cl

Document Sample
cl Powered By Docstoc
					                                Corporate Liquidity

                      Filippo Ippolito                     Ander Perez1
                      Bocconi University              Universitat Pompeu Fabra
                 …lippo.ippolito@unibocconi.it          ander.perez@upf.edu


                                        15 February 2011


                                            Abstract

      The existing literature on corporate liquidity has almost uniquely focused on cash
holdings, ignoring other important sources of liquidity such as outstanding lines of credit
("undrawn credit"). In this paper we provide the …rst large sample evidence on undrawn
credit in U.S. public corporations and use it to re-examine the existing evidence on corporate
liquidity, which we compute as undrawn credit plus cash. First, we …nd that …rms with a
line of credit are signi…cantly larger, more leveraged, and more pro…table than …rms without
one and, importantly, that they hold much less liquidity (27.5% of assets vs. 40.5%). Sec-
ond, cash holdings and undrawn credit are associated with an opposite sign with most …rm
characteristics, such as leverage, market-to-book, tangibility, capex, dividend payer status
and pro…tability. Third, we …nd that cash and undrawn credit are negatively related and
can be regarded as substitutes. In contrast to what existing theory suggests, the degree of
substitution between the two is highest for small high-growth unrated …rms.

       JEL Classi…cations: G30, G31, D22
       Keywords: cash holdings, undrawn credit, liquidity, …nancial constraints




   1
     Contact details: Departament d’Economia i Empresa, Universitat Pompeu Fabra, Ramon Trias Fargas,
25-27, 08005 Barcelona. Email: ander.perez@upf.edu. Phone: +34 935421180
                             Corporate Liquidity
                                      15 February 2011



                                         Abstract

      The existing literature on corporate liquidity has almost uniquely focused on cash
holdings, ignoring other important sources of liquidity such as outstanding lines of credit
("undrawn credit"). In this paper we provide the …rst large sample evidence on undrawn
credit in U.S. public corporations and use it to re-examine the existing evidence on corporate
liquidity, which we compute as undrawn credit plus cash. First, we …nd that …rms with a
line of credit are signi…cantly larger, more leveraged, and more pro…table than …rms without
one and, importantly, that they hold much less liquidity (27.5% of assets vs. 40.5%). Sec-
ond, cash holdings and undrawn credit are associated with an opposite sign with most …rm
characteristics, such as leverage, market-to-book, tangibility, capex, dividend payer status
and pro…tability. Third, we …nd that cash and undrawn credit are negatively related and
can be regarded as substitutes. In contrast to what existing theory suggests, the degree of
substitution between the two is highest for small high-growth unrated …rms.

     JEL Classi…cations: G30, G31, D22
     Keywords: cash holdings, undrawn credit, liquidity, …nancial constraints
1    Introduction

As a result of lack of data availability, the existing literature on corporate liquidity has
almost uniquely focused on cash holdings, ignoring other important sources of liquidity such
                                                ).
as outstanding lines of credit (“undrawn credit” Approximately two thirds of public U.S.
corporations have undrawn credit and, for these …rms, undrawn credit amounts on average
to 13.4% of assets and is of the same magnitude as cash holdings (14.1% of assets). Thus,
in order to study corporate liquidity it seems essential to take into account undrawn credit.
In this paper, we provide the …rst large sample evidence on undrawn credit in the U.S., and
use it to re-examine the existing evidence on corporate liquidity.
     We address three important questions. First, we investigate how …rms with and without
a credit line di¤er. What …rm characteristics are associated with having a credit line? How
does liquidity depend on the presence of a credit line? Second, we explore what are the
determinants of corporate liquidity. What common factors drive cash and undrawn credit?
And, what factors a¤ect them di¤erently? Third, we study whether undrawn credit and
cash can be regarded as substitutes. Are …rms indi¤erent between holding liquidity in the
form of cash or undrawn credit? We address these questions using a sample of U.S. public
…rms composed of 23,013 …rm-years for the period 2002-2008. The sample is representative
of the Compustat population and contains information about the availability of credit lines
and on the amount of undrawn credit.
     Our …rst …nding is that …rms with and without a credit line di¤er signi…cantly in
important respects. Firms with a credit line are on average more than three times larger
than …rms without a credit line. They have higher leverage, tangibility, pro…tability, capital
expenditures, and they are more likely to be rated, to pay dividends, and engage more in
acquisition activities. On the other hand, they have a lower market-to-book ratio, lower
R&D expenses as a share of sales, and, importantly, much lower cash holdings and total
liquidity. The sum of cash plus undrawn credit for a …rm with a credit line is 27.5% of
assets, while the mean cash holding of a …rm without a credit line is 40.5% of assets. Using
rating, the dividend payout ratio and size as measures of …nancing constraints, this evidence
suggests that …rms with a line of credit are less …nancially constrained. This result is in
line with Su… (2009), who reaches a similar conclusion using the test of …nancial constraints
of Almeida, Campello and Weisbach (2004). In addition, we …nd that size and rating of a
…rm are related to signi…cant variations of corporate liquidity holdings only for …rms that
do not have a line of credit. If a credit line is available, cash, undrawn credit and liquidity
do not vary signi…cantly with rating or size, while if there is no credit line, cash changes
dramatically across both dimensions.
     Our second set of …ndings concerns the determinants of corporate liquidity. We sort
…rms with respect to their cash holdings and their level of undrawn credit. Sorting by cash
holdings produces large variations in …rm characteristics. Low cash …rms are bigger, more
pro…table and better rated, and they are very likely to have a credit line. High cash …rms
have a high market-to-book ratio and negative pro…ts. They tend not to have a credit line,
and despite this fact their liquidity holdings are very large (on average 61.7% of their assets
for the highest-cash quartile). When we sort …rms according to undrawn credit (conditional
on having a credit line), we observe little variation in …rm characteristics. This evidence
reinforces our previous …nding that whether a …rm has a line of credit or not has important
implications for liquidity and other …rm characteristics, but that the size of the credit line
does not matter as much.
     Next, we explore how cash and undrawn credit are related to …rm characteristics.
We run three separate regressions in which cash, undrawn credit and liquidity respectively
represent the dependent variables, as a function of …rm characteristics, which are taken
from the existing work on cash holdings (see Bates, Kahle and Stulz (2009)). We …nd
that all the signi…cant coe¢ cients have opposite sign across the cash and undrawn credit
regressions. This suggests that cash and undrawn credit are related to …rm characteristics
in a signi…cantly di¤erent way. The coe¢ cients of the regression in which liquidity is the
dependent variable almost always have the same sign as in the cash regression, and similar


                                              4
magnitude.2
       Finally, we examine the relationship between undrawn credit and cash to evaluate
whether these two sources of liquidity are substitutes. We start by regressing one on the
other, also controlling for a set of other …rm characteristics. We …nd that the coe¢ cient
on undrawn credit is negative and signi…cant for cash, and that the coe¢ cient on cash
is negative and signi…cant for undrawn credit. This …nding strongly suggests that cash
and undrawn credit are substitutes. We then investigate how the degree of substitutability
varies across ratings and …nd that it increases as rating worsens, and that it is highest for
unrated …rms. Consequently, we examine substitution speci…cally in the subset of unrated
…rms and show that the degree of substitution increases drastically as we move from low
to high market-to-book …rms. Taken together, these …ndings suggest that the degree of
substitution is highest for …rms that are more …nancially constrained and stand in contrast
with the common intuition on the relationship between cash, undrawn credit and …nancial
constraints.
       In relation to existing research, our paper is situated between the literature on cash
holdings and that on undrawn credit. The literature on cash holdings focuses on explaining
why …rms hold cash. The above cited Opler, et al. (1999) (OPSW from now on) …nd
that a main reason to accumulate cash reserves is the ability to fund future investments
when cash ‡ows are low or external …nance is very expensive. This precautionary motive
is suggested by the fact that smaller, unrated …rms with more volatile cash-‡ows and more
growth opportunities tend to hold larger cash reserves. Bates et al. (2009) builds on the
evidence of OPSW and shows that U.S. corporations have signi…cantly increased their cash
holdings since 1980 to the point that in 2006 the average …rm could retire all debt obligations

   2
      This would seem to suggest that it is reasonable to draw inferences about the e¤ect that …rm character-
istics have on liquidity by looking at the coe¢ cients of a regression in which cash is the dependent variable.
In this respect, our results con…rm the …ndings of Bates et al. (2009) and Opler, et al. (1999). This result
however depends crucially on whether we consider all cash holdings to be the relevant measure of available
liquidity in the form of cash, or whether it would make more sense to distinguish between operational and
non-operational cash. If non-operational cash turns out to be empirically a small fraction of total cash, the
behavior of total liquidity would resemble the behavior of undrawn credit rather than that of cash.


                                                      5
with its cash holdings.
       The literature on undrawn credit has focused on what drives the choice between cash
                                                                                            s
and undrawn credit. In principle, both cash and lines of credit can be used to satisfy a …rm’
demand for liquidity and, furthermore, existing theory suggests that credit lines should be
unambiguously favored over cash.3 Firms however use cash extensively to manage their
liquidity needs, prompting the question of what drives this decision. Su… (2009) argues
that lines are frequently revoked due the violation of the covenants that they carry, which
                          ow
are mainly based on cash-‡ measures. Firms with poor past or expected cash ‡ows
face a high probability of losing access to undrawn credit and have to rely more on cash
holdings. As a result, cash and undrawn credit cannot be considered substitutes for low
cash-‡ …rms.4 Yun (2009) gives a corporate governance explanation for the choice between
      ow
cash and undrawn credit. Firms with worse governance prefer cash because it is free of
                                                                s
monitoring. Acharya, Almeida and Campello (2010) show that a …rm’ exposure to aggregate
risk determines the choice between cash and undrawn credit. Lins, Servaes and Tufano (2010)
in turn suggest that the choice is driven by the di¤erent uses of cash and undrawn credit.
Undrawn credit is used to invest in future business opportunities, while cash serves a more
                                                            ow
general purpose and is used to hedge against negative cash ‡ shocks. The CFOs that are
most likely to consider cash and undrawn credit as substitutes typically belong to pro…table
…rms with low agency costs that operate in developed …nancial markets. Finally, Campello,
Giambona, Graham and Harvey (2011) study how …rms managed their liquidity needs during
the …nancial crisis of 2008-2009 and …nd that the …rms that hold the largest lines of credit
as a share of total assets in their sample are the small, unpro…table and non-publicly traded
   3
      There are at least two reasons why credit lines should be unequivocally preferred over cash. On the one
hand, lines of credit can be drawn on on a state-contigent basis, as opposed to cash which is carried through
to every possible future state of the world, making it more expensive from an opportunity-cost viewpoint.
On the other hand existing theory suggests that lines of credit exist precisely to deal with the agency costs
that give rise to the need to hoard liquidity (Boot, Thakor and Udell (1987), DeMarzo and Sannikov (2006),
Holmstrom and Tirole (1998)).
    4
      Jimenez, Lopez and Saurina (2008) also study the patterns of credit line drawdowns, and …nd that …rms
that eventually default are heavy users of lines of credit, while large and pro…table …rms draw down on their
lines of credit less. Ex-post, …rms that have su¤ered from …nancial distress in the past do not (or are not
allowed to) access their lines of credit often.


                                                     6
…rms.5 They also document, in slight contrast to Su…’ …ndings, that higher cash ‡
                                                    s                            ows
are associated with larger lines of credit only for …rms with scarce cash holdings, and that
covenant violations rarely lead to line cancellations. Finally, they …nd evidence that …rms
tend to use their lines of credit only when internal sources of liquidity, namely cash ‡ows
and cash holdings, are low.
         The rest of the paper is structured as follows. In Section 2 we discuss the sample
construction and provide a description of the data. In Section 3 we compare …rms with and
without undrawn credit and examine the determinants of the presence of a credit line. In
Section 4 we examine corporate liquidity and show what factors a¤ect cash, undrawn credit
and liquidity. In Section 5 we analyze the relationship between cash and undrawn credit,
and show how this varies across ratings. Finally, Section 6 concludes.



2        Sample Construction and Description

We obtain …rm-level data from the Capital IQ (CIQ) and Compustat databases for the period
of 2002-2008. We restrict ourselves to U.S. …rms covered on both databases and traded on
AMEX, NASDAQ, or NYSE. We remove utilities (SIC codes 4900-4999) and …nancial …rms
(SIC codes 6000-6999). Following Bates et al. (2009), we further remove …rm-years with
negative revenues, and negative or missing assets, obtaining in the end a sample of 23,013
…rm-years involving 4,248 unique …rms.
         CIQ compiles detailed information on capital structure and debt structure by going
through …nancial footnotes contained in …rms’ 10K Securities and Exchange Commission
(SEC) …lings. Most importantly for our purposes, …rms provide detailed information on the
drawn and undrawn portions of their lines of credit in the liquidity and capital resources
section under the management discussion, or in the …nancial footnotes explaining debt oblig-
ations, and CIQ compiles this data. 10K …lings typically also contain information on pricing
and maturity of lines of credit, but this data is not collected by CIQ. We use the information
    5
        See also the related paper of Ivashina and Scharfstein (2009) on bank lending during the …nancial crisis.


                                                         7
of CIQ to construct a dummy for the presence of a credit line, which is equal to one if the
…rm has a positive amount of undrawn credit reported in the 10K. Following Su… (2009) we
also construct a measure of the amount of undrawn credit expressed as a percentage of book
assets (Compustat item 6). As in Bates, et al. (2009), Opler, et al. (1999), Almeida, et al.
(2004) and Acharya, et al. (2010) we compute the ratio of cash and investments (item 1)
over total assets (item 6). We then add undrawn credit from CIQ and cash and investments
(item 1) and divide the sum by assets (item 6) to obtain our main measure of liquidity.
     Following Bates, et al. (2009), we also compute the following variables that are known
to be relevant for cash holdings behavior. Size is the logarithm of assets (item 1), where
assets are expressed in millions of 2001 dollars de‡ated by the consumer price index. Net
working capital to assets is computed as the di¤erence between working capital (item 179)
and cash and investments (1) divided by assets (item 6). R&D expenses are computed as the
ratio of research and development expenses (item 46) over sales (item 12). Book leverage
is debt in current liabilities (item 34) plus long-term debt (item 9) over assets (item 6).
               ow                                             ow
Industry cash-‡ risk (named industry sigma) is the mean cash-‡ volatility computed
                             ow
by two-digit SIC code. Cash-‡ volatility is the standard deviation of operating income
before depreciation (item 13) calculated over the previous twelve quarters and scaled by
assets (item 6). Dividend payout dummy is a dummy that takes value of one if common
stock has paid dividends (item 21). Acquisition expenses are computed as acquisitions (item
129) over assets (item 6).
     Following Lemmon, et al. (2008), we compute the M/B ratio as the sum of the market
value of equity, total debt, and preferred stock at liquidating value (item 10), minus deferred
taxes and investment tax credit (item 35), all divided by assets (item 6). Market value
of equity is computed as stock price (item 199) times number of common shares used to
calculate the earnings per share (item 54). Total debt is current liabilities (item 34) plus
long-term debt (item 9). Industry pro…tability is the average pro…tability computed by two-
digit SIC code. Pro…tability is operating income before depreciation (item 13) over assets


                                              8
(item 6) . Tangibility is net property, plant and equipment (item 8) over assets (item 6).
      We also compute …rm–year rating as the average monthly rating by S&P (item 280),
after converting the S&P rating into numbers. Credit spread is the spread on U.S. corporate
                         s
bond yields between Moody’ AAA and BAA provided by Datastream, based on averages of
seasoned issues. Finally, following standard procedures, all variables are winsorized at the
0.5% in both tails of the distribution. A summary of these variables can be found in Table
A1.


2.1    Sample Overview

Table 1 provides a sample overview and a comparison of characteristics of …rms with and
without a credit line. The …rst two columns provide information on the entire sample, while
columns 3-4, and 5-6 respectively provide information for the sub-samples of …rms with and
without a credit line. The former sub-sample contains 15,596 …rm-years, while the latter
has 7,417 …rm-years. The main objective of this table is twofold; to show that the samples
of …rm-years with and without a credit line are signi…cantly di¤erent and to show that cash
holdings are not a su¢ cient measure of liquidity given the extensive use and quantitative
importance of undrawn credit.
      The main picture that emerges from the table is that …rms with a credit line are larger,
more leveraged, more pro…table, have fewer growth opportunities and more tangible assets,
and are more likely to be rated and to pay dividends. More precisely, …rms with a credit line
are on average three times larger in size ($2.67bn vs. $0.85bn) as measured by the book value
of assets (CPI de‡ated in 2001 dollars), and have leverage of 23.6% versus 15.1% of …rms
without a credit line. This observation is consistent with the view that access to a credit
line is a good measure of whether a …rm is …nancially constrained (Su… (2009)). According
to this interpretation leverage in …rms without a credit line is lower because raising external
…nance for these …rms is costlier than for …rms with a credit line. Also along these lines
we observe that only 8.5% of …rms without a credit line are rated compared to 34.7% of


                                              9
…rms with a credit line. To measure growth opportunities we employ the M/B ratio, R&D
expenditures, and acquisition activity. Firms with a credit line have a lower M/B ratio (1.575
vs. 2.308), a lower ratio of R&D expenses over sales (a median of 0% vs. 11.8%), and higher
acquisition expenses (3% vs. 2%).6 The fact that …rms with a credit line display lower R&D
but higher acquisition expenditures may suggest that these …rms tend to grow externally via
acquisitions rather than organically, as opposed to …rms without access to a line of credit.
Pro…tability is measured by the ratio of cash ‡ows to assets, which is positive (6.3%) for
…rms with a credit line, and negative otherwise ( 9.9%). The information on pro…tability
is supported by the data on dividend payment behavior. Firms with a credit line are often
dividend payers (36%), while this is not the case for non-credit line holders (10.7%). These
…ndings lend support to the claim in Su… (2009) that …rms that su¤er from poor operating
performance are unlikely to be able to obtain a line of credit, and, should they already have
one, are more likely to see it revoked.
       To test formally for the di¤erences between these two samples for each of the eleven
variables analyzed above, we perform a t-test for unpaired data with unequal variances and
a two-sample Wilcoxon rank-sum (Mann-Whitney) test. Both the parametric and the non-
parametric tests show that the two samples are di¤erent along all of the eleven dimensions
with a 1% signi…cance level.
       Finally, another dimension along which these two samples strongly di¤er is cash hold-
ings. Firms with a credit line have a signi…cantly lower cash to assets ratio (14.1%) than
…rms without a credit line (40.5%). This …nding suggests that cash and credit lines are to
some extent substitutes for the purpose of corporate liquidity management. It also reinforces
the notion that access to a line of credit could be an accurate measure of …nancial constraints
as …rms without a credit line tend to hoard high levels of cash, possibly to be able to have
access to funds in the future when external …nance may not be available for them. Adding

   6
     For R&D expenses over sales, we compare medians rather than means because the mean of this ratio
for …rms without a credit line is likely to be in‡uenced by the extremely low values of sales. Speci…cally,
there are 407 …rm-years with sales below 1 million dollars in the sample of …rms without a credit line.


                                                    10
more evidence in this direction, …rms without a line of credit have on average a negative
ratio of net working capital to assets, which suggests that they might rely to a large extent
on trade credit given that other sources of …nance may not be available.



3    Undrawn Credit

In this section we provide an overview of the distribution of undrawn credit. Given that
this is the …rst paper to provide large sample evidence on undrawn credit in the Compustat
population of U.S. …rms, we consider these descriptive statistics interesting for the general
reader. We examine undrawn credit along three dimensions: size, industry and rating.
     Figure 1 examines the relationship between size and undrawn credit. The upper panel
displays the distribution of undrawn credit across di¤erent brackets of size for the entire
sample of …rm-years, while the lower panel focuses solely on the sub-sample of …rm-years with
a credit line. Focusing on the line depicting the ratio of undrawn credit over assets (rhomboid
marker), we observe that in both panels it shows relatively little variation, ‡uctuating around
10%, and being by construction higher in the sub-sample of …rms with a credit line. However,
the line that represents undrawn credit as a percentage of total liquidity (violet line with
cross marker), computed as the sum of undrawn credit plus cash, shows signi…cant variation
across size. In the top panel, the percentage of liquidity provided by undrawn credit increases
from slightly below 20% for …rms of size smaller than $100 million to approximately 50%
for the largest …rms in the sample (larger than $5 billion), and the source of this pattern is
not the increase in undrawn credit for larger …rms but the decrease in cash holdings. The
di¤erences for the sub-sample of …rms with a line of credit are more moderate, the ratio
increasing from slightly above 40% to almost 60% for the largest …rms, suggesting that the
e¤ect of size on the source of corporate liquidity depends crucially on the availability of a
line of credit, an issue that we discuss below in detail.
     Panel A of Table 2 illustrates the distribution of undrawn credit across industries.



                                              11
The …rst column reports the percentage of …rms with a credit line and shows that there
is signi…cant variation across sectors. Construction, wholesale and retail trade have the
highest percentage of …rms with a credit line (respectively, 89.8%, 84.8%, and 83.6%), while
manufacturing and services have the lowest percentages (respectively, 65.3% and 60.3%).
Conditional on having a credit line the di¤erences in the percentage of undrawn credit over
assets also varies signi…cantly across sectors, with transportation, communication, electric
gas and sanitary services (10.5%) having the lowest percentage, and wholesale trade the
highest (16.2%). The sectors with the lowest proportion of …rms with a credit line are
also those for which cash represents the largest share of assets. The ratio of cash to assets
for manufacturing and services is respectively 46.6% and 40.3%, which is four time that of
construction (11.3%). This is the second piece of evidence of a negative relationship between
cash and undrawn credit.
     Panel B of Table 2 displays the distribution of undrawn credit across ratings. A total
of 6,038 …rm-years are rated, while 16,975 are unrated. We consider a …rm-year as rated if
S&P has assigned a rating for at least one month of the year. If there are di¤erent ratings
for months of the same year we take the equal weighted average of these ratings to compute
the yearly rating. Observing the …rst column, there is a striking di¤erence in the presence of
a credit line between …rms with a rating equal to or above B     and …rms with a rating below
this threshold or without a rating. For the …rst group, the percentage of …rms with a credit
line ranges between 84% and 94%, while for the second group the range is between 60%
(unrated) and 68.3% (CCC+ or below). We take this as an indicator of a strong correlation
between rating and access to credit lines. The causality can go both ways as on one hand
rating agencies take into consideration whether a …rm has access to a line of credit in order to
evaluate its liquidity position and credit rating, and on the other hand having a good rating
by S&P may make it more likely to be granted a line of credit by a bank. For …rms with rating
equal to or above B ; the distribution of credit lines is non monotonic, reaching a maximum
for …rms with BBB+/ . In particular, the highest rated …rms in the sample (AAA) do not


                                              12
have the highest proportion of credit lines. Presumably, this does not happen because these
…rms are denied a credit line, but because they have very easy access to external capital,
including commercial paper, and therefore do not need to hoard liquidity in any form. This
small set of highly rated …rms without a credit line also holds the lowest percentage of cash
to assets (6.1%) in the whole sample. The ratio of cash to assets is highest for unrated …rms
(17.2%), which are also the group with the smallest average size ($454.1 million). This ratio
is almost twice as much as that of any other subset of rated …rms, for which cash to assets
is in the range of 8-9%. The ratio increases sharply for the group of …rms without a rating
or a line of credit (42.3%) which suggests that for these …rms, who might be likely to face
…nancing constraints, the precautionary motive to hoard cash is strongest. Surprisingly, the
AAA group of …rms with a credit line also holds a relatively large percentage of cash to
assets (14%). This group is composed of only six …rms, namely Automatic Data Processing,
Exxon Mobil, GE, Johnson and Johnson, P…zer and UPS. Compared to the average …rm in
the sample, these …rms have larger cash ‡ows to assets (8.6%), negative net working capital
( 1.4%), and lower capex and R&D expenditures (respectively 3.4% and 6.3%). One possible
interpretation is that these …rms are cash generators with limited growth opportunities for
which the potential dividend (Free Cash-Flow To Equity (FCFE)) is larger than the actual
dividend paid to shareholders.


3.1    The Determinants of the Presence of a Credit Line

In this section we study the extensive margin of credit line demand by analyzing which …rm
characteristics are associated with the existence of a line of credit. We conduct a Probit
analysis in which the dependent variable is a dummy that indicates the presence of a line
of credit and in which the explanatory variables are drawn mostly from existing empirical
studies trying to explain cash holdings of …rms. In particular we use the estimation model
introduced in Opler, et al. (1999) and Bates, et al. (2009), and we also draw on the very
few existing empirical papers studying undrawn credit, such as Su… (2009), who uses a set


                                             13
of explanatory variables that is very similar and slightly smaller than those used in the cash
literature. We estimate the following regression:


   Line of Credit (dummy)i,t =            0   +   1 Sizei,t   +   2 Book   Leveragei,t +    3 M/Bi,t


                                        +     4 Tangibilityi,t    +   5 CashFlow/Assetsi,t


                                        +     6 Net   Working Capital/Assetsi,t +           7 Capex/Assetsi,t


                                        +     8 R&D/Salesi,t      +   9 Dividend   Payer Dummyi,t

                                        +     10 Industry     Sigmai,t +     11 Industry   Pro…tabilityi,t

                                        +     12 Credit   Spreadt + Industry FE + Rating FE + "i;t ;


For robustness, we run the Probit analysis including and excluding ratings and exchange
…xed e¤ects and also consider lagged regressors. We do not include year …xed e¤ects because
they may introduce a problem of multicollinearity given the presence of the variable Credit
Spread, which is calculated as an annual average for all …rms with public debt. We cluster
errors at the …rm level.
       Table 3 provides the results of our analysis. Our …ndings suggest that …rms that are
less likely to su¤er from …nancing constraints, such as large …rms with strong cash-‡ows, low
M/B ratios, high degree of asset tangibility and net working capital, that are regular dividend
payers, and that belong to industries with high average pro…tability and low volatility, are
more likely to access a credit line. Leverage is strongly positively associated with access to
undrawn credit. In part, this relationship might be mechanical because bank debt …nancing
is often originated in the form of a credit line which is then drawn down.7 In part, however,
the positive relationship between leverage and credit lines may be due to the fact that
leverage is higher for better rated …rms. As these …rms also have more undrawn credit, we
…nd a positive relationship between these two variables. Interestingly, the level of capital


   7
     Colla et al. (2011) illustrate that for smaller and less rated …rms draw downs on credit lines represent
a larger percentage of leverage than for other …rms.


                                                       14
expenditures is strongly positively associated with credit line availability. Firms that are
more highly exposed to risk (as captured by either the degree of R&D/Sales or Industry
Sigma), and that, according to established theory, should have a higher demand for liquidity
protection through credit lines, are actually less likely to have access to one.
     While here we primarily provide an explanation based on the demand for lines of credit,
it is also possible that supply factors play an important role. In particular, a worsening of
banks’ balance sheets following an episode such as the …nancial crisis of 2008-2009 might
contract supply of lines of credit to …rms. Although here we do not thoroughly investigate
the role of macroeconomic or business cycle related factors we attempt to capture possible
supply side e¤ects by including among the explanatory variables the credit spread between
AAA and BAA seasoned issues. We …nd that average corporate debt spreads are negatively
and signi…cantly associated with the presence of a line of credit. This coe¢ cient may have
two possible explanations. On the one hand, higher credit spreads may be associated with
worsening conditions for the pro…tability of …rms, which in turns reduces the willingness of
banks to provide credit. On the other, higher credit spreads may be related to a rising cost
of …nancing for banks who might then reduce the supply of lines of credit.



4    The Determinants of Corporate Liquidity

In this section we examine corporate liquidity, de…ned as liquid resources that a …rm can
access in a swift and ‡exible manner. We consider two items to fall under this de…nition; on
one hand cash and marketable securities, here represented by our variable "cash", and on the
other the undrawn portions of lines of credit. Due to lack of large sample data on undrawn
credit, the existing literature has focused almost exclusively on cash as a main measure of
liquidity. Of course cash reserves and undrawn credit are not the only sources of liquidity
available to a …rm. Future expected cash ‡ows and non-cash assets are also important
in this respect, but they cannot be considered to be readily available or to guarantee a



                                              15
certain amount of future liquidity. We believe these important di¤erences justify limiting
our analysis to cash and undrawn credit.
     We explore which factors drive the accumulation of cash holdings, undrawn credit and
total corporate liquidity, and show that drawing inference about liquidity from a regression
in which cash is the dependent variable can lead to inaccurate conclusions. The e¤ect that
commonly employed variables, such as size, M/B or pro…tability have on cash is very di¤er-
ent from their e¤ect on undrawn credit. In most cases, the sign of the coe¢ cients on these
explanatory variables is the opposite for these two liquidity components. To examine liquid-
ity, we begin by looking at cash and undrawn credit separately, in a univariate framework.
Then, we run multivariate regressions to explain cash holdings, undrawn credit and liquidity,
using the standard regression employed in the literature to explain cash holdings (Opler, et
al (1999), Bates, et al. (2009)), and compare the di¤erences in the coe¢ cients. We then run
again the same regression for liquidity. Subsequently, we condition on …rms having a credit
line and show how the above regressions change. Finally, we consider having a credit line a
form of self-selection à la Heckman, according to which …rms …rst choose whether to have a
line or not, and then choose the desired level of liquidity. We re-run the above regressions
as a second stage in which the Probit estimated in Table 3 represent the …rst stage.
     One important comment is in order. In this paper, and in the vast majority of the
empirical cash literature, we consider total cash holdings as a source of liquidity, while it
can be argued that a fraction of cash reserves, operating cash, is held for regular working
capital commitments such as wage payments and is thus not available for other uses. Lins.
et al (2010) …nd using survey evidence that on average less than half of cash holdings can
be considered non-operational so this point becomes crucial when constructing the liquidity
measure. For lack of a well-established procedure to extract the non-operational component
of cash reserves we use total cash holdings but we take this concern into account when
interpreting our results about liquidity.




                                             16
4.1    Sorting Firms by Cash and Undrawn Credit

Table 4 provides a univariate comparison of means and medians of …rm characteristics across
di¤erent quartiles of cash-holdings. Each quartile contains approximately 5,753 …rm-years.
In the bottom quartile …rms have an average of 1.6% of cash over assets, while in the top
quartile this ratio is 58.7%. The main purpose of this table is to show that cash is an
important sorting variable and that …rms with di¤erent levels of cash di¤er also in many
other dimensions. The table is meant to be directly compared with Table 3 of OPSW on
which it is based. The main di¤erence between our sample and theirs is that their sample is
from 1971 to 1994 while ours starts in 2002. In light of the …ndings of Bates et al. (2009),
the role of cash seems to have changed dramatically over recent decades so we should not
expect these two tables to look the same. In OPSW size varies little across quartiles, while
in our sample …rms in the …rst quartile are almost six times bigger than those in the last
                                            ow
quartile. A similar pattern holds for cash ‡ to assets, and net working capital to assets.
In our sample, …rms with high cash to assets have very low pro…tability and net working
capital, compared to the OPSW sample. Given that the …rms in the fourth quartile are
also less leveraged and have less undrawn credit, this evidence suggests that these …rms
may be …nancing their cash holdings with new equity issues. The two tables are similar
with respect to the trends in M/B, R&D expenses, and acquisition activity. In addition to
what is reported by OPSW, our Table 4 provides information on undrawn credit, and shows
that undrawn credit decreases signi…cantly as we move from the lowest to the highest cash
quartile.
      Table 5 takes up on the latter observation, and displays a univariate comparison of …rm
characteristics across di¤erent quartiles of undrawn credit. The sample here is restricted to
…rms that have a credit line. In the …rst quartile …rms have 3.2% of undrawn credit over
assets, while this ratio increases to 28.4% in the last quartile. The point of this table is to
show that sorting by undrawn credit helps understand the dynamics of cash. This sorting
generates monotonic patterns only for size, net working capital, dividend payer status and,


                                              17
importantly, cash. For all other variables, there is no clear trend across quartiles. With
respect to size, smaller …rms have a larger percentage of undrawn to assets. This …nding
complements the evidence provided by Figure 1. The percentage of cash almost halves
from the …rst to the fourth quartile. However, overall liquidity barely changes in the …rst
three quartiles remaining in the range of 22-26%, and then increases dramatically in the
last quartile. This indicates a signi…cant skewness in the distribution of undrawn credit. In
sum, the picture that emerges is that for …rms with a credit line, cash and undrawn credit
substitute for each other in a way that for most …rms leaves the total amount of liquidity
relatively constant.


4.2    Factors that A¤ect Liquidity

We now examine how …rm characteristics in‡uence liquidity. The main aim of this section is
to illustrate that the various measures of liquidity are associated to the explanatory variables
in signi…cantly di¤erent ways. We start with the standard cash regression that is used in the
literature (Opler, et al. (1999), Bates et al. (2009)) in which the ratio of cash to assets is the
dependent variable and …rm characteristics are the independent variables. We then extend
this regression to di¤erent measures of liquidity. We consider two other measures: the ratio
of undrawn credit to assets, and the ratio of liquidity to assets, where liquidity is de…ned as
the sum of cash and undrawn credit. The explanatory variables are taken from Bates et al.
(2009). They include a number of …rm characteristics, as well as the standard deviation of
industry cash ‡ows (industry sigma). We also include year, rating, and stock exchange …xed
e¤ects in all speci…cations. The rating …xed e¤ects are based on 22 rating dummies and the
unrated dummy.
      Our …ndings are displayed in Table 6. In columns 1-3 we run regressions on the entire
sample; in columns 4-6 we examine the sub-sample of …rms that have positive amounts of
undrawn credit; and in column (7) we look at the sub-sample of …rms without a line of
credit. The main message that emerges from the table is that the coe¢ cients of most …rm


                                               18
characteristics vary across columns. In particular, most coe¢ cients change sign respectively
from column 1 to 2, and from 4 to 5. This indicates that the determinants of cash and
undrawn credit are completely di¤erent, both within the entire sample, as well as within the
sub-sample of …rms with a credit line. The coe¢ cients in the liquidity regressions (columns
3 and 6) look similar to those of the cash regressions (respectively, columns 1 and 4). As
was mentioned above in the introduction to this section we are including total cash holdings
in liquidity and not only non-operating cash, which exaggerates the in‡uence of cash on
liquidity, so these results have to be interpreted in light of this observation. A second
message that the table delivers is that the coe¢ cients of the cash regressions are di¤erent
for …rms with and without a credit line (columns 4 and 7). With the exception of size, all
coe¢ cients keep the same sign, but change in magnitude, sometimes quite dramatically, as
is the case for tangibility, dividend payer status, and acquisition activity.
     More precisely, with reference to columns (1) and (2), larger cash-‡ows, net working
capital, and capex, all scaled by assets, are associated with lower cash holdings but higher
levels of undrawn credit. The opposite holds for R&D expenditures and industry sigma.
Dividend payers tend to hold less cash and more undrawn credit. A pattern that emerges
from columns (1-3) is the negative association between liquidity and …rm size: larger …rms
tend to hold less of any form of liquidity as a fraction of total assets. A doubling of the
size of a …rm is associated with a decrease in undrawn credit and liquidity equivalent to
around 1% of total assets. This observation complements the (univariate) …ndings of Figure
1. It is also consistent with the transactions motive identi…ed early on in the theoretical
literature (Baumol (1952)) which established a relationship between …rm size and cash hold-
ings. However, it is likely that this motive also applies to undrawn credit, which would then
explain the negative sign in the relationship between size and liquidity. In sum, columns
(1-3) suggest that the factors that a¤ect cash are di¤erent from those that a¤ect undrawn
credit. However, insofar as cash acts as a proxy for liquidity, as is the case in Bates et al.
(2009) and Opler, et al. (1999), the sign of the coe¢ cients of the cash regression, with only


                                               19
few exceptions, can be used to draw inferences on liquidity. This is probably due to the fact
that in most cases cash represents a larger share of liquidity than undrawn credit.
        An examination of columns (4-6) allows us to show how the coe¢ cients change for the
sub-sample of …rms with a credit line, and to provide a comparison between the extensive and
intensive margins of undrawn credit. First, in this sub-sample several factors are strongly
associated with cash holdings but have no relevance for the amount of undrawn credit over
assets. These include book leverage, M/B, R&D/sales, and acquisition activity. Second, we
compare the extensive and intensive margins of undrawn credit by looking at the coe¢ cients
of column (5) of Table 6 in relation to those of column (1) of Table 3. The latter shows that
larger …rms with high tangibility are more likely to have a line of credit. The former shows
that lines of credit tend to be smaller for more tangible …rms. Book Leverage, M/B and
R&D/Sales are all signi…cantly associated with the presence of a line of credit, but they are
not related to the level of undrawn credit.


4.2.1     Robustness Checks

We perform two separate robustness checks on the regressions of Table 6. First we look
at how a Heckman selection process a¤ects the regression in column (5). Second, we re-
run the entire set of regressions of Table 6 on lagged regressors, so to account for possible
endogeneity.
        With respect to the …rst robustness check, we examine to what extent column (5) is
informative about the entire population, given that we cannot observe the demand for credit
line for …rms that do not have one. Econometrically, we deal with this self-selection process
by employing a two-stage regression model as the one suggested by Heckman (1979). The
…rst-stage represents the selection process, while the second stage illustrates the e¤ect of
…rm characteristics on the level of credit lines accounting for the potential bias introduced
by the selection process. An estimation of the …rst stage is already provided in Table 3.
While exclusion restrictions are not necessary in the Heckman selection model because it is


                                              20
identi…ed by non-linearity, in the …rst stage of the model we include industry pro…tability
and credit spread as independent variables, which are not part of the second-stage regression.
For the second stage, our (untabulated) results show that the inverse Mills ratios is negative
and signi…cant, thus suggesting that the selection process is relevant. However, none of the
coe¢ cients of column (5) in Table 6 are signi…cantly a¤ected. We then conclude that the
results of column (5) are reliable also after controlling for the selection bias.
     The second robustness check requires the lagging of all the independent variables em-
ployed in Table 6. Lagging partially helps in addressing the potential endogeneity of the
independent variables. We borrow this approach from Lemmon, et al. (2008), Flannery
and Rangan (2006), and Fama and French (2002). We report our results in Table A2. An
inspection of the coe¢ cients in the various columns with the respective columns of Table 6
does not show any signi…cant di¤erence between the two sets of regressions.



5     Are Cash and Undrawn Credit Substitutes?

In this section we examine how cash holdings and undrawn credit are related. Our prelim-
inary evidence in Tables 4 and 5 suggests that cash and undrawn credit may be regarded
by …rms as two substitute forms of liquidity. We …rst test the substitutes hypothesis more
formally by examining whether the coe¢ cient on undrawn credit (cash) is negative and sig-
ni…cant in a regression in which cash (undrawn credit) is the dependent variable. We then
study in section 5.1 how the degree of substitutability varies with certain …rm characteristics.
     To test the substitutes hypothesis we augment the regressions of columns (1) and (4)
and columns (2) and (5) of Table 6 with undrawn credit and cash respectively as right hand
side variables. We do this exercise for contemporaneous and lagged regressors. We resort to
lagging as a way to deal with the possible endogeneity in the choice of cash and undrawn
credit and consider the lagged regressions a robustness check for the contemporaneous re-
gressions. Our results are displayed in Table 7. In columns (1-2) and (5-6) we examine the



                                               21
entire sample, while in columns (3-4) and (7-8) we restrict our attention to the sub-sample
of …rms with a credit line. The main …nding of this table is that across all samples and
speci…cations, undrawn credit has a negative and signi…cant coe¢ cient in the regression in
which cash is the dependent variable, and vice-versa. This provides strong support for the
substitutes hypothesis. In column (1), a decrease in undrawn credit equivalent to 1% of
total assets increases the cash to assets ratio by 0.363%. While, in column (2) a decrease
in the cash ratio by 1% leads to an increase in the undrawn credit ratio of 0.144%. One
possible explanation for why the coe¢ cient of undrawn credit is twice that of cash is that
it is easier to decrease cash once a credit line has been granted, than vice-versa. In support
of this interpretation, it is worth noticing that the two coe¢ cients are much more similar in
the subsample of …rms that have a credit line. In columns (4-5), the coe¢ cient for undrawn
credit is   0:167 and that of cash is   0:125. The regressions with lagged coe¢ cients con…rm
the …ndings obtained for the contemporaneous coe¢ cients.


5.1     Variation in the Degree of Substitutability

The degree of substitutability should be higher for …rms that can obtain a line of credit and
that, conditional on having one, do not expect it to be revoked. While there is no evidence
on which …rms are more likely to be granted a line of credit, we know from the empirical
studies of spot loan supply (Gertler and Gilchrist (1994), Kahle and Stulz (2010) and Iyer, et
al. (2010)) that large, rated …rms with low volatility are more likely not to be credit rationed
by banks. Similar mechanisms may be at play for lines of credit. On the other hand, Su…
(2009) suggests that only …rms with high past and future expected cash ‡ows can expect
their lines not to be revoked. Finally, the survey evidence in Lins, et al. (2010) suggests that
           ow,
high cash-‡ low agency problem …rms in …nancially developed countries are more likely to
consider cash and undrawn as substitutes. Putting this evidence together, we should expect
large, pro…table, well-rated …rms with good access to capital markets to display a larger
degree of substitutability. We conduct the analysis for size, pro…tability, credit ratings, and


                                               22
market-to book, and report below the results for the two dimensions along which results are
most striking: credit ratings and market-to-book.
     We …rst consider the e¤ect of the credit rating. We rank …rms according to their rating
by considering the same rating classes employed in Panel B of Table 2, and exclude …rms
with rating below B    as these are likely to be in a situation of …nancial distress. For the
entire sample and for the sub-sample of …rms with a line of credit we …rst estimate the
correlation coe¢ cient between cash and undrawn credit. Then, we re-run the regressions of
Table 7 and report the coe¢ cients for cash and undrawn credit. In the regressions we do
not include rating …xed e¤ects. From the regression analysis, we exclude the sub-sample of
AAA …rms as there are too few observations for a reliable multivariate analysis.
     Our results are reported in Table 8. The …rst …nding is that the correlations between
cash and undrawn credit are always negative, but non-monotonic across di¤erent ratings.
For the entire sample (left column), in absolute values the correlations are highest for AAA,
then they decrease as we move to BBB+/BBB , and then they increase again as the rating
worsens, peaking again for unrated …rms. A similar pattern holds for the sub-sample of
…rms with a line of credit (right column). The main di¤erence is that, with the exception
of AAA, correlations are generally smaller than in the left column. Overall, the analysis
of correlations provides partial support to the idea that substitutability is higher for less
…nancially constrained …rms.
     The second main …nding of Table 8 is that across the entire sample (left column) the
coe¢ cient for undrawn credit is increasing (in absolute terms) as rating quality drops. This
result is counter-intuitive as it indicates that for less …nancially sound …rms an increase in
undrawn credit leads to a larger drop in the amount of cash, than for more …nancially sound
…rms. When we compare the coe¢ cients of undrawn credit across columns, we …nd that
these are smaller (in absolute terms) for the sub-sample of …rms with a credit line than for
the entire sample. This is also a counter-intuitive result, which however …nds support in the
evidence provided by Table 7. In sum, Table 8 provides support to the substitutes hypothesis


                                             23
across each rating sub-sample, as the sign of the correlations and of the coe¢ cients for cash
and undrawn is always negative when signi…cant. However, the table also shows that the
common intuition according to which the degree of substitution is higher for …rms that have
easier access to …nance (i.e. better rated and with a credit line) is not con…rmed by the data.
     Next we examine in Table 9 whether for unrated …rms the relationship between cash
and undrawn credit varies signi…cantly across di¤erent values of M/B. We restrict ourselves
to the sample of unrated …rms. Table 8 shows that these …rms have the highest degree of
substitution between undrawn credit and cash (the only exception is AAA, which however
is subject to the usual caveats) and this suggests that the relationship between cash and
undrawn credit may show special features for unrated …rms (ELABORATE ON WHY ).
For each quartile we report the correlation coe¢ cient between cash and undrawn credit, as
well as the regression coe¢ cients for cash and undrawn credit. These have been obtained
following the same procedure as in Table 8.
     Based on existing theory and empirical evidence we would expect higher Market-to-
Book …rms to display a lower degree of substitutability, for a number of reasons. First, M/B
proxies for growth opportunities, which means that high M/B …rms will generate most of
their value further into the future and for that reason are riskier and more opaque (Strahan
(1999)). These …rms will …nd it harder to obtain a line of credit and face a higher probability
of seeing it revoked, so they should not consider cash and lines of credit as perfect substitutes.
Second, MB has been identi…ed in the asset pricing literature (Fama and French (1993)) as a
                 s
proxy for the …rm’ exposure to an aggregate risk factor. If high MB …rms are more exposed
to certain aggregate risk factors, and to the extent that this exposure might increase the
price at which these …rms are o¤ered lines of credit (Almeida, et al. (2011)), then high MB
…rms should display a lower degree of substitutability. Finally, high M/B ratios tend to be
associated with growth …rms that are risky, have low current pro…tability and are small, and
which for these reasons tend to be …nancially constrained. For this reason too we should
expect high M/B ratios to be associated with a low degree of substitution.


                                               24
     The main …nding of Table 9 is that the degree of substitution is increasing when we
move from low M/B to high M/B …rms. This evidence is con…rmed by the correlation
coe¢ cients which vary from    29.2% to    39.5% for the entire sample (left column); by the
cash coe¢ cients which increase (in absolute terms) from     12.6% to    14.0%; and, by the
coe¢ cients for undrawn credit which increase (in absolute terms) from     22.7% to    52.1%.
Similar patters can be observed for the limited sample of …rms with a line of credit (right
column). However, in each row of the limited sample the coe¢ cients are smaller (in absolute
terms) with respect to the analogous coe¢ cients of the entire sample, suggesting that the
extensive margin (whether a …rm has a line of credit or not) plays a big role. When we carry
out a similar (untabulated) study for rated …rms, we …nd that there is no clear pattern in
the correlations and in the coe¢ cients across di¤erent values of M/B. This suggests that the
relationship between cash and undrawn credit changes signi…cantly from rated to unrated
…rms. Overall, the analysis of Table 9 indicates that more …nancially constrained …rms
exhibit a higher degree of substitution between undrawn credit and cash.



6    Conclusion

The two main tools used by …rms to satisfy their demand for liquidity are cash and lines of
credit. The existing literature has mainly focused on cash as a key component of liquidity.
This has been mainly the result of data limitations on the availability of credit lines. In
this paper we …ll a gap in the literature by re-examining the existing evidence on liquidity,
inclusive of undrawn credit.
     We document three key …ndings. First, the presence of a line of credit has important
implications for most …rm characteristics. Firms which are less likely to be liquidity con-
strained, such as large, pro…table, tangible, …rms that pay dividends and are rated are much
more likely to have a credit line. However, conditional on having a line of credit, the amount
of undrawn credit does not generate important variations in any dimension. In other words,



                                             25
the extensive margin of lines of credit seems to be much more important than the intensive
margin. Second, the motives driving undrawn credit and cash accumulation seem to be very
di¤erent, to the point that only size matters for both in the same way. Finally, we …nd that
cash and undrawn credit can be considered to be substitute forms of liquidity, but that,
contrary to common intuition, the degree of substitutability is highest for high-growth and
unrated …rms.
     Two of our …ndings challenge the current theory on corporate liquidity management.
First, we …nd that the …rms with best access to capital markets, and hence with less need to
plan for future liquidity, are more likely to have a line of credit. Second, cash and undrawn
credit are treated as closer substitutes by …rms that are less likely to have a line of credit.
These …ndings raise unsolved questions for future research.




                                              26
References

 [1] Acharya, Viral V., Heitor Almeida and Murillo Campello, 2010. "Aggregate Risk and
    the Choice between Cash and Lines of Credit," NBER Working Papers 16122.

 [2] Almeida, Heitor, Murillo Campello and Dirk Hackbarth, 2011. "Liquidity Mergers,"
    NBER Working Papers 16724.

 [3] Almeida, H., M. Campello, and M. Weisbach (2004). "The Cash Flow Sensitivity of
    Cash," The Journal of Finance, 59(4), 1777-1804

 [4] Bates, Thomas W., Kathleen M. Kahle and René M. Stulz, 2009. "Why Do U.S. Firms
    Hold So Much More Cash than They Used To?," Journal of Finance, vol. 64(5).

 [5] Baumol, W. J., 1952, The transactions demand for cash: An inventory theoretic ap-
    proach, Quarterly Journal of Economics 66, 545-556.

 [6] Boot, A., A.V. Thakor, and G.F. Udell. (1987). "Competition, risk neutrality and loan
    commitments," Journal of Banking & Finance, 15, 605-632.

 [7] Campello, Murillo, Erasmo Giambona, John R. Graham and Campbell R. Harvey. "Liq-
    uidity Management and Corporate Investment During a Financial Crisis," forthcoming
    Review of Financial Studies.

 [8] DeMarzo, Peter M. and Sannikov, Yuliy, (2006), Optimal Security Design and Dynamic
    Capital Structure in a Continuous-Time Agency Model, Journal of Finance, 61, 6, 2681-
    2724.

 [9] Fama, Eugene F. and French, Kenneth R., 1993. "Common risk factors in the returns
    on stocks and bonds," Journal of Financial Economics, vol. 33(1), pages 3-56, February.

[10] Fama, Eugene F. and French, Kenneth R., 2002, "Testing trade-o¤ and pecking order
    predictions about dividends and debt," Review of Financial Studies 15, 1-33.


                                           27
[11] Flannery, Mark J. and Rangan, Kasturi P., 2006. "Partial adjustment toward target
    capital structures," Journal of Financial Economics, vol. 79(3).

[12] Gertler, Mark and Simon Gilchrist, 1993, "Monetary policy, business cycles, and the
    behavior of small manufacturing …rms", The Quarterly Journal of Economics 109(2),
    309-40.

[13] Heckman, James J, 1979. "Sample Selection Bias as a Speci…cation Error," Economet-
    rica, vol. 47(1), pages 153-61, January.

[14] Holmstrom, B., and J. Tirole. (1998). "Private and Public Supply of Liquidity," Journal
    of Political Economy, 106(1), 1-40

[15] Ivashina, V., and D.S. Scharfstein. (2009). "Bank Lending During the Financial Crisis
    of 2008,". EFA 2009 Bergen Meetings Paper.

[16] Jiménez, G. , J.A. López and J. Saurina. (2008). "Empirical analysis of corporate credit
    lines", Review of Financial Studies, 22(12), 5069-5098

[17] Iyer, Rajkamal, Samuel Lopes, José-Luis Peydró, and Antoinette Schoar, (2010), "Inter-
    bank liquidity crunch and the …rm credit crunch: Evidence from the 2007-2009 crisis",
    unpublished, MIT.

[18] Kahle, Kathleen M. and René M. Stulz. "Financial Policies and the Financial Crisis:
    How Important Was the Systemic Credit Contraction for Industrial Corporations?"
    NBER, and ECGI Dice Center WP 2010-13

[19] Lemmon, Michael L., Michael R. Roberts and Jaime F. Zender, 2008. "Back to the
    Beginning: Persistence and the Cross-Section of Corporate Capital Structure," Journal
    of Finance, vol. 63(4).




                                               28
[20] Lins, K.V., H. Servaes, and P. Tufano. "What Drives Corporate Liquidity? International
    Evidence from Survey Data on Strategic Cash and Lines of Credit,”Journal of Financial
    Economics, Vol 98 (2010) 160-176

[21] Opler, Tim, Pinkowitz, Lee, Stulz, Rene and Williamson, Rohan, 1999. "The determi-
    nants and implications of corporate cash holdings," Journal of Financial Economics,
    vol. 52(1).

[22] Strahan, Philip E. (1999) "Borrower Risk and the Price and Nonprice Terms of Bank
    Loans" Banking Studies Function October 1999

[23] Su…, Amir. (2009). "Bank Lines of Credit in Corporate Finance: An Empirical Analy-
    sis", Review of Financial Studies, 22(3), 1057-1088

[24] Yun H. (2009). "The Choice of Corporate Liquidity and Corporate Governance," Review
    of Financial Studies, 22(4), 1447-1475




                                             29
                                                     Entire Sample
           0.6

           0.5

           0.4

           0.3
Ratios




           0.2

           0.1

         6E-16
                      <100m            100m-500m            500m-1bn             1bn-5bn             5bn
          -0.1




                                     Sample of Firms with a Credit Line
         0.6

         0.5

         0.4
Ratios




         0.3

         0.2

         0.1

          0
                    <100m           100m-500m             500m-1bn           1bn-5bn              5bn

                                                       Firm Size
                    Undrawn Credit/Assets                          Cash/Assets
                    (Undrawn Credit + Cash)/Assets                 Undrawn Credit/(Undrawn Credit + Cash)


               Figure 1: The Relationship Between Size and Various Liquidity Measures
This picture shows the patterns of liquidity across different firm sizes. The sample consists of non-utilities
(excluding SIC codes 4900-4949) and non-financials (excluding SIC codes 6000-6999) U.S. firms covered by
both Capital IQ and Compustat from 2002 to 2008. We have removed firm- years with 1) negative
revenues, and 2) negative or missing assets. After the above filtering, there are 23,013 firm-year
observations involving 4,248 unique firms in the sample. Table A1 provides a full description of the
variables listed below. All variables are winsorized at the 0.5% in both tails of the distribution. Assets are
expressed in millions of 2001 dollars deflated by the consumer price index.




                                                                                                            1
                                       Table 1
        Sample Overview: Comparison of Firms with and without Undrawn Credit
This table provides summary statistics respectively for the entire sample, the sample of firms with a credit
line, and the sample of firms without a credit line. The entire sample consists of non-utilities (excluding
SIC codes 4900-4949) and non-financials (excluding SIC codes 6000-6999) U.S. firms covered by both
Capital IQ and Compustat from 2002 to 2008. We have removed firm- years with 1) negative revenues,
and 2) negative or missing assets. After the above filtering, there are 23,013 firm-year observations
involving 4,248 unique firms in the sample. Table A1 provides a full description of the variables listed
below. All variables are winsorized at the 0.5% in both tails of the distribution. Assets are expressed in
millions of 2001 dollars deflated by the consumer price index. The last two columns test for differences
between samples with and without undrawn credit using the unequal t-test and the two-sample Wilcoxon
rank-sum (Mann-Whitney) test.

                          Entire Sample       Sample of Firms      Sample of Firms          Test of Difference
                                             with a Credit Line    without a Credit         with vs. without a
                                                                        Line                    Credit Line
                          (1)      (2)         (3)       (4)        (7)       (8)             (9)        (10)
                         Mean     Median      Mean      Median     Mean      Median          t-test     MW-
                                                                                              (p-        test
                                                                                            value)        (p-
                                                                                                        value)

Cash/Assets              0.226      0.131     0.141      0.078      0.405          0.386    76.397    70.993
                                                                                            (0.000)   (0.000)
Undrawn                  0.091      0.061     0.134      0.107        -              -         -         -
Credit/Assets
Size                    2086.7      281.9     2673.9     470.0      852.0          100.5    -23.278   -51.556
                                                                                            (0.000)   (0.000)
Book Leverage            0.208      0.147     0.236      0.197      0.151          0.015    -24.890   -42.623
                                                                                            (0.000)   (0.000)
M/B                      1.799      1.303     1.575      1.205      2.308          1.621    25.145    26.630
                                                                                            (0.000)   (0.000)
Tangibility              0.243      0.163     0.275      0.202      0.176          0.091    -32.666   -40.200
                                                                                            (0.000)   (0.000)
Cash Flow/Assets         0.012      0.067     0.063      0.078     -0.099          0.018    -38.976   -45.031
                                                                                            (0.000)   (0.000)
NWC/Assets               0.047      0.039     0.082      0.071     -0.026          -0.017   -37.129   -39.597
                                                                                            (0.000)   (0.000)
Capex/Assets             0.053      0.032     0.057      0.035      0.045          0.023    -13.482   -27.773
                                                                                            (0.000)   (0.000)
R&D/Sales                0.734      0.005     0.106      0.000      2.065          0.118    20.432    58.710
                                                                                            (0.000)   (0.000)
Dividend Payer           0.280      0.000     0.360      0.000      0.107          0.000    -47.635   -39.423
                                                                                            (0.000)   (0.000)
Acquisition Activity     0.027      0.000     0.030      0.000      0.020          0.000    -11.575   -22.285
                                                                                            (0.000)   (0.000)
Rating Dummy             0.263      0.000     0.347      0.000      0.085          0.000    -52.301   -42.178
                                                                                            (0.000)   (0.000)
 Observations                 23013                15596                    7417




                                                                                                                2
                                          Table 2
                   Patterns of Undrawn Credit across Industries and Ratings
This table illustrates the patterns of undrawn credit across different industries and S&P ratings. The
sample consists of non-utilities (excluding SIC codes 4900-4949) and non-financials (excluding SIC codes
6000-6999) U.S. firms covered by both Capital IQ and Compustat from 2002 to 2008 for a total of 23,013
firm-years. We have removed firm- years with 1) negative revenues, and 2) negative or missing assets.
Table A1 provides a full description of the variables listed below. In both panels, the columns respectively
contain 1) the percentage of firms with a credit line, 2) the percentage of undrawn credit over assets
conditional on having a credit line, 3) the percentage of cash over assets conditional on having a credit line,
4) the percentage of cash over assets for firms without a credit line, 5) size (for the whole sample), 6)
observation in each category. Panel A illustrates the patterns of undrawn credit across industries. Panel B
reports the same information across ratings. Two-digit SIC codes classifications are as follows: Agriculture,
Forestry and Fishing (1-9), Mining (10-14), Construction (15-17), Manufacturing (20-39), Transportation,
Communication, Electric, Gas and Sanitary Services (40-49), Wholesale Trade (50-51), Retail Trade (52-
59), Services (70-89), Non-classifiable (99). Data on ratings ranging from AAA to D are from Compustat
(280).

Panel A: Pattern of Undrawn Credit across Industries
                         % of Firms      Undrawn             Cash            Cash             Size       Obs.
                          with CL      Credit / At          /Assets         /Assets         (Entire
                                        (with CL)          (with CL)       (without         sample)
                                                                             CL)
                              Mean           Mean           Mean             Mean           Mean
                                             Median         Median          Median          Median

Agriculture et al. (1-9)      0.765           0.137          0.099           0.170          1567.6        68
                                              0.106          0.074           0.169           348.7
Construction (15-17)          0.898           0.150          0.122           0.113          2212.9        275
                                              0.124          0.094           0.093           968.5
Manufacturing (20-39)         0.653           0.137          0.153           0.466          2070.1       11715
                                              0.111          0.089           0.463           224.0
Mining (10-14)                0.731           0.130          0.060           0.178          2064.1       1292
                                              0.105          0.028           0.096           503.9
Non-classifiable (99)         0.416           0.167          0.185           0.382          9802.7        101
                                              0.050          0.182           0.203           271.4
Retail Trade (52-59)          0.836           0.140          0.112           0.203          2257.0       1792
                                              0.117          0.064           0.131           444.2
Services (70-89)              0.603           0.127          0.188           0.403          1061.0       5308
                                              0.096          0.118           0.397           187.3
Transp et al. (40-49)         0.769           0.105          0.089           0.240          5518.4       1548
                                              0.079          0.045           0.191           827.6
Whole. Trade (50-51)          0.848           0.162          0.076           0.184          1290.1        914
                                              0.139          0.036           0.081           480.8
All                           0.678           0.134          0.141           0.405          2086.7       23013
                                              0.107          0.078           0.386           281.9




                                                                                                               3
Panel B: Pattern of Undrawn Credit across Rating Classes
                         % of Firms      Undrawn          Cash       Cash       Size    Obs.
                          with CL      Credit / At       /Assets    /Assets   (Entire
                                        (with CL)      (with CL)   (without   sample)
                                                                     CL)
                           Mean          Mean          Mean          Mean     Mean
                           Median        Median        Median       Median    Median

        AAA                 0.840         0.169         0.140       0.061     51445.9    50
                                          0.070         0.139       0.054     64215.8
       AA +/–               0.926         0.106         0.099       0.218     25592.1    121
                                          0.082         0.078       0.145     18228.8
        A +/–               0.926         0.145         0.093       0.108     14825.1    807
                                          0.117         0.062       0.062     8157.0
      BBB +/–               0.940         0.143         0.078       0.118     8382.9    1516
                                          0.121         0.050       0.062     3800.1
       BB +/–               0.915         0.121         0.078       0.187     3267.1    1887
                                          0.101         0.045       0.149     1779.7
        B +/–               0.853         0.110         0.083       0.301     2367.1    1310
                                          0.089         0.055       0.223      883.7
   CCC+ or below            0.683         0.118         0.094       0.253     2031.9     347
                                          0.086         0.051       0.185      665.3
       Unrated              0.600         0.137         0.172       0.423      454.1    16975
                                          0.109         0.106       0.408      139.2
         All                0.678         0.134         0.141       0.405     2086.7    23013
                                          0.107         0.078       0.386      281.9




                                                                                              4
                                           Table 3
                         What Determines the Presence of a Credit Line
Probit specifications for the demand for credit lines. The sample consists of non-utilities (excluding SIC
codes 4900-4949) and non-financials (excluding SIC codes 6000-6999) U.S. firms covered by both Capital
IQ and Compustat from 2002 to 2008 for a total of 23,013 firm-years. We have removed firm-years with 1)
negative revenues, and 2) negative or missing assets. The dependent variables in the three columns is a
dummy for the presence of a credit line. The first two columns are identical except for the inclusion of
fixed effects. In the third column the regressors are all lagged by one period. Rating fixed effects are based
on 22 rating dummies and the unrated dummy. Robust standard errors clustered at the firm level are
reported in parentheses. ***, **, and * denote statistical significance at the 1%, 5%, and 10% levels,
respectively.


                                        Dependent Variable: Presence of a Credit Line (Dummy)
                                                                                      Lagged Regressors
Size                                0.056***                   0.128***                      0.021
                                     (0.016)                    (0.012)                     (0.018)
Book Leverage                       0.967***                   1.016***                    1.046***
                                     (0.098)                    (0.095)                     (0.117)
M/B                                -0.051***                  -0.051***                   -0.071***
                                     (0.012)                    (0.012)                     (0.012)
Tangibility                           0.185                     0.275**                     0.278*
                                     (0.121)                    (0.121)                     (0.145)
Cash Flow/Assets                    0.825***                   0.810***                    0.941***
                                     (0.114)                    (0.119)                     (0.138)
NWC/Assets                          1.554***                   1.629***                    1.729***
                                     (0.109)                    (0.109)                     (0.129)
Capex/Assets                        0.884***                    0.748**                     0.974**
                                     (0.336)                    (0.337)                     (0.417)
R&D/Sales                           -0.017**                   -0.019**                    -0.017**
                                     (0.007)                    (0.007)                     (0.008)
Div. Payer Dummy                    0.347***                   0.456***                    0.377***
                                     (0.048)                    (0.046)                     (0.058)
Industry Sigma                      -3.508**                  -4.378***                   -4.393***
                                     (1.458)                    (1.433)                     (1.572)
Industry Profitability              1.809***                   1.766***                    1.771***
                                     (0.187)                    (0.186)                     (0.222)
Credit Spread                      -0.167***                  -0.157***                   -0.208***
                                     (0.028)                    (0.028)                     (0.066)
Constant                              0.457                   -0.360***                   1.298***
                                     (0.449)                    (0.097)                     (0.502)

Rating FE                             Yes                          No                          Yes
Exchange FE                           Yes                          No                          Yes
Observations                         20,903                      20,903                      16,896
Pseudo R-squared                     0.246                        0.234                       0.282




                                                                                                            5
                                               Table 4
                                        Cash Holdings Quartiles
This table displays a univariate comparison of means and medians of measures of firm characteristics
across different quartiles of cash holdings. The sample consists of non-utilities (excluding SIC codes 4900-
4949) and non-financials (excluding SIC codes 6000-6999) U.S. firms covered by both Capital IQ and
Compustat from 2002 to 2008 for a total of 23,013 firm-years. We have removed firm-years with 1)
negative revenues, and 2) negative or missing assets. Each quartile contains approximately 5753 firm-years.
Table A1 provides a full description of the variables listed below. The t-statistic in the last column is for a
difference of means test from the first to the fourth quartile.

                      First Quartile        Second           Third Quartile     Fourth Quartile    t-statistics
                                           Quartile                                                 (p-value)
                      Mean    Median     Mean Median         Mean    Median     Mean     Median

Cash/Assets           0.016     0.015    0.077     0.074     0.225     0.217    0.587     0.557      -252.411
                                                                                                      (0.000)
Und. Credit/At        0.129     0.106    0.118     0.094     0.086     0.055    0.030     0.000       55.623
                                                                                                      (0.000)
Liquidity             0.144     0.121    0.195     0.169     0.312     0.295    0.617     0.596      -174.087
                                                                                                      (0.000)
Size                 2733.9     572.2    3295.2    564.8    1852.4     265.8    464.9      93.4       21.365
                                                                                                      (0.000)
Book Leverage         0.327     0.296    0.267     0.222     0.152     0.070    0.088     0.001       61.140
                                                                                                      (0.000)
M/B                   1.181     0.995    1.425     1.135     1.899     1.450    2.684     1.992       -45.239
                                                                                                      (0.000)
Tangibility           0.365     0.292    0.299     0.234     0.209     0.155    0.100     0.069       69.779
                                                                                                      (0.000)
Cash Flow/Assets      0.069     0.074    0.057     0.077     0.041     0.076    -0.121    -0.003      39.548
                                                                                                      (0.000)
NWC/Assets            0.095     0.068    0.083     0.075     0.051     0.055    -0.039    -0.024      38.462
                                                                                                      (0.000)
Capex/Assets          0.069     0.039    0.058     0.035     0.052     0.033    0.034     0.021       29.106
                                                                                                      (0.000)
R&D/Sales             0.026     0.000    0.089     0.000     0.187     0.015    2.637     0.173       -21.558
                                                                                                      (0.000)
Dividend Payer        0.401     0.000    0.378     0.000     0.252     0.000    0.089     0.000       41.324
                                                                                                      (0.000)
Acq. Activity         0.040     0.000    0.033     0.000     0.024     0.000    0.011     0.000       22.446
                                                                                                      (0.000)
Rating Dummy          0.399     0.000    0.395     0.000     0.217     0.000    0.041     0.000       51.485
                                                                                                      (0.000)




                                                                                                             6
                                            Table 5
                Quartiles of Undrawn Credit Conditional on Having a Credit Line
This table displays a univariate comparison of means and medians of measures of firm characteristics
across different quartiles of undrawn credit. The sample consists of non-utilities (excluding SIC codes 4900-
4949) and non-financials (excluding SIC codes 6000-6999) U.S. firms covered by both Capital IQ and
Compustat from 2002 to 2008 for a total of 15,596 firm-years. Each quartile contains 3899 firm-years. We
have removed firm- years 1) with negative revenues, 2) with negative or missing assets, and 3) without a
credit line. Table A1 provides a full description of the variables listed below. The t-statistic is for a
difference of means test from the first to the fourth quartile.

                      First Quartile       Second           Third Quartile     Fourth Quartile    t-statistics
                                          Quartile                                                 (p-value)
                     Mean     Median    Mean Median         Mean    Median     Mean    Median

Cash/Assets          0.191     0.120    0.138     0.075     0.122    0.069     0.114     0.061      20.283
                                                                                                    (0.000)
Und. Credit/At       0.032     0.033    0.082     0.082     0.139    0.137     0.284     0.247     -146.050
                                                                                                    (0.000)
Liquidity            0.223     0.153    0.220     0.160     0.261    0.209     0.397     0.346     -41.4034
                                                                                                    (0.000)
Size                 3560.8    435.9    3375.7    563.6    2474.0    565.1    1285.1     354.4       13.353
                                                                                                    (0.000)
Book Leverage        0.246     0.197    0.254     0.220     0.228    0.200     0.216     0.172        5.712
                                                                                                    (0.000)
M/B                  1.656     1.206    1.492     1.163     1.521    1.203     1.633     1.255        0.705
                                                                                                    (0.481)
Tangibility          0.273     0.185    0.286     0.213     0.277    0.209     0.266     0.203       1.409
                                                                                                    (0.159)
Cash Flow/Assets     0.042     0.061    0.063     0.074     0.072    0.084     0.073     0.089       -8.842
                                                                                                    (0.000)
NWC/Assets           0.027     0.018    0.073     0.062     0.103    0.096     0.124     0.125      -23.228
                                                                                                    (0.000)
Capex/Assets         0.057     0.034    0.057     0.035     0.057    0.036     0.058     0.036       -0.862
                                                                                                    (0.389)
R&D/Sales            0.252     0.001    0.069     0.000     0.036    0.000     0.069     0.000        3.854
                                                                                                    (0.000)
Divid. Payer         0.260     0.000    0.361     0.000     0.396    0.000     0.423     0.000      -15.346
                                                                                                    (0.000)
Acq. Activity        0.032     0.000    0.032     0.000     0.031    0.000     0.027     0.000        3.359
                                                                                                    (0.001)
Rating Dummy         0.299     0.000    0.415     0.000     0.379    0.000     0.297     0.000        0.248
                                                                                                    (0.804)




                                                                                                            7
                                                                 Table 6
                                                  The Determinants of Corporate Liquidity
This table presents OLS regression results to explain the determinants of liquidity. The dependent variable is Cash/Assets for Columns (1), (4) and
(7), Undrawn Credit/Assets for Columns (2) and (5) and Liquidity for Columns (3) and (6). Definitions of the variables are provided in Table A1.
All specifications include year, rating and exchange fixed effects. Rating fixed effects are based on 22 rating dummies and the unrated dummy. The
sample in Columns (1), (2) and (3) is the entire sample, the sample in columns (4), (5) and (6) consists of firms with positive amounts of undrawn
credit, and the sample in column (7) consists of firms with no undrawn credit. Robust standard errors clustered at the firm level are reported in
parentheses. All regressions include a (non-reported) constant, year, rating , and exchange fixed effects. ***, **, and * denote statistical significance
at the 1%, 5%, and 10% levels, respectively.

                                       Entire Sample                                       Firms with a Credit Line                    Firms w/out CL
                           (1)                (2)                (3)                (4)                (5)                 (6)                 (7)
                     Cash/Assets      Undrawn Cr/At          Liquidity         Cash/Assets     Undrawn Cr/At           Liquidity         Cash/Assets
Size                     -0.002          -0.009***           -0.011***            -0.003*         -0.017***            -0.020***            0.007**
                        (0.002)            (0.001)             (0.002)            (0.002)           (0.001)              (0.002)            (0.003)
Book Leverage         -0.280***           0.043***           -0.236***          -0.247***             0.011            -0.236***          -0.208***
                        (0.014)            (0.006)             (0.013)            (0.011)           (0.007)              (0.013)            (0.025)
M/B                    0.021***          -0.002***            0.019***           0.028***             0.000             0.029***           0.011***
                        (0.002)            (0.001)             (0.002)            (0.002)           (0.001)              (0.002)            (0.002)
Tangibility           -0.283***             -0.001           -0.284***          -0.175***         -0.024***            -0.199***          -0.505***
                        (0.012)            (0.007)             (0.013)            (0.009)           (0.009)              (0.012)            (0.024)
CF/Assets             -0.131***           0.056***           -0.075***          -0.116***          0.046***            -0.070***          -0.086***
                        (0.015)            (0.006)             (0.014)            (0.022)           (0.013)              (0.025)            (0.016)
NWC/Assets            -0.375***          0.117***            -0.258***          -0.286***         0.083***             -0.203***          -0.314***
                        (0.014)            (0.009)             (0.014)            (0.013)           (0.011)              (0.016)            (0.026)
Capex/Assets          -0.097***          0.100***               0.003           -0.071***         0.112***                0.040              -0.082
                        (0.032)            (0.022)             (0.033)            (0.024)           (0.027)              (0.034)            (0.064)
R&D/Sales              0.006***          -0.001***            0.006***           0.008***            -0.001             0.008***           0.005***
                        (0.001)            (0.000)             (0.001)            (0.001)           (0.001)              (0.001)            (0.001)
Div.Payer             -0.022***           0.030***             0.008*            -0.009**          0.025***             0.016***           -0.030**
                        (0.005)            (0.003)             (0.005)            (0.004)           (0.004)              (0.005)            (0.012)
Acqu. Activity        -0.393***             0.017            -0.376***          -0.263***            -0.007            -0.270***          -0.679***
                        (0.018)            (0.012)             (0.018)            (0.015)           (0.013)              (0.018)            (0.040)
Industry Sigma         0.882***          -0.448***            0.434***           0.797***         -0.361***             0.436***             0.004
                        (0.160)            (0.081)             (0.149)            (0.142)           (0.097)              (0.151)            (0.288)
Observations            20,903             20,903              20,903             14,463            14,463               14,463              6,440
R-squared                0.551              0.155               0.429              0.459             0.089                0.355              0.482




                                                                                                                                                       8
                                                              Table 7
                                         The Relationship between Cash and Undrawn Credit
This table examines the relationship between cash and undrawn credit. The dependent variable is Cash/Assets for Columns (1), (3), (5) and (7)
and Undrawn Credit/Assets for Columns (2), (4), (6) and (8). Definitions of the variables are provided in Table A1. The sample in Columns (1),
(2), (5) and (6) contains all firms, while the sample in columns (3), (4), (7) and (8) consists of firms with positive amounts of undrawn credit.
Regressors are contemporaneous in columns (1-4), and lagged in columns (5-8). Errors are clustered at the firm level. Standard errors are reported
in parentheses. ***, **, and * denote statistical significance at the 1%, 5%, and 10% levels, respectively.

                                    Contemporaneous Regressors                                             Lagged Regressors
                           Entire Sample             Firms with Credit Line                Entire Sample                Firms with Credit Line
                        (1)              (2)            (3)            (4)              (5)               (6)             (7)              (8)
                   Cash/Assets      Und.Cr/At     Cash/Assets     Und.Cr/At        Cash/Assets      Und.Cr/At        Cash/Assets     Und. Cr/At
Cash/Assets                          -0.144***                     -0.125***                         -0.144***                        -0.108***
                                       (0.006)                       (0.012)                           (0.007)                          (0.012)
Undrawn Cr/At       -0.362***                      -0.167***                        -0.319***                         -0.137***
                      (0.017)                        (0.016)                          (0.017)                           (0.015)
Size                -0.005***        -0.009***     -0.006***       -0.017***        -0.006***        -0.009***        -0.006***       -0.016***
                      (0.002)          (0.001)       (0.002)         (0.001)          (0.002)          (0.001)          (0.002)         (0.001)
Book Leverage       -0.264***           0.003      -0.245***        -0.020**        -0.245***        0.024***         -0.230***         0.015*
                      (0.013)          (0.006)       (0.011)         (0.008)          (0.014)          (0.006)          (0.012)         (0.008)
M/B                  0.020***           0.001       0.028***        0.004***         0.020***           -0.000         0.026***           0.001
                      (0.002)          (0.001)       (0.002)         (0.001)          (0.002)          (0.001)          (0.002)         (0.001)
Tangibility         -0.283***        -0.042***     -0.179***       -0.046***        -0.234***        -0.050***        -0.145***       -0.051***
                      (0.012)          (0.008)       (0.009)         (0.009)          (0.012)          (0.008)          (0.010)         (0.009)
CF/At               -0.111***         0.037***     -0.108***         0.031**        -0.152***         0.047***        -0.114***        0.048***
                      (0.014)          (0.006)       (0.022)         (0.013)          (0.017)          (0.006)          (0.023)         (0.012)
NWC/Assets          -0.333***        0.063***      -0.272***        0.047***        -0.298***        0.062***         -0.230***       0.044***
                      (0.013)          (0.009)       (0.013)         (0.011)          (0.014)          (0.009)          (0.014)         (0.011)
Capex/Assets         -0.061**        0.086***       -0.053**       0.103***         -0.131***        0.084***         -0.114***       0.074***
                      (0.031)          (0.021)       (0.024)         (0.027)          (0.032)          (0.023)          (0.027)         (0.028)
R&D/Sales           0.006***         0.001***       0.008***          0.000         0.006***         0.001***         0.007***           0.000
                      (0.001)          (0.000)       (0.001)         (0.001)          (0.001)          (0.000)          (0.002)         (0.001)
Div. Payer (D)       -0.011**         0.027***        -0.004        0.024***        -0.017***         0.028***         -0.010**        0.024***
                      (0.004)          (0.003)       (0.004)         (0.003)          (0.004)          (0.003)          (0.004)         (0.004)
Acqu. Activity      -0.387***        -0.040***     -0.264***       -0.040***        -0.350***           -0.018        -0.233***          -0.019
                      (0.017)          (0.011)       (0.015)         (0.013)          (0.018)          (0.012)          (0.015)         (0.013)
Industry Sigma       0.720***        -0.321***      0.737***       -0.261***         0.838***        -0.338***         0.784***       -0.269***



                                                                                                                                                  9
                (0.152)    (0.075)    (0.139)    (0.095)    (0.152)    (0.080)    (0.142)    (0.101)
Constant       0.287***   0.235***   0.241***   0.367***   0.289***   0.269***   0.234***   0.337***
                (0.037)    (0.059)    (0.031)    (0.066)    (0.038)    (0.068)    (0.033)    (0.070)

Rating FE         Yes        Yes        Yes        Yes        Yes        Yes        Yes        Yes
Exchange FE       Yes        Yes        Yes        Yes        Yes        Yes        Yes        Yes
Year FE           Yes        Yes        Yes        Yes        Yes        Yes        Yes        Yes
Observations    20,903     20,903     14,463     14,463     16,895     16,896     12,249     12,250
R-squared        0.575      0.199      0.470      0.108      0.547      0.211      0.417      0.100




                                                                                                       10
                                           Table 8
                      The Degree of Substitution Across Different Ratings
This table shows how the degree of substitution between cash and undrawn credit varies across ratings.
Data on ratings ranging from AAA to D are from Compustat (280). The coefficients for undrawn credit are
obtained from a regression in which cash/assets is the dependent variable, and undrawn credit, size book
leverage, M/B, tangibility, cash flow/assets, NWC/assets, capex/assets, R&D/sales, a dividend payer
dummy, acquisition activity, industry sigma are independent variables. The regressions also include
exchange and year fixed effects. The correlations are Pearson pair-wise correlations. ***, **, and * denote
statistical significance at the 1%, 5%, and 10% levels, respectively.

                             Entire Sample                             Firms with Line of Credit

                              Correlations                                    Correlations
AAA                             -0.484***                                       -0.729***
AA+/AA-                          -0.191**                                          -0.122
A+/A-                            -0.081**                                         -0.077*
BBB+/BBB-                          -0.026                                           0.014
BB+/BB-                         -0.195***                                       -0.108***
B+/B-                           -0.300***                                       -0.124***
Unrated                         -0.362***                                       -0.166***
                          Coefficients for Cash                           Coefficients for Cash
AAA                                   -                                               -
AA+/AA-                            -0.001                                           0.047
A+/A-                             -0.153*                                         -0.175*
BBB+/BBB-                       -0.163***                                       -0.147***
BB+/BB-                          -0.095**                                          -0.073
B+/B-                           -0.123***                                          -0.022
Unrated                         -0.139***                                       -0.125***
                    Coefficients for Undrawn Credit                 Coefficients for Undrawn Credit
AAA                                   -                                               -
AA+/AA-                            -0.010                                           0.018
A+/A-                             -0.084*                                         -0.100*
BBB+/BBB-                       -0.095***                                        -0.088**
BB+/BB-                          -0.108**                                          -0.074
B+/B-                           -0.247***                                          -0.027
Unrated                         -0.408***                                       -0.192***




                                                                                                        11
                                            Table 9
                          The Degree of Substitution for Unrated Firms
This table shows how the degree of substitution between cash and undrawn credit varies across M/B
quartiles in the sample of unrated firms. Data on ratings ranging from AAA to D are from Compustat
(280). The coefficients for undrawn credit are obtained from a regression in which cash/assets is the
dependent variable, and undrawn credit, size book leverage, M/B, tangibility, cash flow/assets,
NWC/assets, capex/assets, R&D/sales, a dividend payer dummy, acquisition activity, industry sigma are
independent variables. The regressions also include exchange and year fixed effects. The correlations are
Pearson pair-wise correlations. ***, **, and * denote statistical significance at the 1%, 5%, and 10% levels,
respectively. ***, **, and * denote statistical significance at the 1%, 5%, and 10% levels, respectively.

                   Entire Sample of Unrated Firms                  Unrated Firms with Line of Credit

                             Correlations                                     Correlations
Low MB                         -0.292***                                         -0.080*
2                              -0.351***                                         -0.134*
3                              -0.427***                                         -0.257*
High MB                        -0.395***                                         -0.253*
                         Coefficients for Cash                            Coefficients for Cash
Low MB                         -0.126***                                        -0.083***
2                              -0.137***                                        -0.090***
3                              -0.146***                                        -0.144***
High MB                        -0.140***                                        -0.162***
                   Coefficients for Undrawn Credit                  Coefficients for Undrawn Credit
Low MB                         -0.227***                                        -0.075***
2                              -0.333***                                        -0.119***
3                              -0.447***                                        -0.258***
High MB                        -0.521***                                        -0.314***




                                                                                                          12
                                             Table A1
                                      Description of Variables

Variable               Construction

Acquisition Activity   Acquisitions (129) / Total Assets (6)
/ Assets
Book Leverage          Total Debt / Total Assets (6)
BV Equity              Total Assets (6) – Total Liabilities (181) – Deferred Taxes and Investment Tax
                       Credit (35) – Preferred Stock
Cash Flow /Assets      Operating Income Before Depreciation (13) – Interest Expenses (15) – Income
                       Taxes (16) – Dividends (21)
Cash/Assets            Cash and Investments (1) / Total Assets (6)
CF Volatility          Standard Deviation of Operating Income Before Depreciation (13) over Previous 12
                       Quarters Scaled by Total Assets (6)
Credit Spread          Spread on U.S. corporate bond yields between Moody’s AAA and BAA, averages of
                       seasoned issues (from Datastream)
Dividend Payer         A dummy variable that takes the value of one if common stock dividends (21) are
                       positive, and zero otherwise
Firm Size              Logarithm of Book Value of Total Assets (6)
Industry               Mean profitability by two-digit SIC code
Profitability
Industry Sigma         Mean CF Volatility by two-digit SIC code
Liquidity              (Undrawn Credit (from CIQ) + Cash and Investments (1) )/ Total Assets (6)
M/B                     (Market Value of Equity + Total Debt + Preferred Stock Liquidating Value (10)
                       – Deferred Taxes and Investment Tax Credit (35)) / Total Assets (6)
Market Value of        Stock Price (199) × Common Shares Used to Calculate EPS (54)
Equity
Net Working            (Working Capital (179) – Cash and Investments (1)) / Total Assets (6)
Capital/Assets
Profitability          Operating Income Before Depreciation (13) / Total Assets (6)
R&D/Sales              Research and Development Expense (46) / Sales (12)
Rated                  A dummy variable that takes the value of one if the firm is rated by the S&P, and
                       zero otherwise
Rating                 Monthly S&P ratings (280)
Tangibility            Net Property, Plant, and Equipment (8) / Total Assets (6)
Total Debt             Debt in Current Liabilities (34) + Long-Term Debt (9)
Undrawn Credit /       Undrawn Credit (from CIQ) / Total Assets (6)
Assets




                                                                                                      13
                                                             Table A2
                                          The Determinants of Liquidity (Lagged Regressors)
This table reproduces Table 6, which presents OLS regression results to explain the determinants of liquidity, lagging firm characteristics by one
year. All specifications include year, rating and exchange fixed effects. Robust standard errors clustered at the firm level are reported in
parentheses. All regressions include a (non-reported) constant. ***, **, and * denote statistical significance at the 1%, 5%, and 10% levels,
respectively.

                                      Entire Sample                                    Firms with a Credit Line                  Firms w/out CL
                          (1)               (2)                (3)              (4)                (5)                 (6)              (7)
                      Cash/At        Undrawn Cr/At        Liquidity        Cash/Assets     Undrawn Cr/At          Liquidity        Cash/Assets
L1.Size                 -0.004*        -0.009***          -0.013***          -0.004**         -0.016***           -0.019***            0.004
                        (0.002)          (0.001)            (0.002)           (0.002)           (0.001)             (0.002)           (0.004)
L1.Book Lev.          -0.259***         0.064***          -0.195***         -0.233***          0.042***           -0.191***         -0.169***
                        (0.014)          (0.006)            (0.014)           (0.012)           (0.008)             (0.013)           (0.027)
L1.M/B                 0.021***        -0.003***           0.018***          0.026***            -0.002            0.025***          0.011***
                        (0.002)          (0.001)            (0.002)           (0.002)           (0.001)             (0.002)           (0.002)
L1.Tangibility        -0.233***           -0.009          -0.242***         -0.143***         -0.031***           -0.174***         -0.440***
                        (0.012)          (0.008)            (0.013)           (0.010)           (0.009)             (0.013)           (0.027)
L1.CashFlow/At        -0.171***         0.066***          -0.104***         -0.123***         0.060***             -0.063**         -0.124***
                        (0.017)          (0.007)            (0.016)           (0.023)           (0.012)             (0.026)           (0.020)
L1.NWC/Assets         -0.336***         0.118***          -0.218***         -0.243***         0.076***            -0.167***         -0.267***
                        (0.015)          (0.009)            (0.015)           (0.014)           (0.011)             (0.017)           (0.029)
L1.Capex/Assets       -0.164***         0.098***            -0.067*         -0.126***         0.081***               -0.044          -0.157**
                        (0.033)          (0.024)            (0.035)           (0.027)           (0.028)             (0.036)           (0.067)
L1.R&D/Sales           0.006***          -0.001*           0.006***          0.007***            -0.001            0.006***          0.005***
                        (0.001)          (0.000)            (0.001)           (0.002)           (0.001)             (0.002)           (0.001)
L1.Div. Payer         -0.027***         0.032***              0.005         -0.013***          0.025***             0.012**          -0.032**
                        (0.005)          (0.004)            (0.005)           (0.004)           (0.004)             (0.005)           (0.015)
L1.Acqu. Act.         -0.356***         0.040***          -0.316***         -0.232***            0.009            -0.222***         -0.697***
                        (0.019)          (0.012)            (0.019)           (0.015)           (0.013)             (0.018)           (0.053)
L1.Ind. Sigma          0.990***        -0.474***           0.516***          0.836***         -0.360***            0.476***            0.119
                        (0.159)          (0.086)            (0.151)           (0.144)           (0.103)             (0.159)           (0.297)
Observations            16,895           16,896              16,895           12,249            12,250              12,249             4,646
R-squared                0.527            0.167              0.394             0.409             0.085                0.300            0.409




                                                                                                                                               14

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:4
posted:4/16/2012
language:
pages:43