The simplest explanation is that firms did better during the financial crisis if they applied conservative accounting (Watts & Zuo, 2012; Francis et al., 2013), so they will do so in the future is a possibility, but the question of the capital expenditures will remain unanswered. The theory that seems to fit best, from the equity market point of view, is the theory from Ramalingegowda and Yu (2012) because long-term institutional investors demand conservatism of the firms, which would satisfy them and make them invest more. Because of these investments, the conservative firms can increase their firm value more. Furthermore, when these long-term investors helped their firms through the financial crisis they might slow down with the investments, leading to less capital expenditures or at least more at the level of the non-conservative competitors. The possibility of better qualified or skilled managers is plausible, but according to the study of Watts and Zuo (2012) it is not as good of an explanation as the point made by Ramalingegowda and Yu (2012). The last option is that because it is easier for conservative firms to attract loans, they can invest that to increase their firm value in the long-term (Zhang, 2008). Especially, because even debt contracts with covenants are easier to get for conservative firms (Nikolaev, 2010). Debt covenants are more regular and stricter in times as the financial crisis. For the firms that are borrowing more during the financial crisis the interest rate is also higher. This could be a reason to slow down the
28 capital expenditures in the long-term after crisis. Overall, these studies (Nikolaev, 2010; Zhang, 2008) would provide a theory from the debt market side for the results. Of course, it is also possible that the real reason lies in the combination of the different theories outlined here from both the equity as well as the debt market side.
29
6 Conclusion
The purpose of this study is to investigate the relationship of conservatism during the financial crisis with the firm’s long-term value. Prior research show that there are different effects of conservatism, both positive and negative ones. Some studies contradict other studies, this makes it unclear what the absolute benefits are of accounting conservatism. The International Accounting Standards Board (IASB) is negative about accounting conservatism, the IASB even removed the definition from its Conceptual Framework. The reason the IASB gives for the removal of conservatism in its Conceptual Framework is that it would not deal with uncertainty adequately (Hellman, 2008). To test whether the IASB is right to do so, this studies tests what conservatism leads to in time of uncertainty (the period of the financial crisis).
There are two main reasons why this study is relevant. Firstly, because it is never done before. There is only research done into the short-term effect of conservatism on the firm value at the time of a financial crisis, but nothing is known of the long-term effect. Secondly, especially for the stakeholders it is important to know what accounting policy will yield the most long-term benefit for them. Specifically, in times of uncertainty, as the financial crisis was. The influence of conservatism is a much-debated topic. Some studies argue that it increases firm value (Watts, 2003; Lafond and Roychowdhury, 2008), other studies argue that it could destroy firm value (Penman & Zhang, 2002; Gigler et al., 2009). So, the research question of this study is: What effect does a conservative accountancy policy, during a financial crisis, have on the firm’s long-term value?
To answer this question this research takes the theory from Penman and Zhang (2002) as starting point. Penman and Zhang’s (2002) theory is that conservatism can be misleading to investors and that makes the investors unsatisfied. Conservatism is used to mislead the earnings forecast, because conservatism allows firms to make firms to make reserves, which they can use in times the firm is underperforming (Penman & Zhang, 2002). By using the reserves in tough times for the firm it shows a better result than it should, leading to wrongly forecasted earnings. Investors are unhappy with this ambiguity in the earnings forecast and lose trust in those conservative firms (Penman & Zhang, 2002). Bottazzi et al. (2011) find in their study that investors invest less in the firms they have less trust. Less investments would eventually lead to a lower firm value. This theory of Penman and Zhang (2002) should be reflected in the results of this study. The theory of Gigler et al. (2009), about the inefficiency of debt contracting of conservative reporting firm, is also a possible explanation for the predicted results of the hypotheses.
