Citation
Dogs That Bark: Why Are Bank Loan Announcements Newsworthy?

Material Information

Title:
Dogs That Bark: Why Are Bank Loan Announcements Newsworthy?
Creator:
Gonzalez Alana, Laura
Publisher:
University of Florida
Publication Date:
Language:
English

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Business Administration
Finance, Insurance and Real Estate
Committee Chair:
James, Christophe M.
Committee Members:
Nimalendran, Mahendrara
Houston, Joel F.
Naranjo, Andy
Ward, Ronald W.
Graduation Date:
8/9/2008

Subjects

Subjects / Keywords:
Assets ( jstor )
Bank loans ( jstor )
Debt ( jstor )
Library users ( jstor )
Loan financing ( jstor )
Loans ( jstor )
National debt ( jstor )
Net income ( jstor )
News media ( jstor )
Participating loans ( jstor )
announcement, bank, commercial, incremental, loan

Notes

General Note:
My dissertation focuses on the information content of bank loans among public firms. I examine why reporting of bank loans in the press is associated with a positive share price reaction despite the widespread use of loans among public firms. I find that loan restructurings following covenant violations and larger loans relative to firm size are more likely to be reported in the press. Moreover, I find that although all loans are disclosed, the market reacts only to loans reported before activation. In addition, I study how credit risk and asymmetric information proxies determine loan reporting and incremental financing decisions, and observe that a focus on press reported loans, as previous studies have done, overestimates the importance of these proxies. My findings suggest that reported loans are more noteworthy because they are more informative about the potential of the borrower than non-reported loans, and that the information content of the average reported bank loan decreases during the 1996 through 2004 period.

Record Information

Source Institution:
UFRGP
Rights Management:
All applicable rights reserved by the source institution and holding location.
Embargo Date:
8/31/2010

Downloads

This item has the following downloads:


Full Text





studies the likelihood of a higher z-value (average standardized prediction errors times the square

root of the number of observations) in relation to whether the loan is reported and other potential

determinants of market reaction identified in the study of reporting determinants (Table 3-1).

Similarly, Table 3-2 presents the results of the OLS analysis of the market reaction for public

debt and non-bank private debt issues.

As shown in Table 3-1, the market reaction is driven by the size of the issue and market

timing considerations. More specifically, I find that the reaction to loan news increases with

lower measures of cumulative returns of the loan borrower relative to the market during the 12

months preceding the earliest of the news or loan activation of reported loans. This is consistent

with previous work by Hadlock and James (2002), who find evidence in support of a market

timing argument by which firms tend to choose announced bank lending over public debt when

their returns decrease with respect to the market.

The analysis of the market reaction for public debt and non-bank private debt news and

issues is, as expected, inconclusive given that no significant market reaction is observed, in

general, for these two types for issues surrounding news articles or activation of non-reported

issues. As such, and in the case of public debt issues, I find no evidence in support of the market

timing hypothesis, like suggested by the firm characteristics in Table 2-3. However, the size of

the issue relative to firm size still appears to have a positive effect on the market reaction,

especially in the case of public debt issues reported after activation, although the significance

level is a little over 10%. Finally, and although the market reaction is not significant either for

reported and non-reported non-bank private debt issues, I find that tangibles relative to assets can

increase the market reaction, although this firm characteristic does not differentiate reported

from non-reported non-bank private debt borrowers.










LIST OF TABLES


Table page

2-1 Annual Distribution of Loans by Year of Deal Activation ................. .......................28

2-2 Loan Deal Purpose ................. ...............29................

2-3 Firm and Loan Summary Statistics for Reported and Unreported Loans..........................30

2-4 Firm and Deal Summary Statistics for Reported and Unreported Public Debt and
Non-Bank Private Debt ................. ...............32........... ....

2-5 Market Reaction Summary Statistics ................. ...............33...............

3-1 Determinants of the Likelihood and Timing of Bank Loan Reporting in the Press ..........42

3-2 Marginal Effects of the Lowest and Highest Quartile Determinants on Loan
Rep orti ng ................. ...............43........... ....

3-3 Determinants of the Likelihood of Public and Non-Bank Private Debt Reporting ...........44

4-1 Determinants of the Stock Price Reaction Surrounding Loan Reports and Activation.....47

4-2 Determinants of the Stock Price Reaction Surrounding Public Debt and Non-Bank
Private Debt Reports and Activation .............. ...............48....

5-1 Multinomial Logistic Regressions on Incremental Financing Decisions ..........................53

6-1 Median Peer-Adjusted Measures of Subsequent Operating Performance ................... ......57










Table 2-5. Market Reaction Summary Statistics


Mean Median Mean Median Mean Median Mean Median
Loans N=53 N=3 7 N=150 N=1071

CAR (%) 2.56*+ 0.13 0.033 0.26 0.46 0.16 0.25 -0.06
Average SPE 0.32 0.009 0.077 0.05
Z value 2.33 0.05 0.96 1.64

Public Debt N=52 N=111 N=3 6 N=146

CAR (%) 0.58 0.33 -0.11 -0.1 1.21 0.56 -0.18 -0.001
Average SPE 0.21 -0.03 0.39 -0.07
Z value 1.51 -0.31 2.34 -0.85

Non-Bank N=24 N=3 N=11 N=43

Private Debt
CAR (%) 1.12 0.003 -3.49 -4.54 -0.38 -0.42 1.67 -0.38
Average SPE 0.11 -0.67 -0.11 0.05
Z value 0.55 -1.2 -0.36 0.33

The sample consists of 240 loan news and 1071 non-reported bank loan agreements, 199 reported and 146 non-reported public debt
issues, and 38 reported and 43 non-reported non-bank private debt issues activated between January 1996 and December 2004, with
all news being free of contamination. Non-confounding news means that the loan information in the press articles are not bundled with
other non-loan related information such as earnings, dividends, changes in management or completion of acquisitions. I identify
reported and non-reported issues searching Dow Jones sources for articles containing the issue amount and/or the key words used in
the literature of bank loan, public debt and non-bank private debt announcements. CAR and SPE refer to the two-day Cumulative
Abnormal Returns and Standardized Prediction Errors of a 2-day window event study centered on the earliest of news date and deal
activation date. Z value is calculated as the square root of the number of observations times the average Standardized Prediction Error.
* Statistically different at 10% level from non-reported subsample.
+ Statistically different at 10% level from reported subsample that does not precede loan activation.


Non-Confounding News
Before Loan Activation


Non-Confounding News
On Loan Activation


Non-Confounding News
After Loan Activation


Non-reported Deals









of bank borrowing over public debt borrowing. In consequence, one could argue that non-

reported loans require a less close relation between lender and borrower and are, therefore, more

transaction oriented, although somehow still special and different from public debt.

In general, and although the loans in the sample are disclosed, the market reaction is only

significant in the case of reported loans, especially when the earliest news precedes loan

activation. This reaction to loan reporting when the news precedes loan activation is more

intense when loans are bigger and when the performance of the borrower' s stock is poor with

respect to the market during the year prior. Thus, although borrowing is more difficult to obtain

in the case of reported loans, which makes them more surprising, and therefore newsworthy, loan

reporting also appears to time the market. Still, and consistent with the view that bank loans are

more informative about firm potential when they are reported in the press, reported borrowers

improve their operating performance with respect to non-reported ones over the three years

following the activation of the loan.










asymmetries associated with selling securities. Thus, excluding confounding reporting may bias

the results and conclusions.

The news information on bank loans is supplemented with information from filings with

the Securities and Exchange Commission (SEC). More specifically, I collect data from the SEC

filings whenever available on whether the loans constitute a new agreement, renewal or a

restructure deal following a covenant violation and subsequent waiver. This manual search

within SEC filings covers the two years prior to loan activation because Roberts and Sufl (2007)

find in their study of renegotiation of financial contracts that the average effective maturity of

bank loans is half the average stated maturity, which Deal Scan reports to be of around four years

for the loans in the sample. More specifically, I search the SEC filings for specific expressions

used in previous studies, and check each passage to ensure that the expressions indeed refer to

financial covenant violations, waivers, and loan restructurings.

The data on issuing firm characteristics, stock price reaction, and analysts' forecasts, is

obtained from Compustat, the Center for Research of Security Prices (CRSP), and the

Institutional Brokers' Estimate System (IBES) data tape. As a result, the overall sample consists

of 1,375 bank loan deals, 355 public debt issues and 92 non-bank private debt deals. The news

search process identifies 304 loans that are reported in the financial press, 166 additional loans

that are identified in the wire but not reported in the press, 905 loans that are not reported in wire

or press articles, 207 reported public debt issues and 42 reported non-bank private debt issues. In





1 The specific terms are those also used by Roberts and Sufi (2007): "in violation of covenant," "in violation of a
covenant," "in default of covenant," "in technical violation of covenant," "in technical violation of a covenant," "in
violation of financial covenant," "in violation of a financial covenant," "in technical violation of a financial
covenant," "in technical violation of financial covenant," "in technical default of a financial covenant," "in technical
default of financial covenant," "not in compliance," "out of compliance," "received waiver," receiver a waiver,"
"obtained a waiver," "obtained waiver."










Diamond, Douglas, 1991, Monitoring and reputation, the choice between bank loans and directly
placed debt, Journal of Political Economy 99, 688-721.

Eckbo, Espen, Ronald Masulis, and Oyvind Norli, 2007, Security offerings, Handbook of
corporate Finance: Empirical Corporate Finance, Vol. 1, North-Holland/Elsevier,
Handbooks in Finance series, Ch. 6, forthcoming.

Fama, Eugene, 1985, What' s different about banks?, Journal of2~onetazy Economics 17, 239-
249.

Gomes, Armando, and Gordon Phillips, 2005, Why do public firms issue private and public
securities?, SSRN working paper.

Guner, A. Burak, Loan sales and the cost of corporate borrowing, Review ofFinancial Studies,
forthcoming.

Hadlock, Charles, and Christopher James, 2002, Do Banks Provide Financial Slack?, Journal of
Finance 57, 1383-1419.

Houston, Joel and Christopher James, 1996, Bank information monopolies and the mix of private
and public debt claims, Journal ofFinance 28, 149-171.

Ivashina, Victoria, and Zheng Sun, Institutional stock trading on loan market information, SSRN
working paper.

James, Christopher, 1987, Some evidence on the uniqueness of bank loans, Journal ofFinancial
Economics 19, 217-235.

James, Christopher, and David Smith, 2000, Are banks still special? New evidence on their role
in the corporate capital-raising process, Journal ofApplied Corporate Finance 13, 395-
422.

Johnson, Shane, 1997, An empirical investigation of corporate debt ownership structure, Journal
ofFinancial and Quantitative Analysis 32, 47-69.

Jung, Kooyul, Yong-Cheol Kim, and Rene Stulz, 1996, Timing, investment opportunities,
managerial discretion, and the security issue decision, Journal ofFinancial Econontics 42,
159-185.

Lummer, Scott, and John McConnell, 1989, Further evidence on the bank lending process and
the capital-market response to bank loan agreements, Journal ofFinancial Econontics 25,
99-122.

Marsh, Paul, 1979, equity rights issues and the efficiency of the UK stock market, Journal of
Finance 34, 839-862.

Mikkelson, Wayne H. and M. Megan Partch, 1986, Valuation effects of securities offerings and
the issuance process, Journal ofFinancial Economics 15, 3 1-60.










decisions should also consider cases of non-reported loans. In this context, I Eind that although

reported loans and borrowers differ significantly from non-reported loans and borrowers, the

factors that determine the choice between bank debt, public debt and non-bank private debt do

not change whether reported loans or all loans are considered. In consequence, one could argue

that non-reported loans are somewhat special and different from public debt.

Besides a more effective screening, the "specialness" of bank loans relative to other types

of Einancing also has to do with more intense monitoring, which would arguably be more

important for loans with higher covenant intensity, i.e., the reported loans. Furthermore, the

access to private information and contracting conditions that protect banks from the higher

potential default on reported loans could arguably lead or not to an improvement in operating

performance. In this context, I Eind that although reported borrowers present poorer measures of

operating performance during the year prior to loan activation (relative to non-reported

borrowers), there is relative improvement in the operating performance of reported borrowers

with respect to non-reported borrowers two years following loan activation. This result is

consistent with the recent literature on loan covenants (Roberts and Sufi (2007), Nini et al.

(2008)) that studies the relation between loan covenant violations, loan restructurings, and

subsequent performance

The remainder of the paper is organized as follows. Section II provides a description of

the data sources for the unique samples of bank debt, public debt, and non-bank private debt. It

also provides summary statistics of the deals and borrowers, as well as market reaction. Section

III examines the determinants of bank loans, public debt, and non-bank private debt reporting in



3 Besides Nini et al. (2007), Brophy, Ouimet and Sialm (lI r 14) show that the degree of protection granted to
investors has a significant effect on subsequent performance, especially in the case of firms that borrow from
institutional investors as opposed to hedge funds.




































To my parents and Gauray










simultaneous disclosure of firm-specific news to Dow Jones and Reuters, and the American

Stock Exchange requires simultaneous disclosure to Dow Jones, Reuters, Associated Press,

United Press International, the Wall Street Journal, the New York Times, Standard & Poor' s, and

Moody's Investor Service (Thompson et al. (1987)). Nevertheless, both practitioners and

academics rely on Dow Jones as the primary source of news existence and timing, most likely

because of its longer tradition and wider dissemination of information. 6 In any case, when I

select a random subsample of loans and search for press articles without limiting the source, I

find that Dow Jones captures all cases in which the loans are reported.

The nature of the publishing process raises several questions. First, can financial press

editors influence returns and trading volume through editing? Mitchell and Mulherin (1994) find

that the number of news stories and market activity are directly related and that this relation is

robust to the size of the headlines and macroeconomic announcements. In addition, they find that

the association between larger size headlines and higher market returns does not have a

significant effect on trading volume.' Second, how much news is there in successive

announcements? Rippington and Taffler (1996) find that only preliminary announcements and

interim statements convey substantial amounts of new information. Thus, in light of the findings

in previous work, one could argue that multiple incomplete announcement returns could be

aggregated in the study of the market reaction to loan news. However, the loans in the sample are

reported in most cases through unique articles, and in the few cases where there is more than one



6 Dow Jones maintains five wire services in addition to publishing the Wall Street Journal and Barron 's. Dow Jones
also maintains the News Retrieval Service, which contains selected articles from the Dow Jones News Service, now
called Factiva, and the Wall Street Journal. In terms of coverage, the Journal appears to include the initial news
release data for approximately 96% of the items listed in the NYSE/ASE Index (Thompson et al. (1987)).

SMitchell and Mulherin (1994) find April to be the month with the smallest number of announcements per day. In
addition, Thompson et al. (1987) also document that firms are less likely to issue news releases on Fridays and in
December.










the borrower. The reason is that bank lending is credited for providing tailored and flexible

lending to firms that present more opaque riskier profiles (Houston and James (1996), Johnson

(1997), and Hadlock and James (2002)). Hence, since practically all public firms borrow from

banks, heterogeneity in the measures of opaqueness and credit risk within the loan sample may

explain why reported loans and their borrowers may be deemed different from non-reported

ones.

As shown in Table 2-3, firms with reported loans are on average smaller (in terms of

assets and sales), and have higher surprise and dispersion of earnings forecasts, higher leverage,

and poorer operating performance during the fiscal year preceding the activation of the loan as

measured by the ratio of earnings before interest, taxes, depreciation and amortization (EBITDA)

to assets. The debt to EBITDA ratio, covenant included in about 50% of loan contracts

(Demiroglu and James (2007)), is also significantly higher during the year preceding the loan for

borrowers with reported loans. Thus, reported borrowers appear to be on average more opaque

and riskier than non-reported borrowers. This is insightful, because those same factors determine

as well the choice of bank debt over other sources of incremental financing, as previous work

finds. More specifically, previous studies document that firms with higher information

asymmetries and lower credit risk are more likely to choose reported bank debt over public debt

(Hadlock and James (2001)). Hence, the higher information content of reported bank loans

compared to non-reported ones can explain the need for reporting.





SLike Gomes and Phillips (2006), I compute the quarterly analyst earnings surprise as the absolute value of the
difference between the median quarterly earnings estimate and the actual quarterly earnings per share, normalized by
the stock price at the fiscal quarter end. Similarly, I compute the analyst earnings dispersion measure as the standard
deviation of outstanding earnings forecasts normalized by the stock price requiring a minimum of two outstanding
earning forecasts. Additionally, since firms may have incentives to disclose more information prior to a public issue
vs. prior to a private issue. Thus, the earnings surprise and dispersion measures use the average of the last four
quarters ending a quarter before the issue date.










Bank debt and non-bank private debt differ in terms of regulatory requirements, maturity,

costs, placement structure, and creditor concentration (Carey et al. (1998)).4 Overall, since bank

and non-bank private debt offer higher flexibility of renegotiation, it should be optimal for the

riskiest borrowers to borrow privately instead of issuing public debt. 5 In this context,

Chemmanur and Fulghieri (1994) predict that firms with high and low credit ratings use public

debt, while firms with intermediate ratings use bank loans. However, Denis and Mihov (2003)

document that while firms with the highest credit quality are more likely to issue public debt, and

firms with medium credit quality establish bank loans, those with the lowest credit quality

establish non-bank private debt agreements. 6 Thus, as Raj an (1992) argues, the cost of bank

monitoring may outweigh the benefits for low-quality firms, and in consequence, low-quality

firms are more likely to issue non-bank private debt than bank debt.

To examine the determinants of the choice between bank debt, public debt and non-bank

private debt in a multivariate setting, I use a series of four multinomial logistic regressions. Table

5-1 reports the regression results, and Table 5-2 the marginal effects. The first multinomial

regression studies the choice between bank borrowing that is reported, public, and non-bank

private debt, while the third and fourth regressions do the same considering all loans and only

non-reported loans, respectively. The second regression studies the choice between reported

issues of bank, public and non-bank private debt.




4 See Bayless and Chaplinsky (1992), Chaplinsky and Hansen (1993), and Jung, Kim, and Stulz (1996) for the
choice between public debt and equity.

5 Sufi (2005) documents that firms with higher lagged earnings to assets are more likely to establish bank loan
agreements as opposed to arm's length debt.

6 More specifically, Denis and Mihov (2003) find that bank and non-bank private borrowers present significantly
lower returns on assets, lower proportions of rated debt and credit quality when credit ratings are available, and
higher growth opportunities in terms of growth in capital expenditures, sales, and number of employees.










Table 2-1. Annual Distribution of Loans by Year of Deal Activation
Year of All Sample Loans Loans DJ & PR Wire Non-
Deal Bank Reported in Reported in Reported reported
Activation Loans Press Before Press On or Loans Loans
N=1375 Activation After N=166 N=905
N=84 Activation
N=220
1996 11.78% 8.33% 8.64% 11.45% 12.93%
1997 15.56% 15.48% 14.21% 15.66% 15.69%
1998 12.51% 20.24% 11.36% 8.43% 12.82%
1999 11.27% 16.67% 10.9% 7.83% 11.49%
2000 11.85% 7.14% 11.36% 11.45% 12.49%
2001 11.05% 9.52% 14.09% 7.23% 11.16%
2002 10.25% 14.29% 10% 11.45% 9.72%
2003 9.38% 4.76% 11.36% 12.05% 8.84%
2004 6.35% 3.57% 7.27% 14.45% 4.86%

The sample consists of 1474 randomly selected syndicated bank loan agreements activated
between January 1996 and December 2004. The number of loans randomly selected each year
was determined so that each year the proportion of loans in my sample equals the corresponding
annual proportion of loans in Deal Scan database. I include loans involving US banks and only
US publicly traded firms at the time of loan activation. I identify reported and unreported loans
searching the Wire and Dow Jones News Retrieval Service for articles and headlines containing
the deal amount and/or the key words "bank," "line" "credit," "loan." I require the financial press
and wire information to be published between three months prior to one month after the
activation of the loan as reported in Deal Scan.










CHAPTER 5
DETERMINANTS OF INCREMENTAL FINANCING OPTIONS

Previous empirical studies of the determinants of incremental financing choices are based

on samples of bank loans that get reported in the press. Denis and Mihov (2003), Hadlock and

James (2002), Houston and James (1996), Johnson (1997), and Cantillo and Wright (2000) find

that smaller, younger, more levered firms are more likely to raise capital via bank debt. 1. In

addition, Hadlock and James (2002) document that higher volatility of returns and lower market

to book and market adjusted cumulative returns during the year prior increase the probability of a

bank agreement over a public debt issue. More recently, Bradley and Roberts (2004) report that

bank borrowers include more covenants in their debt agreements and have greater growth

opportunities and more volatile cash flows than firms that issue public debt.2 These findings are

insightful because they differentiate bank borrowers from bond borrowers using the same firm

characteristics that I have found differentiate borrowers with reported loans from those without.

Furthermore, if as I find, reported loans are not representative of the entire universe of loans, and

non-reported loans are suspected to have a more transactional nature, the determinants of the

choice between bank loans, public debt, and non-bank private could change when non-reported

loan borrowers are also considered.3





i In relation to agreement size, Smith (1986) specifies that firms are expected to use bank lines of credit until an
efficient public issue size is reached. The reason for this size criterion is that flotation costs for public debt issues
have a larger fixed component and more pronounced economies of scale than does bank debt. In addition, non-bank
private loans tend to have lower flotation costs than public issues and custom-designed covenants.
2 Although banks are considered the most efficient liquidity providers in the economy (Kasyap et al. (2002) and
Gatev et al. (2006)), firms are restricted access to bank credit if they do not maintain high enough cash flows (Sufi
(2007)). One explanation for this finding could be the use of marginal vs. book firm characteristics.

3 Regarding the choice between debt and equity, Gomes and Phillips (2005) find that firms with the greatest
asymmetric information problems tend to borrow from private equity, while the most transparent firms seem to issue
public equity, and those that are relatively more opaque prefer bank debt to public debt.









CHAPTER 3
DETERMINANTS OF LOAN, PUBLIC DEBT, AND NON-BANK PRIVATE DEBT
ANNOUNCEMENT S

The substantial heterogeneity in firm and loan characteristics shown in Table 3-1

evidences that reported loans and their borrowers constitute a non-representative subsample of

the entire Deal Scan universe of loans and their public borrowers. Furthermore, the differences

between reported and non-reported loans may arguably involve differences in loan information

content that would explain the noteworthiness of reported loans and help identify the criteria

used by press editors. In addition, the widespread use of bank loans (94% of public firms borrow

from banks) and the fact that about 60% of public firms issue public debt (Sufl (2007)) also

raises the question of whether press editors follow the same firm and deal characteristics related

criteria when deciding which public debt and non-bank private debt issues to publish, as

compared to loan reporting.

Determinants of Bank Loan Announcements

A large theoretical literature in banking focuses on banks as screeners that reduce ex ante

information asymmetries when compared with public "arm's length" debt (Diamond (1991),

Fama (1985), and Ramakrishnan and Thakor (1984)). Furthermore, the banking literature argues

that loans constitute a unique source of financing because, among other reasons, banks have

access to information that is not available to other lenders and market participants while

screening loans and monitoring borrowers (Raj an (1992)). So, if non-reported loans require less

access to private information due to their less severe information asymmetries, one could argue

that the lenders of non-reported loans do not know significantly more than other market

participants about their non-reported borrowers. Consequently, those loans would constitute a

weaker signal of firm quality and be, therefore, less noteworthy.









LIST OF REFERENCES


Andrade, Gregor, and Steven Kaplan, 1998, How costly is financial (not economic distress?
Evidence from highly leveraged transactions that became distressed, Journal ofFinance
53, 1443-1493.

Barber, Brad M., and John D. Lyon, 1996, Detecting abnormal operating performance: The
empirical power and specification of test statistics, Journal ofFinancial Economics 41,
359-399.

Bayless, Mark, and Susan Chaplinsky, 1991, Expectations of security type and the information
content of debt and equity offers, Journal of Financial Intermediation 1, 195-214.

Best, Ronald, and Hang Zhang, 1993, Alternative information sources and the information
content of bank loans, Journal ofFinance 48, 1507-1522.

Billett, Matthew, Mark Flannery and Jon Garfinkel, 2006, Are banks special? Evidence on the
post-announcement performance of bank borrowers, Journal ofFinancial and Quantitative
Analysis 41, 733-751.

Bradley, Michael, and Michael Roberts, The structure and pricing of corporate debt covenants,
SSRN working paper.

Brophy, David, Paige P. Ouimet, and Clemens Sialm, 2004, PIPE Dreams? The Performance of
Companies Issuing Equity Privately, NBER Working Papers 11011, National Bureau of
Economic Research, Inc.

Cantwello, Miguel and Julian Wright, 2000, How do firms choose their lenders? An empirical
investigation, The Review ofFinancial Studies 13, 155-189.

Carey, Mark, Mitch Post, and Steven A. Sharpe, 1998, Does corporate lending by banks and
finance companies differ? Evidence on specialization in private debt contracting, Journal
ofFinance 53, 845-878.

Chaplinsky, Susan, and Robert Hansen, 1993, Partial anticipation, the flow of information and
the economic impact of corporate debt sales, Review ofFinancial Studies 6, 709-732.

Chemmanur, Thomas, and Paolo Fulghieri, 1994, Reputation, renegotiation, and the choice
between bank loans and publicly traded debt, Review ofFinancial Studies, 7, 475-506.

Denis, David, and Vassil Mihov, 2003, The choice among bank debt, non-bank private debt, and
public debt: evidence from new corporate borrowings, Journal ofFinancial Economics 70,
3-28.

Demiroglu, Cem, and Christopher James, The information content of bank loan covenants,
SSRN working paper.









CHAPTER 4
STOCK PRICE REACTION TO LOAN NEWS. MULTIVARIATE ANALYSIS

Among the several explanations for the general pattern of relative stock price effects in

event studies, Smith (1986) cites the unanticipated announcement hypothesis, which states that

the more predictable the issuance and its structure, the smaller the associated stock price

change. Another explanation mentioned in the study of the differences in announcement effects

across security types is the importance of asymmetric information problems. For example, since

debt and preferred stock are more senior claims, their valuation is expected to be less sensitive

than common stock to changes in firm value. In addition, James and Smith (2000) suggest a third

explanation for why bank loan announcements by large and medium-size companies elicit a

positive stock price reaction. According to their results, commitment-based financing is used by

larger companies when they believe themselves to be undervalued or when interest rates are too

high.2

In relation to loan announcements, and given that loan news articles are very brief, it

could be also argued that the articles per se are not informative. It could be argued that loan

articles mainly signal the existence of a noteworthy loan without giving further information

about the circumstances that make the loan noteworthy. Thus, the study of the magnitude of the

market reaction to loan news should consider the factors that make reported loans noteworthy,

i.e., credit risk, information asymmetries, loan characteristics and potential market timing

hypothesis.

The analysis of the stock price reaction to the earliest loan news vs. the reaction to the loan

activation of non-reported deals is presented in Table 4-1. It uses an OLS regression model that

SSee Marsh (1982) for evidence on the use of short-term debt and how to predict public debt issues.

2 For example, and at the time of the Asian crisis in the last quarter of 1998, $10 billion of commercial paper was
retired and $20 billion of net new commercial loans were booked.









smallest biggest size quartile in terms of assets (the regression results are not reported in the

paper). As expected, I find that when the loans are surprising and informative given loan

characteristics, information asymmetries, and borrower credit risk but the borrower is a small

firm, the wire article that is not followed by a press article is usually a press release wire. Thus,

once could argue that loans to smaller borrowers are less likely to draw attention from the press

because those firms are of less interest to the general investor. On the other hand, in the case of

the biggest firms, which receive greater analyst coverage, I find that not all their loans are

systematically reported. The loans obtained by the biggest firms are reported when credit risk

and loan characteristics justify it, given that these borrowers are generally less opaque.

The second probit model reported in Table 3-1 studies whether loan-reporting

determinants also influence the timing of the loan news with respect to loan activation. Overall, I

find no evidence supporting the hypothesis that firm characteristics affect the timing of the loan

news. However, the size of the loan commitment relative to borrower size increases the

likelihood that the loan is reported before loan activation.

Overall, only the loans that seem more informative about the potential of the borrower

are reported. This evidences deal variety in the loan market that had not been documented

before. Furthermore, the heterogeneity in information asymmetries and credit risk that

distinguishes reported from non-reported cases is insightful because it also determines the choice

of reported bank debt over public debt (Hadlock and James (2001)). In consequence, one could

argue that non-reported loans are more transactional, contrary to the belief that bank loans

always involve a more or less close relationship between borrower and lender.

Determinants of Public Debt and Non-Bank Private Debt

The news search for the samples of 355 public debt issues and 92 non-bank private debt

issues indicates that most public debt deals are reported in the press vs. 57% of the non-bank










Table 3-3. Determinants of the Likelihood of Public and Non-Bank Private Debt Reporting

Probit Reporting Non-Bank Private Public Debt
Debt
Marginal Marginal
Effects Effects
Time trend 0.04 0.01 0.16 0.06
(0.61) (5.24)
Tangibles/Assets -0.54 -0.22 -0.36 -0.14
(-0.96)(-.7
Debt/Assets 0.84 0.33 0.25 0.10
(1.13) (0.50)
EBITDA/Assets 1.72 0.69 -2.21 -0.85
(0.86) (-2.05)
Proceeds/Assets 3.12 1.25 0.46 0.18
(2.3 8) (0.69)
Medium-size firm 0.08 0.03 0.18 0.07
(0.28) (0.63)
Firm's past 12-month -0.13 -0.05 0.57 0.22
cumulative return
(-0.48) (2.81)
Market's past 12-month -0.36 -0.14 -0.13 -0.05
cumulative return
(-0.51) (-0.38)
Constant -0.91 -0.30
(-1.49) (-0.86)
Pseudo R2 0.11 0.10

N 92 355

Table 3-3 provides the estimates of a probit model relating firm and debt characteristics to the
likelihood that the press deems a public debt or non-bank private debt noteworthy through a
press articles or a DJ wire. The analysis is based on a sample of 92 non-bank private debt
agreements and 355 public debt agreements activated between January 1996 and December
2004. I identify reported and non-reported agreements searching the Dow Jones sources for
articles containing the key words used in the announcements literature. M~edium-size firm is a
dummy variable that equals one if the firm has less than $1 billion in assets and more than $400
million in assets. Z statistics are reported in parenthesis.










Table 2-3. Firm and Loan Summary Statistics for Reported and Unreported Loans
Loans Reported in Loans Reported in Press DJ & PR Wire Non-reported Loans
Press Before On or After Activation Reported Loans N=905
Activation N=84 N=220 N=166
Mean Median Mean Median Mean Median Mean Median
Assets (millions) 3680.28 504.87* 2355.06* 500.84* 4119.3 480.32* 4688.93 885.13
Sales (millions) 2261.74* 362.13* 1861.3* 499.22* 3066.3 530.59* 4027.94 819.96
Tangible/Assets 0.36 0.30 0.33 0.26 0.30* 0.24* 0.36 0.29
Earnings Surprise 0.85 0.38* 1.04 0.26* 0.53 0.18 0.55 0.14
(%)
Earnings 0.40 0.17*+ 0.25 0.10* 0.30 0.11* 0.28 0.07
Dispersion (%)
Debt/Assets 0.35* 0.36* 0.33* 0.34* 0.29 0.27 0.28 0.27
Debt/EBITDA 3.84* 3.08* 3.31 2.57* 3.25 2.22 3.01 2.07
EBITDA/Assets 0.12* 0.12* 0.12* 0.12* 0.12* 0.13 0.14 0.13
Credit Rating 0.45 0 0.39* 0 0.43 0 0.47 0
Investment Credit 0.23* 0 0.18* 0 0.30* 0 0.38 0
Rating
Firm's 12-month 0.096 0.03 0.27 0.07 0.26 0.1 0.19 0.11
cum. return
Commitment 448.92 216.73* 392.55 175 565.79 192.22 400.33 150
(millions)
Commitment/ 0.69*+ 0.44*+ 0.44* 0.36* 0.42* 0.29* 0.26 0.16
Assets
Maturity (months) 51.65* 53.86* 46* 42.49* 43.60* 37.5* 35.09 31.43
Covenant Intensity 4.42* 5* 4.53* 4* 3.71* 3* 3.08 3
Index
Restructuring 0.30* 0 0.32* 0 0.16 0 0. 12 0
All-in-drawn 234.2*+ 238.6*+ 204.31* 181.88* 181.88* 160* 133.67 100
spread

The loan sample consists of 304 loans reported in the financial press, 166 loans reported in the wire, and 905 non-reported syndicated
bank loans activated between January 1996 and December 2004. The public debt sample consists of 355 issues, and the non-bank
private debt sample of 92 issues, also activated between January 1996 and December 2004. The number of security issues randomly
selected is determined so that each year the proportion of security issues in the sample equals the corresponding annual proportion of









article about a loan (in most cases just two articles) they are either dated the same day with

practically verbatim information about the loan or one article is a summary of the key points in

the other one, or distant enough in time not to affect the two-day event study results.

The two-day-window event studies of bank, public and non-bank private debt

announcements use the earliest of the first news or the issue activation date as event date, a

standard market model estimation period that ends 46 trading days before the event date, and the

standard methodology by Mikkelson and Partch (1986) to measure the stock price reaction. In

addition, since there is "contamination" of loan, public debt and non-bank private debt news, I

control for it and find similar results when confounding news articles are included. Thus, I report

in this paper only the non-confounding results given that previous studies exclude confounding

press articles in their analysis of market reaction surrounding announcements. I find that for

78.9% (240 out of 304) reported loans, and 63.1% (53 out of 84) of loans reported before the

activation date, there is no bundling of loan information with other non-loan related news. I

classify as bundling or contamination of the earliest loan related article any information

within the same article that can affect the stock price. The nature of the bundling of information

can range, for example, from an earnings announcement that does not appear to be a repetition of

previous news, to the completion of an acquisition.

The cumulative abnormal returns (CARs), average standardized prediction errors (SPE)

and Z values associated to loan, public debt and non-bank private debt reporting and activation

are reported in Table 2-5 and present four interesting findings. First, although the market can

gain access to information about the existence and deal activation date of the non-reported loans,

there is no significant market reaction surrounding the deal activation for the non-reported









borrowers remain weaker than those of non-reported borrowers for two years following loan

activation, I observe no significant difference at the 5% level in EBITDA to assets during the

third fiscal year following loan activation. Furthermore, it is important to note that the

improvement in the operating performance of reported borrowers would be more pronounced if

the firms whose loans are reported mainly on the basis of loan size or maturity were excluded

from the study sample.









private debt deals. In addition, the news search shows that almost 20% of non-bank private deals

are reported by the borrower, while in the case of public debt reporting by the firm is rare and

usually related to debt repayment. Thus, in the case of non-bank private deals, either the

borrowers or the lenders seem to want investors to know about the deals. This is interesting

given that previous studies document a negative market reaction to non-bank private debt (Eckbo

(2007)).

The borrower and lender incentives to report non-bank private debt despite the adverse

reaction of the market could be explained through the findings of Denis and Mihov (2003) and

Carey el al. (1998) that associate non-bank private lenders to the financing of the riskiest

borrowers. It seems that the good news of getting financing should increase the likelihood that

non-bank private debt is announced. However, only banks are considered to be special in terms

of access and generation of private information and efficient monitoring (Fama (1985), Raj an

(1992) and Denis and Mihov (2003)). On the other hand, in the case of reporting of public debt

issues, if those borrowers are significantly less risky and opaque than bank borrowers, one may

argue that the likelihood of reporting would decrease at least among the best performers. Thus,

how do the determinants of bank loan reporting differ from those of public debt and non-bank

private debt? The analysis of the likelihood of public debt and non-bank private debt reporting

through a press article or a Dow Jones wire article is shown in Table 3-3.

The probit analysis for non-bank private debt shows that, besides information

asymmetries, credit risk does not seem to play a significant role in the likelihood of reporting of

non-bank private debt issues (although it plays it in the choice of non-bank private debt

borrowing, as Dennis and Mihov find (2003)). The reason could be that although this type of

financing serves the niche of highest risk borrowers that cannot obtain financing from loans, the










bank loan reporting determinants includes proxies for information asymmetries, credit risk,

market timing, loan restructuring following violations, loan size, and loan maturity. In addition,

one could argue that the largest firms are less opaque and the smallest firms of less interest to the

general investor, and that, consequently, their loans could be less likely to be reported. Thus,

given that the likelihood of reporting is suspected to have a non-linear dependency with respect

to borrower size, the analysis also considers a medium-firm-size dummy variable. This dummy

equals one if the firm has assets of less than $1 billion and more than $400 million in order to

include non-reported borrowers with below-median firm size and reported borrowers with

above-median firm size.

The analysis of loan-reporting determinants is based on a series of probit models. Table

3-1 reports the regression results and marginal effects for the two main probit models. The first

model studies the likelihood that a loan is deemed noteworthy through a press article or a Dow

Jones wire article that is not made press news. The dependent variable equals one if the bank

loan is reported in the financial press or on the Dow Jones wire but not in the press, and zero

otherwise. The second regression analyzes how the factors that determine the likelihood of

reporting can also affect the timing of the reporting, i.e., the likelihood that a loan is reported

before its activation date.

As shown in Table 3-1, higher measures of information asymmetries and credit risk

increase the likelihood of loan reporting. These determinants are measured by tangibles relative

to assets and EBITDA to assets, although other proxies used in the literature, such as volatility,

and surprise and dispersion in earnings forecasts, lead to analogous results. In addition, loan

restructurings, loans with longer maturities, and bigger loans relative to assets are also more

likely to be reported in the press. More specifically, I find that, evaluated at the sample means of









CHAPTER 6
LONG-TERM PERFORMANCE OF BANK BORROWERS

As commented before, reported loans involve a closer relationship between lenders and

borrowers, while non-reported loans are more transaction oriented and, to a certain extent, more

equivalent to public debt. In consequence, one could argue that reported borrowers may improve

their operating performance with respect to the performance of non-reported borrowers. In a

related study, and since reported loans are more likely to be restructured loans following

covenant violations, Nini et al. (2008) show that, following covenant violations, there is an

effective reduction of capital expenditures that leads to higher performance and valuation.

In the study of long-term operating performance, I use two measures: EBITDA to assets,

and debt to EBITDA. I focus on these two ratios because they are closely linked to the

borrower' s ability to service both current and future bank borrowings. While stock returns and

net income are also important measures of performance used in previous studies, they are more

removed from the banker's principal focus. In addition, the debt to EBITDA ratio is present in

about half the loans that include financial covenants. Thus, firms have an incentive to improve

this ratio.

To limit the effect of outliers, I examine the medians of operating performance measures

for a period of seven years that covers four fiscal years preceding loan activation and the three

subsequent ones. The summary statistics for a fixed sample of firms are presented in Table 6-1 .

Year 0 refers to the fiscal year prior to the activation of the loan.

Overall, I observe in the balanced panel that reported borrowers present poorer operating

performance with respect to non-reported loans during the last three fiscal years prior to the

activation of the loan. However, and although the operating performance measures of reported

II find similar results when I exclude unannounced loans for which there is 8-K filing.










may be excluding a significant proportion of smaller firms that borrow from one lender. Those

firms, as Houston and James (1996) find, are more likely to be subj ect to information

monopolies, and have greater incentives to time the market due to more severe asymmetric

information problems. This could explain the lack for evidence in the analysis for the market

timing hypothesis.










renegotiations in trading by institutional investors that are members of loan syndicates. The

profitability of this trading supports the argument that the private information revealed and

produced during the renegotiations is not only good, but also highly valuable, which would give

the borrower a particularly strong incentive to report it.

The disclosure literature documents how managers time voluntary disclosures in a

manner that maximizes insider trading profits while minimizing potential litigation costs

associated with disclosure (Cheng and Ko (2006). Additionally, the literature on the

announcement effects of financing events which spans for over two decades reports a

significant positive stock price reaction to announcements of bank loans, a negative reaction to

non-bank private debt placements, and an insignificant reaction to straight public debt issues

(James (1987), and Eckbo et al. (2007) for a recent review of this literature). Consistent with this

evidence, I find a significant positive market reaction to reported loans, but only when the news

precedes loan activation, and mainly driven by loan size and prior borrower stock performance

with respect to the market. This suggests that reported loans are considered more informative

when the news precedes the activation of the loan because it is arguably more surprising for

reported borrowers to obtain financing and timing intensifies the surprise factor.

Denis and Mihov (2003) find that credit risk is a significant determinant of the choice

between bank debt, public debt and non-bank private debt. However, their analysis considers

only the subset of loans that are reported, like other previous studies on incremental financing

decisions (Hadlock and James (2002)). In other words, the conclusions of previous work on

incremental financing derive from comparing the subset of reported bank loans to full samples of

other transactions such as public debt and non-bank private debt. Therefore, if reported loans are

a non-representative subset of the loans in Deal Scan, the analysis of the incremental financing










DealScan offers affordable real-time web access to agreed-upon but not-yet-active loans2. Hence,

the loans reported in Dow Jones that the financial press views as noteworthy are a subset of a

pool of disclosed loans.

This distinction between "reported" and "disclosed" loans is subtle but nevertheless

important. If there is disclosure about the deals but, as I find, no significant market reaction

surrounding the activation of non-reported bank loans, and one assumes reported loans to be

significant, it can be argued that the market views reported loans differently than non-reported

loans.

SEC regulation requires firms to disclose any "material" event that can affect the stock

price and shareholders should know about. Furthermore, the literature on discretionary disclosure

describes how firms are more forthcoming when the news is good (Miller (2002)). Thus, given

that banks are considered superior screeners (Diamond (1991) and Fama (1985)) and effective

monitors (Nini et al. (2008)), and bank financing a credible signal of firm quality (Schenone

(2004), and James and Wier (1990)), one would expect firms to try to disclose all bank loans.

Accordingly, I find that the source of loan news is usually the borrower. On the other hand, the

source of news for non-bank private debt issues is usually "market participants."

Why would firms not report all of their bank loans? Bagnoli and Watts (2007) find that

firms are more likely to disclose higher values of private information when financial reports do

not contain sufficient good news and performance is significantly different than expected. In a

related study, Verrecchia and Weber (2006) document that firms are less likely to withhold

information in material contract filings when they issue long-term debt. Moreover, prior work

shows that riskier more opaque firms are more likely to choose reported bank debt over public


2 The subscription cost to the DealScan database and real-time website is $2,000 per month for corporations with
one user ($330/month is more than one user) and $7,304 per year for academic institutions.









ACKNOWLEDGMENTS

I would like to sincerely thank Christopher James, my dissertation chair, as well as Joel

Houston, Mahendraraj ah Nimalendran, Andy Naranj o, and Ronal Ward for all their guidance,

encouragement, and advice in making sure I successfully completed this work.









January 1996 to December 2004 period. Overall, and as expected for syndicated loan borrowers,

the firms in this paper are bigger. In terms of firm assets, Denis and Mihov (2003) study bank,

public debt and non-bank private debt borrowers that are significantly smaller than the borrowers

in my sample (firm asset medians of $162.24, $2,43 5, and $246.2 million respectively). James

(1987) also uses smaller public private debt borrowers (firm assets medians of $3,322.27), as

well as longer maturity bank loans (median of 72 months). In terms of leverage, Hadlock and

James (2002) study the choice between bank borrowing and public debt borrowing with higher

debt to assets ratios (median ratios of 0.52 and 0.44 respectively). Consistently with the higher

leverage, the bank and public debt firms in Hadlock and James (2002) study borrow less (median

commitment to assets ratios of 0.26 and 0.047 respectively). However, I observe no significant

difference in terms of firm assets when comparing the size of the bank borrowers in previous

studies with the subsample of reported loans in my sample. This is insightful, because previous

studies consider only reported loans. Hence, the similarities between my subsample of reported

borrowers and the samples of reported borrowers used previously in the literature allow a

comparison of results.

Summary Statistics of Market Reaction to Loan News and Loan Activation

How do loans come to be published through Dow Jones? The issuance of a press release

by the borrower constitutes one more source of information in the first step of the chain that

conveys the information to the market through the press. Thompson et al. (1987) describe the

process. Once reporters or firms transmit the story to the Dow Jones News Service, the editors

summarize them, weight their importance, and determine whether to make them press news. This

process takes place within hours or, in most cases, not more than one day.

One should note that the dissemination of news information through wire services does

not target only Dow Jones. In fact, the New York Stock Exchange regulations require










Table 3-1. Determinants of the Likelihood and Timing of Bank Loan Reporting in the Press

Likelihood of loan Likelihood of reporting
reporting before loan activation
Marginal Marginal
Effects Effects
Time trend 0.06 0.02 -0.01 -0.004


(3.12)
-0.39
(-2.18)
0.30
(1.40)
-1.2
(-2.54)
0.56
(5.32)
0.51
(3.O1)
0.01
(4.52)
0.32
(3.84)
-0.03

(-0.56)
0.01

(0.08)
-1.32
(-8.18)
0.11


(-0.31)
-0.13 0.27
(0.78)
0.10 0.09
(0.22)
-0.41 -0.94
(-1.09)
0.21 -0.03
(-0.17)
0.17 0.41
(2.47)
0.003 0.003
(1.06)
0.11 -0.16
(-0.90)
-0.01 -0.09


Tangibles/Assets

Debt/Assets

EBITDA/Assets

Restructure Loan

Commitment Amount/Assets

Maturity

Medium-size firm

Firm's past 12-month
cumulative return


Market's past 12-month
cumulative return


0.09

0.03

-0.3

-0.01

0.13

0.001

-0.05

-0.03



-0.19


(-1.22)
-0.61

(-1.63)
-0.71
(-2.14)
0.05

294


0.01


Constant

2
Pseudo R


1176


Table 3-1 provides the estimates of two probit models that relate firm and loan characteristics to
the likelihood that the press deems a loan noteworthy through a press article or a DJ wire, and
the likelihood that when there is loan news the press article precedes the loan activation date.
Both the results and marginal effects of the probit regressions are reported. The analysis is based
on a sample of 304 press-reported, 166 PR or DJ wire reported loans and 905 non-reported loans
activated between January 1996 and December 2004. I identify reported and non-reported loans
searching the Dow Jones sources for articles containing the deal amount and/or the key words
used in the literature on loan announcements. Conanitnzent Amount is the size of the loan
commitment. Maturity is the maturity in months of the loan agreement as reported in Dealscan.
M~edium-size firm is a dummy variable that equals one if the firm has less than $1 billion in assets
and more than $400 million in assets. Z statistics are reported in parenthesis.











TABLE OF CONTENTS


page

ACKNOWLEDGMENTS .............. ...............4.....


LIST OF TABLES .........__.. ..... .__ ...............6....

AB S TRAC T ..... ._ ................. ............_........7


CHAPTER

1 INTRODUCTION ................. ...............8.......... ......


2 SAMPLE SELECTION, DATA AND SUMMARY STATISTICS ................ ................. 14


Sample Selection and Data ................. ...............14................
Sum m ary Statistics ................ .. .... ............. ... ...............1
Summary Statistics of Market Reaction to Loan News and Loan Activation ................... .....23

3 DETERMINANTS OF LOAN, PUBLIC DEBT, AND NON-BANK PRIVATE DEBT
ANNOUNCE MENT S .............. ...............3 4....


Determinants of Bank Loan Announcements ............................. ............... 34.....
Determinants of Public Debt and Non-B ank Private Debt ........................... ...............3 8


4 STOCK PRICE REACTION TO LOAN NEWS. MULTIVARIATE ANALYSIS .............45

5 DETERMINANTS OF INCREMENTAL FINANCING OPTIONS............... ................4


6 LONG-TERM PERFORMANCE OF BANK BORROWERS ................. ......................55

7 SUMMARY AND CONCLUSIONS ................ ...............58................


LIST OF REFERENCES ................. ...............60........... ....

BIOGRAPHICAL SKETCH ............. ..............63.....










and Private Placement Reporter. Thus, it may have become less costly over time to report loans

due to advancements in information technology in the case of Dow Jones, but not in the case of

other smaller news sources that do not benefit from economies of scale.









the other explanatory variables, the likelihood of a loan being reported is 13 percentage points

lower with higher tangibles relative to assets, 41 percentage points lower with higher measures

of EBITDA to assets, 21 percentage points higher in the case of loan restructurings, 17

percentage points higher when the commitment amount relative to firm size is bigger, and 11

percentage points higher when the borrower is a medium-size firm. It is important to note that

these are also the determinants that reporters of loan news in Dow Jones cite when asked on the

phone about the criteria followed to determine whether a particular loan is newsworthy. Thus,

both the analysis and anecdotal evidence suggest that reported loans are more surprising and also

informative about firm potential.

Beyond the average marginal effects reported in Table 3-2 reports how the lowest and

highest quartile measures of loan-reporting determinants in the 1,375 loan sample alter the

likelihood of reporting evaluated at the sample means of the other explanatory variables. More

specifically, the likelihood of reporting decreases by 10 percentage points for the firms within

the highest quartile of tangibles to assets, increases by 8 percentage points (decreases by 7

percentage points) for the firms with EBITDA to assets within the lowest (highest) quartiles of

the sample, and increases by 20 percentage points (decreases by 17 percentage points) for the

firms with loan commitment to assets within the highest (lowest) quartiles of the sample.

The analysis of the loan reporting determinants also shows that the likelihood of loan

reporting is increases throughout the study period. This could be arguably be due to lower

reporting costs thanks to developments in information technology, and suggests that the

information content of the average reported loan decreases during the 1996 through 2004 period.

In order to confirm the non-linear dependency of loan reporting with respect to firm size,

I apply the reporting likelihood probit model to the firms in the sample that fall within the










up over the 6-month LIBOR that includes recurring fees associated with the credit facility.5 The

spread is used as a measure of per dollar cost of borrowing in a number of previous empirical

studies on loan pricing (for example, Bradley and Roberts (2004), Guner (2006) and Moerman

(2005)). It is interesting to note here that, besides the loan size relative to the firm size, all-in-

drawn spreads measures also differentiate, among reported loans, those that are reported before

loan activation with respect to those reported on or after activation, as shown in Table 2-3. More

specifically, I find that loans reported in the press before activation are larger (relative to firm

size) and present higher all-in-drawn spreads than those reported on or after activation, which

would arguably make their announcements even more noteworthy.

Overall, the firm and loan summary statistics in Table 2-3 show that reported loans and

their borrowers are not representative of the universe of loans and borrowers. Furthermore, the

firm characteristics that differentiate reported loans from non-reported loans also differentiate

bank borrowing from public borrowing in previous studies (Hadlock and James (2002)). In

consequence, it could be argued that non-reported bank borrowing constitutes an intermediate

debt source between reported bank borrowing and public debt borrowing.

Table 2-4 presents firm and issue characteristics for non-bank private debt and public debt

deals. Like bank loans, both in the case of public debt and non-bank private debt issues, bigger

deals (relative to firm assets or not) are more likely to be reported. However, there are important

differences in terms of frequency of reporting and potential determinants of press reporting

between bank loans, public debt and non-bank private debt issues. First, it is important to note

that about 40% of the non-bank private debt issues are reported and that 60% of the public debt

issues are reported. Thus, bank loans are much less likely to get reported than public debt and


5 LPC computes the spread for non-LIBOR based loans by converting index used to price the loan into a LIBOR
equivalent using the historical relationship between the index and the LIBOR.









loans in Deal Scan database, and public debt and non-bank private debt deals in Securities Data Corporation (SDC). In addition, the
randomly selected issues maintain the overall proportions of total number of security issues for each Einancing type with respect to the
other two, as reported in Deal Scan and (SDC). I identify reported and non-reported deals searching Dow Jones sources for news and
wire articles that contain the deal amount and/or the key words used in the literature on loan, public debt and non-bank private debt
announcements. Tangible assets are plant, property and equipment divided by assets. Earnings Surprise is defined as the absolute
value of the difference between the median quarterly earnings estimate and the actual earnings per share, normalized by the stock
price. Earnings Dispersion is defined as the standard deviation of earning forecasts normalized by the stock price. Credit Rating is an
indicator variable equal to one if the firm has a debt rating, zero otherwise. The Investment Grade Rating is an indicator variable equal
to one if the firm has an existing debt rating of BBB or higher, zero otherwise. Conanitnzent is the size of the loan commitment.
Maturity is the maturity of the loan agreement as reported in Dealscan. The covenant intensity index equals the sum of six covenant
indicators (collateral, dividend restriction, more than two Einancial covenants, asset sales sweep, equity issuance sweep, and debt
issuance sweep) when none of the six indicators is missing. All-in-drawn spread is calculated and reported by Dealscan as the total
borrowing cost of the drawn portion of a loan over and above LIBOR. The maturity and Deal Scan all-in-drawn spread are calculated
as the weighted average of the maturity and spread of each facility in the package loan.
* Significantly different from non-reported loan issue sample at the 0.1 level.
+ Significantly different from sample of reported loans on or after deal activation date at the 0. 1 level.






































O 2008 Laura Gonzalez Alana









Table 5-1. Multinomial Logistic Regressions on Incremental Financing Decisions

(1.a) (1.b) (2.a) (2.b) (3.a) (3.b)
Time trend -0.14 -0.07 -0.06 -0.05 -0.08 -0.02
(-2.92) (-1.35) (-1.08) (-0.67) (-2.88) (-0.41)
Tangibles/Assets 1.02 1.21 0.92 1.2 0.28 0.93
(2.17) (2.0) (1.79) (1.42) (0.85) (1.75)
Debt/Assets 1.2 -1.85 1.02 -1.66 1.52 -0.54
(1.84) (-1.87) (1.58) (-.)(4.25) (-0.82)
EBITDA/Assets 7.65 3.65 6.94 2.98 4.81 1.44
(4.87) (1.92) (3.9) (1.22) (5.28) (1.03)
Deal Size/Assets -4.81 -4.74 -4.76 -2.7 -6.03 -4.53
(-2.04) (-2.58) (15)(-1.48) (-2.18) (-2.36)
Medium-size firm -1.81 0.94 -1.75 0.7 -1.17 1.39
(-4.99) (2.51) (-3.97) (1.52) (-4.31) (4.47)
Firm's past 12-month 0.07 0.09 0.11 0.02 0.07 0.05
cumulative return
(0.74) (0.78) (1.13) (0.13) (0.74) (0.45)
Market' s past 12-month -0.46 -0.64 -0.51 -1.12 -0.55 -0.46
cumulative return
(-1.05) (-1.11) (-1.07) (-1.35) (-1.88) (-0.93)
Constant 0.66 -0.42 -0.02 -1.37 -0.91 -2.72
(1.37) (-0.82) (-0.03) (-2.06) (-2.27) (-6.47)
Pseudo R 0.30 0.30 0.19
14 755 557 1822

Table 5-1 provides the estimates of a series of four multinomial logistic regressions relating the
choice of bank debt over public debt and non-bank private debt, as well as the likelihood and
timing of loan reporting in the context of incremental financing decisions. The analysis is based
on a sample of 304 press-reported, and 1071 non-reported syndicated bank loans, as well as 355
public debt and 92 non-bank private debt agreements, all activated between January 1996 and
December 2004. The public debt and non-bank private debt random samples keep the proportion
of total bank loans in Deal Scan to public debt and non-bank private debt issues in SDC, as well
as the yearly proportions in the original SDC samples. I identify reported and non-reported loan,
public and non-bank issues searching the Dow Jones sources for articles containing the deal
amount and/or the key words used in the announcements literature. Deal Size refers to the
commitment amount in the case of bank loans and proceeds in the case of public debt and non-
bank private debt. Z statistics are reported in parenthesis.

(1.a) Public Debt vs. Reported Bank Loans
(1.b) Non-Bank Private Debt vs. Reported Bank Loans
(2.a) Reported Public Debt vs. Reported Bank Loans
(2.b) Reported Non-Bank Private Debt vs. Reported Bank Loans
(3.a) Public Debt vs. Bank Loans
(3.b) Non-Bank Private Debt vs. Bank Loans















Mean Median Mean Median Mean Median Mean Median
Assets (millions) 2303.2 481.16* 15983.7 867.53 22389.9 6852.44 19453.5 7609.65

Sales (millions) 2393.96 585.95 6877.85 673.2 18460.8 6120.5 17626.3 8358.2
Tangible/Assets 0.4 0.33 0.42 0.39 0.4 0.38 0.43 0.41
Debt/Assets 0.29 0.27 0.28 0.28 0.35 0.33 0.34 0.32
Debt/EBITDA 2.73 1.62 2.29 2.11 2.96 2.29 2.48 2.05
EBITDA/Assets 0.14 0.13 0.15 0.13 0.15* 0.14 0.17 0.16
Credit Rating 0.36* 0 0.57 0 0.97* 0 0.93 0
Investment Credit Rating 0.14* 0 0.43 0 0.83* 0 0.75 0
Firm's 12-month cum. return 0.16 0.12 0.24 0.17 0.21* 0.14 0.12 0.09
Commitment (millions) 55.86* 48.25* 37.94 25 383.92* 250 110.62 36
Commitment/Assets 0.23* 0.1* 0.07 0.03 0.10 0.04 0.05 0.004

The loan sample consists of 304 loans reported in the Einancial press, 166 loans reported in the wire, and 905 non-reported syndicated
bank loans activated between January 1996 and December 2004. The public debt sample consists of 355 issues, and the non-bank
private debt sample of 92 issues, also activated between January 1996 and December 2004. The number of security issues randomly
selected is determined so that each year the proportion of security issues in the sample equals the corresponding annual proportion of
loans in Deal Scan database, and public debt and non-bank private debt deals in Securities Data Corporation (SDC). In addition, the
randomly selected issues maintain the overall proportions of total number of security issues for each financing type with respect to the
other two, as reported in Deal Scan and (SDC). I identify reported and non-reported deals searching Dow Jones sources for news and
wire articles that contain the deal amount and/or the key words used in the literature on loan, public debt and non-bank private debt
announcements. Tangible assets are plant, property and equipment divided by assets. Credit Rating is an indicator variable equal to
one if the firm has a debt rating, zero otherwise. The lnvestnzent Grade Rating is an indicator variable equal to one if the firm has an
existing debt rating of BBB or higher, zero otherwise. Conanitnzent is the size of the loan commitment.
* Significantly different from non-reported issue sample at the 0.1 level.


Table 2-4. Firm and Deal Summary


Statistics for Reported and Unreported Public Debt and Non-Bank Private Debt


Reported Non-Bank
Private Debt
N=42


Non-reported Non-
Bank Private Debt
N=50


Reported Public Debt
N=207


Non-reported Public
Debt
N=148









lenders do not necessarily base the agreements on private information about firm potential, but

on control rights.

In relation to measures of risk, Table 3-3 also shows that although the size of the

proceeds relative for firm size does not affect the likelihood of reporting for public debt

borrowers, non-bank private debt issues are 125 percentage points more likely to be reported as

proceeds relative to firm size increase when evaluated at the sample means of the other

explanatory variables. The reason could be that public debt borrowers are not expected to default

on the agreements, while non-bank private borrowers could. Thus, bigger loans granted to riskier

borrowers may be considered good news because the lenders have positive private information

about the borrower, while bigger lending to the riskiest firms is negative news, since it may just

imply additional risk and a higher probability of default.

In addition, the likelihood of public debt reporting is found to be 22 percentage points

higher in the case of firms with higher cumulative returns when evaluated at the sample means of

the other explanatory variables. This is consistent with market sources echoing the new public

debt issues of the "market darlings," which would not require revelation of any particular insight

regarding firm potential, and would lead to an average null market reaction to the news.

Moreover, and as expected, the likelihood of public debt reporting is 85 percentage points lower

in the case of firms with better operating performance, and therefore, less risky, given that, as

mentioned before, public debt borrowers are less opaque.

Finally, the probit analysis shows that public debt issues, like loans, are more likely to be

reported over time, but that there is no significant time trend in the reporting of non-bank private

debt. The reason could be that while public debt issues are reported in Dow Jones, non-bank

private issues are announced mainly in specialized publications such as Private Placement Letter









Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

DOGS THAT BARK: WHY ARE BANK LOAN ANNOUNCEMENTS NEWSWORTHY?

By

Laura Gonzalez Alana

August 2008

Chair: Christopher James
Major: Business Administration

Virtually all publicly traded firms borrow from banks. However, despite their widespread

use, the reporting of bank loan agreements in the financial press is associated with a positive

share price reaction. In order to address this phenomenon, I empirically examine the frequency

and determinants of bank loan reporting, and find that the credit risk and asymmetric information

proxies behind incremental financing decisions determine loan reporting. In addition, loan

restructurings following covenant violations, and larger loans -relative to firm size- are also more

likely to be reported in the press. Moreover, and although all loans are disclosed, I find that the

market reacts to reported loans only when the news precedes loan activation. Overall, the

evidence suggests that reported loans are more noteworthy because they are more informative

about the potential of the borrower than non-reported loans, and that the information content of

the average reported bank loan decreases during the 1996 through 2004 period.









The results in Table 5-1 and Table 5-2 show that, evaluated at the sample means of the

other explanatory variables, reported loans are 26 percentage points less likely to be chosen over

public and non-bank private debt with higher measures of tangibles to assets, and 23 percentage

points more likely to be chosen over public and non-bank private debt if the borrower is a

medium-size firm. In terms of credit risk, the choice of bank borrowing over public and non-

bank private debt borrowing decreases with higher measures of EBITDA to assets independently

of whether the loans are reported or not. More specifically, I find that, evaluated at the sample

means of the other explanatory variables, reported loans are 170 percentage points less likely to

be chosen over public and non-bank private debt with increases in EBITDA to assets. In

addition, I find that the likelihood that more leveraged firms relative to assets choose bank debt

is 10 percentage points lower for bank loans.

In terms of the effect of proceeds relative to assets on incremental financing choices, and

evaluated at the sample means of the other explanatory variables, firms with higher financial

needs relative to assets are 120 percentage points more likely to choose bank debt over public

and non-bank private debt if only reported loans are considered.

Overall, the study of incremental financing decisions does not vary significantly when

considering only reported loans instead of both reported and non-reported loans. In other words,

whether reported deals or all deals are considered, the results are similar and consistent with

those of previous studies by Hadlock and James (2002), and Denis and Mihov (2003). Riskier

more opaque borrowers are more likely to borrow from banks instead of public debt, and non-

bank private debt has a niche among the riskier borrowers. This suggests that, even if they are

less noteworthy and more transactional, non-reported loans are still somehow special and

different from public debt. Moreover, since Deal Scan covers only syndicated loans, this study










addition, I find that for 78.9% (240 out of 304) of reported loans there is no bundling of loan

information with other non-loan related news.

Summary Statistics

Table 2-1 provides the annual percentage of loans reported in the financial press, those

listed on the wire services but not reported in the press, or those not reported at all. More

specifically, I find that 22% (304 out of 1,375) of the sample loans are reported in the financial

press, while an additional 12% (166 out of 1,375) of loans are reported in the wire but are not

reported subsequently in the press. In addition, about 8% (not reported in the table) of the loans

that are not identified in wire or press articles are made public through 8-K filings.

The Form 8-K filings follow the SEC regulation that requires firms to report any

"material" event that may affect the stock price, or definitive agreement "not made in the

ordinary course of the registrant' s business."2 As Nini et al. (2008) note, as private agreements,

loans are not legal securities and, thus, are not subj ect to direct SEC regulation. However, the

SEC precedent has established a requirement that public companies include copies of all

"material" contracts, including bank loan agreements, with relevant SEC disclosures. These

contracts typically appear as exhibits at the end of a 10-K or 10-Q report, or as an attachment to

an 8-K filing.

Table 2-2 presents some initial evidence of how press-reported loans differ from non-

reported loans in terms of primary loan purpose, and introduces the hypothesis that the timing of



2 An agreement is deemed as material definitive when it "provides for obligations that are material to and
enforceable against the registrant or rights that are material to the registrant and enforceable by the registrant against
one or more parties to the agreement, in each case whether or not subject to conditions." The Form 8-K requires the
agreement date, the identity of the parties and a brief description of any material relationship between the registrant
or its affiliates and any of the parties, other than in respect of the material definitive agreement. Moreover, the
Additional Form 8-K, effective August 2004, expands the number of events that are reportable on Form 8-K and
shortens the filing deadline for most items to four business days. These amendments are said to further the goals of
Section 409 of the Sarbanes-Oxley Act.










Table 4-2. Determinants of the Stock Price Reaction Surrounding Public Debt and Non-Bank
Private Debt Reports and Activation


Non-Bank Private Debt
Non-Confounding News
vs. Non-Reported Issues
-0.32
(-0.21)
-0.01
(-0.05)
7.95
(2.44)
-1.61
(-0.3)
-14.06
(-1.54)
-4.1
(-1.56)
-2.18
(-1.36)
-2.73


Public Debt
Non-Confounding News vs.
Non-Reported Issues
2.66
(1.58)
-0.53
(-1.58)
-2.43
(-0.7)
-2.1
(-0.41)
2.73
(0.26)
1.23
(1.69)
1.7
(0.61)
0.45


Press-reported

Time Trend

Tangible/Assets

Debt/Assets

EBITDA/Assets

Commitment Amount /Assets

Medium-size firm

Firm's past 12-month cumulative
return

Market' s past 12-month
cumulative return

Constant


(-1.46)
-3.7

(-1.03)
1.9
(0.52)
81


(0.23)
1.31

(0.31)
2.84
(0.69)
343


Table 4-2 provides the estimates of two OLS regressions relating the magnitude of the market
reaction to public debt and non-bank private debt news to reporting and firm and loan
characteristics The market reaction is measured as the Standardized Prediction Errors for a
two day event window centered on the earliest of the announcement date or activation date.
The analysis is based on a sample of 199 non-confounding public debt news, 146 non-reported
public debt issues, 38 non-confounding non-bank private debt news, and 43 non-reported non-
bank private debt issues activated between January 1996 and December 2004. I identify
reported and non-reported agreements searching the Dow Jones sources for articles containing
the key words used in the announcements literature. Conanitnzent Amount is the size of the
loan commitment. Maturity is the maturity in months of the loan agreement as reported in
Dealscan. M~edium-size firm is a dummy variable that equals one if the firm has less than $1
billion in assets and more than $400 million in assets. t statistics are reported in parenthesis.









DOGS THAT BARK: WHY ARE BANK LOAN ANNOUNCEMENTS NEWSWORTHY?


By


LAURA GONZALEZ ALANA














A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2008










loans More specifically, and as discussed in the introduction to this paper, Deal Scan uses as

sources of loan information public SEC Eilings and market sources, and maintains a real-time

website that reports the existence of not-yet active loans. Thus, it can be argued that the sample

loans are disclosed.

Second, I Eind that the significant market reaction to reported loans is driven by the

reaction to news articles that precede the activation of the loans. This is consistent with the

differences shown in Table 2-5 between loans reported before activation and loans reported on or

after activation. As mentioned before, loans reported before activation show a higher Deal Scan

spread and commitment size relative to assets that would arguably make them more noteworthy

among all the loans that are reported. In consequence, the reporting in the press of those loans

would more likely trigger a larger market reaction, especially if the timing of the article that

precedes loan activation cannot be foreseen and increases the surprise by which the news takes

the market. Sections III and IV discuss in more detail the determinants of loan reporting and

market reaction in relation to the timing of the news.

Third, the market reaction to public debt announcements is not significant when all

reported issues are considered. However, when press articles are classified by their timing with

respect to deal activation, the announcements following activation are associated with a

significant positive market reaction. This is interesting, because I Eind no difference between

public debt deals reported before activation and after activation in terms of borrower opaqueness

and credit risk, or in terms of deal characteristics.





8 Previous work by Thompson et al. (1987) also observes no significant market reaction to announcements in the
wire services that are not followed by announcements in the financial press, but their study does not distinguish bank
loan reporting from all other news types.









Previous empirical studies document that more opaque firms are more likely to use bank

debt (Houston and James (1996), Johnson (1997), and Sufi (2007))' when the returns of the

borrower decrease with respect to the market (Hadlock and James (2002)), and that relatively

riskier firms also choose bank debt over public debt (Denis and Mihov (2003)). Moreover, and in

relation to disclosure decisions, Bagnoli and Watts (2007) find that firms are more likely to

disclose private information when the operating performance in Einancial reports falls below

expectations and/or does not contain sufficient good news. Thus, besides measures of

asymmetric information, market timing and leverage, EBITDA to assets ratio could also

constitute a significant determinant of reporting likelihood.

Roberts and Sufi (2007) report that, although over 90% of long term loan contracts are

renegotiated prior to their stated maturity, only 16% of the renegotiations are due to default

events such as covenant violations. In addition, a related study by Nini et al. (2008) shows that

following loan covenant violations and subsequent waivers there is an efficient reduction in firm

investment and subsequent increases in market valuation and operation performance. Thus, one

could argue that loan contracts following covenant violations could also be viewed as a

significant signal of firm potential and be, therefore, more noteworthy. Finally, and in relation to

loan characteristics that could make reporting more likely, Verrecchia and Weber (2006) find

that firms are less likely to withhold information in material contract filings when they issue

long-term debt. This suggests that, besides larger loans, loans with longer maturity could also be

more newsworthy.

Therefore, since summary statistics show heterogeneity in all the above-mentioned

borrower and loan characteristics suspected to determine reporting likelihood, the analysis of


SRecent studies by Sufi (2007) find increased access to capital of less informed investors following the introduction
of bank loan ratings as well as more concentrated syndicated loans in the case of opaque borrowers.










three months prior to one month after the effective date of the issue. I specifically look for

articles and headlines that contain the issue size and/or the usual key terms used in previous

studies. In the case of bank loan announcements the key terms are "line of credit," "credit line,"

"credit facility," "credit agreement," "credit extension," "new loan," "loan agreement," "loan

renewal," "loan revision," "loan extension," "finance company loan," "term loan," "commercial

loan," and "bank loan." Once the news and wire articles are selected, I collect data for the loan,

public debt or non-bank debt issue samples on the frequency of wire and press articles, timing of

the earliest article with respect to the issue date, news or wire source, and bundling of

information with other non-issue-related news in the earliest article. Furthermore, this hand-

collected news information identifies articles in which the only loan-related information is the

agreement size, those in which bank lending is inferred through terms such as "loan," and those

that specify it is a bank agreement, whether the identity of one or more members of the loan

syndicate is reported or not.

In 64 out of 304 cases of reported loans (out of 1,375 sample loans), the earliest news is

accompanied by other news concerning dividends, earnings, or control activity. Most empirical

studies of loan announcements exclude these confounding announcements so as to focus solely

on the information content of the financing news. In this paper, confounding press articles are

excluded in the analysis of the market reaction to loan announcements (I obtain similar results

when confounding articles are included). However, confounding press-reported loans are

considered in the examination of incremental financing choices because those decisions are made

earlier. Confounding articles are included as well as in the examination of reporting likelihood

given that earnings or dividend announcements, for example, may reduce information










The loan characteristics reported also in Table 2-3 show that reported loans are larger -

relative to the size of the borrower and have longer maturities than non-reported loans. In

addition, reported loans present significantly higher measures of the covenant intensity index

(4.5 vs. 3) that would arguably involve closer bank monitoring. The covenant intensity index, as

defined in previous studies (Bradley and Roberts (2004)), is an aggregate measure of covenant

structure. It equals the sum of six covenant indicators: collateral, dividend restrictions, asset sales

sweep, debt issuance sweep, equity issuance sweep, and the existence of more than two Einancial

covenants. 4 Thus, the value of the intensity index ranges between 0 and 6, and implicitly

assumes that each covenant is equally restrictive for borrowers. Moreover, following also the

criteria of previous studies, the index is set to missing when the value of one of the indicators is

missing (64% of the reported loans and 83% of non-reported loans).

Reported loans are also more likely than non-reported loans (30% vs. 12% of the cases)

to constitute restructurings following loan covenant violations and subsequent waivers during the

two years preceding the loan agreement object of study. However, the percentage of reported

loans that include in the syndicate lenders from which the firm has borrowed previously is lower

than in the case of non-reported loans (37% of reported loans have at least a common lender with

previous syndicates, vs. 44% in the case of non-reported loans). Thus, reported loans are more

likely to be both loan restructurings and new loans with new lenders.

Reported loans also present significantly higher all-in-drawn spreads, a measure that

further reflects the riskiness of borrowers with reported loans suggested by the firm summary

characteristics. In particular, Deal Scan expresses the all-in-drawn spread as a basis point mark-



4 Dealscan includes covenant information on dividend payment restrictions under certain conditions, collateral
requirements and prepayments requirements (so called sweeps that mandate that a portion of the loan be repaid out
of excess cash flows, debt and equity financings, or asset sales proceeds).










the press. Section IV studies the determinants of stock price reaction surrounding press reporting

or activation of non-reported issues. Section V examines the relationship between the

determinants of incremental financing options and loan reporting. Section VI studies long-term

operating performance. Section VII presents a summary and the conclusions.










Table 3-2. Marginal Effects of the Lowest and Highest Quartile Determinants on Loan
Reporting


Table 3-2 reports how the lowest and highest quartiles of some firm and loan characteristics in
the sample significantly affect the marginal effects of a loan being deemed noteworthy by the
press through a press article or a DJ wire, as examined in the first Probit model of Table 3-1.
*Significant at the 0. 1 level.


Effect of Lowest
Ouartile
On Reporting
Likelihood


Effect of Highest
Ouartile
On Reporting Likelihood


Tangibles/Assets

Debt/Assets

EBITDA/Assets

Restructure Loan

Commitment
Amount/Assets

Maturity


Over $1Billion vs.
Below $400 Million


0.05


-0.06*

0.08*

-0.21*

-0.17*


-0.1*

-0.01

-0.07*

0.21*

0.2*



-0.04


-0.11*


-0.05


-0.04










Table 5-2. Multinomial Logistic Regressions. Marginal Effects


(1.a)
Time trend 0.03
(2.85)
Tangibles/As -0.26
sets


_(1.b)_
-0.03
(-3.0)
0.19

(1.73)
0.38
(2.49)
1.69


_(1.c)_
-0.001
(-0.14)
0.07

(1.14)
-0.24
(-2.10)
0.01


_(2.a)_
0.01
(1.26)
-0.22


-0.08
(-0.61)
-1.31


_(2.b)_
-0.01

0.15

(1.27)
0.22
(1.85)
1.22


_(2.c)_
-0.002
(-0.47)
0.07

(1.04)
-0.14
(-1.38)
0.08


3.a)_
0,01
(2.5)
-0.04


_(3.b)_
-0.01
(-2.61)
0.02

(0.64)
0.13
(2.37)
0.39


_(3.c)
-0.001
(-0.23)
0.03

(1.37)
-0.02
(-0.93)
0.03


(-2.49)
-0.14
(-0.94)
-1.7


(-1.32)
-0.1
(-1.84)
-0.42


Debt/Assets

EBITDA/Ass
ets

Proceeds/Ass
ets

Medium-size
firm


Firm's past
12-month
cumulative
return


Market's past
12-month
cumulative
return


(-4.57) (5.27) (0.04) (-4.28) (4.24) (0.48) (-3.06) (2.81) (0.78)
1.2 -0.95 -0.24 0.93 -0.83 -0.1 0.6 -0.48 -0.13

(2.27) (-.)(-2.02) (2.34) (-2.29) (-0.96) (8.17) (-8.35) (-3.98)
0.23 -0.42 0.2 0.21 -0.29 0.09 0.02 -0.09 0.06

(2.84) (-5.43) (6.23) (1.66) (-2.27) (2.87) (0.53) (19)(5.49)
-0.02 0.01 0.01 -0.02 0.02 -0.001 -0.01 0.01 0.001


(-0.84) (0.62) (0.55) (-1.03) (1.31) (-0.05) (-0.86) (0.79) (0.39)


0.12


-0.08 -0.04 0.14 -0.07 -0.07 0.06 -0.04 -0.013


(1.21) (-0.84) (-0.78) (1.39) (-0.78) (13) (1.68) (-1.45) (-0.77)


Table 5-2 provides the


marginal effects


of a series of four multinomial logistic regressions


relating the choice of bank debt over public debt and non-bank private debt. The analysis is
based on a sample of 304 press-reported, and 1071 non-reported syndicated bank loans, as well
as 355 public debt and 92 non-bank private debt agreements, all activated between January 1996
and December 2004. The public debt and non-bank private debt random samples keep the
proportion of total bank loans in DealScan to public debt and non-bank private debt issues in
SDC, as well as the yearly proportions in the original SDC samples. I identify reported and non-
reported loan, public and non-bank issues searching the Dow Jones sources for articles
containing the deal amount and/or the key words used in the announcements literature. Deal Size
refers to the commitment amount in the case of bank loans and proceeds in the case of public
debt and non-bank private debt. Z statistics are reported in parenthesis.
(1.a) Public Debt vs. Reported Bank Loans
(1.b) Non-Bank Private Debt vs. Reported Bank Loans
(2.a) Reported Public Debt vs. Reported Bank Loans
(2.b) Reported Non-Bank Private Debt vs. Reported Bank Loans
(3.a) Public Debt vs. Bank Loans
(3.b) Non-Bank Private Debt vs. Bank Loans






























The sample consists of 1474 randomly selected syndicated bank loan agreements activated
between January 1996 and December 2004. The number of loans randomly selected each year
was determined so that each year the proportion of loans in my sample equals the corresponding
annual proportion of loans in Deal Scan database. I include loans involving US banks and only
US publicly traded firms at the time of loan activation. I identify reported and unreported loans
searching the Wire and Dow Jones News Retrieval Service for articles and headlines containing
the deal amount and/or the key words "bank," "line" "credit," "loan." I require the financial press
and wire information to be published between three months prior to one month after the
activation of the loan as reported in Deal Scan.


Debt. Repay 24.44 21.43% 32.73% 21.69% 23.2%

Working 14.62 8.33% 21.36% 19.88% 12.6%
Capital %
Takeover 10.4% 23.81% 11.36% 12.65% 8.51%
Corporate 28.8% 23.81% 19.09% 24.7% 32.38%
Purposes
Acquis. Line 4.51% 7.14% 4.09% 8.43% 3.65%
CP Backup 12.0% 5.95% 5.9% 5.42% 15.25%
Other 5.24% 9.53% 5.47% 7.23% 4.41%


Table 2-2. Loan Deal Purpose
Percent
tage wi
within
All
Loans


Percentage
thin Loans
Reported
Before
Activation


Percentage
within Loans
Reported On
or After
Activation


Percentage
within
DJ&PR Wire
Reported
Loans


Percentage
within Non-
reported
Loans










the loan news relative to loan activation may not be random. In particular, takeover is the most-

common primary purpose of the loans reported in the news before activation (24% vs. 1 1% for

loans reported on or after activation, and 8% for non-reported loans). Debt repayment is also

stated as primary purpose of loans reported in the press before activation, but in a lower

proportion than for loans reported in the press on or after activation (21% vs. 32% of the cases

for loans reported or on after activation, and 23% for non-reported loans). On the other hand,

working capital is more frequently the main purpose of loans reported in the press on or after

activation (21% vs. 8% in the case of loans reported before activation, and 12% in the case of

non-reported loans). And, interestingly, non-reported loans are more likely to include 364-day

facilities (not specified in Table 2-2). These 364-day facilities are a maj or source of short-term

financing, given that non-reported loans are established significantly more often for corporate

purposes (32% vs. 24% for loans reported before activation, and 19% for loans reported on or

after activation). Thus, it may be argued that the frequent corporate purposes of non-reported

loans signal a more transaction oriented type of lending that would be deemed less noteworthy

by press editors.

Table 2-3 reports firm and loan summary statistics for the fiscal year preceding loan

activation. The results identify significant differences between reported and non-reported cases,

as well as some potential reasons for the timing of the earliest news with respect to loan

activation. The reported summary statistics include the cases in which loan reporting is bundled

with non-loan-related news information, but I obtain similar results if I exclude confounding

articles.

The summary statistics of the sample firms, shown in Table 2-3, have been used

frequently in the banking literature as proxies for information asymmetries and ex ante risk of










non-bank private debt issues. This would explain why the summary statistics of non-bank private

debt borrowers and public debt borrowers show less statistical differences between reported and

non-reported borrowers than in the case of bank loans. More specifically, reported non-bank

private debt borrowers are smaller than their non-reported counterparts, but this is not the case of

public debt borrowers. Also, reported public debt borrowers present lower operating

performance than their non-reported counterparts, unlike non-bank private debt borrowers.

Table 2-4 introduces two other interesting facts related to credit rating and cumulative

returns. Like in the case of bank loans, non-bank private debt issues are reported more frequently

when the borrower does not have credit rating or its credit rating is below BBB. This is

consistent with the Eindings of Denis and Mihov (2003), who show that non-bank private debt

lenders have a niche among the riskiest borrowers. Furthermore, it can be argued that borrowers

without credit rating are not only riskier, but also more opaque, characteristics that also

differentiate reported borrowers from non-reported ones. On the other hand, the borrowers whose

public debt issues get reported are credit rated more often than non-reported borrowers, and the

credit rating is at least BBB more often than in the case of non-reported borrowers. Furthermore,

reported public debt borrowers also have higher cumulative returns during the year prior to the

issue. This is insightful because it suggests that the reporting of public debt issues in the press

may time the market.

How do the firm characteristics of bank, public debt and non-bank private debt borrowers

in the sample compare to those of previous studies of incremental Einancing decisions? To

address this issue I use the seasonally adjusted monthly Consumer Price Index (CPI), as provided

by the U. S. Department of Labor. I control for inflation by bringing the summary statistics of

previous studies from the mid-month of their study periods to June 2000, the mid-month of my










Fourth, although the reaction to non-bank private debt news is overall negative, I find no

significant market reaction to non-bank private debt news. The reason could be associated to the

bigger size of the non-bank private debt borrowers in my sample. Bigger firms tend to be less

opaque and therefore, less risky.









Mitchell, Mark L., and J. Harold Mulherin, 1984, The impact of public information on the stock
market, Journal ofFinance 49, 923-950.

Moerman, Regina Wittenberg, The impact of information asymmetry on debt pricing and
maturity, University of Chicago working paper, 2005.

Nini, Greg, David Smith, and Amir Sufi, 2007, Creditor Control Rights and Firm Investment
Policy, Journal ofFinancial Economics, forthcoming.

Rajan, Raghuram G., 1992, Insiders and outsiders: the choice between informed and arm's
length debt, Journal ofFinance 47, 1367-1400.

Ramakrishnan, R.T.S and Anjan Thakor, 1984, Information reliability and security design,
Review of Economic Studies 51, 415-432.

Rippington, Frederick, and Richard J. Taffler, 1996, The information content of firm financial
disclosures, Journal of business Finance & Accounting 22, 345-362.

Roberts, Michael, and Amir Sufi, 2007, Control Rights and Capital Structure: An Empirical
Investigation, SSRN working paper.

Roberts, Michael, and Amir Sufi, 2007, Contingency and renegotiation of financial contracts:
Evidence from private credit agreements, SSRN working paper.

Smith, Clifford, 1986, Investment banking and the capital acquisition process, Journal of
Financial Economics 15, 3-29.

Sufi, Amir, 2005, Banks and flexibility: Empirical evidence on the mix of equity, bonds and
bank debt, MIT working paper.

Sufi, Amir, 2007, The Real Effects of Debt Certification: Evidence from the Introduction of
Bank Loan Ratings, Review ofFinancial Studies, forthcoming.

Sufi, Amir, 2007, Information Asymmetry and Financing Arrangements: Evidence from
Syndicated Loans, Journal ofFinance 62, 629-668.

Sufi, Amir, 2007, Bank Lines of Credit in Corporate Finance: An Empirical Analysis, Review of
Financial Studies, forthcoming.

Thompson, Robert B., Chris Olsen and J. Richard Dietrich, 1987, Attributes of news about firms:
An analysis of firm-specific news reported in the Wall Street Journal Index, Journal of
Accounting Research 25, 245-274.









CHAPTER 7
SUMMARY AND CONCLUSIONS

An important strand of the banking literature holds that banks play a special role in the

capital acquisition process through the close relationship they establish with their borrowers.

More specifically, banks can gain access to private information that is not available to other

credit claimants and provide more flexible lending conditions, which is of particular importance

in the case of private firms. However, it is not clear why the more established less opaque public

firms would need to establish the same type of "unique" relationships with banks in all cases.

Also, it is puzzling why there is a significant positive stock price reaction surrounding the press

reporting of bank loans, given that practically all public firms have bank loans.

To address this puzzle, I begin by examining the frequency and determinants of bank

loan reporting in the Einancial press. I find that only 22% of bank loans are reported, and that the

sub sample of reported loans and borrowers is not representative of the entire population of

syndicated loans and public borrowers. More specifically, I find that, besides bigger loans

relative to assets, longer maturity loans, and loan restructuring following loan covenant

violations are significantly more likely to be reported in the press. In addition, consistently also

with a greater need for lender access to private information, I find that more opaque riskier

borrowers are more likely to have their loans reported.

Reported loans seem more informative about the potential of the borrower than non-

reported ones, an heterogeneity that, to my knowledge, has not been documented before.

However, since the cost of reporting has decreased with technological developments, there is an

increase in the frequency of reporting during the 1996 through 2004 study period that makes the

information content of the average loan decrease overtime. In any case, the firm and deal factors

that explain what makes a loan more likely to be reported are also those that explain the choice










Table 4-1. Determinants of the Stock Price Reaction Surrounding Loan Reports and Activation


Non-Confounding Non-Confounding Loans
Loan News vs. Non- News Before Loan Activation
Reported Loans vs. Non-Reported Loans
-0.51 -0.56
(-0.26) (-0.23)
-0.03 -0.09
(-0.09) (-0.2)
0.08 -0.21
(0.02) (-0.05)
4.76 5.86
(1.04) (1.11)
-3.28 0.72
(-0.35) (0.07)
-3.59 -4.45
(-1.15)(-1.08)
5.96 5.95
(3.01) (2.16)
-0.03 -0.02
(-0.66) (-0.35)
1.75 1.62
(0.52) (0.38)
-1.84 -3.3


Press-reported

Time Trend

Tangible/Assets

Debt/Assets

EBITDA/Assets

Restructure

Commitment Amount /Assets

Maturity

Medium-size firm

Firm's past 12-month cumulative
return

Market's past 12-month cumulative
return


(-1.72)
9.43

(2.13)
-0.88
(-0.27)
1196


11.11

(2.12)
-1.53
(-0.4)
1014


Constant


Table 4-1 provides the estimates of two OLS regressions relating the magnitude of the market
reaction to loan news to reporting and firm and loan characteristics The market reaction is
measured as the Standardized Prediction Errors for a two day event window centered on the
earliest of the announcement date or activation date. The analysis is based on a sample of 240
non-confounding press news (53 precede the activation of the loan) and 1071 non-reported loans
activated between January 1996 and December 2004. I identify reported and non-reported loans
searching the Dow Jones sources for articles containing the key words used in the
announcements literature. Commitment Amount is the size of the loan commitment. Maturity is
the maturity in months of the loan agreement as reported in Dealscan. Medium-size airm is a
dummy variable that equals one if the firm has less than $1 billion in assets and more than $400
million in assets. t statistics are reported in parenthesis.









debt (see, for, example, Hadlock and James (2002)). Thus, in light of these findings, one might

argue that loans obtained by borrowers with relatively more costly access to financing higher

information asymmetries and credit risk could be deemed more surprising and, therefore, be

more newsworthy and likely to get reported. This is exactly what I find.

I find that only 22% of syndicated Deal Scan loans obtained by public firms are reported

in the financial press, and that the borrowers whose loans are reported in the press are not

representative of the entire sample of bank borrowers. Reported borrowers are more highly

leveraged, present lower earnings, and higher surprise and dispersion of earnings forecasts.

Press-reported loans are also different from non-reported loans. More specifically, I find that

reported loans are larger, have longer maturities, higher covenant intensity indexes, and

constitute more often loan restructurings following covenant violations. Consistent with these

findings, I also find that information asymmetries, credit risk, loan size and maturity, and

whether or not the loan is a restructured loan determine the likelihood of loan reporting.

Thus, the distinction between reported and non-reported loans provides new insight into

the "specialness" of bank lending by documenting significant heterogeneity in the information

content of loans depending on whether the loan is reported or not. The firm and loan

characteristics that determine reporting require more lender access to private information about

the potential of the firm, a task at which banks have a comparative advantage with respect to

other lenders (Fama (1985) and Rajan (1992)). Thus, the reporting of a loan suggests the

presence of positive private information, although I find a time trend in reporting that leads to a

decrease in the information content of reported loans throughout the 1996 to 2004 period.

The higher likelihood of reporting among restructured loans complements recent work by

Ivashina and Sun (2007), who find 8.8% annualized profits during the month following loan









BIOGRAPHICAL SKETCH

Laura Gonzalez Alana earned her Bachelor of Science in electrical engineering from

Superior College of Engineering of Bilbao (Spain) simultaneously with Bachelors of Arts in

music education and music performance from J.C. Arriaga, Bilbao, Spain. In 2002 and 2003

respectively, she earned her Master of Business Administration and Master of Arts in foreign

languages and literatures from Southern Illinois University-Carbondale. The requirements for

the degree of Doctor of Philosophy in finance were completed at the University of Florida during

the spring of 2008. Upon graduation from the University of Florida, she will j oin Fordham

University, NY, as an assistant professor in finance. Laura' s research interests include Einancial

markets and institutions, corporate Einance, and international finance.









CHAPTER 1
INTTRODUCTION

In a recent study, Sufl (2007) documents that 94% of publicly traded firms have bank

lending relationships. In light of this fact, the reporting in the financial press of bank loan

agreements should be a routine, predictable event. Nevertheless, a number of previous studies

document a positive and statistically significant return associated to press articles on loan

agreements. This evidence suggests that market participants view press-reported bank loans as

material events (Mikkelson and Partch (1986), James (1987), Lummer and McConnell (1989),

Best and Zhang (1993), James and Smith (2000)). Furthermore, the positive share price reaction

to bank loan announcements (in contrast to the negative stock price reaction to most other

financing events) has been widely interpreted as evidence that banks play a unique or special role

in the capital acquisition process (Fama (1985) and Raj an (1992)). However, if all firms have

banking relationships, and if most bank loans are not reported in the press, then the share price

reaction may have more to do with the circumstances in which loan agreements are reported than

with the uniqueness of bank borrowing as a financing source (Smith (1986), and James and

Smith (2000)).

To further explore these issues, I begin by examining the frequency and determinants of

bank loan reporting in Dow Jones during the 1996 through 2004 time period. It is important to

note that the Loan Pricing Corporation (LPC) Deal Scan loans from which the sample is drawn -

are disclosed in some way, and consist of both press-reported and non-reported loan agreements.

More specifically, DealScan cites as sources of loan information Securities and Exchange

Commission (SEC) 8-K filings other public SEC filings, and industry sources. In addition,



1 The Securities and Exchange Commission (SEC) regulations require public firms to report any "material" event
through a Form 8-K. This form is described in more detail in Section III.a. of this paper.











Table 6-1. Median Peer-Adjusted Measures of Subsequent Operating Performance


EBITDA/Assets (%)
Press Reported Unreported
0.13 0.143


Debt/EBITDA (%)
Press Reported Unreported
2.15 1.8


0.13


0.12*


0.12*


0.12*


0.12


0.12

179


0.138


0.136


0.134


0.129


0.124


0.124

593


2.2*


2.8*


1.86


1.93


2.07


2.44


2.78*


3.02*


Year -3


Year -2


Year -1


Year 0


Year 1


Year 2


Year 3


Table 6-1 presents summary statistics for a balanced panel of firms. Year refers to the number of
fiscal years following the loan activation date. Year zero refers to the last fiscal year prior to the
activation date.
* Significantly different from the unreported loan median at the .05 level.
+ Significantly different from the unreported loan median at the .1 level.









CHAPTER 2
SAMPLE SELECTION, DATA AND SUMMARY STATISTICS

Sample Selection and Data

The sample consists of 1,375 randomly selected loans that were activated between 1996

and 2004. I choose 1996 as the start year of the study period because since January 1996 the

Securities and Exchange Commission (SEC) requires all firms to submit their filings

electronically, and the analysis uses information on loan restructurings from 8-K, 10-Q and 10-K

filings. The number of loans per year is initially determined so that each year the proportion of

loans in the sample equals the corresponding annual proportion of loans in the DealScan

database. The reason for this sampling is the analysis of incremental financing choices between

bank debt, public debt, and non-bank private debt. This analysis requires three random samples,

one for each type of financing source, that maintain the yearly proportions of the universe of

issues reported in Deal Scan (bank loans) and Securities Data Corporation (SDC) (public debt and

non-bank private debt). In addition, the random sample maintains the overall proportions of total

number of issues for each financing choice with respect to the other two, and requires the

borrowers to be publicly traded at the time of loan activation. Moreover, the loan sample includes

only completed loans involving U.S. banks with roles other than participant, and excludes loans

granted to financial institutions.


DealScan is the source of data regarding the identity and role of all members of the loan

syndicate, loan maturity, type and purpose, credit risk measures, and covenant information.

Securities Data Corporation provides information on the proceeds and payment conditions of

public debt and non-bank private debt issues.

For the classification of reported and non-reported loans, public and non-bank private

debt issues, I do a Dow Jones search for news and wire articles and headlines published between




Full Text

PAGE 1

1 DOGS THAT BARK: WHY ARE BANK LOAN ANNOUNCEMENTS NEWSWORTHY? By LAURA GONZALEZ ALANA A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2008

PAGE 2

2 2008 Laura Gonzalez Alana

PAGE 3

3 To my parents and Gaurav

PAGE 4

4 ACKNOWLEDGMENTS I would like to sincerely thank Christopher Jam es, my disserta tion chair, as well as Joel Houston, Mahendrarajah Nimalendran, Andy Naranjo, and Ronal Ward for all their guidance, encouragement, and advice in making sure I successfully completed this work.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4LIST OF TABLES................................................................................................................. ..........6ABSTRACT.....................................................................................................................................7 CHAP TER 1 INTRODUCTION....................................................................................................................82 SAMPLE SELECTION, DATA AN D SUMMARY STATISTICS ...................................... 14Sample Selection and Data.....................................................................................................14Summary Statistics.................................................................................................................17Summary Statistics of Market Reaction to Loan News and Loan Activation........................ 233 DETERMINANTS OF LOAN, PUBLIC DE BT, AND NON-BANK P RIVATE DEBT ANNOUNCEMENTS............................................................................................................34Determinants of Bank Loan Announcements......................................................................... 34Determinants of Public Debt and Non-Bank Private Debt..................................................... 384 STOCK PRICE REACTION TO LOAN NEWS. MULTIVARIATE ANALYSIS .............. 455 DETERMINANTS OF INCREMENTAL FINANCING OPTIONS..................................... 496 LONG-TERM PERFORMANCE OF BANK BORROWERS.............................................. 557 SUMMARY AND CONCLUSIONS.....................................................................................58LIST OF REFERENCES...............................................................................................................60BIOGRAPHICAL SKETCH.........................................................................................................63

PAGE 6

6 LIST OF TABLES Table page 2-1 Annual Distribution of Loans by Year of Deal Activation ................................................ 282-2 Loan Deal Purpose.......................................................................................................... ...292-3 Firm and Loan Summary Statistics for Reported and Unreported Loans.......................... 302-4 Firm and Deal Summary Statistics fo r Reported and Unreported Public Debt and Non-Bank Private Debt......................................................................................................322-5 Market Reaction Summary Statistics.................................................................................333-1 Determinants of the Likelihood and Timi ng of Bank Loan Reporting in the Press.......... 423-2 Marginal Effects of the Lowest and Highest Quartile Determinants on Loan Reporting............................................................................................................................433-3 Determinants of the Likelihood of Pub lic and Non-Bank Private Debt Reporting........... 444-1 Determinants of the Stock Price Reacti on Surrounding Loan Reports and Activation..... 474-2 Determinants of the Stock Price Reac tion Surrounding Public Debt and Non-Bank Private Debt Reports and Activation.................................................................................485-1 Multinomial Logistic Regressions on Incremental Financing Decisions.......................... 536-1 Median Peer-Adjusted Measures of Subsequent Operating Performance......................... 57

PAGE 7

7 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy DOGS THAT BARK: WHY ARE BANK LOAN ANNOUNCEMENTS NEWSWORTHY? By Laura Gonzalez Alana August 2008 Chair: Christopher James Major: Business Administration Virtually all publicly traded firms borrow from banks. However, despite their widespread use, the reporting of bank loan agreements in the financial press is associated with a positive share price reaction. In order to address this phenomenon, I em pirically examine the frequency and determinants of bank loan re porting, and find that the credit risk and asymmetric information proxies behind incremental financing decisions determine loan reporting. In addition, loan restructurings following covenant violations, and larg er loans -relative to firm sizeare also more likely to be reported in the press. Moreover, a nd although all loans are disclosed, I find that the market reacts to reported loans only when th e news precedes loan ac tivation. Overall, the evidence suggests that reported loans are more noteworthy because they are more informative about the potential of the borrowe r than non-reported loans, and th at the information content of the average reported bank loan decr eases during the 1996 through 2004 period.

PAGE 8

8 CHAPTER 1 INTRODUCTION In a recent study, Sufi (2007) documents that 94% of publicly traded firms have bank lending relationships. In light of this fact, the reporting in th e financial press of bank loan agreements should be a routine, predictable even t. Nevertheless, a number of previous studies document a positive and statistically significant re turn associated to press articles on loan agreements. This evidence suggests that market participants view pres s-reported bank loans as material events (Mikkelson and Partch (1986) James (1987), Lummer and McConnell (1989), Best and Zhang (1993), James and Smith (2000)). Furthermore, the positive share price reaction to bank loan announcements (in contrast to the negative stock price reaction to most other financing events) has been widely interpreted as evidence that banks play a unique or special role in the capital acquisition process (Fama (1985) and Rajan (1992)). However, if all firms have banking relationships, and if most bank loans are not reported in th e press, then the share price reaction may have more to do with the circumstan ces in which loan agreements are reported than with the uniqueness of bank borrowing as a financing source (Smith (1986), and James and Smith (2000)). To further explore these issues, I begin by examining the frequency and determinants of bank loan reporting in Dow Jones during the 199 6 through 2004 time period. It is important to note that the Loan Pricing Corporation (LPC) Deal Scan loans from which the sample is drawn are disclosed in some way, and consist of both press-reported and non-repor ted loan agreements. More specifically, DealScan cites as sources of loan information Securities and Exchange Commission (SEC) 8-K filings1, other public SEC filings, and industry sources. In addition, 1 The Securities and Exchange Commission (SEC) regulations require public firms to report any material event through a Form 8-K. This form is described in more detail in Section III.a. of this paper.

PAGE 9

9 DealScan offers affordable real-time web access to agreed-upon but not-yet-active loans2. Hence, the loans reported in Dow Jones that the financial press views as noteworthy are a subset of a pool of disclosed loans. This distinction between reported and disclosed loans is s ubtle but nevertheless important. If there is disclosure about the deals but, as I find, no significant market reaction surrounding the activation of non-reported bank loans, and one assumes reported loans to be significant, it can be argued that the market views reported loans differently than non-reported loans. SEC regulation requires firms to disclose a ny material event that can affect the stock price and shareholders should know about. Furthermore, the literatu re on discretionary disclosure describes how firms are more forthcoming when the news is good (Miller (2002)). Thus, given that banks are considered superior screener s (Diamond (1991) and Fama (1985)) and effective monitors (Nini et al. (2008)), and bank financin g a credible signal of firm quality (Schenone (2004), and James and Wier (1990)), one would exp ect firms to try to disclose all bank loans. Accordingly, I find that the source of loan news is usually the borrower. On the other hand, the source of news for non-bank private debt i ssues is usually market participants. Why would firms not report a ll of their bank loans? Bagno li and Watts (2007) find that firms are more likely to disclose higher values of private info rmation when financial reports do not contain sufficient good news and performance is significantly different than expected. In a related study, Verrecchia and Weber (2006) docum ent that firms are less likely to withhold information in material contract filings when they issue long-term debt. Moreover, prior work shows that riskier more opaque firms are more likely to choose reported bank debt over public 2 The subscription cost to the DealScan database and real-time website is $2,000 per month for corporations with one user ($330/month is more than one user) and $7,304 per year for academic institutions.

PAGE 10

10 debt (see, for, example, Hadlock and James (2002 )). Thus, in light of these findings, one might argue that loans obtained by borrowers with relati vely more costly access to financing higher information asymmetries and credit risk could be deemed more surprising and, therefore, be more newsworthy and likely to get re ported. This is exactly what I find. I find that only 22% of syndicated DealScan loans obtained by public firms are reported in the financial press, and that the borrowers whose loans are reporte d in the press are not representative of the entire sample of ba nk borrowers. Reported borrowers are more highly leveraged, present lower earnings and higher surprise and disp ersion of earnings forecasts. Press-reported loans are also di fferent from non-reported loans. More specifically, I find that reported loans are larger, have longer maturitie s, higher covenant intensity indexes, and constitute more often loan restructurings follo wing covenant violations. Consistent with these findings, I also find that information asymmetrie s, credit risk, loan size and maturity, and whether or not the loan is a restructured lo an determine the likelihood of loan reporting. Thus, the distinction between reported and non-reported loans provide s new insight into the specialness of bank lending by documenting significant heterogeneity in the information content of loans depending on whether the loan is reported or not. The firm and loan characteristics that determine reporting require mo re lender access to private information about the potential of the firm, a task at which banks have a comparative advantage with respect to other lenders (Fama (1985) and Rajan (1992)). Thus, the reporting of a loan suggests the presence of positive private information, although I find a time trend in reporting that leads to a decrease in the information content of re ported loans throughout the 1996 to 2004 period. The higher likelihood of reporting among restru ctured loans complements recent work by Ivashina and Sun (2007), who fi nd 8.8% annualized profits during the month following loan

PAGE 11

11 renegotiations in trading by institutional investors that are members of loan syndicates. The profitability of this trading supports the argument that th e private information revealed and produced during the renegotiations is not only good, but also highl y valuable, which would give the borrower a particularly str ong incentive to report it. The disclosure literature documents how managers time voluntary disclosures in a manner that maximizes insider trading profits while minimizi ng potential litigation costs associated with disclosure (Cheng and Ko (2006). Additionally, the literature on the announcement effects of financing events wh ich spans for over two decades reports a significant positive stock price reaction to announ cements of bank loans, a negative reaction to non-bank private debt placements, and an insignifi cant reaction to straight public debt issues (James (1987), and Eckbo et al. (2007) for a recent re view of this literature). Consistent with this evidence, I find a significant positive market reac tion to reported loans, but only when the news precedes loan activation, and mainly driven by loan size and prior borrower stock performance with respect to the market. This suggests that reported loans are considered more informative when the news precedes the activ ation of the loan because it is arguably more surprising for reported borrowers to obtain financing and timing intensifies the surprise factor. Denis and Mihov (2003) find that credit risk is a significant determinant of the choice between bank debt, public debt and non-bank priv ate debt. However, their analysis considers only the subset of loans that are reported, like other previous studies on incremental financing decisions (Hadlock and James (2002)). In other words, the conclusions of previous work on incremental financing derive from comparing the s ubset of reported bank loans to full samples of other transactions such as public debt and non-bank private debt. Th erefore, if reported loans are a non-representative subset of the loans in DealScan, the analysis of the incremental financing

PAGE 12

12 decisions should also consider ca ses of non-reported loan s. In this context, I find that although reported loans and borrowers differ significantly from non-reported loans and borrowers, the factors that determine the choice between bank debt, public de bt and non-bank private debt do not change whether reported loans or all loans are considered. In conse quence, one could argue that non-reported loans are somewhat sp ecial and different from public debt. Besides a more effective screening, the speci alness of bank loans re lative to other types of financing also has to do with more intense monitoring, which would arguably be more important for loans with higher covenant intens ity, i.e., the reported loans. Furthermore, the access to private information and contracting co nditions that protect banks from the higher potential default on reported loans could arguably lead or not to an improvement in operating performance. In this context, I find that although reported borrow ers present poorer measures of operating performance during the year prior to loan activat ion (relative to non-reported borrowers), there is relative improvement in the operating performance of reported borrowers with respect to non-reported bo rrowers two years following loan activation. This result is consistent with the recent literature on loan covenants (Roberts and Su fi (2007), Nini et al. (2008)) that studies the relation between loan co venant violations, loan restructurings, and subsequent performance3. The remainder of the paper is organized as follows. Section II provides a description of the data sources for the unique samples of bank debt, public debt, and non-bank private debt. It also provides summary statistics of the deals and borrowers, as well as market reaction. Section III examines the determinants of bank loans, pub lic debt, and non-bank private debt reporting in 3 Besides Nini et al. (2007), Brophy, Ouimet and Sialm (2004) show that the degree of protection granted to investors has a significant effect on subsequent performance, especially in the case of firms that borrow from institutional investors as opposed to hedge funds.

PAGE 13

13 the press. Section IV studies the determinants of stock pri ce reaction surrounding press reporting or activation of non-reported issues. Secti on V examines the relationship between the determinants of incremental financing options an d loan reporting. Secti on VI studies long-term operating performance. Section VII pres ents a summary and the conclusions.

PAGE 14

14 CHAPTER 2 SAMPLE SELECTION, DATA AN D SUMMARY STATISTICS Sample Selection and Data The sam ple consists of 1,375 randomly select ed loans that were activated between 1996 and 2004. I choose 1996 as the start year of the study period because since January 1996 the Securities and Exchange Com mission (SEC) requires all firms to submit their filings electronically, and the analysis uses informati on on loan restructurings from 8-K, 10-Q and 10-K filings. The number of loans per year is initially determined so that each year the proportion of loans in the sample equals the corresponding annual proportion of loans in the DealScan database. The reason for this sampling is the anal ysis of incremental financing choices between bank debt, public debt, and non-ba nk private debt. This analysis requires three random samples, one for each type of financing source, that maintain the yearly proportions of the universe of issues reported in DealScan (bank loans) and S ecurities Data Corporation (SDC) (public debt and non-bank private debt). In addition, the random sample maintains the overall proportions of total number of issues for each financing choice w ith respect to the other two, and requires the borrowers to be publicly traded at the time of loan activation. More over, the loan sample includes only completed loans involving U.S. banks with ro les other than participant, and excludes loans granted to financial institutions. DealScan is the source of data regarding the identity and ro le of all members of the loan syndicate, loan maturity, type and purpose, cr edit risk measures, and covenant information. Securities Data Corporation provides information on the proceeds and payment conditions of public debt and non-bank private debt issues. For the classification of reported and non -reported loans, public and non-bank private debt issues, I do a Dow Jones search for news and wire articles and headlines published between

PAGE 15

15 three months prior to one month after the effective date of the issue. I specifically look for articles and headlines that contain the issue si ze and/or the usual key terms used in previous studies. In the case of bank loan announcements th e key terms are line of credit, credit line, credit facility, credi t agreement, credit exte nsion, new loan, loan agreement, loan renewal, loan revision, loa n extension, finance company lo an, term loan, commercial loan, and bank loan. Once the news and wire ar ticles are selected, I collect data for the loan, public debt or non-bank debt issue samples on the frequency of wire and press articles, timing of the earliest article with resp ect to the issue date, news or wire source, and bundling of information with other non-issue -related news in the earliest ar ticle. Furthermore, this handcollected news information identifies articles in which the only loan-related information is the agreement size, those in which bank lending is in ferred through terms such as loan, and those that specify it is a bank agreement, whether the identity of one or more members of the loan syndicate is reported or not. In 64 out of 304 cases of reported loans (out of 1,375 sample loans), the earliest news is accompanied by other news concerning dividends, earnings, or control activity. Most empirical studies of loan announcements exclude these co nfounding announcements so as to focus solely on the information content of the financing news In this paper, confounding press articles are excluded in the analysis of the market reaction to loan announcements (I obtain similar results when confounding articles are included). However, confounding press-reported loans are considered in the examination of incremental financing choices because those decisions are made earlier. Confounding articles are included as well as in the examination of reporting likelihood given that earnings or dividend announcem ents, for example, may reduce information

PAGE 16

16 asymmetries associated with selling securiti es. Thus, excluding conf ounding reporting may bias the results and conclusions. The news information on bank loans is supplem ented with information from filings with the Securities and Exchange Commission (SEC). Mo re specifically, I collect data from the SEC filings whenever available on whether the loans constitute a new agreement, renewal or a restructure deal following a covenant violati on and subsequent waiver. This manual search within SEC filings covers the tw o years prior to loan activation because Roberts and Sufi (2007) find in their study of renegotiation of financial co ntracts that the average effective maturity of bank loans is half the average st ated maturity, which DealScan re ports to be of around four years for the loans in the sample. More specifically, I search the SEC filings for specific expressions1 used in previous studies, and check each passage to ensure that the expressions indeed refer to financial covenant violations, waiv ers, and loan restructurings. The data on issuing firm characteristics, st ock price reaction, and an alysts forecasts, is obtained from Compustat, the Center for Res earch of Security Prices (CRSP), and the Institutional Brokers Estimate Sy stem (IBES) data tape. As a result, the overall sample consists of 1,375 bank loan deals, 355 public debt issues and 92 non-bank private de bt deals. The news search process identifies 304 loan s that are reported in the fina ncial press, 166 additional loans that are identified in the wire but not reported in the press, 905 loans that are not reported in wire or press articles, 207 reported pub lic debt issues and 42 reported nonbank private debt issues. In 1 The specific terms are those also used by Roberts and Sufi (2007): in violation of covenant, in violation of a covenant, in default of covenant, in technical violation of covenant, in technical violation of a covenant, in violation of financial covenant, in violation of a fina ncial covenant, in technical violation of a financial covenant, in technical violation of financial covenant, i n technical default of a financial covenant, in technical default of financial covenant, not in compliance, out of compliance, received waiver, receiver a waiver, obtained a waiver, obtained waiver.

PAGE 17

17 addition, I find that for 78.9% (240 out of 304) of reported loans there is no bundling of loan information with other non-loan related news. Summary Statistics Table 2-1 provides the annual percentage of loans reported in the financial press, those listed on the wire services but not reported in th e press, or those not re ported at all. More specifically, I find that 22% (304 out of 1,375) of th e sample loans are reported in the financial press, while an additional 12% (166 out of 1,375) of loans are re ported in the wire but are not reported subsequently in the press. In addition, about 8% (not reported in the table) of the loans that are not identified in wire or press ar ticles are made public through 8-K filings. The Form 8-K filings follow the SEC regul ation that requires firms to report any material event that may affect the stock pri ce, or definitive agreement not made in the ordinary course of the registrants business.2 As Nini et al. (2008) note, as private agreements, loans are not legal securities a nd, thus, are not subject to dire ct SEC regulation. However, the SEC precedent has established a requirement th at public companies include copies of all material contracts, including bank loan agreements, with relevant SEC disclosures. These contracts typically appear as exhibits at the end of a 10-K or 10-Q report, or as an attachment to an 8-K filing. Table 2-2 presents some initial evidence of how press-reported loans differ from nonreported loans in terms of primary loan purpose, and introduces the hypothes is that the timing of 2 An agreement is deemed as material definitive when it provides for obligations that are material to and enforceable against the registrant or right s that are material to the registrant an d enforceable by the registrant against one or more parties to the agreement, in each case whether or not subject to conditions. The Form 8-K requires the agreement date, the identity of the partie s and a brief description of any materi al relationship between the registrant or its affiliates and any of the parties, other than in re spect of the material definitive agreement. Moreover, the Additional Form 8-K, effective August 2004, expands the nu mber of events that are reportable on Form 8-K and shortens the filing deadline for most items to four business days. These amendments are said to further the goals of Section 409 of the Sarbanes-Oxley Act.

PAGE 18

18 the loan news relative to loan activation may not be random. In pa rticular, takeover is the mostcommon primary purpose of the loan s reported in the news before activation (24% vs. 11% for loans reported on or after activ ation, and 8% for non-reported loan s). Debt repayment is also stated as primary purpose of loans reported in the press before activation, but in a lower proportion than for loans reported in the press on or after activa tion (21% vs. 32% of the cases for loans reported or on after activation, and 23% for non-reported loans). On the other hand, working capital is more frequently the main pur pose of loans reported in the press on or after activation (21% vs. 8% in the cas e of loans reported before ac tivation, and 12% in the case of non-reported loans). And, interest ingly, non-reported loans are mo re likely to include 364-day facilities (not specified in Tabl e 2-2). These 364-day facilities ar e a major source of short-term financing, given that non-reported loans are established significan tly more often for corporate purposes (32% vs. 24% for loans reported befo re activation, and 19% fo r loans reported on or after activation). Thus, it may be argued that the frequent corporate purposes of non-reported loans signal a more transaction oriented type of lending that would be deemed less noteworthy by press editors. Table 2-3 reports firm and loan summary st atistics for the fiscal year preceding loan activation. The results identify si gnificant differences between reported and non-reported cases, as well as some potential reasons for the timing of the earliest news with respect to loan activation. The reported summary st atistics include the cases in which loan reporting is bundled with non-loan-related news information, but I obt ain similar results if I exclude confounding articles. The summary statistics of the sample firms, shown in Table 2-3, have been used frequently in the banking literature as proxies for information asymmetries and ex ante risk of

PAGE 19

19 the borrower. The reason is that bank lending is credited for providing tailored and flexible lending to firms that present more opaque ri skier profiles (Houston and James (1996), Johnson (1997), and Hadlock and James (2002)). Hence, si nce practically all publ ic firms borrow from banks, heterogeneity in the measures of opaquene ss and credit risk within the loan sample may explain why reported loans and their borrowers may be deemed different from non-reported ones. As shown in Table 2-3, firms with reporte d loans are on average smaller (in terms of assets and sales), and have higher surpri se and dispersion of earnings forecasts3, higher leverage, and poorer operating performance during the fiscal ye ar preceding the activation of the loan as measured by the ratio of earnings before interest, taxes, deprec iation and amortization (EBITDA) to assets. The debt to EBITDA ratio, covenant included in about 50% of loan contracts (Demiroglu and James (2007)), is also significantly higher during the year preceding the loan for borrowers with reported loans. Thus, reported bor rowers appear to be on average more opaque and riskier than non-reported borrowers. This is insightful, because those same factors determine as well the choice of bank debt over other source s of incremental financing, as previous work finds. More specifically, previ ous studies document that firms with higher information asymmetries and lower credit risk are more likel y to choose reported bank debt over public debt (Hadlock and James (2001)). Hence, the highe r information content of reported bank loans compared to non-reported ones can explain the need for reporting. 3 Like Gomes and Phillips (2006), I compute the quarterly an alyst earnings surprise as the absolute value of the difference between the median quarterly earnings estimate an d the actual quarterly earnings per share, normalized by the stock price at the fiscal quarter end. Similarly, I com pute the analyst earnings dispersion measure as the standard deviation of outstanding earnings forecas ts normalized by the stock price re quiring a minimum of two outstanding earning forecasts. Additionally, since firms may have incentives to disclose more information prior to a public issue vs. prior to a private issue. Thus, the earnings surprise and dispersion measur es use the average of the last four quarters ending a quarter before the issue date.

PAGE 20

20 The loan characteristics reported also in Ta ble 2-3 show that reported loans are larger relative to the size of the borrower and have longer matu rities than non-reported loans. In addition, reported loans present si gnificantly higher measures of the covenant intensity index (4.5 vs. 3) that would arguably involve closer bank monitoring. The covenant intensity index, as defined in previous studies (Bradley and Robert s (2004)), is an aggregat e measure of covenant structure. It equals the sum of six covenant indicators: collater al, dividend restrictions, asset sales sweep, debt issuance sweep, equity issuance sweep, and the existence of more than two financial covenants.4 Thus, the value of the intensity inde x ranges between 0 and 6, and implicitly assumes that each covenant is equally restrictive for borrowers. Moreover, following also the criteria of previous studies, th e index is set to missing when the value of one of the indicators is missing (64% of the reported loans and 83% of non-reported loans). Reported loans are also more likely than non -reported loans (30% vs. 12% of the cases) to constitute restructurings fo llowing loan covenant violations and subsequent waivers during the two years preceding the loan agreement object of study. However, the percentage of reported loans that include in the syndica te lenders from which the firm has borrowed previously is lower than in the case of non-reported loans (37% of reported loans have at least a common lender with previous syndicates, vs. 44% in the case of nonreported loans). Thus, reported loans are more likely to be both loan restructuri ngs and new loans with new lenders. Reported loans also present significantly hi gher all-in-drawn spreads, a measure that further reflects the riskiness of borrowers with reported loans suggested by the firm summary characteristics. In particular, DealScan expresses the all-in-drawn spread as a basis point mark4 Dealscan includes covenant information on dividend payment restrictions under certain conditions, collateral requirements and prepayments requirements (so called sweeps that mandate that a portion of the loan be repaid out of excess cash flows, debt and equity financings, or asset sales proceeds).

PAGE 21

21 up over the 6-month LIBOR that includes recurri ng fees associated with the credit facility.5 The spread is used as a measure of per dollar cost of borrowing in a number of previous empirical studies on loan pricing (for example, Bradle y and Roberts (2004), Guner (2006) and Moerman (2005)). It is interesting to note here that, besides the loan size relative to the firm size, all-indrawn spreads measures also differentiate, among reported loans, those that are reported before loan activation with respect to those reported on or after activati on, as shown in Table 2-3. More specifically, I find that loans reported in the press before activation are larger (relative to firm size) and present higher all-in-d rawn spreads than those report ed on or after activation, which would arguably make their announ cements even more noteworthy. Overall, the firm and loan summary statistics in Table 2-3 show that reported loans and their borrowers are not representative of the universe of loans and borrowers. Furthermore, the firm characteristics that differentiate reported loans from non-reported loans also differentiate bank borrowing from public borrowing in previo us studies (Hadlock and James (2002)). In consequence, it could be argued that non-reporte d bank borrowing constitutes an intermediate debt source between reported bank borrowing and public debt borrowing. Table 2-4 presents firm and issue characteris tics for non-bank private debt and public debt deals. Like bank loans, both in the case of pub lic debt and non-bank priv ate debt issues, bigger deals (relative to firm assets or not) are more likely to be reported. However, there are important differences in terms of frequency of reporting and potential determinants of press reporting between bank loans, public debt and non-bank privat e debt issues. First, it is important to note that about 40% of the non-bank private debt issues are reported and that 60% of the public debt issues are reported. Thus, bank loans are much less likely to get reported than public debt and 5 LPC computes the spread for non-LI BOR based loans by converting index us ed to price the loan into a LIBOR equivalent using the historical relationship between the index and the LIBOR.

PAGE 22

22 non-bank private debt issues. This would explai n why the summary statistics of non-bank private debt borrowers and public debt borrowers show le ss statistical differences between reported and non-reported borrowers than in the case of ba nk loans. More specifically, reported non-bank private debt borrowers are smaller than their non-re ported counterparts, but th is is not the case of public debt borrowers. Also, reported public debt borrowers present lower operating performance than their non-reported counterpart s, unlike non-bank private debt borrowers. Table 2-4 introduces two other interesting facts related to credit rating and cumulative returns. Like in the case of bank loans, non-bank private debt issues are reported more frequently when the borrower does not have credit rating or its credit rating is below BBB. This is consistent with the findings of Denis and Mihov (2003), who show that non-bank private debt lenders have a niche among the riskiest borrowers. Furthermore, it can be argued that borrowers without credit rating are not only riskier, but also more opaque characteristics that also differentiate reported borrowers from non-reporte d ones. On the other hand, the borrowers whose public debt issues get reported are credit rate d more often than non-reported borrowers, and the credit rating is at least BBB more often than in the case of non-reported borrowers. Furthermore, reported public debt borrowers also have higher cumulative returns during the year prior to the issue. This is insightful because it suggests that the reporting of public de bt issues in the press may time the market. How do the firm characteristics of bank, public debt and non-bank private debt borrowers in the sample compare to those of previous studies of incremental financing decisions? To address this issue I use the seas onally adjusted monthly Consumer Price Inde x (CPI), as provided by the U.S. Department of Labor. I control for inflation by bringing the summary statistics of previous studies from the mid-month of their study periods to June 2000, the mid-month of my

PAGE 23

23 January 1996 to December 2004 period. Overall, and as expected for syndicated loan borrowers, the firms in this paper are bigger. In terms of firm assets, Denis a nd Mihov (2003) study bank, public debt and non-bank private de bt borrowers that are significantly smaller than the borrowers in my sample (firm asset medians of $162.24, $2,435, and $246.2 million respectively). James (1987) also uses smaller public private debt borrowers (firm assets medians of $3,322.27), as well as longer maturity bank loans (median of 72 months). In terms of leverage, Hadlock and James (2002) study the choice between bank borro wing and public debt borrowing with higher debt to assets ratios (median ratios of 0.52 and 0.44 respectively). Consistently with the higher leverage, the bank and public debt firms in Hadlock and James (2002) study borrow less (median commitment to assets ratios of 0.26 and 0.047 re spectively). However, I observe no significant difference in terms of firm assets when compar ing the size of the bank borrowers in previous studies with the subsample of reported loans in my sample. This is insightful, because previous studies consider only reported loans. Hence, th e similarities between my subsample of reported borrowers and the samples of reported borrowers used previously in the literature allow a comparison of results. Summary Statistics of Market Reaction to Loan Ne ws and Loan Activation How do loans come to be published through Dow Jones? The issuance of a press release by the borrower constitutes one more source of information in the first step of the chain that conveys the information to the market through th e press. Thompson et al (1987) describe the process. Once reporters or firms transmit the stor y to the Dow Jones News Service, the editors summarize them, weight their importance, and determine whether to make them press news. This process takes place within hours or, in most cases, not more than one day. One should note that the dissemination of ne ws information through wire services does not target only Dow Jones. In fact, the New York Stock Exchange regulations require

PAGE 24

24 simultaneous disclosure of firm-specific news to Dow Jones and Reuters, and the American Stock Exchange requires simultaneous disclosure to Dow Jones, Reuters, Associated Press, United Press International, the Wall Street Journal the New York Times Standard & Poors, and Moodys Investor Service (Thompson et al. (19 87)). Nevertheless, bo th practitioners and academics rely on Dow Jones as the primary source of news existence and timing, most likely because of its longer tradition and wi der dissemination of information.6 In any case, when I select a random subsample of loans and search for press articles without limiting the source, I find that Dow Jones captures all case s in which the loans are reported. The nature of the publishing process raises several questions. First, can financial press editors influence returns and tr ading volume through editing? Mitchell and Mulherin (1994) find that the number of news stories a nd market activity are directly re lated and that this relation is robust to the size of the headlines and macroeco nomic announcements. In addition, they find that the association between larger size headlines and higher market returns does not have a significant effect on trading volume.7 Second, how much news is there in successive announcements? Rippington and Taffler (1996) fi nd that only preliminary announcements and interim statements convey substantial amounts of new information. Thus, in light of the findings in previous work, one could argue that multip le incomplete announcement returns could be aggregated in the study of the market reaction to loan news. However, the loans in the sample are reported in most cases th rough unique articles, and in the few cases where there is more than one 6 Dow Jones maintains five wire services in addition to publishing the Wall Street Journal and Barrons Dow Jones also maintains the News Retrieval Serv ice, which contains selected articles from the Dow Jones News Service, now called Factiva, and the Wall Street Journal In terms of coverage, the Journal appears to include the initial news release data for approximately 96% of the items listed in the NYSE/ASE Index (Thompson et al. (1987)). 7 Mitchell and Mulherin (1994) find April to be the month with the smallest number of announcements per day. In addition, Thompson et al. (1987) also document that firms are less likely to issue news releases on Fridays and in December.

PAGE 25

25 article about a loan (in most cases just two artic les) they are either dated the same day with practically verbatim information about the loan or one article is a summary of the key points in the other one, or distant enough in time not to affect the two-day event study results. The two-day-window event studies of bank, public and non-bank private debt announcements use the earliest of the first news or the issue activation date as event date, a standard market model estimation period that ends 46 trading days before the event date, and the standard methodology by Mikkelson and Partch ( 1986) to measure the stock price reaction. In addition, since there is contamination of loa n, public debt and non-bank private debt news, I control for it and find similar results when conf ounding news articles are included. Thus, I report in this paper only the non-confounding results given that previous st udies exclude confounding press articles in their analysis of market reaction surrounding announcements. I find that for 78.9% (240 out of 304) reported lo ans, and 63.1% (53 out of 84) of loans reported before the activation date, there is no bundli ng of loan information with other non-loan related news. I classify as bundling or contamination of the earliest loan related article any information within the same article that can affect the stock price. The nature of the bundling of information can range, for example, from an earnings announcemen t that does not appear to be a repetition of previous news, to the completion of an acquisition. The cumulative abnormal returns (CARs), aver age standardized prediction errors (SPE) and Z values associated to loan, public debt and non-bank private debt reporting and activation are reported in Table 2-5 and pr esent four interesting findings. First, although the market can gain access to information about the existence and deal activation date of the non-reported loans, there is no significant market reaction surrounding the deal activation for the non-reported

PAGE 26

26 loans8. More specifically, and as disc ussed in the introduction to this paper, DealScan uses as sources of loan information public SEC filings and market sources, and maintains a real-time website that reports the existen ce of not-yet active loans. Thus, it can be argued that the sample loans are disclosed. Second, I find that the signifi cant market reaction to repor ted loans is driven by the reaction to news articles that precede the activation of the loans. This is consistent with the differences shown in Table 2-5 between loans repo rted before activation and loans reported on or after activation. As mentioned be fore, loans reported before activation show a higher DealScan spread and commitment size relative to assets th at would arguably make them more noteworthy among all the loans that are reported. In conseque nce, the reporting in the press of those loans would more likely trigger a larger market reaction, especially if the timing of the article that precedes loan activation cannot be foreseen and increases the surprise by which the news takes the market. Sections III and IV discuss in more detail the determinants of loan reporting and market reaction in relation to the timing of the news. Third, the market reaction to public debt announcements is not significant when all reported issues are considered. However, when pr ess articles are classified by their timing with respect to deal activation, the announcements following activa tion are associated with a significant positive market reaction. This is in teresting, because I find no difference between public debt deals reported before activation and after activation in term s of borrower opaqueness and credit risk, or in terms of deal characteristics. 8 Previous work by Thompson et al. (1987) also observes no significant market reaction to announcements in the wire services that are not followed by announcements in the financial press, bu t their study does not distinguish bank loan reporting from all other news types.

PAGE 27

27 Fourth, although the reaction to non-bank private debt news is overall negative, I find no significant market reaction to non-bank private debt news. The reason could be associated to the bigger size of the non-bank private debt borrowers in my sample. Bigger firms tend to be less opaque and therefore, less risky.

PAGE 28

28 Table 2-1. Annual Distribution of Loans by Year of Deal Activation Year of Deal Activation All Sample Bank Loans N=1375 Loans Reported in Press Before Activation N=84 Loans Reported in Press On or After Activation N=220 DJ & PR Wire Reported Loans N=166 Nonreported Loans N=905 1996 11.78% 8.33%8.64%11.45% 12.93% 1997 15.56% 15.48%14.21%15.66% 15.69% 1998 12.51% 20.24%11.36%8.43% 12.82% 1999 11.27% 16.67%10.9%7.83% 11.49% 2000 11.85% 7.14%11.36%11.45% 12.49% 2001 11.05% 9.52%14.09%7.23% 11.16% 2002 10.25% 14.29%10%11.45% 9.72% 2003 9.38% 4.76%11.36%12.05% 8.84% 2004 6.35% 3.57%7.27%14.45% 4.86% The sample consists of 1474 randomly selected syndicated bank loan agreements activated between January 1996 and December 2004. The number of loans randomly selected each year was determined so that each year the proportion of loans in my sample equals the corresponding annual proportion of loans in DealScan database I include loans involving US banks and only US publicly traded firms at the time of loan activa tion. I identify reported and unreported loans searching the Wire and Dow Jones News Retrieval Service for articles and headlines containing the deal amount and/or the key words bank, l ine credit, loan. I requ ire the financial press and wire information to be published between three months prior to one month after the activation of the loan as reported in DealScan.

PAGE 29

29 Table 2-2. Loan Deal Purpose Percen tage within All Loans Percentage within Loans Reported Before Activation Percentage within Loans Reported On or After Activation Percentage within DJ&PR Wire Reported Loans Percentage within Nonreported Loans Debt. Repay 24.44 % 21.43%32.73%21.69% 23.2% Working Capital 14.62 % 8.33%21.36%19.88% 12.6% Takeover 10.4% 23.81%11.36%12.65% 8.51% Corporate Purposes 28.8% 23.81%19.09%24.7% 32.38% Acquis. Line 4.51% 7.14%4.09%8.43% 3.65% CP Backup 12.0% 5.95%5.9%5.42% 15.25% Other 5.24% 9.53%5.47%7.23% 4.41% The sample consists of 1474 randomly selected syndicated bank loan agreements activated between January 1996 and December 2004. The number of loans randomly selected each year was determined so that each year the proportion of loans in my sample equals the corresponding annual proportion of loans in DealScan database I include loans involving US banks and only US publicly traded firms at the time of loan activa tion. I identify reported and unreported loans searching the Wire and Dow Jones News Retrieval Service for articles and headlines containing the deal amount and/or the key words bank, l ine credit, loan. I requ ire the financial press and wire information to be published between three months prior to one month after the activation of the loan as reported in DealScan.

PAGE 30

30Table 2-3. Firm and Loan Summary Statistics for Re ported and Unreported Loans The loan sample consists of 304 loans reported in the financia l press, 166 loans reported in the wire, and 905 non-reported syn dicated bank loans activated between January 1996 and December 2004. The public debt sample consists of 355 issues, and the non-bank private debt sample of 92 issues, also activated between Ja nuary 1996 and December 2004. The number of security issues randomly selected is determined so that each year the proportion of security issues in the sample equals the corresponding annual propor tion of Loans Reported in Press Before Activation N=84 Loans Reported in Press On or After Activation N=220 DJ & PR Wire Reported Loans N=166 Non-reported Loans N=905 Mean MedianMeanMedianMeanMedianMeanMedian Assets (millions) 3680.28 504.87*2355 .06*500.84*4119.3480. 32*4688.93885.13 Sales (millions) 2261.74* 362.13*1861 .3*499.22*3066.3530.59*4027.94819.96 Tangible/Assets 0.36 0.300.330.260.30*0.24*0.360.29 Earnings Surprise (%) 0.85 0.38*1.040.26*0.530.180.550.14 Earnings Dispersion (%) 0.40 0.17*+0.250.10*0.300.11*0.280.07 Debt/Assets 0.35* 0.36*0.33*0.34*0.290.270.280.27 Debt/EBITDA 3.84* 3.08*3.312.57*3.252.223.012.07 EBITDA/Assets 0.12* 0.12*0.12*0.12*0.12*0.130.140.13 Credit Rating 0.45 00.39*00.4300.470 Investment Credit Rating 0.23* 00.18*00.30*00.380 Firms 12-month cum. return 0.096 0.030.270.070.260.10.190.11 Commitment (millions) 448.92 216.73*392.55175565.79192.22400.33150 Commitment/ Assets 0.69*+ 0.44*+0.44*0.36*0.42*0.29*0.260.16 Maturity (months) 51.65* 53.86*46*42.49*43.60*37.5*35.0931.43 Covenant Intensity Index 4.42* 5*4.53*4*3.71*3*3.083 Restructuring 0.30* 00.32*00.1600.120 All-in-drawn spread 234.2*+ 238.6*+204.31*181.88*181.88*160*133.67100

PAGE 31

31loans in DealScan database, and public debt and non-bank private debt deals in Securities Data Corporation (SDC). In addition, the randomly selected issues maintain the overall proportions of total number of security issues for each financing type with respect to the other two, as reported in DealScan and (SDC). I identify reported and non-reporte d deals searching Dow Jones sources for news a nd wire articles that contain the deal amount and/or the key words used in the literatur e on loan, public debt and non-bank privat e debt announcements. Tangible assets are plant, property a nd equipment divided by assets. Earnings Surprise is defined as the absolute value of the difference between the median quarterly earnings estimate and the actual earnings per share, normalized by the sto ck price. Earnings Dispersion is defined as the standard deviation of earning forecasts normalized by the stock price. Credit Rating is an indicator variable equal to one if the fi rm has a debt rating, zero otherwise. The Investment Grade Rating is an indicator variable equal to one if the firm has an existing debt rating of BBB or higher, zero otherwise. Commitment is the size of the loan commitment. Maturity is the maturity of the loan agreement as reported in Dealscan. The covenant intensity index equals the sum of six covenant indicators (collateral, dividend restriction, more than two financial covenants, asset sales sweep, equ ity issuance sweep, and debt issuance sweep) when none of the six indicators is missing. All-in-drawn spread is calculated and reported by Dealscan as the total borrowing cost of the drawn portion of a loan over and above LIBOR. The maturity and DealScan all-in-drawn spread are calculated as the weighted average of the maturity and spread of each facility in the package loan. Significantly different from non-reported loan issue sample at the 0.1 level. + Significantly different from sample of reported loans on or after deal activation date at the 0.1 level.

PAGE 32

32Table 2-4. Firm and Deal Summary St atistics for Reported and Unreported P ublic Debt and Non-Bank Private Debt The loan sample consists of 304 loans reported in the financia l press, 166 loans reported in the wire, and 905 non-reported syn dicated bank loans activated between January 1996 and December 2004. The public debt sample consists of 355 issues, and the non-bank private debt sample of 92 issues, also activated between Ja nuary 1996 and December 2004. The number of security issues randomly selected is determined so that each year the proportion of security issues in the sample equals the corresponding annual propor tion of loans in DealScan database, and public debt and non-bank private debt deals in Securities Data Corporation (SDC). In addition, the randomly selected issues maintain the overall proportions of total number of security issues for each financing type with respe ct to the other two, as reported in DealScan and (SDC). I identify reported and non-reporte d deals searching Dow Jones sources for news a nd wire articles that contain the deal amount and/or the key words used in the literatur e on loan, public debt and non-bank privat e debt announcements. Tangible assets are plant, property a nd equipment divided by assets. Credit Rating is an indicator variable equal to one if the firm has a debt rating, zero otherwise. The Investment Grade Rating is an indicator variable equal to one if the firm has an existing debt rating of BBB or higher, zero otherwise. Commitment is the size of the loan commitment. Significantly different from non-repor ted issue sample at the 0.1 level. Reported Non-Bank Private Debt N=42 Non-reported NonBank Private Debt N=50 Reported Public Debt N=207 Non-reported Public Debt N=148 MeanMedianMeanMedianMeanMedianMeanMedian Assets (millions) 2303.2481.16*15983.7 3 867.5322389.96852.4419453.57609.65 Sales (millions) 2393.96585.956877.85673.218460.86120.517626.38358.2 Tangible/Assets 0.40.330.420.390.40.380.430.41 Debt/Assets 0.290.270.280.280.350.330.340.32 Debt/EBITDA 2.731.622.292.112.962.292.482.05 EBITDA/Assets 0.140.130.150.130.15*0.140.170.16 Credit Rating 0.36*00.5700.97*00.930 Investment Credit Rating 0.14*00.4300.83*00.750 Firms 12-month cum. return 0.160.120.240.170.21*0.140.120.09 Commitment (millions) 55.86*48.25*37.9425383.92*250110.6236 Commitment/Assets 0.23*0.1*0.070.030.100.040.050.004

PAGE 33

33Table 2-5. Market Reaction Summary Statistics The sample consists of 240 loan news and 1071 non-reported bank loan agreements, 199 reported and 146 non-reported public debt issues, and 38 reported and 43 non-reported non-bank private debt issues activated between Janua ry 1996 and December 2004, with all news being free of contamination. Nonconfounding news means that the loan information in the press articles are not bundled with other non-loan related information such as earnings, dividends, changes in management or completion of acquisitions. I identify reported and non-reported issues searching Dow Jones sources for articles containing the issue am ount and/or the key words used in the literature of bank loan, pub lic debt and non-bank private debt announcements. CAR and SPE refer to the two-day Cumulative Abnormal Returns and Standardized Prediction Errors of a 2-day window event study centere d on the earliest of news date and dea l activation date. Z value is calculated as th e square root of the number of observations times the average Standardized Predicti on Error. Statistically different at 10% level from non-reported subsample. + Statistically different at 10% le vel from reported subsample that does not precede lo an activation. Non-Confounding News Before Loan Activation Non-Confounding News On Loan Activation Non-Confounding News After Loan Activation Non-reported Deals Mean MedianMeanMedianMeanMedianMeanMedian Loans N=53N=37N=150N=1071 CAR (%) 2.56*+ 0.130.0330.260.460.160.25-0.06 Average SPE 0.32 0.009 0.077 0.05 Z value 2.33 0.05 0.96 1.64 Public Debt N=52 N=111 N=36 N=146 CAR (%) 0.58 0.33-0.11-0.11.210.56-0.18-0.001 Average SPE 0.21 -0.03 0.39 -0.07 Z value 1.51 -0.31 2.34 -0.85 Non-Bank Private Debt N=24 N=3 N=11 N=43 CAR (%) 1.12 0.003-3.49-4.54-0.38-0.421.67-0.38 Average SPE 0.11 -0.67 -0.11 0.05 Z value 0.55 -1.2 -0.36 0.33

PAGE 34

34 CHAPTER 3 DETERMINANTS OF LOAN, PUBLIC DE BT, AND NON-BANK PRIVATE DEBT ANNOUNCEMENTS The substantial heterogeneity in firm a nd loan characteristic s shown in Table 3-1 evidences that reported loans and their borrowers constitute a non-repres entative subsample of the entire DealScan universe of loans and thei r public borrowers. Furthe rmore, the differences between reported and non-reported loans may argua bly involve differences in loan information content that would explain the noteworthiness of reported loans and help identify the criteria used by press editors. In addition, the widespread use of bank loans (94% of public firms borrow from banks) and the fact that about 60% of publ ic firms issue public debt (Sufi (2007)) also raises the question of whether press editors follow the same firm and deal characteristics related criteria when deciding which public debt and non-bank private debt issues to publish, as compared to loan reporting. Determinants of Bank Loan Anno uncements A large theoretical literature in banki ng focuses on banks as screeners that reduce ex ante information asymmetries when compared with public arms length debt (Diamond (1991), Fama (1985), and Ramakrishnan and Thakor (1984)). Furthermore, the banking literature argues that loans constitute a unique source of financing because, among other reasons, banks have access to information that is not available to other lenders and market participants while screening loans and monitoring bor rowers (Rajan (1992)). So, if non-reported loans require less access to private information due to their less se vere information asymmetries, one could argue that the lenders of non-reporte d loans do not know significantly more than other market participants about their non-reported borrowers Consequently, those loans would constitute a weaker signal of firm quality a nd be, therefore, less noteworthy.

PAGE 35

35 Previous empirical studies document that mo re opaque firms are more likely to use bank debt (Houston and James (1996), Johnson (1997), and Sufi (2007))1 when the returns of the borrower decrease with respect to the market (H adlock and James (2002)), and that relatively riskier firms also choose bank debt over public debt (Denis and Mihov (2003)). Moreover, and in relation to disclosure decisions, Bagnoli and Wa tts (2007) find that firms are more likely to disclose private information when the operating performance in financial reports falls below expectations and/or does not contain sufficient good news. Thus, besides measures of asymmetric information, market timing and le verage, EBITDA to assets ratio could also constitute a significant dete rminant of reporting likelihood. Roberts and Sufi (2007) repor t that, although over 90% of l ong term loan contracts are renegotiated prior to their stated maturity, only 16% of the renegotiati ons are due to default events such as covenant viola tions. In addition, a related study by Nini et al. ( 2008) shows that following loan covenant violations and subsequent waivers there is an efficient reduction in firm investment and subsequent increases in market valuation and operation performance. Thus, one could argue that loan contract s following covenant violations could also be viewed as a significant signal of firm potential and be, therefore, more notewor thy. Finally, and in relation to loan characteristics that could make reporting more likely, Verrecchia and Weber (2006) find that firms are less likely to withhold information in material contract filings when they issue long-term debt. This suggests that, besides larger loans, loans with longer maturity could also be more newsworthy. Therefore, since summary statistics show heterogeneity in all the above-mentioned borrower and loan characteristics suspected to determine reporting likelihood, the analysis of 1 Recent studies by Sufi (2007 ) find increased access to capital of less in formed investors follo wing the introduction of bank loan ratings as well as more concentrated syndicated loans in the ca se of opaque borrowers.

PAGE 36

36 bank loan reporting determinants includes proxie s for information asymmetries, credit risk, market timing, loan restructuring following viola tions, loan size, and loan maturity. In addition, one could argue that the largest fi rms are less opaque and the smallest firms of less interest to the general investor, and that, cons equently, their loans could be less likely to be reported. Thus, given that the likelihood of repor ting is suspected to have a no n-linear dependency with respect to borrower size, the analysis also considers a medium-firm-size dummy variable. This dummy equals one if the firm has assets of less than $1 billion and more than $400 million in order to include non-reported borrowers with below-me dian firm size and reported borrowers with above-median firm size. The analysis of loan-reporting determinants is based on a series of probit models. Table 3-1 reports the regression results and marginal effects for the two main probit models. The first model studies the likelih ood that a loan is deemed noteworth y through a press article or a Dow Jones wire article that is not made press news. The dependent va riable equals one if the bank loan is reported in the financia l press or on the Dow Jones wire but not in the press, and zero otherwise. The second regression analyzes how the factors that determ ine the likelihood of reporting can also affect the timing of the repor ting, i.e., the likelihood that a loan is reported before its activation date. As shown in Table 3-1, higher measures of information asymmetries and credit risk increase the likelihood of loan reporting. These de terminants are measured by tangibles relative to assets and EBITDA to assets, al though other proxies used in the literature, such as volatility, and surprise and dispersion in earnings forecasts lead to analogous results. In addition, loan restructurings, loans with longer maturities, and bigger loans relative to assets are also more likely to be reported in the press. More specifically, I find that, evaluated at the sample means of

PAGE 37

37 the other explanatory variables, the likelihood of a loan being reported is 13 percentage points lower with higher tangibles relative to assets, 41 percentage points lower with higher measures of EBITDA to assets, 21 percen tage points higher in the ca se of loan restructurings, 17 percentage points higher when the commitment amount relative to firm size is bigger, and 11 percentage points higher when the borrower is a medium-size firm. It is important to note that these are also the determinants that reporters of loan news in Dow Jones cite when asked on the phone about the criteria followed to determine whether a particular loan is newsworthy. Thus, both the analysis and anecdotal ev idence suggest that reported loan s are more surprising and also informative about firm potential. Beyond the average marginal effects reported in Table 3-2 reports how the lowest and highest quartile measures of loan-reporting de terminants in the 1,375 loan sample alter the likelihood of reporting evaluated at the sample means of the othe r explanatory variables. More specifically, the likelihood of re porting decreases by 10 percentage points for the firms within the highest quartile of tangibles to assets, increases by 8 percentage points (decreases by 7 percentage points) for the firms with EBITDA to a ssets within the lowest (highest) quartiles of the sample, and increases by 20 percentage poin ts (decreases by 17 percentage points) for the firms with loan commitment to assets within the highest (lowest) quartiles of the sample. The analysis of the loan reporting determinan ts also shows that the likelihood of loan reporting is increases throughout the study period This could be arguably be due to lower reporting costs thanks to developments in information technology, and suggests that the information content of the av erage reported loan decreases during the 1996 through 2004 period. In order to confirm the non-lin ear dependency of loan reporti ng with respect to firm size, I apply the reporting likelihood prob it model to the firms in the sample that fall within the

PAGE 38

38 smallest biggest size quartile in terms of assets (the regressi on results are not reported in the paper). As expected, I find that when the loan s are surprising and informative given loan characteristics, information asymmetries, and bo rrower credit risk but the borrower is a small firm, the wire article that is not followed by a pr ess article is usually a press release wire. Thus, once could argue that loans to smaller borrowers ar e less likely to draw attention from the press because those firms are of less interest to the ge neral investor. On the other hand, in the case of the biggest firms, which receive greater analys t coverage, I find that not all their loans are systematically reported. The loans obtained by the biggest firms are reported when credit risk and loan characteristics justify it, given th at these borrowers are generally less opaque. The second probit model reported in Ta ble 3-1 studies whether loan-reporting determinants also influence the timing of the loan news with respect to loan activation. Overall, I find no evidence supporting the hypothesis that firm characteristics affect the timing of the loan news. However, the size of the loan commitme nt relative to borrower size increases the likelihood that the loan is re ported before loan activation. Overall, only the loans that seem more in formative about the potential of the borrower are reported. This evidences deal variety in the loan market that had not been documented before. Furthermore, the heterogeneity in in formation asymmetries and credit risk that distinguishes reported from non-reported cases is insi ghtful because it also determines the choice of reported bank debt over public debt (Hadlock and James (2001)). In consequence, one could argue that non-reported loans ar e more transactional, contrary to the belief that bank loans always involve a more or less close re lationship between borrower and lender. Determinants of Public Debt and Non-Bank Private Debt The news search for the samples of 355 pub lic debt issues and 92 non-bank private debt issues indicates that most public debt deals are reported in the press vs. 57% of the non-bank

PAGE 39

39 private debt deals. In addition, the news search shows that almo st 20% of non-bank private deals are reported by the borrower, while in the case of public debt reporting by the firm is rare and usually related to debt repayment. Thus, in the case of non-bank priv ate deals, either the borrowers or the lenders seem to want investor s to know about the deals. This is interesting given that previous studies document a negative market reaction to non-bank private debt (Eckbo (2007)). The borrower and lender incentives to report non-bank private debt despite the adverse reaction of the market could be explained th rough the findings of Denis and Mihov (2003) and Carey el al. (1998) that associate non-bank private lenders to the financing of the riskiest borrowers. It seems that the good news of gettin g financing should increa se the likelihood that non-bank private debt is announced. However, only banks are considered to be special in terms of access and generation of private information and efficient monitori ng (Fama (1985), Rajan (1992) and Denis and Mihov (2003)). On the other hand, in the cas e of reporting of public debt issues, if those borrowers are significantly less risky and opaque than bank borrowers, one may argue that the likelihood of reporting would decrease at least among the best performers. Thus, how do the determinants of bank loan reporting differ from those of public debt and non-bank private debt? The analysis of the likelihood of public debt and non-bank private debt reporting through a press article or a Dow Jones wi re article is shown in Table 3-3. The probit analysis for non-bank privat e debt shows that, besides information asymmetries, credit risk does not seem to play a significant role in the likelihood of reporting of non-bank private debt issues (although it plays it in the choice of non-bank private debt borrowing, as Dennis and Mihov find (2003)). The r eason could be that a lthough this type of financing serves the niche of hi ghest risk borrowers that cannot obtain financing from loans, the

PAGE 40

40 lenders do not necessarily base the agreements on private information about firm potential, but on control rights. In relation to measures of risk, Table 33 also shows that a lthough the size of the proceeds relative for firm size does not affect the likelihood of reporting for public debt borrowers, non-bank private debt issues are 125 percentage points more likely to be reported as proceeds relative to firm size increase when evaluated at the sample means of the other explanatory variables. The reason co uld be that public debt borrowe rs are not expected to default on the agreements, while non-bank private borrowers could. Thus, bigger loans granted to riskier borrowers may be considered good news because the lenders have positive private information about the borrower, while bigger lending to the ri skiest firms is negative news, since it may just imply additional risk and a higher probability of default. In addition, the likelihood of public debt re porting is found to be 22 percentage points higher in the case of firms with higher cumulative returns when evaluated at the sample means of the other explanatory variables. This is consistent with market sources echoing the new public debt issues of the market darlings, which woul d not require revelation of any particular insight regarding firm potential, and w ould lead to an average null market reaction to the news. Moreover, and as expected, the likelihood of public debt reporting is 85 percentage points lower in the case of firms with better operating perfor mance, and therefore, le ss risky, given that, as mentioned before, public debt borrowers are less opaque. Finally, the probit analysis show s that public debt issues, like loans, are more likely to be reported over time, but that there is no significa nt time trend in the reporting of non-bank private debt. The reason could be that while public debt issues are reported in Dow Jones, non-bank private issues are announced mainly in specialized publications such as Private Placement Letter

PAGE 41

41 and Private Placement Reporter Thus, it may have become less costly over time to report loans due to advancements in information technology in the case of Dow Jones, but not in the case of other smaller news sources that do not benefit from economies of scale.

PAGE 42

42 Table 3-1. Determinants of the Likelihood a nd Timing of Bank Loan Reporting in the Press Table 3-1 provides the estimates of two probit models that relate firm and loan characteristics to the likelihood that the press deems a loan notewor thy through a press article or a DJ wire, and the likelihood that when there is loan news th e press article precedes th e loan activation date. Both the results and marginal effects of the probi t regressions are reported. The analysis is based on a sample of 304 press-reported, 166 PR or DJ wire reported loans and 905 non-reported loans activated between January 1996 and December 2004. I identify reported and non-reported loans searching the Dow Jones sources for articles containing the deal amount and/or the key words used in the literature on loan announcements. Commitment Amount is the size of the loan commitment. Maturity is the maturity in months of the lo an agreement as reported in Dealscan. Medium-size firm is a dummy variable that equals one if the firm has less than $1 billion in assets and more than $400 million in assets. Z st atistics are reported in parenthesis. Likelihood of loan reporting Likelihood of reporting before loan activation Marginal Effects Marginal Effects Time trend 0.060.02-0.01 -0.004 (3.12) (-0.31) Tangibles/Assets -0.39 -0.130.27 0.09 (-2.18) (0.78) Debt/Assets 0.30 0.100.09 0.03 (1.40) (0.22) EBITDA/Assets -1.2 -0.41-0.94 -0.3 (-2.54) (-1.09) Restructure Loan 0.56 0.21-0.03 -0.01 (5.32) (-0.17) Commitment Amount/Assets 0.51 0.170.41 0.13 (3.01) (2.47) Maturity 0.01 0.0030.003 0.001 (4.52) (1.06) Medium-size firm 0.32 0.11-0.16 -0.05 (3.84) (-0.90) Firms past 12-month cumulative return -0.03 -0.01-0.09 -0.03 (-0.56) (-1.22) Markets past 12-month cumulative return 0.01 0.01-0.61 -0.19 (0.08) (-1.63) Constant -1.32 -0.71 (-8.18) (-2.14) Pseudo R2 0.11 0.05 N 1176 294

PAGE 43

43 Table 3-2. Marginal Effects of the Lowest and Highest Quartile Determinants on Loan Reporting Table 3-2 reports how the lowest and highest quartiles of some fi rm and loan characteristics in the sample significantly affect the marginal e ffects of a loan being deemed noteworthy by the press through a press article or a DJ wire, as examined in the first Probit model of Table 3-1. *Significant at the 0.1 level. Effect of Lowest Quartile Effect of Highest Quartile On Reporting Likelihood On Reporting Likelihood Tangibles/Assets 0.05-0.1* Debt/Assets -0.06*-0.01 EBITDA/Assets 0.08*-0.07* Restructure Loan -0.21*0.21* Commitment Amount/Assets -0.17*0.2* Maturity -0.05-0.04 Over $1Billion vs. Below $400 Million -0.04-0.11*

PAGE 44

44 Table 3-3. Determinants of the Likelihood of Public and Non-Bank Private Debt Reporting Table 3-3 provides the estimates of a probit model relating firm and debt characteristics to the likelihood that the press deems a public debt or non-bank private debt noteworthy through a press articles or a DJ wire. The analysis is based on a sample of 92 non-bank private debt agreements and 355 public debt agreements ac tivated between January 1996 and December 2004. I identify reported and non-reported agreem ents searching the Dow Jones sources for articles containing the key words us ed in the announcements literature. Medium-size firm is a dummy variable that equals one if the firm has less than $1 billion in assets and more than $400 million in assets. Z statistics ar e reported in parenthesis. Probit Reporting Non-Bank Private Debt Public Debt Marginal Effects Marginal Effects Time trend 0.040.010.160.06 (0.61) (5.24) Tangibles/Assets -0.54-0.22-0.36-0.14 (-0.96) (-1.17) Debt/Assets 0.840.330.250.10 (1.13) (0.50) EBITDA/Assets 1.720.69-2.21-0.85 (0.86) (-2.05) Proceeds/Assets 3.121.250.460.18 (2.38) (0.69) Medium-size firm 0.080.030.180.07 (0.28) (0.63) Firms past 12-month cumulative return -0.13-0.050.570.22 (-0.48) (2.81) Markets past 12-month cumulative return -0.36-0.14-0.13-0.05 (-0.51) (-0.38) Constant -0.91 -0.30 (-1.49) (-0.86) Pseudo R2 0.11 0.10 N 92 355

PAGE 45

45 CHAPTER 4 STOCK PRICE REACTION TO LOAN NEWS. MULTIVARIATE ANALYSIS Among the several explanations for the genera l pattern of relative stock price effects in event studies, Smith (1986) cites the unanticip ated announcement hypothesis, which states that the more predictable the issuance and its struct ure, the smaller the associated stock price change.1 Another explanation mentioned in the stu dy of the differences in announcement effects across security types is the importance of asymme tric information problems. For example, since debt and preferred stock are more senior claims, their valuation is expected to be less sensitive than common stock to changes in firm value. In addition, James and Smith (2000) suggest a third explanation for why bank loan announcements by large and medium-size companies elicit a positive stock price reaction. According to their re sults, commitment-based financing is used by larger companies when they believe themselves to be undervalued or when interest rates are too high.2 In relation to loan announcements, and given th at loan news articles are very brief, it could be also argued that the articles per se are not informative. It could be argued that loan articles mainly signal the exis tence of a noteworthy loan wit hout giving further information about the circumstances that make the loan not eworthy. Thus, the study of the magnitude of the market reaction to loan news should consider the factors that make reported loans noteworthy, i.e., credit risk, information asymmetries, loan characteristics and potential market timing hypothesis. The analysis of the stock price reaction to the ea rliest loan news vs. the reaction to the loan activation of non-reported deals is presented in Table 4-1. It uses an OLS regression model that 1 See Marsh (1982) for evidence on the use of short-term debt and how to predict public debt issues. 2 For example, and at the time of the Asian crisis in th e last quarter of 1998, $10 billion of commercial paper was retired and $20 billion of net new commercial loans were booked.

PAGE 46

46 studies the likelihood of a higher z -value (average standardized pr ediction errors times the square root of the number of observati ons) in relation to whether the lo an is reported a nd other potential determinants of market reaction identified in the study of repor ting determinants (Table 3-1). Similarly, Table 3-2 presents the results of the OLS analysis of the market reaction for public debt and non-bank pr ivate debt issues. As shown in Table 3-1, the market reaction is driven by the size of the issue and market timing considerations. More specif ically, I find that the reaction to loan news increases with lower measures of cumulative returns of the lo an borrower relative to the market during the 12 months preceding the earliest of the news or loan activation of re ported loans. This is consistent with previous work by Hadlock and James ( 2002), who find evidence in support of a market timing argument by which firms tend to choose announced bank lending over public debt when their returns decrease with respect to the market. The analysis of the market reaction for public debt and nonbank private debt news and issues is, as expected, inconclusive given th at no significant market reaction is observed, in general, for these two types for issues surround ing news articles or activation of non-reported issues. As such, and in the case of public debt issues, I find no evidence in support of the market timing hypothesis, like suggested by the firm charact eristics in Table 2-3. However, the size of the issue relative to firm size still appears to have a positiv e effect on the market reaction, especially in the case of public debt issues reported afte r activation, altho ugh the significance level is a little over 10%. Fina lly, and although the market reacti on is not significant either for reported and non-reported non-bank privat e debt issues, I find that tangi bles relative to assets can increase the market reaction, although this firm characteristic does not differentiate reported from non-reported non-bank private debt borrowers.

PAGE 47

47 Table 4-1. Determinants of the Stock Price Reaction Surrounding Loan Reports and Activation Non-Confounding Loan News vs. NonReported Loans Non-Confounding Loans News Before Loan Activation vs. Non-Reported Loans Press-reported -0.51 -0.56 (-0.26) (-0.23) Time Trend -0.03 -0.09 (-0.09) (-0.2) Tangible/Assets 0.08 -0.21 (0.02) (-0.05) Debt/Assets 4.76 5.86 (1.04) (1.11) EBITDA/Assets -3.28 0.72 (-0.35) (0.07) Restructure -3.59 -4.45 (-1.15) (-1.08) Commitment Amount /Assets 5.96 5.95 (3.01) (2.16) Maturity -0.03 -0.02 (-0.66) (-0.35) Medium-size firm 1.75 1.62 (0.52) (0.38) Firms past 12-month cumulative return -1.84 -3.3 (-1.72) (-2.11) Markets past 12-month cumulative return 9.43 11.11 (2.13) (2.12) Constant -0.88 -1.53 (-0.27) (-0.4) N 1196 1014 Table 4-1 provides the estimates of two OLS regressions relating the magnitude of the market reaction to loan news to reporting and firm and loan characteristics The market reaction is measured as the Standardized Prediction Erro rs for a two day event window centered on the earliest of the announcement date or activation date. The analysis is based on a sample of 240 non-confounding press news (53 precede the activation of the loan) and 1071 non-reported loans activated between January 1996 and December 2004. I identify reported and non-reported loans searching the Dow Jones sources for articles containing the key words used in the announcements literature. Commitment Amount is the size of the loan commitment. Maturity is the maturity in months of the loan agreement as reported in Dealscan. Medium-size firm is a dummy variable that equals one if the firm has less than $1 billion in assets and more than $400 million in assets. t statistics are reported in parenthesis.

PAGE 48

48 Table 4-2. Determinants of the Stock Price Reaction Surrounding Public Debt and Non-Bank Private Debt Reports and Activation Non-Bank Private DebtPublic Deb t Non-Confounding News vs. Non-Reported Issues Non-Confounding News vs. Non-Reported Issues Press-reported -0.32 2.66 (-0.21) (1.58) Time Trend -0.01 -0.53 (-0.05) (-1.58) Tangible/Assets 7.95 -2.43 (2.44) (-0.7) Debt/Assets -1.61 -2.1 (-0.3) (-0.41) EBITDA/Assets -14.06 2.73 (-1.54) (0.26) Commitment Amount /Assets -4.1 1.23 (-1.56) (1.69) Medium-size firm -2.18 1.7 (-1.36) (0.61) Firms past 12-month cumulative return -2.73 0.45 (-1.46) (0.23) Markets past 12-month cumulative return -3.7 1.31 (-1.03) (0.31) Constant 1.9 2.84 (0.52) (0.69) N 81 343 Table 4-2 provides the estimates of two OLS regr essions relating the magn itude of the market reaction to public debt and non-bank private debt news to reporting and firm and loan characteristics The market reaction is measured as the Standardized Prediction Errors for a two day event window centered on the earliest of the announcement date or activation date. The analysis is based on a sample of 199 non-confounding public debt news, 146 non-reported public debt issues, 38 non-confounding non-bank private debt news, and 43 non-reported nonbank private debt issues activated between January 1996 and December 2004. I identify reported and non-reported agreements searching the Dow Jones sources for articles containing the key words used in the announcements literature. Commitment Amount is the size of the loan commitment. Maturity is the maturity in months of th e loan agreement as reported in Dealscan. Medium-size firm is a dummy variable that equals one if the firm has less than $1 billion in assets and more than $400 million in asse ts. t statistics are repo rted in parenthesis.

PAGE 49

49 CHAPTER 5 DETERMINANTS OF INCREMENTAL FINANCING OPTIONS Previous empirical studies of the determinan ts of incremental financing choices are based on samples of bank loans that get reported in the press. Denis and Mihov (2003), Hadlock and James (2002), Houston and James (1996), Johnson (1997), and Cantillo and Wright (2000) find that smaller, younger, more levered firms ar e more likely to rais e capital via bank debt.1. In addition, Hadlock and James (2002) document that higher volatility of returns and lower market to book and market adjusted cumulative returns during the year prio r increase the pr obability of a bank agreement over a public debt issue. More recently, Bradley and Roberts (2004) report that bank borrowers include more covenants in thei r debt agreements and have greater growth opportunities and more volatile cash flow s than firms that issue public debt.2 These findings are insightful because they differentiate bank borrowe rs from bond borrowers using the same firm characteristics that I have found differentiate bor rowers with reported loans from those without. Furthermore, if as I find, reported loans are not repr esentative of the entire universe of loans, and non-reported loans are suspected to have a more transactional nature, the determinants of the choice between bank loans, public debt, and non-bank private could change when non-reported loan borrowers are also considered.3 1 In relation to agreement size, Smith (1986) specifies that firms are expected to use bank lines of credit until an efficient public issue size is reached. Th e reason for this size criterion is that flotation costs for public debt issues have a larger fixed component and more pronounced economies of scale than does bank debt. In addition, non-bank private loans tend to have lower flotation costs than public issues and custom-designed covenants. 2 Although banks are considered the most efficient liquidity providers in the economy (Kasyap et al. (2002) and Gatev et al. (2006)), firms are restrict ed access to bank credit if they do no t maintain high enough cash flows (Sufi (2007)). One explanation for this finding could be the use of marginal vs. book firm characteristics. 3 Regarding the choice between debt and equity, Gomes and Phillips (2005) find that firms with the greatest asymmetric information problems tend to borrow from private equity, while the most transparent firms seem to issue public equity, and those that are relatively more opaque prefer bank debt to public debt.

PAGE 50

50 Bank debt and non-bank private debt differ in terms of regulatory requirements, maturity, costs, placement structure, and credit or concentration (C arey et al. (1998)).4 Overall, since bank and non-bank private debt offer higher flexibility of renegotiation, it should be optimal for the riskiest borrowers to borrow privat ely instead of issuing public debt.5 In this context, Chemmanur and Fulghieri (1994) predict that firm s with high and low credit ratings use public debt, while firms with intermediate ratings us e bank loans. However, Denis and Mihov (2003) document that while firms with the highest credit quality are more likely to issue public debt, and firms with medium credit quality establish bank loans, those with the lowest credit quality establish non-bank private debt agreements.6 Thus, as Rajan (1992) argues, the cost of bank monitoring may outweigh the benefits for low-qual ity firms, and in consequence, low-quality firms are more likely to issue nonbank private debt than bank debt. To examine the determinants of the choi ce between bank debt, public debt and non-bank private debt in a multivariate set ting, I use a series of four multi nomial logistic regressions. Table 5-1 reports the regression results, and Table 5-2 the marginal effects. The first multinomial regression studies the choice be tween bank borrowing that is reported, public, and non-bank private debt, while the third and fourth regres sions do the same consid ering all loans and only non-reported loans, resp ectively. The second regression studi es the choice between reported issues of bank, public and non-bank private debt. 4 See Bayless and Chaplinsky (1992), Chaplinsky and Hansen (1993), and Jung, Kim, and Stulz (1996) for the choice between public debt and equity. 5 Sufi (2005) documents that firms with higher lagged ear nings to assets are more likely to establish bank loan agreements as opposed to arms length debt. 6 More specifically, Denis and Mihov (2003) find that bank and non-bank private borrowers present significantly lower returns on assets, lower proportions of rated debt and credit quality when credit ratings are available, and higher growth opportunities in terms of growth in ca pital expenditures, sales, and number of employees.

PAGE 51

51 The results in Table 5-1 and Table 5-2 show that, evaluated at the sample means of the other explanatory variables, repor ted loans are 26 percentage points less likely to be chosen over public and non-bank private debt with higher measur es of tangibles to assets, and 23 percentage points more likely to be chosen over public a nd non-bank private debt if the borrower is a medium-size firm. In terms of credit risk, the choice of bank borrowing over public and nonbank private debt borrowing decreases with higher measures of EBITDA to assets independently of whether the loans are reported or not. More specifically, I find that, evaluated at the sample means of the other explanatory variables, reported loans are 170 percentage points less likely to be chosen over public and non-bank private de bt with increases in EBITDA to assets. In addition, I find that the likelihood th at more leveraged firms relati ve to assets choose bank debt is 10 percentage points lower for bank loans. In terms of the effect of pr oceeds relative to assets on incr emental financing choices, and evaluated at the sample means of the other expl anatory variables, firms with higher financial needs relative to assets are 120 percentage poi nts more likely to choose bank debt over public and non-bank private debt if only reported loans are considered. Overall, the study of incremental financing decisions does not vary significantly when considering only reported loans instead of both re ported and non-reported loans. In other words, whether reported deals or all deals are consider ed, the results are similar and consistent with those of previous studies by Hadlock and Ja mes (2002), and Denis a nd Mihov (2003). Riskier more opaque borrowers are more likely to borrow from banks instead of public debt, and nonbank private debt has a niche among the riskier bo rrowers. This suggests that, even if they are less noteworthy and more transactional, non-reported loans are sti ll somehow special and different from public debt. Moreover, since Deal Scan covers only syndicated loans, this study

PAGE 52

52 may be excluding a significant pr oportion of smaller firms that borrow from one lender. Those firms, as Houston and James (1996) find, are more likely to be subject to information monopolies, and have greater incentives to time the market due to more severe asymmetric information problems. This could explain the lack for evidence in the analysis for the market timing hypothesis.

PAGE 53

53 Table 5-1. Multinomial Logistic Regressi ons on Incremental Financing Decisions (1.a)(1.b)(2.a)(2.b)(3.a) (3.b) Time trend -0.14-0.07-0.06-0.05-0.08 -0.02 (-2.92)(-1.35)(-1.08)(-0.67)(-2.88) (-0.41) Tangibles/Assets 1.021.210.921.20.28 0.93 (2.17)(2.0)(1.79)(1.42)(0.85) (1.75) Debt/Assets 1.2-1.851.02-1.661.52 -0.54 (1.84)(-1.87)(1.58)(-1.3)(4.25) (-0.82) EBITDA/Assets 7.653.656.942.984.81 1.44 (4.87)(1.92)(3.9)(1.22)(5.28) (1.03) Deal Size/Assets -4.81-4.74-4.76-2.7-6.03 -4.53 (-2.04)(-2.58)(-1.51)(-1.48)(-2.18) (-2.36) Medium-size firm -1.810.94-1.750.7-1.17 1.39 (-4.99)(2.51)(-3.97)(1.52)(-4.31) (4.47) Firms past 12-month cumulative return 0.070.090.110.020.07 0.05 (0.74)(0.78)(1.13)(0.13)(0.74) (0.45) Markets past 12-month cumulative return -0.46-0.64-0.51-1.12-0.55 -0.46 (-1.05)(-1.11)(-1.07)(-1.35)(-1.88) (-0.93) Constant 0.66-0.42-0.02-1.37-0.91 -2.72 (1.37)(-0.82)(-0.03)(-2.06)(-2.27) (-6.47) Pseudo R2 0.30 0.30 0.19 N 755 557 1822 Table 5-1 provides the estimates of a series of four multinomial logistic regressions relating the choice of bank debt over public de bt and non-bank private debt, as well as the likelihood and timing of loan reporting in the context of increm ental financing decisions. The analysis is based on a sample of 304 press-reported, and 1071 non-repor ted syndicated bank loans, as well as 355 public debt and 92 non-bank privat e debt agreements, all activated between January 1996 and December 2004. The public debt and non-bank privat e debt random samples keep the proportion of total bank loans in DealScan to public debt and non-bank private debt issues in SDC, as well as the yearly proportions in the original SDC samples. I identif y reported and non-reported loan, public and non-bank issues searching the Dow J ones sources for articles containing the deal amount and/or the key words used in the announcements literature. Deal Size refers to the commitment amount in the case of bank loans an d proceeds in the case of public debt and nonbank private debt. Z statistics ar e reported in parenthesis. (1.a) Public Debt vs. Reported Bank Loans (1.b) Non-Bank Private Debt vs. Reported Bank Loans (2.a) Reported Public Debt vs. Reported Bank Loans (2.b) Reported Non-Bank Private Debt vs. Reported Bank Loans (3.a) Public Debt vs. Bank Loans (3.b) Non-Bank Private Debt vs. Bank Loans

PAGE 54

54 Table 5-2. Multinomial Logistic Regressions. Marginal Effects (1.a) (1.b) (1.c) (2.a) (2 .b) (2.c) (3.a) (3.b) (3.c) Time trend 0.03 -0.03 -0.001 0.01 -0.01 -0.002 0,01 -0.01 -0.001 (2.85) (-3.0) (-0.14) (1.26) (-1.2) (-0.47) (2.5) (-2.61) (-0.23) Tangibles/As sets -0.26 0.19 0.07 -0.22 0.15 0.07 -0.04 0.02 0.03 (-2.49) (1.73) (1.14) (-1.81) (1.27) (1.04) (-1.32) (0.64) (1.37) Debt/Assets -0.14 0.38 -0.24 -0.08 0.22 -0.14 -0.1 0.13 -0.02 (-0.94) (2.49) (-2.10) (-0.61) (1.85) (-1.38) (-1.84) (2.37) (-0.93) EBITDA/Ass ets -1.7 1.69 0.01 -1.31 1.22 0.08 -0.42 0.39 0.03 (-4.57) (5.27) (0.04) (-4.28) (4.24) (0.48) (-3.06) (2.81) (0.78) Proceeds/Ass ets 1.2 -0.95 -0.24 0.93 -0.83 -0.1 0.6 -0.48 -0.13 (2.27) (-1.9) (-2.02) (2.34) (-2.29) (-0.96) (8.17) (-8.35) (-3.98) Medium-size firm 0.23 -0.42 0.2 0.21 -0.29 0.09 0.02 -0.09 0.06 (2.84) (-5.43) (6.23) (1.66) (-2.27) (2.87) (0.53) (-1.91) (5.49) Firms past 12-month cumulative return -0.02 0.01 0.01 -0.02 0.02 -0.001 -0.01 0.01 0.001 (-0.84) (0.62) (0.55) (-1.03) (1.31) (-0.05) (-0.86) (0.79) (0.39) Markets past 12-month cumulative return 0.12 -0.08 -0.04 0.14 -0.07 -0.07 0.06 -0.04 -0.013 (1.21) (-0.84) (-0.78) (1.39) (-0.78) (-1.13) (1.68) (-1.45) (-0.77) Table 5-2 provides the marginal effects of a se ries of four multinomial logistic regressions relating the choice of bank debt over public debt and non-bank pr ivate debt. The analysis is based on a sample of 304 press-reported, and 10 71 non-reported syndicated bank loans, as well as 355 public debt and 92 non-bank private debt agreements, all activated between January 1996 and December 2004. The public debt and non-bank private debt random samples keep the proportion of total bank loans in DealScan to public debt and non-bank private debt issues in SDC, as well as the yearly proportions in the original SDC samples. I identify reported and nonreported loan, public and non-ba nk issues searching the Dow Jones sources for articles containing the deal amount a nd/or the key words used in the announcements literature. Deal Size refers to the commitment amount in the case of bank loans and proceeds in the case of public debt and non-bank private debt. Z sta tistics are reported in parenthesis. (1.a) Public Debt vs. Reported Bank Loans (1.b) Non-Bank Private Debt vs. Reported Bank Loans (2.a) Reported Public Debt vs. Reported Bank Loans (2.b) Reported Non-Bank Private Debt vs. Reported Bank Loans (3.a) Public Debt vs. Bank Loans (3.b) Non-Bank Private Debt vs. Bank Loans

PAGE 55

55 CHAPTER 6 LONG-TERM PERFORMANCE OF BANK BORROWERS As commented before, reported loans involve a closer relationshi p between lenders and borrowers, while non-reported loans are more trans action oriented and, to a certain extent, more equivalent to public debt. In consequence, one could argue that reported borrowers may improve their operating performance with respect to the performance of non-reported borrowers. In a related study, and since reported loans are more likely to be restructured loans following covenant violations, Nini et al (2008) show that, following cove nant violations, there is an effective reduction of capital expenditures that leads to higher perf ormance and valuation. In the study of long-term operating performan ce, I use two measures: EBITDA to assets, and debt to EBITDA. I focus on these two rati os because they are closely linked to the borrowers ability to service bot h current and future bank borrowings. While stock returns and net income are also important measures of perfor mance used in previous studies, they are more removed from the bankers principal focus. In a ddition, the debt to EBITDA ratio is present in about half the loans that include financial covenants. Thus, firm s have an incentive to improve this ratio. To limit the effect of outliers, I examine the medians of operating performance measures for a period of seven years that covers four fi scal years preceding loan activation and the three subsequent ones. The summary statistics for a fixe d sample of firms are presented in Table 6-11. Year 0 refers to the fiscal year prior to the activation of the loan. Overall, I observe in the balanced panel that reported borrowers present poorer operating performance with respect to non-reported loans du ring the last three fiscal years prior to the activation of the loan. However, and although th e operating performance measures of reported 1 I find similar results when I exclude unannounced loans for which there is 8-K filing.

PAGE 56

56 borrowers remain weaker than those of non-re ported borrowers for two years following loan activation, I observe no significant difference at the 5% level in EBITDA to assets during the third fiscal year following loan activation. Fu rthermore, it is important to note that the improvement in the operating pe rformance of reported borrowers would be more pronounced if the firms whose loans are reported mainly on the basis of loan size or maturity were excluded from the study sample.

PAGE 57

57 Table 6-1. Median Peer-Adjusted Measur es of Subsequent Operating Performance EBITDA/Assets (%) Debt/EBITDA (%) Press Reported Unreported Press Reported Unreported Year -3 0.13 0.143 2.15 1.8 Year -2 0.13 0.138 2.2* 1.86 Year -1 0.12* 0.136 2.8* 1.93 Year 0 0.12* 0.134 2.78* 2.07 Year 1 0.12* 0.129 3.02* 2.44 Year 2 0.12 0.124 2.8 2.5 Year 3 0.12 0.124 2.8 2.5 N 179 593 130 494 Table 6-1 presents summary statistics for a balanced panel of firms. Year refers to the number of fiscal years following the loan activation date. Year zero refers to the last fiscal year prior to the activation date. Significantly different from the unre ported loan median at the .05 level. + Significantly different from the unre ported loan median at the .1 level.

PAGE 58

58 CHAPTER 7 SUMMARY AND CONCLUSIONS An important strand of the banking literature holds that banks play a special role in the capital acquisition process through the close rela tionship they establish with their borrowers. More specifically, banks can gain access to priv ate information that is not available to other credit claimants and provide more flexible lendi ng conditions, which is of particular importance in the case of private firms. However, it is not clear why the more esta blished less opaque public firms would need to establish th e same type of unique relations hips with banks in all cases. Also, it is puzzling why there is a significant positive stock price react ion surrounding the press reporting of bank loans, given that practi cally all public firms have bank loans. To address this puzzle, I begin by examining the frequency and determinants of bank loan reporting in the financial pr ess. I find that only 22% of bank loans are reported, and that the subsample of reported loans and borrowers is not representative of th e entire population of syndicated loans and public bo rrowers. More specifically, I find that, besides bigger loans relative to assets, longer maturity loans, and loan restructurings following loan covenant violations are significantly more likely to be repo rted in the press. In addition, consistently also with a greater need for lender access to privat e information, I find that more opaque riskier borrowers are more likely to have their loans reported. Reported loans seem more informative about the potential of the borrower than nonreported ones, an heterogeneity that, to my knowledge, has not been documented before. However, since the cost of repor ting has decreased with technological developments, there is an increase in the frequency of reporting during the 1996 through 2004 study period that makes the information content of the average loan decrease overtime. In any case, the firm and deal factors that explain what makes a loan more likely to be reported are also those that explain the choice

PAGE 59

59 of bank borrowing over public debt borrowing. In consequence, one could argue that nonreported loans require a less close relation between lender and borrower and are, therefore, more transaction oriented, although somehow still sp ecial and different from public debt. In general, and although the loans in the sample are disclosed, the market reaction is only significant in the case of reported loans, especially when the earliest news precedes loan activation. This reaction to loan reporting when the news precedes loan activation is more intense when loans are bigger and when the perf ormance of the borrowers stock is poor with respect to the market during the year prior. Thus, although borrow ing is more difficult to obtain in the case of reported loans, which makes them more surprising, and therefore newsworthy, loan reporting also appears to time the market. Still, an d consistent with the view that bank loans are more informative about firm potential when they are reported in the pr ess, reported borrowers improve their operating performance with resp ect to non-reported one s over the three years following the activati on of the loan.

PAGE 60

60 LIST OF REFERENCES Andrade, Gregor, and Steven Kaplan, 1998, How co stly is financial ( not econom ic distress? Evidence from highly leveraged tran sactions that became distressed, Journal of Finance 53, 1443-1493. Barber, Brad M., and John D. Lyon, 1996, Detecting abnormal operating performance: The empirical power and specifi cation of test statistics, Journal of Financial Economics 41, 359-399. Bayless, Mark, and Susan Chaplinsky, 1991, Expecta tions of security type and the iinformation content of debt and equity offers, Journal of Financial Intermediation 1, 195-214. Best, Ronald, and Hang Zhang, 1993, Alternative information sources and the information content of bank loans, Journal of Finance 48, 1507-1522. Billett, Matthew, Mark Flannery and Jon Garf inkel, 2006, Are banks special? Evidence on the post-announcement performance of bank borrowers, Journal of Financial and Quantitative Analysis 41, 733-751. Bradley, Michael, and Michael Roberts, The struct ure and pricing of cor porate debt covenants, SSRN working paper. Brophy, David, Paige P. Ouimet, and Clemens Sialm, 2004, PIPE Dreams? The Performance of Com panies Issuing Equity Privately, NBER Working Papers 11011, National Bureau of Econom ic Research, Inc. Cantwello, Miguel and Julian Wright, 2000, How do firms choose their lenders? An empirical investigation, The Review of Financial Studies 13, 155-189. Carey, Mark, Mitch Post, and Steven A. Shar pe, 1998, Does corporate lending by banks and finance companies differ? Evidence on speci alization in private debt contracting, Journal of Finance 53, 845-878. Chaplinsky, Susan, and Robert Hansen, 1993, Partia l anticipation, the flow of information and the economic impact of corporate debt sales, Review of Financial Studies 6, 709-732. Chemmanur, Thomas, and Paolo Fulghieri, 1994, Reputation, renegotiation, and the choice between bank loans and publicly traded debt, Review of Financial Studies 7, 475-506. Denis, David, and Vassil Mihov, 2003, The choice among bank debt, non-bank private debt, and public debt: evidence from new corporate borrowings, Journal of Financial Economics 70, 3-28. Demiroglu, Cem, and Christopher James, The in formation content of bank loan covenants, SSRN working paper.

PAGE 61

61 Diamond, Douglas, 1991, Monitoring and reputation, the choice betw een bank loans and directly placed debt, Journal of Political Economy 99, 688-721. Eckbo, Espen, Ronald Masulis, and Oyvind No rli, 2007, Security offerings, Handbook of corporate Finance: Empirical Corporate Fi nance, Vol. 1, North-Holland/Elsevier, Handbooks in Finance series, Ch. 6, forthcoming. Fama, Eugene, 1985, Whats different about banks?, Journal of Monetary Economics 17, 239249. Gomes, Armando, and Gordon Phillips, 2005, Why do public firms issue private and public securities?, SSRN working paper. Guner, A. Burak, Loan sales and the cost of co rporate borrowing, Review of Financial Studies forthcoming. Hadlock, Charles, and Christopher James, 2002, Do Banks Provide Financial Slack?, Journal of Finance 57, 1383-1419. Houston, Joel and Christopher James, 1996, Bank in formation monopolies and the mix of private and public debt claims, Journal of Finance 28, 149-171. Ivashina, Victoria, and Zheng Sun, Institutional stock trading on loan market information, SSRN working paper. James, Christopher, 1987, Some evid ence on the uniqueness of bank loans, Journal of Financial Economics 19, 217-235. James, Christopher, and David Smith, 2000, Are banks still special? New evidence on their role in the corporate capital-raising process, Journal of Applied Corporate Finance 13, 395422. Johnson, Shane, 1997, An empirical investigation of corporate debt ownership structure, Journal of Financial and Quantitative Analysis 32, 47-69. Jung, Kooyul, Yong-Cheol Kim, and Rene Stul z, 1996, Timing, invest ment opportunities, managerial discretion, and the security issue decision, Journal of Financial Economics 42, 159-185. Lummer, Scott, and John McConnell, 1989, Furthe r evidence on the bank lending process and the capital-market response to bank loan agreements, Journal of Financial Economics 25, 99-122. Marsh, Paul, 1979, equity rights issues and the efficiency of the UK stock market, Journal of Finance 34, 839-862. Mikkelson, Wayne H. and M. Megan Partch, 1986, Va luation effects of securities offerings and the issuance process, Journal of Financial Economics 15, 31-60.

PAGE 62

62 Mitchell, Mark L., and J. Harold Mulherin, 1984, The impact of public information on the stock market, Journal of Finance 49, 923-950. Moerman, Regina Wittenberg, The impact of information asymmetry on debt pricing and maturity, University of Chicago working paper, 2005. Nini, Greg, David Smith, and Amir Sufi, 2007, Creditor Control Rights and Firm Investment Policy, Journal of Financial Econom ics forthcoming. Rajan, Raghuram G., 1992, Insiders and outsiders: the choice between informed and arms length debt, Journal of Finance 47, 1367-1400. Ramakrishnan, R.T.S and Anjan Thakor, 1984, In formation reliability and security design, Review of Economic Studies 51, 415-432. Rippington, Frederick, and Richard J. Taffler, 1996, The information content of firm financial disclosures, Journal of business Finance & Accounting 22, 345-362. Roberts, Michael, and Amir Sufi, 2007, Control Rights and Capital S tructure: An Empirical Investigation, SSRN working paper. Roberts, Michael, and Amir Sufi, 2007, Contingenc y and renegotiation of financial contracts: Evidence from private credit agreements, SSRN working paper. Smith, Clifford, 1986, Investment banking and the capital acquisition process, Journal of Financial Economics 15, 3-29. Sufi, Amir, 2005, Banks and flexibility: Empiri cal evidence on the mi x of equity, bonds and bank debt, MIT working paper. Sufi, Amir, 2007, The Real Effects of Debt Certificati on: Evidence from the Introduction of Bank Loan Ratings, Review of Financial Studies, forthcoming. Sufi, Amir, 2007, Information Asymmetry and Financi ng A rrangements: Evidence from Syndicated Loans, Journal of Finance 62, 629-668. Sufi, Amir, 2007, Bank Lines of Credit in Corporate Finance: An Empirical Analysis, Review of Financial Studies, forthcom ing. Thompson, Robert B., Chris Olsen and J. Richard Dietrich, 1987, Attributes of news about firms: An analysis of firm-specific news re ported in the Wall Street Journal Index, Journal of Accounting Research 25, 245-274.

PAGE 63

63 BIOGRAPHICAL SKETCH Laura Gonzalez Alan a earned her Bachelor of Science in electrical engineering from Superior College of Engineering of Bilbao (Spa in) simultaneously with Bachelors of Arts in music education and music performance from J.C. Arriaga, Bilbao, Spain. In 2002 and 2003 respectively, she earned her Master of Business Administration and Master of Arts in foreign languages and literatures from Southern Illin ois UniversityCarbondale. The requirements for the degree of Doctor of Philosophy in finance were completed at the University of Florida during the spring of 2008. Upon graduation from the Univ ersity of Florida, she will join Fordham University, NY, as an assistant pr ofessor in finance. Lauras res earch interests include financial markets and institutions, corporate fi nance, and international finance.