Debt Structure and Conditional Conservatism
Abstract
We investigate whether characteristics of firms’ debt structure, beyond leverage, are associated with predictable variation in conditional conservatism. The contracting theory of conservatism holds that conditional conservatism is an efficient mechanism employed by an organization to address agency conflict arising from contracts with various parties. For firms with contracts that are associated with more agency conflict (asymmetric information, asymmetric payoffs and differing control rights) the potential benefit of a conservative reporting strategy should increase. We examine the following debt characteristics:(1) convertibility, (2) securitization,(3) seniority, and (4) placement. We find that debt characteristics thought to be associated with an upwards shift in thedemand for conservative reporting, based on the contracting theory framework, are associated with more asymmetric timely loss recognition. Thus, our evidence supports theassertion that the contracting theory links debt structure and conservatism. Also, our evidence suggests that the effect of debt on conservatism can be under or over-stated when firms, or sub-samples of firms, have heterogeneous debt characteristics.
Keywords: Conditional conservatism; debt structure; agency costs
JEL Classifications:M41; D21; D82; G14
1. Introduction
We examine whether conditional conservatism[1] varies in predictable ways based on the characteristics of debt structure, such as the convertibility, securitization,seniority and placement of debt. Using a contracting framework, Watts (2003a) asserts that conservatism is an efficient contracting mechanism that reduces the costs ofagency conflict between the firm’s different stakeholders. Agency conflict arises from contractual relationships in the presence of asymmetric information, asymmetric payoffs and differing control rights. Consequently, debt contracts are associated with a large degree of agency conflict between managers (shareholders)and debtholders due to debtholders generally being at an informational disadvantage, having an asymmetric payoff function and having limited control rights.
Consistent with the contracting theory of conservatism, numerous papers have documented a positive association between debt and conservatism. With a few notable exceptions, research on conservatism has treated debt as a monolithic block when testing the contracting theory of conservatism, considering only total leverage. However, there is heterogeneity in debt contracts that should alter the potential benefit of conservative reportingunder the contracting viewpoint, which should result in a shift in demand for conservative reporting when a firm uses one debt characteristic vs. another. Thus, exploring how heterogeneity in debt contracts affects conservatism should provide both a useful test of the contracting theory of conservatism and help more accurately quantify the impact of debt on conservatism across firms and sub-samples of firms. Our evidence suggests that the characteristics of debt appear to shift the demand for conditional conservatism in predictable ways, even after controlling for the level of leverage. This evidence supports the contracting theory as the mechanism linking debt to conservatism and shows that there is significant variation in conservatism associated with debt heterogeneity that is not captured by examining leverage alone.
Ball et al. (2008) examine the relation between capital structure and conditional conservatism in an international setting and provide evidence that the debt market is a primary source of conditional conservatism. In a discussion of Ball et al., however, Monahan (2008) calls for more evidence on the relation between capital structure and conditional conservatism.[2] He suggests that if the positive relation between debt market size and conditional conservatism documented by Ball et al. (2008) is consistent with a contracting explanation (Watts 2003a), then this interpretation should also apply to the relation between conditional conservatism and other debt attributes such as the placement, securitization, and maturity. The mechanism linking capital structure to the provision of conditional conservatism is ambiguous if we do not observe greater conditional conservatism for types of debt that should be associated with higher demandfor conservative reporting (consistent with a contracting explanation). Our descriptive statistics show that there is a large degree of variation in debt characteristics across firms. Also, Rauh and Sufi (2010) show that firms frequently adjust their debt structure even when total debt remains relatively stable. This observation indicates that studies examining the relationship between leverage and conditional conservatism will not fully capture the influence of debt on conservatism by ignoring debt heterogeneity across firms and sub-samples of firms.
We predict that the level of conditional conservatism will be higher when the characteristics of the firm’s debt structure are associated with a shift in the potential benefits of conservative reporting. Therefore, debt characteristics associated with relatively 1) higher levels of information asymmetry between creditors and managers, 2) greater divergence between creditors and shareholder payoff structures, or 3) lower creditor control rights should be associated with more conservative reporting by the firm.
We examine the following characteristics: (1) convertibility, (2) securitization, (3) seniority, and (4) placement. We find that firms with a greater proportion of debt that should be associated with higher demand for conservative reporting (non-convertible, unsecured, subordinated, and public debt) issue more conservative financial reports. The results hold after controlling for debt covenants. These findings are consistent with the arguments in Watts (2003a) that conservatism is an efficient contracting mechanism that helps mitigate agency conflicts associated with the firm’s contracts.
Due to the possibility of reverse causality (i.e., conservatism drives variation in debt characteristics) we include additional empirical analyses. Causality concerns are commonly faced by studies in the conservatism literature and thus several methods to address causality issues are available. For instance,LaFond & Watts (2008) explore the information role of conservatism, predicting that information asymmetry drives variation in conservatism, however it was their concern that the reverse could be true. To address this issue LaFond & Watts (2008) capitalize on the lead-lag relationship between information asymmetry and accounting conservatism. We employ a similar methodology. If, debt characteristics alter the demand for conservative reporting then we would expect to see the predicted variation in the level of asymmetric timely loss recognitionin the contemporaneous year when there are increases in debt characteristics in the previous year or the contemporaneous year. If, on the other hand, conservatism drives changes in debt characteristics then we would expect to see variation in asymmetric timely loss recognition that leads changes in debt-characteristics. The results of these lead-lag analyses, which we conduct on each of our debt characteristics, all suggest that reverse causality is not driving the associations we observe.
The study most closely related to ours is Khurana & Wang (2015), which examines the consequences of debt maturity for conservatism. Khurana & Wang argue that short maturity debt helps to mitigate the agency costs of debt, and document evidence consistent with short maturity debt being negatively associated with accounting conservatism. First, we corroborate Khurana & Wang’s findings, by showing that other debt features that should be related to the agency costs of debt arealso predictablyassociated with the provision of conservative financial reporting. Further, by examining a broad set of debt features simultaneously we are able to provide evidence on the joint impact of these features on accounting conservatism. This comprehensive test is important as debt features are inter-related and therefore a focus on one individual debt feature could result in misleading findings.[3]We find that Khurana & Wang’s findings related to debt maturity are robust to this approach.
Two other studies bear mentioning, Beatty et al. (2008) and Nikolaev (2010), which both examine the relationship between covenant restrictions and conditional conservatism. Beatty et al. (2008) examine whether more conservative financial reporting is associated with more conservative covenant restrictions and, in some cases, finds that they are. Nikolaev (2010), also examines debt covenants anddocuments a consistent positive relationship between the use of covenants and conditional conservatism. Our paper differs on a theoretical level, as our evidence relates more to the “substitutes” hypothesis discussed in Guay (2008) and Nikolaev (2010). That is, debt covenants are designed to mitigate moral hazard by shifting control rights to debtholders when the firm approaches financial distress, which should reduce the demand for conservative reporting (i.e., debt covenants and conservatism are substitutes). Nikolaev (2010) however, finds support for the “complements” hypothesis as being the dominant force in his setting. Under the complements hypothesis, by explicitly contracting on financial numbers debt covenants can increase the importance of financial numbers in triggering thetransfer of control rights, which increases the demand for conservative reporting.[4]
Our study contributes to the extant literature in several ways. First, we provide evidence supporting Watts (2003a)’s hypothesis that agency conflict arising from the firm’s contracts with outside parties generates demand for conditional conservatism. We base all of our hypotheses on the contracting framework, where shifts in asymmetric information, asymmetric payoffs and control rights related to debt contracts should alter the demand forconservative reporting in predictable ways. It is important to note that firms will issue debt with characteristics that will respond to the existence of the underlying agency problems and thus debt structure is a function of the pre-existing underlying contracting problems.Therefore, it is unclear ex-ante how total agency conflict and debt characteristics will be related. However, it is possible to predict that certain debt characteristics should help mitigate agency conflict more than others, thus shifting the demand for conservative reporting.
Our evidence bears out these hypotheses, as across several dimensions debt contracts that are more likely to mitigate agency conflict are associated with lessasymmetric timely loss recognition. This evidence complements the results reported by Ball et al. (2008) that debt market size is positively associated with asymmetric timely loss recognition. Our study provides greater assurance that the mechanism driving the association between conditional conservatism and capital structure is the variation in demand for conditional conservatism arising from the debt contracts rather than confounding effects associated with un-modeled heterogeneity across countries or firms.
The second contribution of this study is to provide a more complete account of how variation in accounting conservatism is driven by debt. We show the importance of recognizing that, contrary to the majority of extant studies examining capital structure and conditional conservatism, debt heterogeneity is associated with heterogeneity in conditional conservatism. We find that the economic significance of each of our individual debt characteristics is between 3-73% of our control for leverage. Our evidence suggests that the effect of debt on conservatism can be significantly under or over-stated when firms, or sub-samples of firms, have heterogeneous debt characteristics.
2. Prior Literature and Hypothesis Development
2.1. Prior literature
Capital structure and accounting conservatism
Watts (2003a, 2003b) defines accounting conservatism as the differential verifiability required for recognition of profits versus losses (asymmetric treatment of gains and losses). Prior studies advance a number of explanations for why conservative reporting is so pervasive including contracting, shareholder litigation, taxation, and accounting regulations. Watts (2003a, p.209) states that “under the contracting explanation, conservative accounting is a means of addressing moral hazard caused by parties to the firm having asymmetric information, asymmetric payoffs, limited horizons and limited liability.” Debt contracting has many of the characteristics identified by Jensen and Meckling (1976), Myers (1977), Watts (2003a) and many other studies as being associated with agency conflict.
Why firms supply more conservative financial reporting?
If the lenders’ benefits arising from conditional conservatism are not shared then the firm would not have an incentive to provide more timely loss recognition (which is a costly activity).[5] Lenders can share the expected benefits from conservatism with borrowers who pre-commit, or have a reputation[6] for providing more conservative reporting, by lowering the interest rate. If the reduction in the cost of debt is of first order benefit to borrowers, borrowers have an incentive to provide more conservative financial reporting (allowing them to enjoy the benefit of lower interest rates and/or increased debt capacity). Prior studies are consistent with this reasoning, finding that more conservative reporting is associated with a lower cost of debt (Ahmed et al. 2002; Zhang 2008). Firms’ incentive to provide conditional conservatism will vary based on the potential net benefit arising from doing so, and it is likely that firms have the discretion necessary to alter their level of reporting conservatism in response. For instance, Guay (2008, p.177) states that “although all public U.S. firms follow GAAP, this does not imply that all firms supply the same degree of conservatism in their financial reporting.” Many of the accruals that underpin the provision of conditionally conservative reporting, such as asset impairments, involve a high degree of estimation.[7]
Existing evidence on the relationship betweenleverage and conditional conservatism
Research on the relation between the level of debt and conditional conservatism is consistent with the positive association predicted by theory. Ahmed et al. (2002) investigate whether conservatism mitigates the conflicts between bondholders and stockholders over dividend policy. They use leverage as a proxy for conflicts in bondholder-shareholder dividend policy and find that leverage is significantly related to conservatism. They also find that firms with more conservative financial reporting experience a lower cost of debt. Zhang (2008) examines the efficiency gains from conditional conservatism in the debt contracting process and finds that conservatism benefits lenders ex-post by producing a timely signal of default risk in the form of accelerated covenant violations, and benefits borrowersex-ante through lower initial interest rates. Ball et al. (2008) use an international setting to examine whether conservatism is driven by equity markets or by debt markets. They provide evidence that the debt market, and not the equity market, is associated with the timeliness of loss recognition, supporting the significant impact of debt markets on accounting practice. Haw, Lee and Lee (2014) use a sample of private firms that issue public debt, finding that private firms increase their level of conservative reporting after the issuance of debt. In summary, these studies provide evidence that leverage is relevant to the existence and pervasiveness of conservatism in accounting.
A few studies have called into question the mechanism explaining the link between debt and conditional conservatism, as well as the robustness of the empirical evidence supporting the link. Monahan (2008) argues in his discussion of Ball et al. (2008) that the paper’s conclusion that debtholders represent the primary source of demand for timely accounting reports[8] is premature for the following reasons: (1) problems in research design and (2) the lack of additional evidence on the relationship between accounting conservatism and debt-instrument attributes that supports a debt contracting explanation. Several studies question the robustness of cross-sectional tests based on the Basu (1997) model. Most germane to our study is Lawrence et al. (2013), which shows that leverage (and debt covenants) is not significantly associated with conditional conservatism when controls for the beginning book-to-market value are included in the research design. Studies that question the robustness of the Basu model in general include Dietrich et al. (2007) and Patatoukas and Thomas (2011). Ball et al. (2013) addresses these concerns and suggests additional control variables that should be included when estimating the Basu (1997) model to mitigate concerns about cross-sectional bias. Consequently, the design of our empirical analyses follows the recommendations of Ball et al. (2013).
2.2. Hypothesis development
Moving beyondtotal leverage, variation in the firm’s debt structure should be associated with further variation in conservatism. We expect that greaterasymmetries in payoffs and information and lower debtholder control rights will be associated with anupwards shift in the demand for conservative reporting, based on the contracting theory, as these factors are associated with higher agency conflict. However, because debt contracting characteristics are determined at least in part as a response to the underlying agency problems at the firm (i.e., endogenously determined), debt characteristics that should help mitigate agency conflict could actually be positively associated with the level of agency conflict in the cross-section of firms.[9] Therefore, we discuss all of our hypotheses in terms of an expected shift in the demand for conservative reporting that should result from a particular contractual term.
Due to debtholders’ asymmetric payoff function, relative to shareholders, they are much more concerned with the lowerend of the earnings and net asset distributions. Debtholders in general are likely to demand timelier reporting of bad economic news relative to good news because the value of their investment is generally more sensitive to bad news (Ball, et al., 2008).Thus, the asymmetric payoff structure in isolation can result in demand for asymmetric timely loss recognition. As the value of debt become more asymmetrically sensitive to negative changes in the value of the firm, we expect that debtholders will find the timely reporting of bad news incrementally more informative. However, it is also important to consider shareholders’ and managers’ incentives in the presence of asymmetric payoffs, which provide the shareholders of the firm an opportunity to take actions that may reduce firm value, in an attempt to transfer wealth from lenders to shareholders (Myers 1977; Smith and Warner 1979). These debt value-reducing actions include underinvestment[10], asset substitution, dividend payments, and claim dilution (issuance of additional debt). As the payoff structure becomes more asymmetric, the shareholders/managers of the firm will have greater opportunities to expropriate debtholder wealth (Jesen and Meckling, 1976).As these potential agency costs of debt increase, holding constant debtholders’ information and control rights, the potential benefits to debtholders of timely loss recognition should increase. Together, these arguments suggest that there should be an increase in the demand for asymmetric timely loss recognition when debt contracts result in a more asymmetric payoffstructure.