Author's personal copy - Booth School of Businessfaculty.chicagobooth.edu/Regina.Wittenberg...

This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/copyright

http://www.elsevier.com/copyright

Author's personal copy

The role of information asymmetry and financial reporting quality indebt trading: Evidence from the secondary loan market$

Regina Wittenberg-Moerman �

The University of Chicago Graduate School of Business, 5807 South Woodlawn Avenue, Chicago, IL 60637, USA

a r t i c l e i n f o

Article history:

Received 22 August 2006

Received in revised form

30 July 2008

Accepted 18 August 2008Available online 5 September 2008

JEL classification:

M41

D82

G14

G21

L10

Keywords:

Secondary loan trading

Information asymmetry

Timely loss recognition

Accounting conservatism

a b s t r a c t

I explore which firm and loan characteristics decrease or exacerbate information

asymmetry in the trading of private debt. I find that loans of public firms, loans with an

available credit rating, loans of profit firms and loans syndicated by more reputable

arrangers are traded at lower bid–ask spreads, while revolvers, distressed loans and

loans issued by institutional investors are associated with higher information costs.

I also find that timely loss recognition reduces the bid–ask spread. This finding suggests

that conservative reporting decreases information asymmetry regarding a borrower and

increases the efficiency of the secondary trading of debt securities.

& 2008 Elsevier B.V. All rights reserved.

1. Introduction

The U.S. syndicated loan market bridges the private and public debt markets and provides borrowers and lenders with ahighly valuable source of financing and investment. The market consists of a wide-range primary loan market wheresyndicated loans are originated, and an active secondary market where loans are traded after the close of primarysyndication. In the past 20 years, the syndicated loan market has been one of the most rapidly growing and innovative

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/jae

Journal of Accounting and Economics

ARTICLE IN PRESS

0165-4101/$ - see front matter & 2008 Elsevier B.V. All rights reserved.

doi:10.1016/j.jacceco.2008.08.001

$ I am especially grateful to the members of my dissertation committee: Ray Ball (Chair), Philip Berger, Douglas Diamond, Douglas Skinner and Abbie

Smith for many insightful comments and helpful discussions. I also thank Sudipta Basu, Zahi Ben-David, John Core, Steven Crawford, Ellen Engel, Wayne

Guay, Wendy Heltzer, Eugene Kandel, Randall Kroszner, Christian Leuz, Thomas Lys, Edward Riedl, Jonathan Rogers, Darren Roulstone, Sugata

Roychowdhury (the referee), Gil Sadka, Haresh Sapra, Tony Tang, Robert Verrecchia, Ross Watts (the editor), Mark Flannery (the discussant) and

participants at the Bank Structure Conference, Jayanthi Sunder (the discussant) and participants at the AAA 2006 Annual Meeting, participants at the 2007

Center for Accounting Research and Education Conference, and seminar participants at Duke University, Emory University, the Federal Reserve Bank of

Chicago, Harvard University, Massachusetts Institute of Technology, Northwestern University, the Ohio State University, Stanford University, the

University of California, Berkeley, the University of California, Los Angeles, the University of Chicago, the University of Illinois at Urbana-Champaign, the

University of Minnesota, the University of North Carolina at Chapel Hill, and the University of Pennsylvania for valuable comments and suggestions. I

thank the Loan Pricing Corporation for letting me use their loan trading data. I gratefully acknowledge the financial support of the University of Chicago,

Graduate School of Business and the Wharton School, University of Pennsylvania.� Tel.: +1773 8347256.

E-mail address: [email protected]

Journal of Accounting and Economics 46 (2008) 240–260


sectors of the U.S. capital market (Yago and McCarthy, 2004). U.S. firms obtain over $1 trillion in new syndicated loans eachyear, which represents more than 50% of the annual U.S. equity and debt issuance (Weidner, 2000). The trading ofsyndicated loans has expanded from $8 billion in 1991 to $144.6 billion in 2003, a compound annual growth rate of 27%.

I employ a sample of traded syndicated loans to examine information asymmetry in the trading of debt securities. Therole of information asymmetry in trading on secondary markets has long been of interest to researchers in accounting andfinance. The existing literature, however, examines information asymmetry mainly in the context of equity markets,1

leaving information asymmetry in debt markets largely unexplored.The secondary loan market is a promising empirical setting to examine information asymmetry in trading of debt

securities. First, the secondary loan market provides unique information regarding the trading of private debt issues.Second, the secondary loan market involves an exceptionally wide range of debt securities, a range that includes loansissued to public and private firms, as well as investment grade and leveraged (high yield) loans.

Third, the secondary loan market involves informed traders, such as the arranger and syndicate participants, who have aconsiderable information advantage over uninformed traders. Through private communication with a borrower, informedlenders receive quarterly or monthly financial disclosures, covenant compliance information, amendment and waiverrequests, financial projections, and even plans for acquisitions or dispositions. While this information is critical toevaluating a borrower’s financial health, it is usually either unavailable for uninformed market participants or it onlybecomes available after a considerable delay. As a result, informed lenders are able to get a more timely and preciseevaluation of the borrower’s traded loans compared to an uninformed traders’ evaluation, which is based primarily onpublicly available information.

Because information asymmetry plays such an important role in the secondary loan market, it is natural to pose thequestion of how information asymmetry is resolved in loan trading. In this paper, I examine which firm- and loan-specificcharacteristics decrease or exacerbate information asymmetry, as estimated by the bid–ask spread in the loan trade. First, Iexamine firm and loan characteristics suggested by prior research as generally related to information asymmetry. Second, Iexamine the unique characteristics of the information environment of syndicated loans. Third, controlling for these morereadily recognizable determinants of information asymmetry, I examine whether there are aspects of a borrower’s financialreporting quality that decrease information asymmetry in the loan trade. While I address a number of reporting qualitycharacteristics, I place particular emphasis on exploring the effect of timely loss recognition.

It is important to note that the secondary loan market is a natural empirical setting in which to examine the importanceof timely loss recognition. Since debt holders’ returns are mainly determined by the downside region of a borrower’searnings distribution, debt holders are more sensitive to borrowers’ losses than they are to borrowers’ profits. Leftwich(1983), Watts and Zimmerman (1986), Watts (1993, 2003a, b) and Holthausen and Watts (2001) argue that the demand fortimely recognition of losses is driven, at least partially, by the reporting demand of debt markets. By providing evidencethat the degree of financial reporting conservatism increases with the importance of a country’s debt markets, Ball et al.(2008b) also support the significant impact of the debt markets on accounting practice.

I expect timely loss recognition to decrease information asymmetry in loan trading by enhancing corporate governanceof a borrower and by increasing the amount and quality of information available to secondary loan market participants. Asa governance mechanism, timely loss recognition may reduce information asymmetry via two routes. First, timely lossrecognition increases debt contracting efficiency. By triggering ex-post violations of debt covenants in a timely manner,timely loss recognition allows lenders to more rapidly employ their decision rights following economic losses (Watts andZimmerman, 1986; Watts, 1993, 2003a; Ball, 2001). Because syndicated loans are subject to tight and numerous financialcovenants, I expect timely loss recognition to significantly increase the efficiency of syndicated loan contracts.

Second, conservative financial reporting generates an understatement of net asset values and therefore facilitatesmonitoring by debt holders and reduces deadweight agency costs (Watts and Zimmerman, 1986; Watts, 1993, 2003a; Ball,2001). As suggested by Watts (2003a), lenders rely on verifiable lower bound measures of net assets both in assessing apotential loan and in monitoring the borrower’s creditworthiness after a loan issuance. Further, conservative reportingconstrains managers’ opportunistic behavior. This reduces managers’ incentives to transfer wealth to themselves eitherfrom debt holders or from the firm as a whole (Watts, 2003a, b; LaFond and Roychowdhury, 2008; Roychowdhury andWatts, 2007). Relying on these arguments, I anticipate that by enhancing corporate governance of the borrower, timely lossrecognition decreases information asymmetry in secondary loan trading.

It is important to emphasize that conservative financial reporting also fulfills an important role in providing informationto investors (Watts, 2003a; Khan and Watts, 2007; LaFond and Watts, 2008). Because managers have incentives to discloseinformation about unrealized gains but to withhold information regarding losses, conservative financial reporting not onlyprovides a ‘‘hard’’ source of information regarding a firm’s performance, but it also provides investors with a control foralternative ‘‘soft’’ sources of information (Watts, 2003a).

By increasing the amount and quality of public information available to uninformed investors, I expect timely lossrecognition to decrease information asymmetry between informed and uninformed traders in debt securities. Morespecifically, timely loss recognition induces an early revelation of the changes in a borrower’s credit quality and therefore

ARTICLE IN PRESS

1 See, for example, Copeland and Galai (1983), Glosten and Milgrom (1985), Kyle (1985), Amihud and Mendelson (1986), Diamond and Verrecchia

(1991), Botosan (1997), Leuz and Verrecchia (2000), Easley et al. (2002), Easley and O’Hara (2004), Ertimur (2004) and Schrand and Verrecchia (2005).

R. Wittenberg-Moerman / Journal of Accounting and Economics 46 (2008) 240–260 241


increases the borrower’s transparency. As a result, timely loss recognition allows uninformed loan traders who do notpossess private sources of information regarding the borrower to get a more timely and precise evaluation of the borrower’straded loans. Moreover, by enhancing the transparency of a borrower, timely loss recognition allows uninformed loanbuyers to perform more efficient post-sale monitoring of the borrower. Both of these effects reduce adverse selection in thesecondary loan trade and are therefore expected to lead to lower information asymmetry in trading the loans of borrowerswho incorporate economic losses in a timely manner.

The empirical findings are consistent with the predicted relations. There is clear evidence that publicreporting decreases information asymmetry in loan trading. The availability of firm-specific and/or loan-specificcredit ratings also decreases information costs in the loan trade. Further, consistent with the higher uncertaintyassociated with revolvers, I find that revolving loans are traded at higher bid–ask spreads. Empirical evidence alsosuggests that the loans of profitable firms are traded at significantly lower spreads than are the loans of firms experiencinglosses.

The empirical findings also confirm that the bid–ask spread is strongly associated with the unique characteristicsof the information environment of syndicated loans. Consistent with the primary role of the arranger of syndi-cation in resolving information asymmetry regarding the borrower, I find that loans syndicated by more reputablearrangers are traded at significantly lower spreads. In addition, I find that the considerable information advantageof informed traders in the trading of distress loans (loans traded at a bid price below 90% of their par value) is mani-fested in the significantly higher bid–ask spreads of these loans. To further address the distinctive features of syndicatedloans, I distinguish between loans issued by institutional investors and loans issued by banks. Institutional loans,typically issued with long maturities and back-end-loaded repayment schedules, are traded at higher spreads than bankloans are.

Finally, I find evidence that the timely incorporation of economic losses in a borrower’s financial statements reduces thebid–ask spread at which its loans are traded. The effect of timely loss recognition on the loan trading spreads is bothstatistically and economically significant. These results demonstrate that by improving the efficiency of a borrower’s loanagreements and by enhancing corporate governance and transparency of a borrower, timely loss recognition decreasesinformation asymmetry in the secondary loan trade. Overall, the analysis presented in this paper provides unique empiricalevidence that timely loss recognition reduces the information advantage of informed traders and increases the efficiency ofthe secondary trading of debt securities.2

To further examine the impact of financial reporting quality on loan trading, I investigate the relation between thebid–ask spread and timely gain recognition. Even though timely gain recognition improves the timeliness of accountingearnings and therefore is expected to make earnings a more informative measure of a borrower’s performance (Ball andShivakumar, 2006; Guay, 2006; Guay and Verrecchia, 2006), I do not find that it decreases the bid–ask spread. Timely gainrecognition also does not decrease the bid–ask spread of distress loans, for which good news is more important inevaluating lenders’ claims. The results also show that the overall timeliness of a borrower’s financial reporting is notassociated with the bid–ask spread in the loan trade. The insignificant relation between timely gain recognition, overalltimeliness and the bid–ask spread suggests that these attributes of accounting reporting do not decrease informationasymmetry regarding a borrower’s performance and creditworthiness.

The analysis presented in this paper broadens our understanding of how information asymmetry is resolved in thetrading of private debt securities. Prior research addresses loan trading primarily by examining banks’ incentives for loantrading and by testing price formation across the loan, bond and equity markets.3 To the best of my knowledge, this paper isthe first to examine the determinants of the bid–ask spread in the loan trade.

This study also contributes to the literature that examines the influence of accounting conservatism on debt markets.Ahmed et al. (2002) and Zhang (2008) argue that timely loss recognition reduces the cost of debt capital. By providingevidence that timely loss recognition reduces the bid–ask spread at which loans are traded, my paper documents thatconservative reporting decreases information asymmetry regarding the borrower. To the best of my knowledge, this paperis the first to document the efficiency gain from timely loss recognition in the trading of debt securities on secondarymarkets.

Finally, this paper contributes to the literature which explores the relation between accounting conservatism andinformation asymmetry. As suggested by Khan and Watts (2007) and LaFond and Watts (2008), conservative reporting is anequilibrium response to mitigate agency problems between managers and equity investors. This paper extends this line ofresearch and documents that conservative financial reporting helps resolve information asymmetry between informed anduninformed traders in debt securities.

The following section provides a brief description of the secondary loan market. The third section discusses the sourcesof information asymmetry in loan trading. The fourth section describes the data. The fifth section focuses on the researchdesign. The sixth section discusses empirical findings. The seventh section concludes.

ARTICLE IN PRESS

2 Copeland and Galai (1983), Glosten and Milgrom (1985) and Kyle (1985) confirm that information asymmetry between potential buyers and sellers

introduces adverse selection and reduces liquidity in the secondary markets. Following this line of research, by ‘‘more efficient secondary trading’’, I mean

more liquid trading, which is reflected in relatively lower bid–ask spreads.3 See Pavel and Phillis (1987), Pennacchi (1988), Gorton and Pennacchi (1995), Froot and Stein (1998), Demsetz (2000), Cebenoyan and Strahan

(2004), Allen and Gottesman (2005) and Altman et al. (2006).

R. Wittenberg-Moerman / Journal of Accounting and Economics 46 (2008) 240–260242


2. The secondary loan market: background and development

Secondary loan sales occur after the close of primary syndication; loan sales are structured as either assignments orparticipations. When interests in the loan are transferred by assignment, the buyer becomes a direct signatory to the loan.In participation, the original lender remains the holder of the loan and the buyer takes a participating interest in theexisting lender’s commitment (Standard & Poor’s, 2003, 2006). While assignments usually require the consent of both theborrower and the arranger for the loan sale, in participations such consent is almost never required. The majority of loansales in the secondary loan market are performed via assignment. Today, loan sales are arranged through loan trading desksin more than 30 institutions which act as market makers in the secondary loan market (Taylor and Yang, 2004).

The secondary loan market has grown rapidly in recent years, with trading volume increasing from $8 billion in 1991 to$144.6 billion in 2003 (Loan Pricing Corporation (LPC), 2003). The market expanded in both par and distressed loans; thetrading volume of loans traded at par and of distressed loans respectively reached $87 billion and $57 billion in 2003.Leveraged loans (defined by LPC as loans rated below BBB- or unrated and priced at the spread equal or higher than 150 bpsabove Libor) represent the largest and fastest growing part of the secondary loan market. Since 2001, trading of leveragedloans has constituted 80% of the total value of par loan trades.

The involvement of institutional investors in the secondary loan market has increased considerably with the market’sdevelopment. Banks, finance companies, loan participation mutual funds (prime funds) and collateralized loan obligations(CLOs) constitute the main secondary loan market participants. Prime funds are mutual funds that invest in leveragedloans; for the most part, prime funds are continuously offered funds with quarterly tender periods or true closed-end,exchange-traded funds (Standard & Poor’s, 2003, 2006). The CLOs purchase assets subject to credit risk (such as syndicatedloans and mainly leveraged syndicated loans), and securitize them as bonds of various degrees of creditworthiness.Additionally, hedge funds and pension funds are increasing their activity in loan trading (Yago and McCarthy, 2004).

Several reasons contributed to the strong growth in loan sales. New bank regulatory requirements, such as the 1989Highly Leveraged Transaction guidelines and the 1988 Basel Capital Accord, encourage banks to decrease their credit riskexposure (Altman et al., 2006; Barth et al., 2004). Additionally, the adoption of SEC Rule 144A in 1990 provided a safe-harbor relief from the registration requirements of Section 5 of the Securities Act of 1933 for the resale of privately helddebt and equity securities to qualified institutional buyers (QIB) (Yago and McCarthy, 2004). QIB is defined as an institutionthat owns and manages $100 million ($10 million in the case of a registered broker-dealer) or more in qualifying securities.The objective of Rule 144A was to increase the efficiency and liquidity of the U.S. market for equity and debt securitiesissued in private placements by allowing large institutional investors to trade restricted securities more freely with eachother. The foundation of the Loan Syndication and Trading Association (LSTA) in 1995 was an additional factor thatstimulated the development of the secondary loan market (Hugh and Wang, 2004).

Development of the secondary loan market coincided with improvements in the market’s transparency. In 1987,LPC initiated the publication of Gold Sheets, which provide a detailed analysis of market trends, loan price indexes andnews coverage. In the late nineties, LSTA created standard documentation for the primary and secondary loan markets and,jointly with LPC, started providing mark-to-market loan pricing based on dealer quotes (Yago and McCarthy, 2004). Theseinitiatives significantly increased the amount of information available to market participants. In addition, Standard &Poor’s, Moody’s and Fitch-ICBA started rating syndicated loans in 1995. The rapid increase in the number of rated loansreduced information asymmetry in the secondary loan market.

3. The sources of information asymmetry in the secondary loan market

3.1. The role of information asymmetry in loan trading

The secondary loan market involves informed market participants, such as arranger and syndicate participants, whohave access to a firm’s private sources of information regarding its creditworthiness and liquidity. Even though the creditagreement between a loan syndicate and a borrowing firm is available to a loan’s buyer, there is ongoing communicationbetween the borrower and the syndicate lenders beyond the credit agreement. This communication includes quarterly ormonthly financial disclosures, covenant compliance information, amendment and waiver requests, financial projections,and even plans for acquisitions or dispositions (Standard & Poor’s, 2003, 2006). The transfer of this private information ismade in accordance with confidentiality agreements between the borrower and the syndicate and is typically out of thepublic domain.

The information disclosed by the borrower via private communication with syndicate lenders is usually very importantto evaluating the borrower’s financial health; this information becomes available to the general public only when aborrower issues a press release or, if it is a public firm, when it files with the SEC. As a result, syndicate lenders have aconsiderable information advantage in the secondary loan trade compared to uninformed secondary market participants.

It is important to note that uncertainty regarding a borrower and/or a borrower’s traded loans is positively correlatedwith information asymmetry between informed and uninformed traders in the secondary loan market. To illustrate thispoint, consider two borrowers—one with low uncertainty regarding its future profitability and creditworthiness andanother that operates in a high uncertainty environment. Access to private sources of information regarding the

ARTICLE IN PRESS



‘‘low uncertainty’’ borrower does not grant informed traders considerable information advantage because of the lowuncertainty regarding the borrower’s loan contractual repayments. However, private information regarding the‘‘high uncertainty’’ borrower allows informed lenders to get a more timely and precise evaluation of the borrower’straded loans compared to an uninformed trader’s evaluation, which is based primarily on publicly available information. Inother words, higher uncertainty regarding the borrower increases the benefits of private information available to informedtraders and consequently causes higher information asymmetry in the secondary loan trade.

Two additional characteristics of the secondary loan market further underscore the importance of informationasymmetry in the loan trade. First, the vast majority of secondary loan market participants are large institutions, such asbanks and institutional investors. Diamond and Verrecchia (1991) demonstrate that large traders are especially concernedabout adverse selection in the secondary trade.

Second, the majority of loan trading involves leveraged loans. These loans, issued by more risky borrowers, require anextensive bank monitoring (Diamond, 1991). As established by Diamond (1984, 1996) and Lummer and McConnell (1989),banks provide unique services in the form of credit evaluation and monitoring, and therefore decrease the informationasymmetry associated with risky borrowers. However, it is not clear whether a loan originating bank is motivated tocontinue a loan’s monitoring after a portion of the loan has been sold. For a bank to have an incentive to continuemonitoring, it seems necessary that it hold a significant fraction of the loan that it originates. Prior literature questions theefficiency of an originating bank’s post-sale monitoring (Pennacchi, 1988; Gorton and Pennacchi, 1995; Gorton and Winton,2002), which further suggests that the secondary loan trade is subject to high information asymmetry.

Because information asymmetry plays such an important role in the secondary loan market, it is natural to posethe question of how information asymmetry is resolved in loan trading. More specifically, I examine which firm- andloan-specific characteristics decrease information asymmetry, as estimated by the bid–ask spread in the secondary loantrade. First, I examine firm and loan characteristics suggested by prior research as generally related to informationasymmetry. Second, I examine the unique characteristics of the information environment of syndicated loans. Third,controlling for these more readily recognizable determinants of information asymmetry, I investigate whether the financialreporting quality of a borrower is associated with the bid–ask spread in the secondary loan trade. This last researchquestion is the main focus of this paper.

3.2. Determinant of the bid–ask spread generally related to information asymmetry

3.2.1. Publicly reporting versus private firms

When a borrower does not report to the SEC, secondary market participants have less publicly available informationregarding its creditworthiness. In addition, private firms are not subject to rigorous monitoring by market forces, such asthe SEC, auditors, analysts and public exchanges. Private firms are also less subject to litigations related to financialreporting and disclosure. Therefore, the information advantage of informed traders is typically more substantial in thetrading of loans of private borrowers.

In addition, Diamond and Verrecchia (1991), Leuz and Verrecchia (2000) and Verrecchia (2001) establish that acommitment to higher disclosure reduces information asymmetry. Since, compared with private firms, public firms havean inherent commitment to higher disclosure, this evidence further emphasizes the importance of public reporting in loantrading. Consequently, I expect the loans of publicly reporting borrowers to be traded at lower bid–ask spreads.

3.2.2. Availability of credit rating

I expect the availability of a credit rating to decrease information asymmetry regarding a borrower. When estimatingcredit rating availability, I consider the existence of both firm-specific and loan-specific credit ratings. The significance ofthe availability of a credit rating is supported by Diamond’s (1991) theoretical model, which emphasizes the importance ofpublicly available information, such as credit ratings, to the lender–borrower relationship. The importance of loan-specificratings is also supported by Sufi (2007b) who shows that availability of a loan-specific credit rating reduces informationasymmetry between borrowers and uninformed lenders in the syndicated loan market.

3.2.3. Loan size

According to Jones et al. (2005), information asymmetries tend to be less severe for large loans, since any fixed costsassociated with obtaining information about a firm are less of an obstacle for large loans. Bharath et al. (2007) also suggestthat small borrowers have greater information asymmetries, and that a loan’s size is typically positively correlated with itsborrower’s size. Additionally, Diamond and Verrecchia (1991) show that large firms receive a larger benefit from disclosurethan small firms do, which further suggests that the trading of larger loans is associated with lower informationasymmetry.

3.2.4. Revolving loans

A revolving loan, which acts like a credit line, provides the borrower with additional flexibility, but increases uncertaintyfor the lender. Compared to the term loan, the revolver is more likely to be subject to takedown risk because it exposes thelender to sizeable changes in its commitment (Ho and Saunders, 1983). Moreover, borrowers tend to draw down revolvers

ARTICLE IN PRESS



when their performance deteriorates (Berger and Udell, 1995), further increasing uncertainty for the lender. Becausegreater uncertainty leads to higher information asymmetry in loan trading, I expect revolvers to be traded at higherspreads.

3.2.5. Loss versus profit firms

I also hypothesize that the loans of profitable firms are traded at lower spreads on the secondary loan marketrelative to the loans of firms reporting losses. First, loss firms operate in a high uncertainty environment. Second, losses aretypically associated with a deterioration in a borrower’s credit quality and with an increase in the probability ofbankruptcy. This evidence indicates that the loans of loss firms are likely to experience higher adverse selection costs in thesecondary trade.

Equity market studies also support this prediction. Ertimur (2004) and Sadka and Sadka (2004) find that the stocktrading of loss firms is subject to higher information asymmetry. Lang and Lundholm (1993) demonstrate that profit firmsprovide more information to stockholders than loss firms do. The theoretical model of Verrecchia and Weber (2007) alsosuggests that firms that report poor performance experience higher adverse selection costs in the stock market. Becausedebt holders’ returns are determined primarily by the downside region of a firm’s earnings distribution, the impact oflosses on the information environment is likely to be even more pronounced in the loan market.

3.3. Unique characteristics of the information environment of syndicated loans

3.3.1. Reputation of the arranger of syndication

I expect loans syndicated by more reputable arrangers to be associated with lower information asymmetry when tradedon the secondary market. The syndicate participants typically rely on information provided by the arranger (Jones et al.,2005). The arranger negotiates the loan agreement, coordinates the documentation process, recruits loan participants andperforms primary monitoring and enforcement responsibilities (Dennis and Mullineaux, 2000; Lee and Mullineaux, 2004).In addition, Gorton and Haubrich (1990) and Gorton and Pennacchi (1995) emphasize that the bank’s reputation serves asan implicit guarantee in a loan sale with no recourse, which is a common practice in the sale of syndicated loans.4 Theimportance of the arranger’s reputation in resolving information asymmetry is further motivated by the evidence thatmore reputable arrangers are more likely to syndicate loans and are able to sell off a larger portion of a loan to theparticipants (Dennis and Mullineaux, 2000; Panyagometh and Roberts, 2002).5 The literature interprets these findings asconsistent with the proposition that the arranger’s status is a certification of the borrower’s financial conditions.

3.3.2. Distressed versus par loans

To examine information asymmetry in loan trading, it is also important to distinguish between distressed and par loans.According to the conventions of the secondary loan market, distressed loans are loans traded at a bid price below 90% of thepar value. I expect the trading of distressed loans to be associated with stronger adverse selection for the following reasons.First, access to private sources of information regarding a borrower grants informed traders a considerable informationadvantage when the borrower’s loans are in distress. Via private communication with syndicate lenders, a firm disclosesimportant information, such as monthly financial disclosures, covenant compliance information, amendment and waiverrequests and financial projections. While this information is critical for evaluating distress loans, for uninformed traders itis usually either not available or it only becomes available after a delay. Second, Agrawal et al. (2004) show that as a firm’sfinancial condition worsens, informed investors intensify their trading activity, further increasing adverse selection insecondary trading.

3.3.3. Identity of the lender (i.e., institutional investor or bank)

To address the distinctive characteristics of syndicated loans, I also distinguish between loans issued by institutionalinvestors and loans issued by banks. Institutional investors constitute the main participants in the secondaryloan market and institutional loans represent 45% of traded loans during the sample period. I expect loansissued by institutional investors to be traded at higher bid–ask spreads than loans issued by banks are. Institutionalterm loans typically have a longer maturity and back-end-loaded repayment schedules than bank term loans.The considerably longer duration of institutional term loans is likely to cause higher information asymmetry in theloan trade.

ARTICLE IN PRESS

4 These papers analyze the bilateral lender–borrower relationship and therefore refer to the reputation of the selling bank. In the setting of the

syndicated market where the arranger manages a number of syndicate lenders, I conjecture that the reputation of the arranger dominates the reputation

of the other members of the syndication, including the seller, in a specific transaction. Rajan (1998) also suggests that buyers trust the selling bank in a

secondary loan sale, because of the importance of maintaining the bank’s reputation.5 Prior literature suggests that the arranger does not exploit its information advantage to distribute lower quality loans to syndicate participants. A

number of studies find that the arranger holds larger proportions of informationally problematic and riskier loans in its own portfolio (Simons, 1993; Lee

and Mullineaux, 2004; Jones et al., 2005; Sufi, 2007a). In addition, the arranger has been found to syndicate a larger proportion of a loan that is

subsequently upgraded (Panyagometh and Roberts, 2002).



3.3.4. Restructuring purpose loans

I also examine the information environment of restructuring purpose loans, which represent 52.9% of loans traded overthe sample period. I expect restructuring purpose loans—loans with a primary purpose of Takeover, LBO/MBO andRecapitalization—to be associated with higher information asymmetry. Restructuring purpose loans indicate aconsiderable change in a borrower’s capital structure, which is typically associated with a higher uncertainty regardingthe borrower’s future performance. This higher uncertainty should lead to higher information asymmetry betweeninformed and uninformed traders because, through private sources of information, the former are better able to evaluatethe consequences of the restructuring on the firm’s fundamentals.

3.4. The association between financial reporting quality and information asymmetry

In this section, I analyze the association between the quality of financial reporting and information asymmetry inthe secondary loan trade, with particular emphasis on exploring the effect of timely loss recognition on informationasymmetry.

3.4.1. Timely loss recognition

Because the secondary loan market is characterized by substantial information asymmetry between informed lenders(e.g., the arranger and syndicate participants) and uninformed market participants, I expect conservative financialreporting, as measured by the timely incorporation of economic losses in a borrower’s financial statements, to have asignificant impact on loan trading. In addition, debt holders are more sensitive to borrowers’ losses than to borrowers’profits (Leftwich, 1983; Watts and Zimmerman, 1986; Watts, 1993, 2003a), which further supports the prediction thattimely loss recognition should be important in resolving information asymmetry in the loan market.

I expect conservative financial reporting to decrease information asymmetry in loan trading by enhancing corporategovernance of a borrower and by increasing the amount and quality of information available to the secondary loan marketparticipants. Within the governance mechanism, there are two routes through which information asymmetry may bereduced. First, via a contracting route, timely loss recognition triggers ex-post violations of debt covenants in a timelymanner (Watts and Zimmerman, 1986; Watts, 1993, 2003a; Ball, 2001). By triggering debt covenant violations, timely lossrecognition allows lenders to more rapidly employ their decision rights following economic losses, which increases theefficiency of debt agreements.6

Compared to public debt contracts, syndicated loan contracts impose more numerous and tighter covenants (Assender,2000; Dichev et al., 2002; Dichev and Skinner, 2002). The proportion of loans subject to financial covenants is even higheracross traded loans: 94% of the facility-year observations of public borrowers are subject to financial covenants.Furthermore, the vast majority of traded loans have covenants explicitly based on a borrower’s earnings. These covenantsare strongly influenced by the timely incorporation of economic losses in a borrower’s financial statement. Of all the loansof public borrowers on which lenders impose financial covenants, 92% are subject to the interest coverage covenant(Min Interest Coverage and Min Fixed Charge Coverage) and 88% to the debt-to-profitability covenant (Max Debt to EBITDAand Max Senior Debt to EBITDA); 97% are restricted by at least one of these covenants.7 In addition, 15% of the loans ofpublic borrowers are subject to net-worth-based covenants (Min Net Worth and Max Debt to Net Worth). Net-worth-basedcovenants are also directly affected by timely loss recognition, which makes these covenants more binding (Beatty andWeber, 2006; Zhang, 2008). Because of the prevalence of income and net- worth-based covenants in loan agreements,I expect timely loss recognition to significantly influence the contract efficiency of traded syndicated loans.

Via the second governance route, by generating an understatement of net asset values, conservative financial reportingfacilitates monitoring by debt holders and reduces deadweight agency costs (Watts and Zimmerman, 1986; Watts, 1993;Ball, 2001; Watts, 2003a). As discussed in Watts (2003a), lenders rely on verifiable lower bound measures of assets both inassessing a potential loan and in monitoring the borrower’s creditworthiness after a loan issuance.8 Further, conservativereporting constrains managers’ opportunistic behavior. This reduces managers’ incentives to transfer wealth to themselveseither from debt holders or from the firm as a whole (Ball, 2001; Watts, 2003a, b; Roychowdhury and Watts, 2007; LaFondand Roychowdhury, 2008). Thus, timely loss recognition reduces managers’ ability to overstate earnings and receiveovervalued compensation. Timely loss recognition also decreases managers’ incentives to invest in or postpone theabandonment of negative NPV projects (Ball, 2001; Ball and Shivakumar, 2005; LaFond and Roychowdhury, 2008).9

ARTICLE IN PRESS

6 Even if debt holders perform conservative adjustments to reported numbers when covenant compliance is estimated, this does not eliminate the

need for conservative financial reporting. Thus, Beatty et al. (2008) find that conservative covenant adjustments and financial reporting conservatism

complement rather than substitute for each other in satisfying debt holders’ demand for conservatism.7 There is a substantial variation in the definition of covenants in loan contracts. Thus, debt might be defined by the loan agreement as total debt,

long-term debt, senior debt or total debt minus cash, and EBITDA might be related to EBIT, EBITDA, net income or income before extraordinary items.8 It is important to note that an understatement of net asset values generated by conservative reporting is directly linked to the net-worth-based

covenant imposed by syndicated loan contracts. An understatement of net asset values makes a net-worth-based covenant more binding, triggering more

timely covenant violations.9 By investigating the separation of ownership and control, LaFond and Roychowdhury (2008) find that to counter managers’ incentives to invest in or

delay divestment of negative NPV projects, there is greater demand for conservatism when there is less alignment of interests between managers and

shareholders.



Relying on these arguments, I anticipate that by enhancing corporate governance of the borrower, timely loss recognitiondecreases information asymmetry regarding the borrower’s performance and creditworthiness.

It is important to emphasize that conservative reporting also fulfills an important role in providing information toinvestors (Watts, 2003a; Khan and Watts, 2007; LaFond and Watts, 2008). Because managers have incentives to discloseinformation about unrealized gains but to withhold information regarding losses, conservative financial reporting not onlyprovides a ‘‘hard’’ source of information regarding a firm’s performance, but it also provides investors with a control foralternative ‘‘soft’’ sources of information (Watts, 2003a). LaFond and Watts (2008) and Khan and Watts (2007) show thatconservative reporting is an equilibrium response to mitigate agency problems arising from information asymmetry.

By increasing the amount and quality of public information available to uninformed investors, I expect timely lossrecognition to decrease information asymmetry between informed and uninformed loan traders. More specifically, timelyloss recognition induces early revelation of the changes in a borrower’s credit quality and therefore increases theborrower’s transparency. As a result, timely loss recognition allows uninformed loan traders who do not possess privatesources of information regarding the borrower to get a more timely and precise evaluation of the borrower’s traded loans.Increased information flow also allows uninformed loan buyers to perform more efficient post-sale monitoring of theborrower, which further reduces adverse selection in loan trading.

To summarize these arguments, by improving the efficiency of debt agreements and by enhancing corporate governanceand transparency of a borrower, timely loss recognition decreases the information advantage of informed traders.Therefore, I expect timely loss recognition to reduce the bid–ask spread in loans trading.

3.4.2. Timely gain recognition and the overall timeliness

I also examine the impact of timely gain recognition on loan trading. As with timely loss recognition, timelygain recognition improves the timeliness of accounting earnings and therefore is expected to make earnings a moreinformative measure of a borrower’s performance (Ball and Shivakumar, 2006; Guay, 2006; Guay and Verrecchia, 2006).Furthermore, the effect of timely gain recognition on information asymmetry is likely to be more pronounced fordistressed loans. When loans are in distress, the debt holders’ returns are determined not only by the downside regionof a borrower’s earnings distribution, but also by the upside one. In other words, compared to investors in par loans,investors in distress loans are more sensitive to economic gains: therefore good news is more important in evaluating theselenders’ claims.10

In addition, accounting research suggests that the overall timeliness of a borrower’s financial reporting, for both gainsand losses, is associated with the borrower’s transparency and the efficiency of its debt agreements (Ball et al., 2008a).Therefore, I also explore whether the overall reporting timeliness is helpful in resolving information asymmetry insecondary loan trading.

4. Data and descriptive statistics

4.1. Data sources and sample selection

I obtain loan trade data from the Loan Trade Database provided by LPC. The database includes indicative loan bid andask price quotes reported to LPC by trading desks at institutions that make a market in these loans. The Loan TradeDatabase provides bid and ask price quotes aggregated across dealers. Bid and ask prices are quoted as a percent of par(or cents on the dollar of par value). In addition to price coverage, for every traded loan the database provides theborrower’s name, quote date and the number of market makers reporting quotes to LPC. The Loan Trade Database covers2,125,589 observations for the period from June 1998 to December 2003; these observations represent the trading of about4,788 syndicated facilities11 (Table 1).12

I match the Loan Trade Database to the DealScan database, which covers a majority of the syndicated loan issues in theU.S. (this database is also provided by LPC). Connecting these two databases allows for the identification of traded loans onthe primary loan market, including their characteristics, such as the amount, maturity, seniority, securitization, covenantsand syndicate structure. Merging the Loan Trade and DealScan databases results in a sample of 1,732,065 identified tradingobservations related to 3,611 trading facilities (Table 1).

Most market makers report loan price quotes to LPC on a daily, biweekly and weekly basis. To address the time-seriescorrelation and measurement error in the trading data, I perform an empirical analysis based on the average annual

ARTICLE IN PRESS

10 The demand for timely gain recognition may also be driven by the interest-decreasing performance pricing option imbedded in some loan

contracts. According to this option, a borrower becomes eligible for lower interest if some financial benchmark is reached. Because a borrower might

intentionally accelerate gain recognition to benefit from a lower cost of debt, which, given a borrower’s true credit risk, would lead to insufficient

compensation to a lender, the benefits of timely loss recognition to a lender are questionable.11 In the syndicated loan market, a loan is identified as a ‘‘facility’’. Usually, a number of facilities with different maturities, interest rate spreads and

repayment schedules are structured and syndicated as one transaction (deal) with a borrower. The analysis in this paper is performed at the individual

facility level.12 The database coverage is limited in 1998, but it increases sharply in 1999. According to LPC estimates, the Loan Trade Database covers 80% of the

trading volume in the secondary loan market in the U.S.



estimation of the loans’ prices and bid–ask spreads. Because most of the explanatory variables I employ in theanalysis vary annually or remain constant over a facility’s trading period, I presume that annual estimations providebetter specification of the empirical tests. It is important to note that the core results are robust to performing theanalysis based on daily or monthly trading observations. 1,732,065 identified trading observations constitute 10,193facility-year observations (a majority of the 3,611 identified facilities are traded for a number of years over thesample period). Additionally, I drop loans issued to non-U.S. firms or in currencies other than the U.S. dollar; the remain-ing sample contains 9,779 facility-year observations representing 3,464 facilities. These facilities are syndicated to1,435 firms.

I match the sample firms with CRSP and Compustat databases. By using the tickers available on DealScan, I identify 408of the borrowers as publicly traded firms. To improve the identification, I match the rest of the firms with Compustat/CRSPby name, industry and state location. This results in the recognition of an additional 333 borrowers as publicly reportingfirms, 179 of which are also publicly traded. The accuracy of this matching is high, with 79% of the firms being matched onall three parameters.

4.2. Traded versus non-traded loans

The comparison of traded loans to non-traded ones emphasizes the distinctive characteristics of loans traded on thesecondary market. Since the majority (96%) of the traded loans in the sample were syndicated starting in 1997, the U.S. non-traded loans syndicated in the primary loan market over the period from 1997 to 2003 are chosen as the most appropriatecomparison group for the traded sample.

Consistent with the high involvement of institutional investors in the secondary loan market, institutional term loansare heavily traded (Table 2). Loans with the purpose of a Takeover or LBO/MBO represent 41.9% of the traded facilities, whiletheir proportion in the non-traded sample is considerably lower. In contrast, loans for corporate purposes and working

ARTICLE IN PRESS

Table 1Identification of the traded loans

Number of

observations

Number of

facilities

% of total trading

observations (facilities)

Total trading observationsa 2,125,589 4,788b

Trading observations with missing Facility-ID and LINc,d 50,591 266e 2.4

(5.6)

Trading observations with less than 13-digit LINsf 87,274 252 4.1

(5.3)

Trading observations with available identifier—Facility-ID and/or 13-digit LIN 1,987,724 4,270 93.5

(89.2)

Observations successfully matched with the DealScan database 1,732,065 3,611g,h 81.5

(75.4)

a Institutions providing bid and ask prices currently include but are not limited to: Bank of Montreal, The Bank of New York, The Bank of Nova Scotia,

BANK ONE, Bank of America Securities LLC, BankBoston, BT Alex Brown/Deutsche Bank AG, The Chase Manhattan Bank, NA, CIBC World Markets, Citibank,

NA, Credit Lyonnais, Credit Suisse First Boston Corporation, DLJ Capital Funding, INC., First Union Capital Markets Corp., Goldman, Sachs & Company, J.P.

Morgan Securities, Inc., Lehman Brothers, Inc., Merrill Lynch, Pierce Fenner & Smith Incorporated, Morgan Stanley Dean Witter and TD Securities (USA)

Inc.b Because some of the trading observations are not assigned to specific facilities, this number is an approximation of the total number of traded

facilities. This proxy is estimated as the number of distinct facilities identified on the Loan Trade Database (4,522) plus the number of firms (266) with

traded observations without facility identification. For further details, see footnotes c, d and e.c Facility-ID is a number assigned by LPC to each syndicated facility on the primary loan market. LIN (Loan Identification Number) is assigned to each

syndicated facility that is traded on the secondary loan market. The Loan Trade Database and DealScan are merged by the Facility-ID and/or LIN

numbers.d According to LPC, observations missing Facility-ID and LIN identifiers belong to the period when LPC just started covering the secondary loan

market.e Assuming that borrowers do not change the company name during the period of loan trading, there are 266 firms with missing identifiers (Facility-

ID and/or LIN numbers). As a result, there are at least 266 non-identified facilities, because every borrower might have more than one trading facility.f LINs with less than 13 digits cannot be matched with the DealScan database. LINs with less than 13 digits are assigned to traded facilities in the

following circumstances: (a) the traded loan is private and is not covered by DealScan; (b) the traded loan is a ‘‘prorate piece’’—a combination of two

different facilities; since these two facilities are traded as one piece, but were originated as independent facilities in the primary loan market, prorate

pieces cannot be directly connected to the DealScan database. All these observations also lack a Facility-ID number.g The Facility-ID and/or LIN numbers of 659 facilities do not have an appropriate match on the DealScan database.h From the total number of identified facilities, 3,464 facilities are issued to U.S. borrowing firms in U.S. dollars.



capital loans represent a smaller percentage of the secondary market, relative to their fraction of non-traded loans. Most ofthe traded loans are senior and secured.13

Traded loans are also characterized by a longer maturity: the median maturity of the traded loans is six years, while themedian maturity of the non-traded loans is three years. The longer average maturity of the traded loans is explained, atleast partially, by the longer maturity of institutional loans that represent more than 40% of the traded sample. Tradedloans are substantially bigger than non-traded ones: the median size of traded loans reaches $140 million, while themedian size of non-traded loans is $62.5 million.14

4.3. Distinctive characteristics of traded loans

To perform a regression analysis, I exclude observations without available data on price quotes, loan size andmaturity, and the identity of the arranger of syndication. The final sample results in 8,619 facility-year observations

ARTICLE IN PRESS

Table 2Traded versus non-traded loansa

Traded loans % of traded loans % of non-traded loans

Loan type

Institutional term loan Bb 1,208 34.9 4.1

Revolver above one yearc 915 26.4 40.0

Amortizing term loan Ad 477 13.8 2.8

Term loand 356 10.3 15.9

Institutional term loan Cb 200 5.8 0.7

Other 308 8.9 36.6

Loan purpose

Takeover 855 24.7 7.6

Debt repay 691 19.9 19.0

LBO/MBO 595 17.2 4.2

Corporate purposese 422 12.2 32.2

Acquisition linef 217 6.3 6.4

Working capital 225 6.5 11.7

Recapitalizationg 140 4.0 1.2

Other 319 9.2 17.7

Seniority

Senior 3,448 99.5 98.2

Subordinated 7 0.2 1.6

Not available 9 0.3 0.2

Security

Secured 2,552 73.6 35.3

Unsecured 167 4.8 9.9

Not available 745 21.5 54.7

a Loans in the traded sample and in the non-traded sample are restricted to loans issued in U.S. dollars to U.S. borrowers. The majority (96%) of the

traded loans were syndicated on the primary loan market starting in 1997. Therefore, the non-traded sample is limited to loans issued during the period

from 1997 to 2003; the non-traded sample includes 39,600 facilities. The sample of traded loans incorporates 3,464 facilities.b An installment loan issued by institutional investors, characterized by a longer maturity and a back-end-loaded repayment schedule compared to

term loan originated by banks. An installment loan is a loan commitment that does not allow the amounts repaid to be re-borrowed. Because of the

extremely low frequency in the traded sample, institutional term loan D is included in the ‘‘Other’’ category.c A revolving credit line that the borrower may draw down, repay, and re-borrow under. A borrower is charged an annual commitment fee regardless

of usage.d An installment loan issued by banks, characterized by a progressive repayment schedule. An amortizing term loan is typically syndicated along with

revolving credits as part of a large syndication. According to LPC, the majority of the loans in the Term loan category are amortizing term loans issued by

banks.e An all-purpose loan that can be used for various activities related to general operations, working capital and purchases. It may include a roll-over of

maturing debt.f A loan for unspecified asset acquisitions. Though the loan may contain limits on the size and scope of the acquisition, the borrower typically has

latitude over which assets to purchase.g A loan to support a material changes in a company’s capital structure, often made in conjunction with other debt or equity offerings.

13 The analysis of traded versus non-traded loans is robust to using different control samples. The results are almost identical whether the traded

loans are compared to the loans issued in a variety of currencies, or to the loans issued over different time periods.14 While the entire amount of a syndicated facility may be traded on the secondary loan market, it is also possible for only a partial amount to be

traded. The Loan Trade Database does not provide information regarding the relative proportion of a loan that is traded on the secondary market.

According to LPC, the average secondary loan trade size amounted to $2.5 million over the sample period.



representing 3,029 facilities. Table 3 presents traded loans’ characteristics. Loans are traded at relatively highspreads; this is especially true of distressed facilities. Traded loans have a median size of $150 million and a mediantime to maturity of 51 months. The typical market share in the primary loan market of the loan’s arranger is0.85%. Additionally, bid and ask price quotes of the majority of the traded loans are reported to LPC by one or twomarket makers.

Loans of public firms are traded at lower spreads; this relation holds for both par and distressed loans. In addition, thetraded loans of public firms are bigger in size and are syndicated by arrangers with higher market share. Further, publicborrowers have a lower proportion of institutional loans and a higher proportion of revolvers compared to privateborrowers. A lower percentage of public borrowers’ traded loans are issued with a primary purpose of Takeover, LBO/MBOor Recapitalization (restructuring purpose loans). Public firms and/or their loans are more frequently rated by credit rating

ARTICLE IN PRESS

Table 3Characteristics of the traded loans

Facility characteristics Facility-year observations Mean SD Distribution

25% 50% 75%

Total samplea

Bid–ask spreadb 8,619 1.55 1.82 0.50 0.85 1.87

Bid–ask spread—par facilitiesc 6,918 0.89 0.66 0.50 0.67 1.00

Bid–ask spread—distressed facilitiesd 1,701 4.21 2.48 2.43 3.41 5.06

Size of facility (in millions) 8,619 261.88 353.31 75.00 150.00 300.00

Time to maturitye (in month) 8,619 49.69 24.44 32.50 51.00 68.00

Market share of the arrangerf (in %) 8,619 9.84 8.35 0.18 0.85 15.27

Number of market makersg 8,619 2.21 1.80 1.00 1.47 2.77

Institutional facilitiesh (in %) 8,619 45.39

Revolver-line facilitiesi (in %) 8,619 24.89

Restructuring purpose facilitiesj (in %) 8,619 52.89

Facilities with available credit ratingk (in %) 8,619 81.01

Facilities with financial covenantsl (in %) 8,619 73.78

Distressed facilitiesd (in %) 8,619 19.74

Publicly traded samplem

Bid–ask spreadb 2,772 1.04 1.05 0.49 0.63 1.06

Bid–ask spread—par facilitiesc 2,524 0.80 0.53 0.48 0.59 1.00

Bi–dask spread—distressed facilitiesd 248 3.47 1.68 2.02 2.99 4.77

Size of facility (in millions) 2,772 390.42 489.91 125.00 225.00 450.00

Time to maturitye (in month) 2,772 48.97 24.35 32.50 50.50 67.00

Market share of the arrangerf (in %) 2,772 10.86 8.67 0.25 1.19 15.27

Number of market makersg 2,772 2.62 2.27 1.00 1.89 3.25

Institutional facilitiesh (in %) 2,772 41.13

Revolver-line facilitiesi (in %) 2,772 29.00

Restructuring purpose facilitiesj (in %) 2,772 40.98

Facilities with available credit ratingk (in %) 2,772 87.45

Facilities with financial covenantsl (in %) 2,772 94.35

Distressed facilitiesd (in %) 2,772 8.95

a 8,619 facility-year observations have all the data required for the regression analysis. 4,503 facility-year observations are related to publicly

reporting firms and 4,116 observations are related to private firms.b The bid–ask spread is estimated based on bid and ask price quotes aggregated across dealers. Bid and ask prices are quoted as a percent of par (or

cents on the dollar of par value). The bid–ask spread is measured as the average annual bid–ask spread of the traded facility.c Facilities with an annual average bid price equal or above 90% of the par value.d Facilities with an annual average bid price below 90% of the par value.e Time-to-maturity is measured by the number of months between the facility’s trading date on the secondary loan market and the date when the

facility matures. The estimation is based on the annual average of a facility’s traded observations.f The market share is measured by the ratio of the amount of loans that the financial intermediary syndicated as a lead arranger to the total amount of

loans syndicated on the primary loan market over the period from 1998 to 2003. In case of the multiple arrangers, I consider the highest market share

across the arrangers involved in the loan transaction. The market share is presented at percentage value.g Number of market makers that provide a facility’s bid and ask price quotes to LPC. The estimation is based on the annual average of a facility’s traded

observations.h Institutional term loans (Term Loan B, Term Loan C and Term Loan D).i A revolving credit line with a duration above one year; the commitment that the borrower may draw down, repay, and re-borrow under. A borrower

is charged an annual commitment fee regardless of usage.j Loans with the primary purpose of Takeover, LBO/MBO or Recapitalization. A loan with a primary purpose of recapitalization is a loan to support a

material change in a company’s capital structure, often made in conjunction with other debt or equity offerings.k Includes the following credit rating categories: Moody’s Sr. Debt, Moody’s Loan Rating, S&P Sr. Debt, S&P Loan Rating, Fitch LT and Fitch Loan

Rating.l Facilities that are subject to at least one financial covenant.m 2,772 facility-year observations have all the data required for the regression analysis of the bid–ask spread of loans of publicly traded firms.



agencies. Additionally, lenders more often impose financial covenants on public borrowers. The proportion of loans indistress is substantially lower for public borrowers.

5. Research design

5.1. Adverse selection component versus transitory component of the bid–ask spread

Following Copeland and Galai (1983), Glosten and Milgrom (1985) and Kyle (1985), many papers rely on the bid–askspread as the main measure of information asymmetry (e.g., Lee et al., 1993; Yohn, 1998; Leuz and Verrecchia, 2000; Leuz,2003; Kalimipalli and Warga, 2002). While these papers argue that the bid–ask spread successfully captures informationasymmetry, relying on the bid–ask spread may be problematic because it includes both the adverse selection component,related to asymmetric information, and the transitory component, related to the inventory and order-processing costs ofthe market maker.

A number of prior studies unravel the adverse selection component of the stock spread. However, because the tradingvolume and actual transaction data are not available for traded loans, the models suggested by these studies cannot beused to measure the adverse selection component in loan spreads.15 Consequently, the analysis in this paper is performedwithout differentiating between adverse selection and transitory components.

To address the transitory component in the empirical analysis, I control for the determinants of this componentsuggested by prior research. First, I control for the number of institutions making a market in a traded loan. Garbade (1982)and Stoll (1985) suggest that the transitory component of the stock spread is negatively related to the number of marketmakers. Further, Goldstein and Nelling (1999) find that competition among market makers effectively reduces the bid–askspread in the stock market. Second, I incorporate into the analysis the time to maturity of the traded loans. Bond tradingliterature suggests that corporate bonds are more actively traded following issuance and tend to become less liquidwith age (Nunn et al., 1986; Sarig and Warga, 1989; Alexander et al., 2000; Hong and Warga, 2000; Chakravarty andSarkar, 2003).

5.2. Empirical estimation of financial reporting quality

In this section, I focus on the empirical estimation of the attributes of financial reporting quality: timely loss recognition,timely gain recognition, and the overall timeliness of financial reporting. I also address the importance of incorporating themarket-to-book ratio into the analysis.

To measure timely loss recognition, I employ the market-based model suggested by Basu (1997). The modelrelates earnings to contemporaneous stock returns, which serve as a proxy for economic gains and losses. Following Basu(1997), I estimate a piecewise-linear regression of earnings on stock returns: NIit ¼ b0+b1DRit+b2Rit+b3Rit *DRit. Thetimeliness of income in reflecting current-year economic losses (decreases in stock market value) is measured by the sumof b2 and b3.16

The definitions of the variables employed in the model are as follows. NIit is earnings per share for firm i in the fiscal yeart deflated by the opening stock price and adjusted by the average EP ratio for sample firms in fiscal year t. Rit is the return onfirm i from nine months before fiscal year-end t to three months after fiscal year-end t less the corresponding CRSPNYSE/AMEX/NASDAQ market return. DRit is an indicator variable taking the value of one if the firm’s market-adjustedreturns are negative, zero otherwise. Observations falling either in the top or bottom 1% of either price or asset deflatedearnings or returns in each year are excluded.

I estimate Basu’s (1997) model at the 3-digit-industry-year level. Each publicly traded borrower is assigned acorresponding industry-year timely loss recognition measure estimated in the year preceding the year in which the bid–askspread is measured. I also estimate the firm timely loss recognition measure; this estimation is based on firm-specific Basuregressions estimated over the 20-year period preceding the loan trading. To get more reliable measures of timely lossrecognition from industry-year and firm-specific regressions, I require a minimum of 10 observations to be available forindustry-years and firms, respectively. I also employ Basu’s (1997) model to estimate timely gain recognition and theoverall timeliness. Timely gain recognition is measured by b2 coefficient, while the overall timeliness is estimated by R2 ofthe Basu regression.

Following Frankel and Roychowdhury (2007) and Khan and Watts (2007), I also allow the earnings timeliness withrespect to good news and the asymmetric timeliness with respect to bad news to vary with firm-specific characteristics.These characteristics—size, market-to-book and leverage—have been found to be both theoretically and empirically

ARTICLE IN PRESS

15 Glosten and Harris (1988) use trading volume and trade frequency to break the bid–ask spread into a transitory and adverse selection component.

Stoll (1989) and Hasbrouck (1991) estimate the permanent component of the bid–ask spread based on the quoted spread and the actual trade data.

Examining the earnings announcements, Barclay and Dunbar (1996) also rely on the trading volume. Sadka (2004) develops a measure of liquidity/

information asymmetry by using intraday stock trading data. The liquidity measure of Acharya and Pedersen (2005) is based on stock returns and trading

volume. Because LTD provides bid and ask price quotes aggregated across market makers, I also cannot rely on the frequency of the price revisions to

estimate a loan’s liquidity.16 See Ball and Kothari (2007) for an extensive review of the Basu (1997) model.



related to financial reporting conservatism. I estimate the model suggested by Khan and Watts (2007) to obtain the weightsused to aggregate the firm characteristics into annual firm-specific measures of timeliness17:

Eit ¼ b1 þ b2DRit þ Ritðm1 þ m2Sizeit þ m3M=Bit þ m4LEVitÞ þ DRit � Ritðl1 þ l2Sizeit þ l3M=Bit þ l4LEVitÞ

þ b3Sizeit þ b4M=Bit þ b5LEVit þ DRitðb6Sizeit þ b7M=Bit þ b8LEVitÞ

Timeliness with respect to good news (G_Score) and asymmetric timeliness with respect to bad news (C_Score) are linearfunctions of firm-specific characteristics:

G_Score ¼ m1 þ m2Sizeit þ m3M=Bit þ m4LEVit; C_Score ¼ l1 þ l2Sizeit þ l3M=Bit þ l4LEVit .

Timely loss recognition—B_Score—is estimated as the sum of the C_Score and G_Score.The definitions of the variables employed in the model are as follows. E is net income before extraordinary items,

deflated by the market value of equity at the end of the prior fiscal year. Size is the natural log of the market value of equity.LEV is the sum of long-term debt and debt in current liabilities, deflated by the market value of equity. M/B is the marketvalue of equity divided by the book value of equity. The rest of the variables are defined as in Basu’s (1997) model.

I estimate the model using annual cross-sectional regressions over the 20-year period from 1978 to 1997, and calculatethe average weights on size, market-to-book and leverage. These average weights are subsequently used to estimate annualfirm-specific C_Score and G_Score measures over the five-year research sample period from 1998 to 2003. Whenestimating the model, I delete firm years with missing data and a negative book value of equity. I also exclude observationsfalling either in the top or bottom 1% of earnings, returns, size, market-to-book and leverage in each year. Because Khan andWatts (2007) demonstrate that C_Score successfully distinguishes between the firms that will, upon receiving bad news inthe future, be more conservative or less, I estimate timeliness scores in the year preceding the year in which the bid–askspread is measured.

It is important to note that prior research suggests that when the earnings–returns association is employed toinvestigate differences in earnings timeliness it is necessary to control for the market-to-book ratio (Pae et al., 2005). Byapplying stricter verification standards to the recognition of good news as gains than to the recognition of bad news aslosses, conservative reporting leads to an understatement of book asset values relative to their market value (Givoly andHayn, 2000; Watts, 2003a; Beaver and Ryan, 2005; Roychowdhury and Watts, 2007). Therefore, the greater the reportingconservatism, the larger is the market-to-book ratio. Because conservative reporting reduces information asymmetryregarding a borrower, the borrower’s market-to-book ratio is likely to be negatively associated with the bid–ask spread inthe secondary loan trade.

However, a larger market-to-book ratio also indicates a borrower’s greater growth options. Because greater growthoptions are associated with larger information asymmetry (Smith and Watts, 1992), the market-to-book ratio is expected tobe positively associated with the loan trading spread. As a result, while the market-to-book ratio is expected to be relatedto information asymmetry in the secondary loan trade, there is no clear prediction regarding the sign of this relation.

6. Empirical results

6.1. Determinants of information asymmetry in the secondary loan trade

In this section, I analyze firm- and loan-specific characteristics that resolve or exacerbate information asymmetry in thesecondary loan market. First, I address general determinants of information asymmetry. Second, I focus on specificcharacteristics of the syndicated loans’ information environment. Third, controlling for these more readily recognizabledeterminants of information asymmetry, I examine whether there are aspects of a borrower’s financial reporting qualitythat decrease information asymmetry in the secondary loan trade.18

6.1.1. Determinants of the bid–ask spread generally related to information asymmetry

Table 4 presents the results from estimating the loan bid–ask spread for the total sample and for the sample of publiclytraded borrowers. There is clear evidence that public reporting and the availability of a credit rating decrease informationasymmetry in the secondary loan trade. Thus, loans of publicly reporting firms are traded at lower spreads than the loans ofprivate firms are. This result is statistically and economically significant; facilities of publicly reporting firms experiencespreads that are 12.6 cents lower than spreads on facilities of private firms (economic effects are reported as cents/dollarsper $100 of par value). This effect is substantial, given that it constitutes 14.9% of the median bid–ask spread for the totalloan sample. Additionally, rated facilities experience a decrease in the bid–ask spread.19 Facilities with an available credit

ARTICLE IN PRESS

17 See Khan and Watts (2007) for the model’s developments and the validation of the conservatism score.18 A majority of the explanatory variables employed in the empirical analysis are not highly correlated. The Pearson/Spearman rank correlation

coefficients are considerably high only for two pairs of the explanatory variables: Time-to-maturity and Investor (0.42), and Revolver and Investor (�0.53).

I winsorize the bid–ask spread and all the explanatory variables at the 1% and 99% level.19 It could be useful to test whether the rated loans of public borrowers experience a further decrease in the traded spread. The extremely high

correlation (92%) between Public and the interaction term between the Public and Rated variables prevents the incorporation of the interaction term into

the analysis.



rating are traded at spreads that are 16.3 and 28 cents lower then spreads on non-rated facilities, for the total and thepublicly traded samples, respectively.20

For the total sample of traded loans, I also find a negative relation between the bid–ask spread and loan size. An increaseof one standard deviation in Facility-size is associated with a decrease of 13.6 cents in the bid–ask spread. This result isconsistent with the higher amount and quality of information available regarding larger debt issues. However, since largerdebt issues typically have a larger trading volume, there is a concern that a relation between the bid–ask spread and theloan size is at least partially attributed to a loan’s trading volume. Because the Loan Trade Database does not provideinformation regarding trading volume, it is not possible to directly control for the volume’s effects on the bid–ask spread.

ARTICLE IN PRESS

Table 4Determinants of the bid–ask spread in the secondary loan trade

Pred. signs Total sample Publicly traded sample Publicly traded sample Publicly traded sample

Industry-year TLR measure Firm TLR measure Firm B-score measure

Public (b1) � �0.126** � � �

(0.05)

Rating (b2) � �0.163** �0.280*** � �0.256***

(0.08) (0.10) (0.09)

Facility-size (b3) � �0.134*** �0.034 �0.001 �0.028

(0.02) (0.02) (0.03) (0.02)

Revolver (b4) + 0.112*** 0.226*** 0.173*** 0.221***

(0.04) (0.04) (0.07) (0.04)

Arranger-reputation (b5) � �0.075*** �0.055*** �0.037* �0.055***

(0.03) (0.02) (0.02) (0.02)

Distress (b6) + 3.205*** 2.439*** 2.669*** 2.307***

(0.11) (0.15) (0.26) (0.14)

Investor (b7) + 0.111*** �0.023 0.060 �0.028

(0.04) (0.04) (0.06) (0.04)

Primary-purpose (b8) + 0.007 0.030 0.112** 0.023

(0.05) (0.04) (0.05) (0.04)

N-of-market-makers (b9) � �0.052*** �0.021** �0.033** �0.026**

(0.01) (0.01) (0.01) (0.01)

Time-to-maturity (b10) � �0.007*** �0.005*** �0.003*** �0.003***

(0.00) (0.00) (0.00) (0.00)

Income-net (b11) � � �0.273*** �0.227*** �0.248***

(0.05) (0.06) (0.05)

Timely-loss-recognition (b12) � � �0.372** �0.159* �0.208**

(0.15) (0.09) (0.09)

Market-to-book (b13) ? � �0.004 �0.005* �0.001

(0.00) (0.00) (0.00)

Adj R-Sq 57.13% 59.53% 59.97% 60.53%

Number of observations 8,619 2,722 1,178 2,767

Number of clusters 1,252 492 222 503

Spread ¼ a+b1 Public+b2 Rating+b3 Facility-size+b4 Revolver+b5 Arranger-reputation+b6 Distress+b7 Investor+b8 Primary-purpose+b9 N-of-market-makers+

b10 Time-to-maturity+b11 Income-net+b12 Timely-loss-recognition+b13 Market-to-book.

Regressions include year and industry fixed effects. Standard errors are heteroskedasticity robust, clustered at the firm level. Standard errors are reported

in parentheses. ***, **, * denote significance at the 1, 5 and 10 percent level, respectively.

Variables: Spread—the average annual bid–ask spread of the traded facility. Public—an indicator variable taking the value of one if a borrower is a publicly

reporting firm in the year when the facility is traded on the secondary loan market, zero otherwise. Rating—an indicator variable taking the value of one if

a firm and/or facility has an available credit rating, zero otherwise. Facility-size—the size of the facility measured by a logarithm of the facility’s amount.

Revolver—an indicator variable taking the value of one if the facility’s type is Revolver above one year, zero otherwise. Arranger-reputation—reputation of

the arranger of syndication, estimated by the average market share of the facility’s arranger in the primary syndicated loan market. Distress—an indicator

variable taking the value of one if the facility is traded at the annual average bid price below 90% of the par value, zero otherwise. Investor—an indicator

variable taking the value of one if the facility has been originated by an institutional investor, zero otherwise. Primary-purpose—an indicator variable

taking the value of one if the facility’s primary purpose is Takeover, LBO/MBO or Recapitalization, zero otherwise. N-of-market-makers—the average

annual number of market makers that provide a loan’s bid and ask prices to LPC. Time-to-maturity—the number of months between the facility’s trading

date on the secondary loan market and the date when the facility matures. Income-net—an indicator variable taking the value of one if the borrower’s

current year net income is positive, zero otherwise. Timely-loss-recognition—in Column (2) the measure is estimated by the sum of b2 and b3 in a

piecewise-linear industry-year regression of earnings on the contemporaneous stock returns (Basu, 1997): NIit ¼ b0 þ b1DRit þ b2Rit þ b3Rit � DRit; in

Column (3) the measure is estimated by the sum of b2 and b3 in a piecewise-linear firm-specific Basu (1997) model; in Column (4) the measure is

estimated by the B_Score based on the Khan and Watt (2007) model. Market-to-book—the ratio of the firm’s market value to the book value of its common

equity, estimated at the end of the borrower’s fiscal year.

20 Economic effects for the sample of public borrowers are estimated based on the regression analysis that relies on the industry-year timely loss

recognition measure, as reported in Column (2) of Table 4.



For the loans of publicly traded borrowers, there is no evidence that the bid–ask spread is significantly related toFacility-size. This result is potentially explained by a more homogeneous disclosure level among public firms.

Consistent with the higher uncertainty associated with revolvers, these facilities are traded at higher bid–ask spreads.The impact of Revolver on the bid–ask spread is statistically and economically significant for both total and publicly tradedsamples.21

Further, loans of profit firms are traded at significantly lower spreads than loans of firms experiencing losses are. Loansof profit firms experience spreads that are 27 cents lower than the spreads on the loans of loss firms; this effect constitutes43.1% of the median traded spread across loans of publicly traded borrowers. The relation between the bid–ask spread andIncome-net is not driven by the age of a firm. Young sample firms do not experience a higher frequency of losses than moremature firms do. In addition, the results are robust whether a loss indicator variable is based on the current year incomebefore extraordinary items or on the borrower’s net income in the prior year.

6.1.2. Unique characteristics of the information environment of syndicated loans

Empirical findings suggest that loans syndicated by more reputable arrangers are traded at significantly lower bid–askspreads. The effect of Arranger-reputation on the bid–ask spread is economically significant and robust to alternativemeasures of the arranger’s reputation.22 This evidence is consistent with the arranger’s primary role in resolvinginformation asymmetry regarding the borrower and its traded loans.

Another key variable of interest is the loan’s distress status. The considerable information advantage of informed tradersin the trading of distress loans is manifested in the significantly higher spreads of these facilities. Distressed loansexperience spreads that are $3.20 higher than spreads on loans traded at par.23 It is important to emphasize that this effectis attributed primarily to the adverse selection component and not to the transitory component of the bid–ask spread.

There is a concern that the trading of distress loans is associated with higher transitory costs because it is more costlyfor a market maker to hold more risky loans in the inventory. However, secondary loan trading is characterized by a markedpositive correlation between credit risk and liquidity, so more risky loans are associated with a higher trading liquidity(Yago and McCarthy, 2004). This relation is the opposite of that observed in the corporate bond market, where investmentgrade bonds are more liquid than high yield bonds. The large and fast-growing demand for high credit risk loans in thesecondary loan market generates a significantly larger order flow for more risky loans. As suggested by Bhasin and Carey(1999), this order flow effect may more than offset the higher holding costs associated with risky loans, typically causingthe lower transitory component of the bid–ask spread in trading more risky loans on the secondary market. Furthermore,for the sample borrowers, I find that distressed loans have a significantly higher number of market makers than loanstraded at par. This evidence further suggests that the transitory component of the bid–ask spread is lower in trading ofdistress loans.

Empirical findings also suggest that, for the total sample, institutional loans are traded at higher spreads than bankloans are. Thus, the longer duration of institutional loans translates into higher information costs in the loan trade. Theeffect of the longer duration is, however, offset by the higher amount and quality of information available regarding thepublicly traded borrowers. Regarding restructuring purpose loans, there is no evidence that these loans are traded at higherspreads than loans issued for more general purposes, such as debt repayment, working capital and corporate operations.24

6.1.3. Characteristics of the financial reporting quality of a borrower

6.1.3.1. Timely loss recognition. From the results reported in Table 4, it is immediately apparent that the timely incorporationof economic losses in a borrower’s financial statements reduces the bid–ask spread at which its loans are traded. This effectis statistically and economically significant and is robust to different measures of timely loss recognition. For Basu’s (1997)industry-year measure of timely loss recognition, an increase of one standard deviation in Timely-loss-recognition

reduces the bid–ask spread of a traded loan by about 6 cents. This effect is substantial: it constitutes 8.8% of the medianbid–ask spread of the loans of publicly traded borrowers.25 For Basu’s (1997) firm-specific measure and for the B_Score

ARTICLE IN PRESS

21 With the higher spreads for revolving facilities, one concern is the common perception that financial intermediaries usually issue revolvers to more

stable, investment-grade borrowers. However, this banking policy mainly applies to 364-day revolving facilities (Yago and McCarthy, 2004), while the vast

majority of revolvers in the research sample are long-term revolvers (credit lines above one year).22 Alternative measures of the arranger’s reputation include: (1) an estimation of the arranger’s market share in the primary market over an extended

period, from 1990 to 2003; (2) an estimation that accounts for the total market share of all the arrangers involved in the loan (in case of the multiple

arrangers).23 The positive relation between loan distress status and the bid–ask spread indicates that, for debt securities, the ‘‘second-moment’’ effect (i.e.,

variance effect) and the ‘‘first-moment’’ effect (i.e., mean effect) are not independent. Equity theory-based models characterize information asymmetry as

a second-moment effect that is unrelated to means, or first moments (Verrecchia, 2001). This important difference between debt and equity trading

emphasizes that the theory models may have to be modified to incorporate the distinctive features of debt securities. I thank Robert Verrecchia for

drawing my attention to this issue.24 The analysis does not control for financial covenants because 73.8% and 94.4% of observations of the total and the publicly traded samples

respectively have at least one financial covenant.25 As a robustness test, I employ an industry-year timely loss recognition measure estimated based on the Fama–French industries. I find that the

association between timely loss recognition and the bid–ask spread is even more pronounced in this estimation: the reduction in the bid–ask spread

caused by an increase of one standard deviation in Timely-loss-recognition represents 9.5% of the sample median bid–ask spread.



annual firm-specific measure, an increase of one standard deviation in Timely-loss-recognition reduces the bid–ask spreadby about 6 and 5 cents, respectively.26

These results demonstrate that by improving the efficiency of a borrower’s loan agreements and by enhancing corporategovernance and transparency of a borrower, timely loss recognition decreases information asymmetry in the secondaryloan trade. The analysis presented here provides unique empirical evidence that timely loss recognition decreases theinformation advantage of informed traders and therefore increases the efficiency of the trading of debt securities.27

While the empirical findings confirm that timely loss recognition significantly decreases information asymmetrybetween informed and uninformed traders, it is not possible to empirically disentangle the contracting route from theother mechanisms through which timely loss recognition operates. First, because 94% of the facility-year observations ofpublicly traded borrowers are subject to financial covenants, Timely-loss-recognition cannot be interacted with the variableindicating the existence of financial covenants. Second, the vast majority of the sample loans have income-based covenants.Therefore, controlling for covenants involving earnings measures also does not permit quantifying the impact of loancontracting efficiency on information asymmetry.

Additionally, there is no evidence that the market-to-book ratio significantly influences loan spreads. While Market-to-

book is negatively related to the bid–ask spread, and while this relation is consistent with high market-to-book borrowersexperiencing more conservative reporting, it is not statistically significant. The result is potentially explained by theequivocal relation between the market-to-book ratio and information asymmetry in the secondary loan trade. On onehand, the larger the market-to-book ratio, the larger the financial reporting conservatism is. On the other hand, a largermarket-to-book ratio is associated with a borrower’s greater growth options.

6.1.3.2. Timely gain recognition and the overall timeliness. I do not find that timely gain recognition or the overall timelinessof a borrower’s financial reporting increases debt trading efficiency. As evidenced in Table 5, the Timely-gain-recognition

and the Overall-timeliness variables do not affect the bid–ask spread. The results are unchanged when Basu’s (1997)firm-specific measures of timely gain recognition and the overall reporting timeliness (instead of the industry-specificmeasures) and the G_Score measure are employed in the analysis.

I also examine whether timely gain recognition has a stronger effect on the bid–ask spreads of distress loans.Untabulated results show no relation between the trading spread and the interaction term between the Timely-

gain-recognition and Distress variables. As an alternative measure of a loan’s performance, I employ the loan bid price in thesecondary trade. In this specification, I also do not observe that timely gain recognition has a stronger effect on theinformation environment of distressed loans. I realize that the insignificant impact of timely gain recognition onthe trading of distressed loans may result from the low power of the empirical tests, driven by a small number of distressedfacilities across the loans of public borrowers (8.9% of loan-year observations).

I also address the relation between loan trading and unconditional conservatism. Unconditional conservatism isestimated by b0+b1LF in the Basu (1997) model, where LF is the frequency of negative market-adjusted stock returns(Ball et al., 2008b). The untabulated results demonstrate that high levels of unconditional conservatism do not reduce thebid–ask spread in the secondary loan trade. These results are consistent with Basu (2005) and Ball et al. (2008b),who suggest that unconditional conservatism does not provide new information to investors.28

The insignificant relation between the bid–ask spread and timely gain recognition, overall timeliness, and unconditionalconservatism suggest that these attributes of accounting reporting do not decrease information asymmetry regarding aborrower’s performance and creditworthiness. These findings further support the special role timely loss recognition playsin enhancing the corporate governance of a borrower and in increasing information flow to market participants.

6.2. Control variables and robustness tests

The loadings on control variables are consistent with the predicted relations. The significant impact of N-of-market-

makers on the loan spread suggests that this variable successfully captures the transitory component of the bid–ask spread.I realize that the same association between the bid–ask spread and the number of market makers might be observed ifsome institutions were to intentionally avoid making a market in loans with high exposure to private information.However, given data availability, it is not possible to control for endogeneity between the number of market makers and thebid–ask spread. To alleviate this concern, I include in the analysis, to the best of data availability, all the variables that arepotentially associated with the bid–ask spread.

ARTICLE IN PRESS

26 The average weights on size, market-to-book and leverage that I obtain when estimating Khan and Watts’s (2007) model are consistent with the

values reported by Khan and Watts (2007). The mean and median values of C_Score and G_Score and the correlations between them are also consistent

with Khan and Watts’s (2007) findings.27 I have repeated the tests reported in Table 4 with the corresponding asymmetric timeliness measures instead of the timely loss recognition

measures incorporated into the analysis. The timely loss recognition and the asymmetric timeliness measures are highly correlated—the Pearson rank

correlation coefficients range between 71% and 95%, depending on the timeliness estimation method. Consequently, relying on the asymmetric timeliness

measures results in findings that are almost identical to those reported in Table 4.28 I realize that the insignificant relation between unconditional conservatism and the trading spreads may be driven by measurement error in the

empirical estimation of unconditional conservatism; the measurement error is caused by estimating unconditional conservatism over the finite sample

period.



The negative relation between Time-to-maturity and the loan spread suggests that younger loans are more heavilytraded and become less liquid with age. This result is consistent with the corresponding findings in bond trading. Theimpact of the time to maturity on the bid–ask spread is considerable; an increase of one standard deviation in Time-to-

maturity is associated with a decrease of 17 cents in the bid–ask spread.The results are robust to the inclusion of additional control variables, such as loan price, additional dummies for loan

type and purpose, loan maturity,29 specific types of financial covenants, the performance pricing provisions,30 the numberof lenders in the syndication, the number of the firm’s traded loans, the time period between the loan origination and itsfirst trading date, the rating category, the discrepancy between S&P’s and Moody’s credit ratings, leverage, sales growth,capital expenditures, the ratio of R&D expenses to sales and stock return volatility (estimated by a standard deviation ofdaily or monthly holding period returns) and firm size (measured by a logarithm of the annual sales or by a logarithm of thetotal assets).31 I do not control for a loan’s seniority and security because the vast majority of the traded loans are seniorand secured.

ARTICLE IN PRESS

Table 5Incorporating timely gain recognition and the overall timeliness measures

Pred. signs Publicly traded sample Publicly traded sample

Rating (b1) � �0.267*** �0.283***

(0.09) (0.10)

Facility-size (b2) � �0.033 �0.035

(0.02) (0.02)

Revolver (b3) + 0.226*** 0.225***

(0.04) (0.04)

Arranger-reputation (b4) � �0.059*** �0.057**

(0.02) (0.02)

Distress (b5) + 2.442*** 2.450***

(0.15) (0.15)

Investor (b6) + �0.023 �0.022

(0.04) (0.04)

Primary-purpose (b7) + 0.029 0.030

(0.04) (0.04)

N-of-market-makers (b8) � �0.021** �0.021**

(0.01) (0.01)

Time-to-maturity (b9) � �0.004*** �0.004***

(0.00) (0.00)

Income-net (b10) � �0.286*** �0.282***

(0.05) (0.05)

Timely-loss-recognition (b11) � �0.302** �0.489**

(0.13) (0.19)

Timely-gain-recognition (b12) � 0.439 �

(0.29)

Overall-timeliness (b12) � � 0.380

(0.34)

Market-to-book (b13) ? �0.004 �0.004

(0.00) (0.00)

Adj R-Sq 59.69% 59.57%

Number of observations 2,722 2,722

Number of clusters 492 492

Spread ¼ a+b1Rating+b2Facility-size+b3Revolver+b4Arranger-reputation+b5Distress+b6Investor+b7Primary-purpose+b8N-of-market-makers+b9Time-to-

maturity+b10Income-net+b11Timely-loss-recognition+b12Timely-gain-recognition/Overall-timeliness+b13Market-to-book.

Regressions include year and industry fixed effects. Standard errors are heteroskedasticity robust, clustered at the firm level. Standard errors are reported

in parentheses. ***, **,* denote significance at the 1, 5 and 10 percent level, respectively. Variables: Timely-loss-recognition—estimated by the sum of b2 and

b3 in a piecewise-linear industry-year regression of earnings on the contemporaneous stock returns (Basu, 1997): NIit ¼ b0 þ b1DRit þ b2Rit þ b3Rit � DRit .

Timely-gain-recognition—estimated by b2 in Basu’s (1997) model. Overall-timeliness—a measure of the overall timeliness, for both gains and losses,

estimated by R2 of Basu’s (1997) model. For the definition of Spread and the rest of the explanatory variables, see Table 4.

29 Because of the high correlation between the time-to-maturity and maturity variables (71%), I do not include these variables simultaneously in the

regression analysis.30 The 48% of the sample loans are subject to the performance pricing provision, which includes both/either an interest-increasing performance

pricing option and/or an interest-decreasing performance pricing option. The interest-increasing performance pricing option gives lenders the right to

receive higher interest rates if the borrower’s credit quality deteriorates The interest-decreasing performance pricing option reduces interest rates if the

borrower’s creditworthiness improves; thereby this option alleviates loan prepayment risk to the lenders (Asquith et al., 2005). No evidence is found of a

significant relation between the loan bid–ask spread and the inclusion of the performance pricing provision/options in a loan contract.31 A high correlation (75%) between loan size and firm size prevents the simultaneous incorporation of both variables in the regression. An analysis

incorporating firm size instead of loan size provides almost identical results.



Table 6 offers additional specifications to test the robustness of the results. I begin by verifying that the empiricalfindings are not driven by observations based on a single institution’s reporting. First, measurement error is likely to behigher when loan bid and ask price quotes are reported to LPC by a single market maker. Second, the transitory componentof the bid–ask spread is expected to be higher in these cases. Third, the most worrisome concern is that an arranger ofsyndication, who is an informed trader in the loan market, is likely to be a market maker if only one institution makes amarket in a traded loan. To alleviate these concerns, I perform the analysis for loans followed by more than one marketmaker. Despite a substantial reduction in the sample size, the explanatory power of the model increases and theinformation asymmetry variables continue to be significantly related to the loan spreads.32 Additionally, I verify that thefindings are not sensitive to the clustering procedure employed in the analysis: clustering at the year level (instead of firm-level clustering) provides similar results. I interpret these findings as a further verification of the robustness of theempirical analysis.

The model’s explanatory power is high; while it is comparable to the explanatory power of the models explaining equitytrading spreads, it is considerably higher than that of models explaining corporate bond trading spreads. The model’sexplanatory power is not driven by the incorporation of the fixed effects in the regression; estimating the model without 2-digit industry and year fixed-effects results in reduction of Adj R-Sq by only 1.4%–1.8%, depending on a model’sspecification.

ARTICLE IN PRESS

Table 6Robustness tests

Pred. signs Quotes reported by more than one market maker Clustering at the year level

Total Publicly traded Total Publicly traded

Public (b1) � �0.142*** � �0.126** �

(0.05) (0.06)

Rating (b2) � � � �0.163*** �0.280**

(0.03) (0.14)

Facility-size (b3) � �0.169*** �0.049* �0.134*** �0.034

(0.03) (0.03) (0.03) (0.02)

Revolver (b4) + 0.199*** 0.291*** 0.112*** 0.226***

(0.04) (0.04) (0.03) (0.06)

Arranger-reputation (b5) � �0.067** �0.083*** �0.075*** �0.055***

(0.03) (0.03) (0.02) (0.01)

Distress (b6) + 2.447*** 2.155*** 3.205*** 2.439***

(0.11) (0.18) (0.23) (0.29)

Investor (b7) + 0.073 �0.041 0.111** �0.023

(0.05) (0.04) (0.05) (0.03)

Primary-purpose (b8) + �0.050 0.019 0.007 0.030

(0.05) (0.05) (0.03) (0.03)

N-of-market-makers (b9) � �0.033*** �0.015 �0.052*** �0.021**

(0.01) (0.01) (0.01) (0.01)

Time-to-maturity (b10) � �0.005*** �0.004*** �0.007*** �0.005***

(0.00) (0.00) (0.00) (0.00)

Income-net (b11) � � �0.259*** � �0.273***

(0.06) (0.02)

Timely-loss-recognition (b12) � � �0.509*** � �0.372***

(0.18) (0.11)

Market-to-book (b13) ? � �0.004 � �0.004*

(0.00) (0.00)

Adj R-Sq 61.28% 63.68% 57.13% 59.53%

Number of observations 4,281 1,572 8,619 2,722

Number of clusters 781 324 6 6

Spread ¼ a+b1 Public+b2 Rating+b3 Facility-size+b4 Revolver+b5 Arranger-reputation+b6 Distress+b7 Investor+b8 Primary-purpose+b9 N-of-market-makers+b10

Time-to-maturity+b11 Income-net+b12 Timely-loss-recognition+b13 Market-to-book.

Regressions include year and industry fixed effects. Standard errors are heteroskedasticity robust. Standard errors are clustered at the firm level for the

regression analysis of the traded facilities with price quotes reported by more than one market maker. Standard errors are reported in parentheses. ***, **,*

denote significance at the 1, 5 and 10 percent level, respectively. Variables: Timely-loss-recognition—estimated by the sum of b2 and b3 in a piecewise-

linear industry-year regression of earnings on the contemporaneous stock returns (Basu, 1997): NIit ¼ b0 þ b1DRit þ b2Rit þ b3Rit � DRit . For the definition

of Spread and the rest of the explanatory variables, see Table 4.

32 I exclude the Rating indicator variable from this analysis because more than 90% of the facilities followed by more than one market maker have a

public credit rating.



7. Conclusions

In this paper, I employ a sample of traded syndicated loans to explore firm- and issue-specific characteristics thatresolve or exacerbate information asymmetry in the trading of debt securities. First, I examine variables suggestedby prior research as generally related to information asymmetry. I find that public reporting and the availabilityof a credit rating decrease information asymmetry in the loan trade. Further, consistent with the higheruncertainty associated with revolvers, I find that revolvers are traded at higher bid–ask spreads. Empirical evidencealso suggests that loans of profit firms are traded at significantly lower spreads than loans of firms experiencinglosses are.

Second, I examine unique characteristics of the information environment of syndicated loans. Consistentwith the arranger’s primary role in resolving information asymmetry in the syndicated loan market, I find that loanssyndicated by more reputable arrangers are traded at significantly lower spreads. Further, empirical findingssuggest that the considerable information advantage of informed traders in the trading of distress loans is mani-fested in significantly higher spreads of these facilities. I also find that institutional loans, which are typicallyissued with long maturities and back-end-loaded repayment schedules, are traded at higher spreads than bankloans are.

Third, I examine whether there are aspects of the financial reporting quality of a borrower that decreaseinformation asymmetry in the secondary loan trade. I find that timely loss recognition reduces the bid–ask spread atwhich loans are traded. The effect of timely loss recognition on the loan trading spreads is both statisticallyand economically significant and is robust to different measures of timely loss recognition. These results demonstratethat by enhancing corporate governance of a borrower and by increasing the amount and quality of informationavailable to uninformed market participants, timely loss recognition decreases information asymmetry in the secondaryloan trade. Overall, the analysis presented in this paper provides unique empirical evidence that timely loss recognitionreduces the information advantage of informed traders and increases the efficiency of the secondary trading of debtsecurities.

Appendix A. Variable definitions

Variables Description

Arranger-reputation As suggested by Lee and Mullineaux (2004) and Sufi (2007a, b), the arranger’s reputation is estimated by the arranger’s

average market share in the primary syndicated loan market. The market share is measured by the ratio of the amount of

loans that the financial intermediary syndicated as a lead arranger to the total amount of loans syndicated on the primary

loan market over the period from 1998 to 2003. In the case of multiple arrangers, I consider the highest market share across

the arrangers involved in the loan transaction.

Bid–ask spread The bid–ask spread is estimated based on bid and ask price quotes aggregated across dealers. Bid and ask prices are quoted

as a percent of par (or cents on the dollar of par value). The bid–ask spread is measured as the average annual bid–ask spread

of the traded facility.

Distress An indicator variable taking the value of one if a loan is traded at the annual average bid price below 90% of the par value,

zero otherwise.

Facility-size A logarithm of the facility amount.

Income-net An indicator variable taking the value of one if the borrower’s current year net income is positive, zero otherwise.

Investor An indicator variable taking the value of one if the loan’s type is term loan B, C or D (institutional term loans), zero otherwise.

Market-to-book The ratio of the firm’s market value to book value of common equity, estimated at the end of the borrower’s fiscal year.

Number-of-market-

makers

The average annual number of market makers that provide a loan’s bid and ask prices to LPC.

Overall-timeliness The overall timeliness of financial reporting, for both gains and losses, estimated by R2 of Basu’s (1997) model. The measure

is based on the industry-year estimation on the model.

Public An indicator variable taking the value of one if the borrower is a publicly reporting firm in the year when the loan is traded

on the syndicated loan market, zero otherwise.

Primary-purpose An indicator variable taking the value of one if the facility’s primary purpose is Takeover, LBO/MBO or Recapitalization, zero

otherwise. A loan with a primary purpose of recapitalization is a loan to support a material change in a firm’s capital

structure, often made in conjunction with other debt or equity offerings.

Rating An indicator variable taking the value of one if a firm and/or facility has an available credit rating, zero otherwise. Credit

rating categories include Moody’s Sr. Debt, Moody’s Loan Rating, S&P Sr. Debt, S&P Loan Rating, Fitch LT and Fitch Loan

Rating.

Revolver An indicator variable taking the value of one if the facility’s type is revolver above one year, zero otherwise.

Time-to-maturity The number of months between the facility’s trading date on the secondary loan market and the date when the facility

matures.

Timely-gain-

recognitionThe measure of timely gain recognition in financial reporting, estimated by b2 in Basu’s (1997) model. The measure is based

on the industry-year estimation on the model.

Timely-loss-recognition The measure of timely loss recognition in financial reporting. This measure is estimated as one of the following: (1) the sum

of b2 and b3 in Basu’s (1997) model, based on the industry-year model’s estimation; (2) the sum of b2 and b3 in Basu’s (1997)

model, based on the firm-specific model’s estimation; (3) B_Score based on Khan and Watt’s (2007) model.

ARTICLE IN PRESS



References

Acharya, V.V., Pedersen, L.H., 2005. Asset pricing with liquidity risk. Journal of Financial Economics 77, 375–410.Agrawal, V., Kothare, M., Rao, R.K.S., Wadhwa, P., 2004. Bid–ask spreads, informed investors, and the firm’s financial conditions. Quarterly Review of

Economics and Finance 44, 58–76.Ahmed, A., Billings, B., Morton, R., Harris, M., 2002. The role of accounting conservatism in mitigating bondholder–shareholder conflict over dividend

policy and in reducing debt cost. The Accounting Review 77, 867–890.Alexander, G.J., Edwards, A.K., Ferri, M.G., 2000. The determinants of trading volume of high-yield corporate bonds. Journal of Financial Markets 3,

177–204.Allen, L., Gottesman, A.A., 2005. The informational efficiency of the equity market as compared to the syndicated bank loan market. Working Paper, Baruch

College.Altman, E.I., Gande, A., Saunders, A., 2006. Bank debt versus bond debt: evidence from secondary market prices. Working Paper, New York University.Amihud, Y., Mendelson, H., 1986. Asset pricing and the bid–ask spread. Journal of Financial Economics 17, 223–249.Asquith, P., Beatty, A.L., Weber, J.P., 2005. Performance pricing in bank debt contracts. Journal of Accounting and Economics 40, 101–128.Assender, T., 2000. Growth and importance of loan ratings. In: Rhodes, T., Clark, K., Campbell, M. (Eds.), Syndicated Lending: Practice and Documentation.

Euromoney, third ed., London, UK.Ball, R., 2001. Infrastructure requirements for an economically efficient system of public financial reporting and disclosure. Brookings-Wharton Papers on

Financial Services, pp. 69–127.Ball, R., Kothari, S.P., 2007. Econometrics of the Basu asymmetric timeliness coefficient and accounting conservatism. Journal of Accounting and

Economics, forthcoming.Ball, R., Shivakumar, L., 2005. Earnings quality in UK private firms: comparative loss recognition timeliness. Journal of Accounting and Economics 39,

83–128.Ball, R., Shivakumar, L., 2006. The role of accruals in asymmetrically timely gain and loss recognition. Journal of Accounting Research 44,

204–242.Ball, R., Bushman, R., Vasvari, F., 2008a. The debt-contracting value of accounting information and loan syndicate structure. Journal of Accounting

Research 46, 247–287.Ball, R., Robin, A., Sadka, G., 2008b. Is financial reporting shaped by equity markets or by debt markets? An international study of timeliness and

conservatism. Review of Accounting Studies 13, 168–205.Barclay, J.B., Dunbar, C.G., 1996. Private information and the costs of trading around quarterly earnings announcements. Financial Analysts Journal 52,

75–84.Barth, J., Goldberg, L., Nolle, D., Yago, G., 2004. Financial supervision and crisis management: U.S. experience and lessons for emerging market economies.

Mimeo.Basu, S., 1997. The conservatism principle and the asymmetric timeliness of earnings. Journal of Accounting and Economics 24, 3–37.Basu, S., 2005. Discussion of conditional and unconditional conservatism: concepts and modeling. Review of Accounting Studies 10, 311–321.Beatty, A., Weber, J., 2006. Accounting discretion in fair value estimates: an examination of SFAS goodwill impairments. Journal of Accounting Research 44,

257–288.Beatty, A., Weber, J., Yu, J., 2008. Conservatism and debt. Journal of Accounting and Economics 45, 154–174.Beaver, W.H., Ryan, S.G., 2005. Conditional and unconditional conservatism: concepts and modeling. Review of Accounting Studies 10, 269–276.Berger, A.N., Udell, G.F., 1995. Relationship lending and lines of credit in small firm finance. Journal of Business 68, 351–381.Bharath, S.T., Dahiya, S., Saunders, A., Srinivasan, A., 2007. So what do I get? The bank’s view of lending relationships. The Journal of Financial Economics

85, 368–419.Bhasin, V., Carey, M., 1999. The determinants of corporate loan maturity. In: Federal Reserve Bank of Chicago Bank Structure Conference Proceedings.Botosan, C., 1997. Disclosure level and the cost of equity capital. The Accounting Review 72, 323–349.Cebenoyan, A.S., Strahan, P.E., 2004. Risk management, capital structure and lending at banks. Journal of Banking & Finance 28, 19–43.Chakravarty, S., Sarkar, A., 2003. Trading costs in three U.S. bond markets. The Journal of Fixed Income 13, 39–48.Copeland, T.E., Galai, D., 1983. Information effects on the bid–ask spread. The Journal of Finance 38, 1457–1469.Demsetz, R.S., 2000. Bank loan sales: a new look at the motivations for secondary market activity. Journal of Financial Research 23, 192–222.Dennis, S.A., Mullineaux, D.J., 2000. Syndicated loans. Journal of Financial Intermediation 9, 404–426.Diamond, D.W., 1984. Financial intermediation and delegated monitoring. The Review of Economic Studies 51, 393–414.Diamond, D.W., 1991. Monitoring and reputation: the choice between bank loans and directly placed debt. The Journal of Political Economy 99,

689–721.Diamond, D.W., 1996. Financial intermediation as delegated monitoring: a simple example, debt maturity structure and liquidity risk. Federal Reserve

Bank of Richmond Economic Quarterly 82/3, 51–66.Diamond, D.W., Verrecchia, R.E., 1991. Disclosure, liquidity, and the cost of capital. The Journal of Finance 46, 1325–1359.Dichev, I.D., Skinner, D.J., 2002. Large-sample evidence on the debt covenant hypothesis. Journal of Accounting and Economics 40, 1091–1123.Dichev, I.D., Beatty, A.L., Weber, J.P., 2002. The role and characteristics of accounting-based performance pricing in private debt contracts. Working Paper,

University of Michigan.Easley, D., O’Hara, M., 2004. Information and the cost of capital. The Journal of Finance 59, 1553–1583.Easley, D., Hvidkjaer, S., O’Hara, M., 2002. Is information risk a determinant of asset returns? The Journal of Finance 58, 2185–2210.Ertimur, Y., 2004. Accounting numbers and information asymmetry: evidence from loss firms. Working Paper, Duke University.Frankel, R.M., Roychowdhury, S., 2007. Predicted asymmetric timeliness and the persistence of special items. Working Paper, Washington University,

St. Louis.Froot, K.A., Stein, J.C., 1998. Risk management, capital budgeting, and capital structure policy for financial institutions: an integrated approach. Journal of

Financial Economics 47, 55–82.Garbade, K., 1982. Securities Markets. McGraw-Hill, New York.Givoly, D., Hayn, C., 2000. The changing time-series properties of earnings, cash flows and accruals: has financial reporting become more conservative?

Journal of Accounting and Economics 29, 287–320.Glosten, L.R., Harris, L.E., 1988. Estimating the components of the bid/ask spread. Journal of Financial Economics 21, 123–142.Glosten, L.R., Milgrom, P.R., 1985. Bid, ask and transaction prices in a specialist market with heterogeneously informed traders. Journal of Financial

Economics 14, 71–100.Goldstein, M.A., Nelling, E.F., 1999. Market making and trading in Nasdaq stocks. The Financial Review 34, 27–45.Gorton, G.B., Haubrich, J.G., 1990. The loan sales market. Research in Financial Services 2, 85–135.Gorton, G.B., Pennacchi, G.G., 1995. Banks and loan sales: marketing nonmarketable assets. Journal of Monetary Economics 35, 389–411.Gorton, G.B., Winton, A., 2002. Financial intermediation. Working Paper No. w8928, NBER.Guay, W., 2006. Discussion of Ball and Shivakumar (2006): the role of accruals in asymmetrically timely gain and loss recognition. Journal of Accounting

Research 44, 243–255.Guay, W., Verrecchia, R., 2006. Discussion of an economic framework for conservative accounting and Bushman and Piotroski (2006). Journal of

Accounting and Economics 42, 149–165.Hasbrouck, J., 1991. Measuring the information content of stock trades. The Journal of Finance 46, 179–207.

ARTICLE IN PRESS



Ho, T., Saunders, A., 1983. Fixed-rate loan commitments, takedown risk, and the dynamics of hedging with futures. Journal of Financial and QuantitativeAnalysis 18, 499–516.

Holthausen, R.W., Watts, R.L., 2001. The relevance of the value-relevance literature for financial accounting standard setting. Journal of Accounting andEconomics 31, 3–75.

Hong, G., Warga, A., 2000. An empirical study of bond market transactions. Financial Analyst Journal 58, 32–46.Hugh, T., Wang, Z., 2004. The integration of bank syndicated loan and junk bond markets. Journal of Banking & Finance 28, 299–329.Jones, J.D., Lang, D.W., Nigro, P.J., 2005. Agent bank behavior in bank loan syndications. The Journal of Financial Research 28, 385–402.Kalimipalli, M., Warga, A., 2002. Bid–ask spread, volatility, and volume in the corporate bond market. The Journal of Fixed Income 11, 31–42.Khan, M., Watts, R.L., 2007. Estimation and validation of a firm-year measure of conservatism. Working Paper, Massachusetts Institute of Technology.Kyle, A., 1985. Continuous auctions and insider trade. Econometrica 53, 1315–1335.LaFond, R., Roychowdhury, S., 2008. Managerial ownership and accounting conservatism. Journal of Accounting Research 46, 101–135.LaFond, R., Watts, R.L., 2008. The information role of conservatism. Accounting Review 83, 447–479.Lang, M., Lundholm, R., 1993. Cross-sectional determinants of analyst ratings of corporate disclosures. Journal of Accounting Research 31, 246–271.Lee, S.W., Mullineaux, D.J., 2004. Monitoring, financial distress, and the structure of commercial lending syndicates. Financial Management 33, 107–130.Lee, C., Mucklow, B., Ready, M., 1993. Spreads, depths and the impact of earnings information: an intraday analysis. The Review of Financial Studies 6,

345–374.Leftwich, R., 1983. Accounting information in private markets: evidence from private lending agreements. The Accounting Review 58, 23–42.Leuz, C., 2003. IAS versus U.S. GAAP: information asymmetry-based evidence from Germany’s new market. Journal of Accounting Research 41, 445–472.Leuz, C., Verrecchia, R.E., 2000. The economic consequences of increased disclosure. Journal of Accounting Research 38, 91–124.Loan Pricing Corporation, 2003. Gold Sheets.Lummer, S.L., McConnell, J.J., 1989. Further evidence on the bank lending process and the capital-market response to bank loan agreements. Journal of

Financial Economics 25, 99–122.Nunn, K.P., Hill Jr., J., Schneeweis, T., 1986. Corporate bond price data sources and return/risk measurement. Journal of Financial and Quantitative Analysis

21, 197–208.Pae, J., Thornton, D.B., Welker, M., 2005. The link between earnings conservatism and the price-to-book ratio. Contemporary Accounting Research 22,

693–717.Panyagometh, K., Roberts, G., 2002. Private information, agency problems and determinants of loan syndications: evidence from 1987–1999. Working

Paper, York University.Pavel, C.A., Phillis, D., 1987. Why commercial banks sell loans: an empirical analysis. Economic Perspectives, Federal Reserve Bank of Chicago.Pennacchi, G.G., 1988. Loan sales and the cost of bank capital. The Journal of Finance 43, 375–396.Rajan, R., 1998. The past and future of commercial banking viewed through an incomplete contract lens. Journal of Money, Credit, and Banking 30,

524–550.Roychowdhury, S., Watts, R.L., 2007. Asymmetric timeliness of earnings, market-to-book and conservatism in financial reporting. Journal of Accounting

and Economics 44, 2–31.Sadka, R., 2004. Liquidity and asset pricing. EFA 2004 Maastricht Meetings Paper No. 5290.Sadka, G., Sadka, R., 2004. The post-earnings-announcement drift and liquidity risk. Working Paper, University of Chicago.Sarig, O., Warga, A., 1989. Bond price data and bond market liquidity. Journal of Financial and Quantitative Analysis 24, 367–378.Schrand, C.M., Verrecchia, R.E., 2005. Information disclosure and adverse selection explanations for IPO underpricing. Working Paper, University of

Pennsylvania.Simons, K., 1993. Why do banks syndicate loans? New England Economic Review Jan/Feb, 45–53.Smith, C.W., Watts, R.L., 1992. The investment opportunity set and corporate financing, dividend and compensation policies. Journal of Financial

Economics 32, 263–292.Standard & Poor’s, 2003, 2006. A Guide to the U.S. Loan Market.Stoll, H.R., 1985. Alternative views of market making. In: Amihud, Y., Ho, T., Schartz, R. (Eds.), Market Making and the Changing Structure of the Securities

Industry. Lexington Health, MA, pp. 67–92.Stoll, H.R., 1989. Inferring the components of the bid–ask spread: theory and empirical tests. Journal of Finance 44, 115–134.Sufi, A., 2007a. Information asymmetry and financing arrangements: evidence from syndicated loans. The Journal of Finance 62, 629–668.Sufi, A., 2007b. The real effects of debt certification: evidence from the introduction of bank loan ratings. Review of Financial Studies, forthcoming.Taylor, A., Yang, R., 2004. The Evolution of the Corporate Loan Asset Class. Loan Syndications and Trading Association, New York.Verrecchia, R.E., 2001. Essays on disclosure. Journal of Accounting and Economics 32, 91–180.Verrecchia, R.E., Weber, J., 2007. Firm performance measures and adverse selection costs. Working paper, University of Pennsylvania.Watts, R.L., 1993. A proposal for research on conservatism. Working Paper, University of Rochester.Watts, R.L., 2003a. Conservatism in accounting Part I: explanations and implications. Accounting Horizons 3, 207–221.Watts, R.L., 2003b. Conservatism in accounting Part II: evidence and research opportunities. Accounting Horizons 4, 287–301.Watts, R.L., Zimmerman, J.L., 1986. Positive Accounting Theory. Prentice-Hall, Englewood Cliffs, NJ.Weidner, D., 2000. Syndicated lending closes out 90s on a tear. American Banker January 10th, 1.Yago, G., McCarthy, D., 2004. The U.S. Leveraged Loan Market: A Primer. Milken Institute.Yohn, T.L., 1998. Information asymmetry around earnings announcements. Review of Quantitative Finance and Accounting 11, 165–182.Zhang, J., 2008. The contracting benefits of accounting conservatism to lenders and borrowers. Journal of Accounting and Economics 45, 27–54.

ARTICLE IN PRESS


Author's personal copy - Booth School of Businessfaculty.chicagobooth.edu/Regina.Wittenberg...

Documents

Transcript of Author's personal copy - Booth School of Businessfaculty.chicagobooth.edu/Regina.Wittenberg...