Asset specificity and conditional accounting conservatism

• Published date: 01 July 2018
• Author: Qingyuan Li,Li Xu
• DOI: http://doi.org/10.1111/jbfa.12308
• Date: 01 July 2018

DOI: 10.1111/jbfa.12308

Asset specificity and conditional accounting

conservatism

Qingyuan Li1Li Xu2

1Schoolof Economics and Management, Wuhan

University,Luojia Hill, Wuchang District 430072,

China

2Collegeof Business, Washington State Univer-

sity,Vancouver,WA 98686, USA

Correspondence

LiXu, College of Business, Washington State

University,Vancouver,WA 98686, USA.

Email:li.xu3@wsu.edu

JELClassification: M41, D21, G14

Abstract

Asset specificity, the redeployability of an asset to alternative uses,

is a key determinant of an asset's resale value. Asset specificity

has a direct impact on a firm's ongoing fair value determination,

bankruptcy risk, liquidation value, and abandonment option. We

document a significant negative association between asset speci-

ficity and conditional conservatism. Further tests reveal that this

inverse relation manifests as bad news being less quickly incorpo-

rated in earnings as asset specificity increases. We find no differ-

ence in the extent to which good news is delayed in earnings for

firms conditional on asset specificity. In addition, the documented

association is stronger when asset specificity arises from lower

within-industry acquisition activity.The association is also more pro-

nouncedfor firms that are in less competitive industries, have institu-

tionalinvestors, have limited access to the public debt market, and/or

have more unsecured debt. Our findings are noteworthy for regula-

tors and researchers given the recent interest in the determinants of

conservatism.

KEYWORDS

asset specificity, conditional conservatism, timely loss recognition

1INTRODUCTION

Asset specificity, a keydeterminant of asset values, captures the extent to which an asset is fungible or redeployable

to alternative uses outside of the firm (Williamson, 1985). General-purpose assets, such as land, havealternative uses

and typically experienceliquid markets. On the other hand, highly specific assets (such as steel mills) are tailored to the

needs of the firm and have limited uses other than their intended objective.Thus, they may not be easily transferrable.

Moreover, highly specific assets provide less flexibility in adjusting business operations, and as a result, increase

bankruptcy risk; they are also difficult for outsiders to value and monitor effectively(Klein, Crawford, & Alchian, 1978;

Vincente-Lorente, 2001; Williamson, 1975, 1985). In addition, assets that are more specific have lower liquidation or

J Bus Fin Acc. 2018;45:839–870. wileyonlinelibrary.com/journal/jbfa c

2018 John Wiley & Sons Ltd 839

840 LI ANDXU

abandonment values, and are likely to be worth far less in liquidation or abandonment than when the firm is a going

concern (Ronen & Sorter,1972; Williamson, 1988).1

Timely loss recognition (TLR), a common financial reporting practice (Watts, 2003a), is especially germane to the

study of specific assets given their lower liquidation or abandonment values and the possibility of significant write-

downs. On the one hand, prior research documents that firms with specific assets are associated with large information

asymmetry because of their relative uniqueness (Aboody & Lev,2000; Kothari, Laguerre, & Leone, 2002). To mitigate

value reductions resulting from information asymmetries between managers and outside equity investors,the demand

for conservatism increases (LaFond& Watts, 2008). On the other hand, for firms with more specific assets, conditional

conservatism is more likely to communicate (false) low signals when in fact the future is favorable. The costs of false

signals are also high for firms with more specific assets. In addition, the low liquidation value feature of specific assets

can attenuate agency-based demand for TLR (Myers & Rajan, 1998). Moreover,firms with more specific assets are less

likely to write down the assets due to lack of readily availableresale values as a verifiable benchmark for current asset

values (Schonberger, 2015). All the above mayreduce demand and supply for conservatism. Thus, whether and how

asset specificity is associated with the demand for conservative financial reports is an empirical question. However,to

the best of our knowledge, the relation between asset specificity and TLR remains largely unclear.

Tofill this void, in this paper we investigate the association between asset specificity and conditional conservatism,

which requires a higher standard of verifiability for recognizing good news as gains than for recognizing bad news

as losses (Basu, 1997). In contracting theory,conservatism evolves as an efficient contracting mechanism to mitigate

agency friction between contracting parties (Ahmed, Billings, Morton, & Stanford-Harris, 2002).

We examine and find a robust significant negative association between asset specificity and conditional conser-

vatism. This finding conforms to the notion that asset specificity may reduce the demand for conservative financial

reporting. This result holds after controlling for other factors that are often associated with the demand for con-

servatism such as the market-to-book ratio (Roychowdhury & Watts,2007), leverage (Ahmed et al., 2002; LaFond &

Roychowdhury, 2008), size (Givoly, Hayn, & Natarajan, 2007), product competition (Dhaliwal, Huang, Khurana, &

Pereira, 2014), firm age, standard deviation of stock returns (Khan & Watts, 2009), and firm/yearfixed effects. Fur-

ther tests reveal that the inverseassociation between asset specificity and conditional conservatism manifests as bad

news being less quickly incorporated into earnings as asset specificity increases. We find no difference in the extent

to which good news is delayed for firms conditional on their prior year asset specificity.Furthermore, we find that this

effect is stronger when asset specificity arises from lower within-industry acquisition activity.

We also examine the cross-sectional difference in the association between asset specificity and conservatism, and

conjecture that the negative association is more pronounced for firms in less competitive industries, with institutional

investors,with access to the public debt market, and with non-secured debts based on the following. First, with intense

industry competition, managers, shareholders, and creditors can easily find comparable firms and use their account-

ing numbers as benchmarks, which reduces the likelihood of false signals caused byconditional conservatism. Second,

the presence of institutional investors may also affect the association between asset specificity and conditional con-

servatism. Institutional investors are more sophisticated (e.g., Sias, Starks, & Titman, 2006; Walther,1997), and are

more likely to understand the costs of false signals due to overly conservative accounting for firms with high specific

assets. Next, a firm's access to the public debt market may affect the association between asset specificity and con-

ditional conservatism. Creditors may demand less conservatism when the cost of false signals associated with con-

ditional conservatism is high, and can demand more conservatism when agency costs are high. Gaining access to the

publicdebt market may significantly increase the agency costs of debt, which may be more severe for firms with specific

assets since the unique features of specific assets make them difficult to monitor.Last, for non-secured debtholders,

no assets serve as collateral for debt. Therefore, managers havegreat discretion over firm assets, which increases the

risk of wealth expropriation. When asset specificity increases, it becomes much costlier for managers to sell assets.

1Despitethese potentially unattractive features, firm-specific assets offer benefits. For instance, they can generate quasi-rents for the firm (Klein et al., 1978),

enhance shareholder value (Teece,1986), and provide a competitive advantage in the marketplace (Ireland, Hitt, & Sirmon, 2003; Lippman & Rumelt, 1982;

Rumelt,1991).

LI ANDXU 841

Thus, managers are less likely to expropriate value since they gain little compared with the benefits they gain from

operating these assets (Myers & Rajan, 1998; Morelli, 2001). Consistent with the above conjectures, we find that the

negative association between asset specificity and conditional conservatism is more accentuated when firms are in

less competitive industries, have institutional investors,have limited access to the public debt market, and have more

unsecured debt. The findings are robust to alternative measures for conditional conservatism and asset specificity,as

well as estimations using the Fama and MacBeth (1973) approach.

Our study makes at least three major contributions to prior literature.First, our study adds to our understanding of

the consequences of asset specificity. Although asset specificity has been studied in finance, economics, and manage-

ment literature, it has rarely been exploredin an accounting context. For example, prior work in the finance literature

investigatingasset specificity addresses its effects on capital structure (e.g., Shleifer & Vishny, 1992) or the cost of cap-

ital (e.g., Benmelech & Bergman, 2009; Ortiz-Molina & Phillips, 2014). Tothe best of our knowledge, we are the first to

provide direct evidence that links asset specificity,as a salient characteristic of a firm's assets, to its financial reporting

practices.

Second, prior studies in accounting have been predominantly focused on examining the association between con-

servatism and bondholder–shareholder conflicts regarding dividend policy (Ahmed et al., 2002); information asymme-

try (LaFond & Watts, 2008);institutional ownership (Ramalingegowda & Yu, 2012); managerial ownership (LaFond &

Roychowdhury, 2008); managerial overconfidence (Ahmed & Duellman, 2013); and product market competition

(Dhaliwalet al., 2014). Our findings highlight how asset specificity is inversely associated with conditional conservatism

and identifies conditions under which this association may be accentuated (attenuated).

Last, findings reported in this study extend the understanding of the nature of the economic forces associated with

the demand for conservatism (Watts, 2003a, b). The results in this paper support the claim of policymakers and aca-

demics that certain asset characteristics may contribute to the variations in practice of conservative financial report-

ing. As discussed in Kausar and Lennox(2017), there is a lack of conservatism in the balance sheet. We show that asset

specificity can be associated with the lack of conservatism.

This paper is organized as follows. Section 2 develops the hypotheses. Section 3 describes the sample. Section 4

reports the results of the primary empirical tests. Section 5 reports the resultsof additional empirical tests. We present

robustness tests in Section 6. Section 7 concludes.

2HYPOTHESES DEVELOPMENT

In contracting theory (Ahmed et al., 2002), conservatism evolves as an efficient contracting mechanism to mitigate

agency friction associated with (1) asymmetries in information and loss functions of the contracting parties, and

(2)an inability to verify the more informed parties'private information. Conditionally conservative accounting requires

a higher degree of verification for gains than for losses, and accelerates the recognition of bad news in earnings. This

provides assurance to external creditors and shareholders that gains will not be overstated and losses understated.2

Therefore, creditors and shareholders may demand more conditional conservatism if they face larger agency costs

(LaFond & Watts, 2008). However, in spite of the usefulness of conditional conservatism in reducing agency costs, it

can also increase the likelihoodof false communication by sending low signals when in fact the future is favorable. This

can impose real economic costs on firms (Leuz, 2001). For example, conservativeaccounting may falsely trigger debt

covenants to generate unnecessary renegotiation costs between the firm and its lenders, and unnecessarily restrict

2Because managers’ liability is limited bytheir wealth, shareholders prefer to defer compensation payments until there is verifiable evidence of the result of

themanagers’ actions. By recognizing e conomic losses of bad projects in a more timely manner (i.e., during managers’ tenure),conservative accounting sends

anearly warning signal to shareholders to investigate these losses. The resulting revelation of bad projects can damage the managers’ reputation and threaten

their job security.In addition, to the extent that managerial compensation is tied to earnings, conservatism penalizes managers for their failures (economic

losses)in a timely manner but defers rewards for their successes (economic gains) until benefits are realized, thereby reducing managers’ incentive and ability

tooverstate the value they create. Debtholders also demand conservatism in financial reports, as conditional conservatism can improve contracting efficiency

byproviding a lower bound estimate of net assets for debt contracting (Watts, 2003a).