Access To External Finance Theory and Evidence

Journal of Banking & Finance 30 (2006) 199–227
www.elsevier.com/locate/jbf
Access to external finance: Theory and

evidence on the impact of monetary policy
and firm-specific characteristics
a,*
Spiros Bougheas , Paul Mizen a, Cihan Yalcin b
a
University of Nottingham, University Park, Nottingham NG7 2RD, UK
b
Central Bank of Turkey, Ankara, Turkey
Received 18 May 2004; accepted 31 January 2005

Available online 18 March 2005
Abstract
This paper examines firmsÕ access to bank and market finance when allowance is made for
differences in firm-specific characteristics. A theoretical model determines the characteristics
such as size, risk and debt that would determine firmsÕ access to bank or market finance; these
characteristics can result in greater (or lesser) tightening of credit when interest rates increase.
An empirical evaluation of the predictions of the model is conducted on a large panel of UK
manufacturing firms. We confirm that small, young and risky firms are more significantly
affected by tight monetary conditions than large, old and secure firms.
2005 Elsevier B.V. All rights reserved.
JEL classification: E32; E44; E51
Keywords: Credit channel of monetary transmission; External finance
*
Corresponding author. Tel.: +44 115 8466108.
E-mail address: spiros.bougheas@nottingham.ac.uk (S. Bougheas).
0378-4266/$ - see front matter 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.jbankfin.2005.01.002
200 S. Bougheas et al. / Journal of Banking & Finance 30 (2006) 199–227
1. Introduction
A considerable body of literature has explored the credit channel of monetary trans-
mission under imperfect information including papers by Bernanke and Blinder (1988),
Romer and Romer (1990), Friedman and Kuttner (1993), Bernanke (1995) to mention
just a few. The influence of this channel is felt through the balance sheet (Gertler and
Gilchrist, 1994), the effects of bank lending on those firms that are particularly bank
dependent (Kashyap et al., 1993) and through the stimulation of endogenous cycles
or accelerator effects (Fuerst, 1995; Kiyotaki and Moore, 1997; Bernanke et al., 1999).
Institutions that use information from companyÕs balance sheets as market signals
of creditworthiness are referred to by Bernanke (1995) as offering credit through the
balance sheet channel. Rationing is exercised by pricing the loans to reflect the ob-
served risks in balance sheet information, driving a wedge in the relative price of
lending to alternative sources of external funds. The price is based on factors that
are easily monitored such as the profitability, financial wealth, previous loan pay-
ments history (see Leland and Pyle, 1977; Fama, 1984) as well as outstanding debt,
and will be influential in determining the eligibility of a company for access to loans.
For small and medium sized enterprises banks play a crucial role in the provision
of external finance and this gives rise to the bank lending channel, see Bernanke and
Blinder (1988). It is assumed that bank loans and alternative sources of finance are
imperfect substitutes and that persistent differentials in the spreads emerge because
there is imperfect arbitrage. Imperfection in substitutability arises because small
and medium sized firms may be unable to access other markets for funds and there-
fore have a certain dependence on banks for external sources of funds (see Kashyap
and Stein, 1994). It also arise because of imperfect substitutability on the supply side
since banks themselves might not regard bank loans and securities as perfect substi-
tutes in their own portfolios, and therefore the response of the banking sector to a
monetary tightening has a direct effect on the provision of loans, which affects small
and medium sized firms disproportionately.
Financially constrained firms are more exposed than unconstrained firms to mon-
etary cycle through both channels, and this implies that monetary policy is unlikely
to have uniform effects across firms. A significant literature has developed to attempt
to detect the impact of financial constraints on real activity by exogenously classify-
ing firms into constrained and unconstrained groups on the basis of their size, div-
idend payouts and capital structure (cf. Fazzari et al., 1988, and the survey by
Hubbard, 1998).1 This approach has raised a number of criticisms. Kaplan and
1
The original approach by Fazzari et al. (1988) classified firms according to whether they were likely to
be financially constrained. Once classified, the firms were analysed within the high-dividend, medium
dividend and low-dividend groupings to assess the sensitivity of investment to net worth measured by cash
flow. The highest sensitivities were found for firms categorized as financially constrained, and this was
taken to indicate that financial constraints were binding in this case. However, Kaplan and Zingales (1997,
2000) have made use of more detailed information in financial statements from annual reports to classify
the same firms over the same sample period into three categories Ôfinancially constrainedÕ, Ôpossibly
financially constrainedÕ and Ônot financially constrainedÕ. Using this classification they find that financially
constrained firms have the lowest sensitivity of investment to cash flow.
S. Bougheas et al. / Journal of Banking & Finance 30 (2006) 199–227 201
Zingales (1997, 2000) have argued that the classification adopted by Fazzari et al.
(1988) tends to assign firms incorrectly, and after re-classification they find substan-
tial differences in the degree of sensitivity of investment to financial constraints be-
tween firms. A very recent paper by Atanasova and Wilson (2004) also points out
that the classifications are based on variables that are endogenous to the firm, and
are highly restrictive since firms cannot move between classifications once they have
been assigned. Their approach endogenously classifies firms with the use of a disequi-
librium model, allowing switching between regimes.
We follow a different but complementary strategy to the exogenous categorization
and disequilibrium approaches. We make use of the seminal contribution by Kash-
yap et al. (1993) that isolated the influence of monetary policy contractions on bank
lending by measuring the relative changes of bank lending to non-bank sources of
funds.2 Kashyap et al. (1993) constructed a ÔmixÕ variable defined as the ratio of bank
lending to total external finance (bank lending plus commercial paper), allowing the
effect of the interest rate channel on all types of finance to be distinguished from a
credit channel on bank lending alone. When they used US data, they showed that
the mix between bank lending and market-based finance declined with a monetary
contraction and thus they provided strong support for the credit channel in general
and the bank lending channel in particular.
Although this research drew a reaction from Oliner and Rudebusch (1996) who
criticized the findings of Kashyap et al. (1993) because the sources of external finance
they considered were relatively narrow and no distinction was made between small
and large firms, the re-estimations of Oliner and Rudebusch were still supportive
of the credit channel.3 After taking into account the criticisms, Oliner and Rude-
busch (1996) found evidence in favor of the broad credit channel does exist, but more
muted support for the bank lending channel. In reply Kashyap et al. (1996) recalcu-
lated the effects for small and large firms using their own definition of the mix, and
found support for their original results.
The interchange between Kashyap et al. (1993, 1996) and Oliner and Rudebusch
(1996) is far from a minor dispute. It touches on an important issue for this paper –
the influence of firm-specific characteristics on the response to monetary contrac-
tions. If factors such as the size of the firm – to take the characteristic chosen by
2
Early attempts to measure the influence of policy tightening on the level of bank lending did not
distinguish between demand-side influences, operating through the liabilities side of banks balance sheets
(via the interest rate channel), and supply shifts, and therefore could not establish beyond doubt that there
was a separate credit channel.
3
While they were convinced by the use of a mix variable to capture the relative adjustment in the
financial portfolio, they were unsure whether Kashyap et al. (1993) had used the correct mix. They argued
that the original mix variable did not take into account a sufficiently wide range of alternative sources of
finance and did not account for differential effects on small as opposed to large firms. Small firms are
almost entirely bank dependent and therefore their mix is likely to be invariant to the monetary policy
stance. Kashyap et al. (1996) have responded to this critique by arguing that the re-interpretation of Oliner
and Rudebusch (1996) is misleading. The implication that the mix does not respond to monetary policy
when the data are disaggregated, they argue, is entirely expected for small firms (because they are bank
dependent at all times) and an artefact of the different measure of the mix for large firms.
Oliner and Rudebusch (1996) – can have an influence on the composition of finance,
then other characteristics may also alter access to credit. In other words, why con-
sider only size? In their conclusion Kashyap et al. (1996) note that there is Ômore
to be learned from careful analysis of a variety of micro-data, at the level of both
individual banks and individual firmsÕ (p. 313), and we agree. In fact, studies pio-
neered by Altman (1969) and Merton (1974) and supplemented recently by Shum-
way (2001), Hillegeist et al. (2004) and Atanasova and Wilson (2004) have used a
wider range of firm-specific variables to assess access to external finance. Now that
micro-data are accessible on a wide range of firm characteristics, such as their real
assets, perceived riskiness and indebtedness, in panels spanning periods of both tight
and benign monetary policy, we can bring these literatures closer together. The influ-
ence of the above factors on firmsÕ access to bank versus market-based finance, after
a change in monetary policy, is the point that the present paper addresses.
We begin by presenting a simple theoretical model that allows us to derive a tax-
onomy of firms according to their source (if any) of external finance based on their
characteristics. Then we examine access to credit under this taxonomy. Our model-
ling approach follows Bernanke and Gertler (1989), Diamond (1984) and William-
son (1987) who adopt TownsendÕs (1979) costly-state verification framework.
Banks in this environment have the ability to monitor their clients and thus verify
the returns of their projects.4 In contrast, capital markets (bondholders) are unable
to do so because of the free-rider problem. As a result, only firms with healthy bal-
ance sheets are able to borrow from the capital market. Under-capitalized firms are
forced either to borrow from banks and raise funds at higher interest rates that re-
flect the cost of monitoring, or self-finance their projects. Monetary policy can affect
the access of firms to external finance because it alters the cost of funds. Crucially,
from the point of view of our paper, we are interested to know how these effects de-
pend on those firm characteristics that credit providers use to identify creditworthy
applicants. Examples of the kind of characteristics that we have in mind are size,
total assets, the ratio of tangible to intangible assets, credit ratings, profitability
and gearing. The predictions from our model are evaluated for a panel of 16,000
manufacturing firms in the UK. We show that the more financially vulnerable firms
– smaller, younger, more risky and more indebted firms – are more severely affected
by monetary tightening as credit supply is withheld. Thus we offer empirical support
for the theoretical model, and can quantify the effects of particular characteristics on
access to external finance. Our results are very similar in character to those of Whited
(1992), Kashyap et al. (1993), Gertler and Gilchrist (1994), Kashyap and Stein
(1994), and Atanasova and Wilson (2004) since we find that collateral assets, per-
ceived riskiness, debt levels and monetary conditions influence access to external
finance.
4
The ability to monitor is also what distinguishes banks from capital markets in Besanko and Kanatas
(1993), Boot and Thakor (1997), Diamond (1991), Holmstrom and Tirole (1997), Hoshi et al. (1993) and
Repullo and Suarez (2000). However, in these models monitoring allows banks to alleviate a moral hazard
problem related to the choice of technologies by firms.
The paper is organized as follows. Section 2 presents our theoretical model that is
used to explore the influence of firm-specific characteristics on the variation in the
composition of external finance as a consequence of contractions and expansions
in monetary policy. The data sources and empirical methodology are discussed in
Section 3 followed by the estimations. Section 4 concludes.
2. The theoretical model
We assume that firms own assets which consist of tangible collateral assets (C)
and intangible assets. The liability side of their balance sheets consists of equity
and debt (DS). The latter is senior relative to any new (junior) debt (DJ), in the sense
that it will be paid off first in the case of default. Since any new finance will be junior
to all existing debt raised previously, at the margin the availability of market finance
will rest on the likelihood that junior debt will be repaid.
New projects require an initial investment F (project size) and generate financial
payoffs kF (k > 1) with probability p and zero with probability 1 p. We assume
that the projects are socially efficient, i.e. pkF > F. Firms need to raise external funds
to finance new projects. Firm owners and potential creditors are risk neutral. All
financial markets are competitive. The opportunity cost of funds is given by the risk-
less interest rate r, which is also the operational variable of monetary policy used by
the Bank of England to target inflation.
Following Townsend (1979), we assume that only firm owners can costlessly ob-
serve project returns. Monitoring the activities of firms allows creditors to verify the
returns reported by firm owners, but monitoring is costly and only banks find it prof-
itable to monitor their clients.5
When firms default on their debt obligations, creditors can liquidate their tangible
assets. We assume that the liquidation value of these assets is uncertain at the time
when liquidation decisions are taken.6 More specifically, with probability p the value
of tangible assets is equal to CH and with probability 1 p is equal to CL (<CH) . Let
pCH + (1 p)CL = C (i.e. the expected liquidation value is equal to the value of the
tangible assets at the time when the financial contract is agreed). Finally, we assume
that when liquidation takes place firms also lose any expected discounted continua-
tion payoffs V.7 For simplicity, we impose the following restrictions on the continu-
ation payoffs and the liquidation values:
5
See Diamond (1984) for a costly-state verification model where financial intermediaries arise
endogenously.
6
In equilibrium liquidation values will be affected by the stance of monetary policy. As Schleifer and
Vishny (1992) have pointed out it is during periods of recessions, when most of the bankruptcies take
place, that liquidation values are at their lowest levels.
7
To keep things simple we have restricted our attention to a static model. This is without any loss of
generality, as long as the optimal financial decision at any time is independent of future investment
opportunities. Nevertheless, we have introduced a continuation payoff in order to allow for risky capital
market financing.
Condition 1: CL > DS.

Condition 2: V > (1 p)(DS + DJ CL).
The first condition states that senior debt is riskless. The second sets a minimum
level for the continuation payoffs. As we will show below, there is no loss of gener-
ality by imposing these restrictions. We introduce them in order to eliminate some
cases that would not further add any new insights to the analysis of our model.
2.1. Market finance
In this section, we consider the case where the only source of external finance is
the capital market. Let DJC denote the amount of debt raised in the capital market.
In the absence of monitoring, firms might have an incentive to misreport their true
payoffs. We begin with the following lemma:
Lemma 1. If DS þ DJC > C H there exists a cut-off value V* for the continuation
payoffs such that if V < V* firms will always default independently of their projectÕs
payoff.
Proof. The expected profit of a firm that truthfully reports its projectÕs payoff is
equal to:
pðkF DS DJC þ V Þ þ ð1 pÞðpðmaxfC H C DS DJC; CgÞ
þ ð1 pÞ maxfC L C DS DJC; CgÞ: ð1Þ
In contrast, the expected payoff of a firm that always defaults is given by:
pkF þ pðmaxfC H C DS DJC; CgÞ þ ð1 pÞ maxfC L C DS DJC; Cg:

ð2Þ
In deriving the above expressions we take into account that payoffs are affected by
whether or not liquidation takes place and by changes in the value of tangible assets.
When firms default there are two possibilities: either the liquidation value is suffi-
ciently high so that the proceeds cover the total debt obligations or the total payoff
to the debtors is restricted by the liquidation value. There are three cases to consider.
Case 1: C L > DS þ DJC : In this case all debt is riskless. Since V is strictly positive we
find that in this case firms never default when the project payoff is positive.
Case 2: C H > DS þ DJC > C L : Subtracting (1) from (2) we find that condition 2
implies that in this case firms never default when the project payoff is
positive.
Case 3: DS þ DJC > C H : Subtracting (1) from (2) we find that if V > DS þ DJC C
firms will not default when the project payoff is positive.
Then V ¼ DS þ DJC C. h
The above lemma has established those conditions under which firms have the
incentives to truthfully reveal their payoffs. From the proof it becomes clear that
the imposition of conditions 1 and 2 is without any loss of generality. If we remove
condition 1 then we will have to consider the effects of senior debt on the incentives
of firms to default but would not change the qualitative comparative statics derived
below. Condition 2 implies that only in case 3 firms might have an incentive to mis-
report their payoffs. The following proposition follows directly from the above
lemma:
Proposition 1. If DS þ DJC > C H and V < V* firms will not be able to fund new
projects in the capital market.
Proof. Lemma 1 implies that when the above inequalities hold firms will default with
certainty. In this case the expected liquidation proceeds will be less than the total
debt obligations. Because senior debt is paid first the expected payoff of any new debt
will be negative. h
Proposition 1 sets a maximum value on the amount of junior debt that firms can
raise in the capital market. Obviously, the amount of junior debt is endogenous and
it depends on the size of the project, the market interest rate and the value of collat-
eral. Next, we calculate DJC; the available new finance from the capital market, and
the rate of interest on new debt. The zero profit condition for creditors requires that:
pDJC þ ð1 pÞðpðminfDJC ; C H DS gÞ þ ð1 pÞ minfDJC; C L DS gÞ ¼ ð1 þ rÞF :
ð3Þ
Lemma 1 implies that we need to consider three cases. Solving for DJC we get:
Case 1: If C L > DS þ DJC then

DJC ¼ ð1 þ rÞF :
Even when the liquidation proceeds are low they are still sufficiently high to cover all
debt and therefore junior debt covers the initial investment and the interest, where
the effective interest rate on debt DJC =F is equal to the gross riskless interest rate. This
last point is not surprising given that in this case debt is riskless.
Case 2: If C H > DS þ DJC > C L then the left-hand side of (3) is equal to
pDJC þ ð1 pÞðpDJC þ ð1 pÞðC L DS ÞÞ:
The inequality C L DS < DJC implies that DJC > ð1 þ rÞF and therefore the effective
interest rate on debt DJC =F is higher than the gross riskless interest rate.
Case 3: If DS þ DJC > C H and V > V* then the left-hand side of (3) is equal to
pDJC þ ð1 pÞðpðC H DS Þ þ ð1 pÞðC L DS ÞÞ:
Again, the inequality C H DS < DJC implies that DJC > ð1 þ rÞF and therefore
the effective interest rate on debt DJC =F is higher than the gross riskless interest rate.
Indeed the effective interest rate is higher than the corresponding rate in case 2.
Up to this point, we have established under what conditions firms can raise funds
in the capital market, we have determined the risk level (if any) of these loans and
have calculated the corresponding interest rates. Next, we turn our attention to inter-
mediary finance.
2.2. Intermediary finance
In contrast to capital markets, banks can monitor the activities of their clients and
thus verify project returns. Townsend (1979) has shown that when monitoring is
costly the optimal deterministic contract is the standard debt contract.8 In our
model, this means that banks verify project returns only when firms report that their
projects have failed. Under the supposition that banks can impose sufficiently high
penalties when firms misreport project returns, firms always have the incentive to re-
port truthfully. We assume that the cost of monitoring M is an increasing function of
the size of the project; i.e. M = m(F), m 0 (F) > 0.9
Monitoring costs make bank credit more expensive than credit from the capital
market, therefore, the only firms that seek bank loans will be those that do not have
access to the capital market. As we have demonstrated in the previous section, these
are firms whose balance sheets satisfy the following two conditions:
DS þ DJC > C H and V > V ¼ DS þ DJC C:
Let DJB denote the amount of new debt owed to banks. The bankÕs zero profit con-
dition implies that the condition (3) in the previous section should be amended to
include the expected monitoring costs, (1 p)m(F), such that:
pDJB þ ð1 pÞðpðC H DS Þ þ ð1 pÞðC L DS ÞÞ ¼ ð1 þ rÞF þ ð1 pÞmðF Þ:
Solving for DJB we get
DJB ¼ fð1 þ rÞF þ ð1 pÞmðF Þ ð1 pÞðpðC H DS Þ þ ð1 pÞðC L DS ÞÞg=p:
ð4Þ
Comparing the interest rates charged by banks to those offered by the capital market
we find that the former are higher by the value of the expected monitoring costs.
8
Townsend (1979) suggested that by expanding the set of admissible contracts to include stochastic ones
we can improve the welfare of participants. Boyd and Smith (1994) compare and contrast the two types of
contracts. Since there are only two states in our model without loss of generality we restrict attention to
deterministic contracts. Allowing for stochastic contracts would mean that banks verify project returns,
when firm owners report the low state, only probabilistically. This would reduce expected monitoring
costs, and thus the interest rates that banks charge on their loans, but it would require that banks can
impose very high penalties on those firms that do not report truthfully. If there is an upper limit on the
penalties that banks can charge (even if you would never observe them in equilibrium) then stochastic
contracts may not be feasible.
9
If we assume that monitoring costs do not vary with the size of project then the effective interest rate on
debt will be decreasing in the size of the loan.
Notice that since the expected payoff of firms is decreasing in DJB if m00 (F) > 0 then
there might be firms that are unable to break-even because of the high interest pay-
ments. Firms indifferent between investing and being inactive are those whose char-
acteristics satisfy the following equality:
pðkF DS DJB þ V Þ þ ð1 pÞpðC H C DS DJB Þ ¼ 0: ð5Þ
These will be firms with very low values of collateral, high risk of default, large pro-
jects relative to their size, and high levels of accumulated debt. Firms with these char-
acteristics might not be able to get access to external finance.
2.3. Model predictions
We consider two sets of predictions arising from the model. First, we wish to
know what the model predicts about access to credit at the margin based on obser-
vable firm-specific characteristics. This should tell us what creditors infer from fac-
tors such as size, profitability, risk, collateral and the debt to equity ratio about
the viability of extending (further) short-term and long-term debt.10 Second, we wish
to know what the model predicts about the effect of monetary policy on the overall
availability of external debt and how this effect varies with firm characteristics.
We derive the first set of predictions from Proposition 1. We can infer that the
higher the level of debt (either existing or new debt), the lower the level of future
profitability (captured by V), and the lower the value of intangible assets (collateral),
the more likely it is that new investments will be financed through short-term bank
loans rather than by long-term debt from the markets. If existing debt levels indicate
extreme vulnerability then firms may not obtain credit from either source.
The value of new debt is endogenous and can be derived using the zero-profit con-
dition for creditors given by (3) (or (4)). This condition implies that the value of new
debt is positively correlated with the size of the project, the level of interest rates, the
level of risk (captured by the inverse of p), and the value of existing debt. The zero
profit conditions in each case imply that the value of new debt is negatively corre-
lated with the level of collateral, and the level of economic activity (when the state
of the economy is good there is a higher likelihood i.e. larger value of p that the value
of collateral will be high).
Our theoretical model also predicts that, other things equal, smaller firms are
more likely to use bank loans to finance their projects. This is because larger firms
might be expected, on average, to be characterized by higher collateral and debt
values than smaller firms, therefore according to our model they will also be able
to finance, on average, larger projects using market finance. Put differently, if we
10
Since short-term debt of maturity 1 to 5 years is dominated by bank loans (because the market for
commercial paper is not as well developed as it is in the US) we can make a distinction between finance
raised predominantly from banks and finance from the capital market. Given also that adjustment at the
margin affects the accumulated bank borrowing and market debt in relation to scalars such as total debt,
total liabilities or turnover we consider ratios when we put these theoretical predictions to the test.
fix the size of investments then larger firms are more likely to finance them using
market funds. We summarize the above predictions in the following proposition:
Proposition 2. Financing investments with bank loans, rather than raising funds in the
capital market, is more likely when (a) the level of existing debt is high, (b) the level of
collateral is low, (c) the level of risk is high, (d) the level of future profitability is low,
and (e) the level of economic activity is low. Our model also predicts that, other things
equal, larger firms are more likely to finance their projects with funds raised in the
capital market.
The second set of predictions relates to the impact of monetary policy on the over-
all availability of external debt. The following proposition is a direct consequence of
the break-even condition (5):
Proposition 3. The volume of market finance will be lower during periods of a tight
monetary policy (high interest rates) relatively to periods of loose monetary policy, and
the effects of a tightening of monetary policy will be stronger during periods of low
economic activity. Firms that are most likely to be affected are those with (a)
low expected profits (captured by either/both high risk (low p) or/and low profitability
(low k)), (b) low collateral, and (c) high debt levels.
In the remaining of this paper we test the above predictions and evaluate their rel-
ative impact on external credit in quantitative terms.
3. An application to UK manufacturing firms
3.1. Data and methodology
The FAME database covers all UK registered companies offering up to 11 years

of detailed information (modified accounts) for about 500,000 large, small and med-
ium sized UK companies. We construct a sample from the FAME Database that al-
lows us some flexibility in analysing the monetary transmission mechanism and
corporate sector finance. The sample is extracted on the following criteria:11
• Firms whose primary activity is classified as manufacturing according to the 1992

SIC UK Code in England, Scotland, Wales and Northern Ireland.12
11
The sample is based on figures that were downloaded in October and November 2001. A sample
selected at a different time but still using the same criteria is likely to be different because of monthly
revisions of firm accounts.
12
The software also includes 940 firms (5.7% of the total sample) whose secondary rather that the
primary activity is classified in the manufacturing sector.
Table 1
Summary statistics
Observations Mean Std. dev. Min Max
Total debt/total liabilities (%) 124,532 41.83 23.38 0.00 100.00
Short-term debt/total debt (%) 124,532 68.34 32.48 0.00 100.00
RATE 145,030 7.48 2.58 5.33 14.75
SPREAD10 145,030 0.31 1.94 3.11 4.15
SIZE 144,983 8.63 1.46 5.34 13.39
RISK SCORE 142,937 57.40 19.53 1.00 96.00
COLLATERAL 140,740 0.33 0.19 0.00 1.00
GEARING 128,628 148.90 254.96 0.25 2561.44
PROFIT 121,439 16.06 31.06 161.00 230.00
AGE 145,030 31.08 23.88 1 136
GDP 145,030 2.27 1.58 1.38 4.66
• Firms that were established prior to 1989 and were still reporting in years 1999
and 2000.13
On this basis we extract 16,000 manufacturing firms with rich information about
firm-specific characteristics. The summary statistics of our variables are reported in
Table 1.
The logarithm of real total assets is used to indicate the impact of SIZE and is
calculated by deflating nominal total assets by the relevant sectoral producer price
index. Size is a vital indicator in the debate between Kashyap et al. (1993) and Oliner
and Rudebusch (1996) and is the key proxy for capital market access for the manu-
facturing sector in Gertler and Gilchrist (1994). Our measure of risk, RISK SCORE,
is the QuiScore measure produced by Qui Credit Assessment Ltd., which assesses the
likelihood of company failure in the 12 months following the date of calculation. It is
based on a statistical analysis of a random selection of companies, and to ensure that
the model is not distorted, three categories are screened out from the initial selection.
These are: major public companies, companies that have insignificant amounts of
unsecured trade credit, and liquidated companies that have a surplus of assets over
liabilities. The score is calculated as a number in the range 0–100, which to facilitate
interpretation is separated into five distinct bands.14 Firms with scores above eighty
are regarded as secure and those below 40 are regarded as high risk. This is a wider
13
In fact, only 3% of the firms in the manufacturing industry stopped reporting during the period of
1990–1999. This may stem from either a failure of the company or because the company entered the
exemption threshold. These drops are prevalent in the first couple of years of the sample period. Therefore,
the sample is not a balanced panel, since firms whose turnover is under the threshold are not observed (the
turnover threshold is £90,000).
14
The bands are: Secure (81–100) where failure is very unusual and normally occurs only as a result of
exceptional changes within the company or its market. Stable (61–80), where company failure is a rare
occurrence and will only come about if there are major company or marketplace changes. Normal (41–60)
containing many companies that do not fail, but some that do. Unstable (21–40) where a significant risk of
company failure exists. and High Risk (0–20) most companies are unlikely to be able to continue trading
unless significant remedial action is undertaken.
definition of perceived financial health or riskiness than the commonly used bond
rating, used by Whited (1992) and Kashyap and Stein (1994), which only applies
to a small fraction of rated firms.
There are four other measures of firm-specific characteristics that we employ. We
introduce AGE as an explanatory variable to measure the importance of a track re-
cord for the change in the composition of firm external finance; the ratio of tangible
assets to total assets to measure COLLATERAL available to support borrowing; the
return to capital, PROFIT, which is a measure of profitability scaled by capital; and
a measure of senior debt that is captured by GEARING, the ratio of total loans to
shareholder funds as an indicator of indebtedness of firms in relation to their equity,
and by DEBT (1), which is debt outstanding at the end of the previous period.
Our explanatory variables includes a non-firm-specific variable that is nevertheless
regarded as an important determinant of credit supply (cf. Atanasova and Wilson,
2004). This is the monetary policy stance variable, which is measured by the level
of the rate of interest set by the Bank of England (the repo rate), and is comparable
to the Fed Funds rate used in US studies as the preferred indicator of monetary con-
ditions by Bernanke and Blinder (1988, 1992), Kashyap et al. (1993), Gertler and
Gilchrist (1994), and Oliner and Rudebusch (1996). We corrobortate our results
using the spread between the base rate and the rate on 10 year government securities
as a further indicator of monetary tightness.
Our sample offers a natural experiment to evaluate the influence of firm-specific
characteristics on the response of corporate finance to monetary policy. The first per-
iod of our sample, 1990–1992, was a tight episode, when monetary policy in the UK
was dedicated towards maintaining the exchange rate within its target zone in the
Exchange Rate Mechanism. The high rates of interest in Germany post-reunification
and the perceived weakness of sterling as a currency contributed to keep UK interest
rates high during this period in order to meet the external policy objective. The per-
iod coincided with a recession and a harsh environment for existing and new corpo-
rate borrowers. The second period, 1993–1999, witnessed a period of sustained
economic growth, falling unemployment and inflation, and interest rates that quickly
approached very low levels in comparison to recent historical experience. The corpo-
rate sector experienced an improvement in net worth and borrowing conditions that
were less constrained. We make use of this natural experiment by interacting the
interest rate (spread) during tight and loose periods of monetary policy with the
other explanatory variables. Our results, reported in the empirical section, show that
there was a marked difference in the response to firm specific characteristics when
interacted with monetary policy.15
To ensure that we can control for the effects of the business cycle and year specific
events we use the GDP growth rate to control for cyclical effects, and time effects
15
We also consider the impact of a tight period dummy variable which we set equal to one for the period
1990–1992. We do not report the results in the empirical section because they are almost identical to the
results based on the interaction with the interest rate (they are available on request).
with dummies for 1992–1999. A literature has indicated that certain types of indus-
tries may be more financially constrained than others (cf. Dedola and Lippi, 2005;
Peersman and Smets, forthcoming) and we control for these effects, although in
our fixed effects results the dummies for the industries drop out when the data are
transformed to remove firm specific effects.16 The inclusion of these variables in
our panel estimates control for remaining demand-side influences, time and industry
effects.
We now turn to the dependent variable – our measure of the financial choice of
the firm. The theoretical model has made predictions about the choice at the margin
between bank debt and marketable debt. This corresponds very closely to Whited
(1992), where the case for looking at bank versus public debt finance rather than
at equity is set out in detail. In the tradition of Kashyap et al. (1993) and Oliner
and Rudebusch (1996), who used ratios of bank loans to total short-term debt, we
derive ratios that change with the decision at the margin. These help to abstract from
demand-side influences to a degree because the factors that influence the uptake of
credit from the demand side affect both numerator and denominator leaving the
ratio relatively unchanged.17 Changes in the ratio are more likely to reflect the influ-
ence of the supply-side.
We have two measures of financial choice based on ratios corresponding to short-
term debt to total debt and total debt to total liabilities.18 The former refers to access
to market finance versus bank finance, where the majority of short-term debt is bank
finance, while the latter refers to the overall availability of external debt (i.e. total
debt).19 We estimate the relationship between the financial choices of firms and their
specific characteristics using a standard panel model that enables us to control for
firm specific unobservable effects and to account for firm heterogeneity. The format
is:
yit ¼ ai þ X it b þ eit ;
where i = 1, 2, . . ., N refers to a cross-section unit (firms in this study), t = 1, 2, . . ., T
refers to time period. yit and Xit denote the dependent variable and the vector of non-
stochastic explanatory variables for firm i and year t, respectively. eit is the error
term, and ai is a vector capturing firm-specific intercepts. A preliminary investigation
that compared estimates from a random effects model against a fixed effects alterna-
tive using the Hausman test rejected the hypothesis of no systematic difference
16
This specification drops the industry dummies as the value of the dummy is one and the average value
of the dummies across years for each industry is also one.
17
We further control for the economic cycle where necessary by including GDP as a regressor and for
year effects using year dummies.
18
Short-term debt is made up of the sum of bank overdrafts, short-term group and director loans, hire
purchase, leasing and other short-term loans, but is predominantly bank finance. Total liabilities is made
of short-term debt, trade credit and total other current liabilities that include some forms of finance
resembling commercial paper or bonds, long-term debt and other long-term liabilities.
19
Our results were unchanged when we considered the ratios of short-term debt and total debt over total
assets.
between coefficients obtained from the two models. Therefore, we report the fixed
effects, which are generally regarded as more efficient.20
3.2. Response to firms-specific characteristics and control variables
In Table 2A and B we evaluate the (supply) response of short-term debt to total

debt and total debt to total liabilities to firm-specific characteristics, where the effect
of monetary policy is captured by the base rate and the spread respectively. The the-
oretical model predicts that the first measure, short-term debt to total debt (which
comprises mostly bank lending) will rise for smaller or riskier firms, those with high-
er levels of debt and less collateral, and with lower profitability, while total debt will
increase for larger, less risky, highly collateralized firms with evidence of a good re-
turn to capital. When we examine the empirical evidence for these effects we find that
the predictions are confirmed in the data. The GDP growth rate controls for cyclical
effects, since an increase in the GDP growth rate encourages firms to shift toward
non-debt liabilities. There are significant time effects with dummies for 1992, 1993,
1996 (all negative, suggesting a shift towards long-term debt as interest rates fell
to lower levels). The inclusion of these variables in our panel estimates control for
remaining demand-side influences.
We find strong positive evidence in favor of the theoretical model from our empir-
ical application in Table 2A and B. First, the logarithm of real assets, taken as an
indicator of firm SIZE, is an important influence on the debt ratios. We observe that
firms with more real assets tend to have greater access to long-term debt and reduce
their short-term debt, hence the signs of the coefficient for the short-term debt to to-
tal debt ratio is negative and for the total debt to total liabilities is positive in re-
sponse to SIZE. This confirms the finding of Gertler and Gilchrist (1994) that size
of the firm is a major determinant of access to bank and marketable debt. Like Oli-
ner and Rudebusch (1996) we find that small firms are heavily dependent on short-
term bank finance.
Second, for the RISK SCORE, we expect a negative sign on this measure indicat-
ing that safer firms will reduce short-term debt (note, our measure based on the Qui-
Score rating takes a higher value the less risky the firm is judged to be over the
following 12 months). This is what we find, and the negative sign on total debt sug-
gests that a good credit rating is also associated with lower debt in general, possibly
because these firms make greater use of non-debt finance. Previous studies such as
Whited (1992) and Kashyap and Stein (1994) have used bond ratings by MoodyÕs
and Standard and PoorÕs respectively, to form a risk assessment on firms. Their argu-
20
Ideally we would like to report dynamic panel GMM estimates, however, the requirement for
instruments under GMM poses a problem for our study. The period when monetary policy was tight
occurs only at the very beginning of our sample. If we were to make use of GMM those observations at the
beginning of our sample would be lost, and the results would be indicative only of access to credit under a
benign period of monetary policy. This would severely undermine the rationale for our empirical work,
and therefore we rely on fixed effects estimates.
Table 2
Firm-specific characteristics and access to credit (Panel A) evidence from rates and (Panel B) evidence
from spreads
Short-term debt/total debt (%) Total debt/total liabilities (%)
Panel A
RATE 0.219*** 0.251*** 0.516*** 0.575***
(2.74) (3.15) (10.33) (11.55)
SIZE 4.534*** 3.755*** 4.081*** 3.982***
(18.89) (16.35) (27.07) (27.73)
RISK SCORE 0.355*** 0.287*** 0.416*** 0.446***
(58.35) (50.29) (108.73) (124.21)
COLLATERAL 38.492*** 37.797*** 14.210*** 13.987***
(41.36) (41.44) (24.35) (24.43)
GEARING 0.003*** 0.003***
(17.48) (32.16)
PROFIT 0.028*** 0.019*** 0.006*** 0.003***
(16.47) (14.24) (5.63) (3.32)
AGE 0.681*** 0.604*** 0.208*** 0.176***
(10.06) (9.00) (4.92) (4.18)
GDP 0.454*** 0.478*** 0.556*** 0.570***
(5.28) (5.59) (10.36) (10.65)
Constant 123.340*** 113.807*** 22.018*** 27.052***
(38.83) (36.70) (11.05) (13.91)
Year92 1.028*** 1.160*** 2.232*** 2.369***
(3.38) (3.84) (11.70) (12.46)
Year93 1.452*** 1.548*** 0.806*** 0.723***
(4.22) (4.52) (3.74) (3.37)
Year95 0.352 0.294 0.192 0.300*
(1.40) (1.17) (1.22) (1.91)
Year96 0.754*** 0.744*** 0.014 0.002
(2.96) (2.93) (0.09) (0.01)
Year97 0.206 0.246 0.744*** 0.845***
(0.74) (0.89) (4.28) (4.86)
Year98 0.232 0.306 1.403*** 1.475***
(0.73) (0.97) (7.04) (7.41)
Observations 105,750 109,900 107,428 112,697
Number of firm 14,750 14,980 14,804 15,062
R-squared 0.06 0.05 0.17 0.17
Panel B
SPREAD10 0.178*** 0.202*** 0.347*** 0.358***
(2.76) (3.14) (8.55) (8.85)
SIZE 4.541*** 3.761*** 4.023*** 3.722***
(18.92) (16.38) (26.61) (25.73)
RISK SCORE 0.355*** 0.287*** 0.411*** 0.442***
(58.34) (50.27) (107.19) (122.78)
COLLATERAL 38.549*** 37.849*** 13.852*** 13.046***
(41.42) (41.49) (23.62) (22.70)
GEARING 0.003*** 0.003***
(17.47) (32.07)
PROFIT 0.028*** 0.019*** 0.007*** 0.003***
(16.46) (14.24) (6.05) (3.42)
(continued on next page)
Table 2 (continued)
Short-term debt/total debt Total debt/total liabilities (%)
(%)
AGE 0.896*** 0.856*** 0.818*** 0.822***
(15.83) (15.21) (22.95) (23.18)
GDP 0.492*** 0.524*** 0.636*** 0.634***
(4.94) (5.28) (10.14) (10.14)
Constant 115.181*** 104.352*** 0.441 6.257***
(47.10) (44.11) (0.29) (4.20)
Year92 0.890*** 1.004*** 2.426*** 2.534***
(3.11) (3.53) (13.45) (14.15)
Year93 0.936*** 0.957*** 2.008*** 2.040***
(3.57) (3.68) (12.15) (12.44)
Year95 0.482* 0.457* 0.442*** 0.370**
(1.81) (1.72) (2.63) (2.21)
Year96 1.072*** 1.129*** 1.119*** 1.139***
(3.34) (3.54) (5.53) (5.66)
Year97 0.096 0.128 0.591*** 0.575***
(0.36) (0.48) (3.51) (3.43)
Year99 0.584 0.710* 2.083*** 2.054***
(1.50) (1.83) (8.48) (8.40)
Observations 105,630 109,777 105,630 109,777
Number of firm 14,728 14,956 14,728 14,956
R-squared 0.06 0.05 0.17 0.17
Absolute value of t-statistics in parentheses.
*
Significant at 10%.
**
Significant at 5%.
***
Significant at 1%.
ment indicates that a bond rating signals those firms that have undergone greater
scrutiny by investors and are less likely to be affected by informational asymmetries
than unrated firms. Our measure is broader than ratings data since all firms are given
a risk score while bond ratings are given for only about 13–20% of the sample
according to Kashyap and Stein (1994).
Third, the ratio of tangible assets in total (COLLATERAL) enhances access to
longer-term debt, reducing the proportion of short-term to total debt while increas-
ing total debt in relation to total liabilities as predicted. Here our findings support
the evidence in Berger and Udell (1990) that collateral is an important factor in
reducing the riskiness of a loan by giving the financial institution a claim on a tan-
gible asset without diminishing its claim on the outstanding debt (Stiglitz and Weiss,
1981). Many firms lack access to marketable debt because they cannot post the nec-
essary collateral to reduce the risk to the lender (Whited, 1992).
Fourth, since senior debt has priority over junior debt in our theoretical model
should the firm liquidate, we therefore predict that firms with more senior debt
are less likely to obtain further access to credit. This may result in less long-term debt
and greater borrowing in the short term from banks, but equally, if the debt level is
high enough to cast doubts over a firmÕs viability, it may reduce all forms of debt. We
measure senior debt using the conventional debt-to-equity ratio (GEARING) but this
has a negative influence on the proportion of short-term to total debt and a positive
influence on total debt to total liabilities21. Both coefficients although small, are
highly significant, and they indicate that short-term debt declines and long-term debt
increases, which accords with neither of the scenarios that we discussed above, but
further investigation of the impact of debt illustrates why this is the case.
Higher levels of existing debt may deter creditors from offering further long-term
credit for firms that are vulnerable on the basis of other characteristics besides
debt.22 For these types of firms long-term debt declines. But for firms that are
healthy on the basis of the other characteristics, creditors may be willing to offer
more debt as they have done in the past. For these firms greater debt does not indi-
cate vulnerability, but rather is the consequence of their success in accessing credit
previously. The type of credit that they are likely to obtain is typically long-term
debt. If the influence of the latter group outweighs the effects of the former on
long-term debt, we would observe growing total debt and declining short-term debt
in response to higher debt levels.23
We illustrate this point in Table 3A and B where we construct interactions be-
tween gearing and indicators that demonstrate that a firm is in the upper or lower
tail of the distribution with respect to other characteristics (respectively firms above
the 25th and below the 75th percentile of the distribution). Thus we can identify the
firms that have high profits, low risk and high collateral, that are strong on other cri-
teria other than debt levels, and compare the impact of gearing in their case with the
firms that are weak on these criteria. Our results show that risky and highly geared
firms in particular are not able to access long-term credit, but can obtain short-term
credit, while firms that are secure and highly geared can access long-term credit.24
Our conclusion is that the impact of debt depends to some degree on the other
indicators of creditworthiness. In Whited (1992) it has been shown that there is an
interaction between riskiness of firms (as measured by the bond rating) and financial
indicators such as the debt-to-total-asset ratio (and also the coverage ratio) when
examining investment equations. Firms without a bond rating that are unscrutinized
have responses in investment equations to indicators of debt-to-asset and coverage
ratios that are roughly twice as large as those for rated firms.
21
Concern that this finding could be the result of the construction of the dependent variable, since both
the dependent variables contain total debt as denominator (numerator) while the explanatory variable also
includes total debt in the numerator, led us to investigate further. The results are unchanged if we make
use of the level of debt outstanding at the end of the previous period TDEBT(1) as the measure of senior
debt so that there is no arithmetic reason for the dependent and the explanatory variables to be linked. The
same results also held if we used other measures of senior debt scaled by real assets or turnover (results not
reported). We conclude that the finding is not simply obtained by the construction of the variables.
22
We explore the characteristics of private and public firms separately later in the paper, since the
existence of debt might have different consequences for private and public firms.
23
Note that this argument cannot be applied to other explanatory variables. Firms may be highly
indebted because they are weak or because they are successful, and have obtained long-term credit in the
past. We cannot think of unprofitable firms, or firms with low collateral in the same way.
24
Other interacted criteria such as size and collateral assets seem less influential, since they have
coefficients that are insignificantly difference from zero. Gearing interacted with low profit levels seems to
reduce access to all types of credit, and vice versa for high profit.
Table 3
Effects of debt interacted with other characteristics (Panel A) evidence from rates and (Panel B) evidence
from spreads
Panel A
RATE 0.216*** 0.214*** 0.216*** 0.509*** 0.519*** 0.511***
(2.71) (2.68) (2.71) (10.24) (10.42) (10.30)
SIZE 4.485*** 4.429*** 4.396*** 4.051*** 4.117*** 4.054***
(18.68) (18.46) (18.32) (26.89) (27.32) (26.93)
RISK SCORE 0.341*** 0.333*** 0.327*** 0.424*** 0.430*** 0.434***
(55.95) (54.30) (53.22) (110.87) (111.48) (112.67)
COLLATERAL 37.953*** 38.710*** 38.276*** 14.297*** 14.454*** 14. 346***
(40.54) (41.11) (40.66) (24.39) (24.50) (24.34)
AGE 0.676*** 0.673*** 0.662*** 0.224*** 0.212*** 0.230***
(10.03) (9.97) (9.84) (5.31) (5.03) (5.46)
GEARING 0.009*** 0.004*** 0.010*** 0.008*** 0.004*** 0.008***
(24.59) (14.43) (19.56) (33.48) (23.17) (25.70)
GEARING * SMALL 0.000 0.000 0.001*** 0.000
(1.22) (0.25) (2.80) (0.95)
GEARING * RISKY 0.008*** 0.007*** 0.005*** 0.004***
(24.27) (18.45) (22.20) (16.23)
GEARING * LCOLL 0.000 0.000 0.001*** 0.001***
(0.54) (0.30) (6.58) (4.27)
GEARING * LPROFIT 0.002*** 0.000 0.002*** 0.003***
(5.09) (1.18) (9.45) (10.81)
GEARING * LARGE 0.000 0.000 0.001*** 0.001***
(0.13) (0.07) (6.86) (5.74)
GEARING * SECURE 0.015*** 0.011*** 0.011*** 0.008***
(21.66) (15.38) (25.36) (18.16)
GEARING * HCOLL 0.001*** 0.001*** 0.001*** 0.001***
(3.92) (3.53) (5.50) (2.83)
GEARING * HPROFIT 0.004*** 0.005*** 0.000 0.002***
(11.66) (11.50) (0.80) (6.75)
PROFIT 0.031*** 0.021*** 0.025*** 0.013*** 0.007*** 0.011***
(16.95) (11.37) (13.25) (11.60) (6.09) (9.08)
GDP 0.444*** 0.450*** 0.442*** 0.562*** 0.557*** 0.563***
(5.19) (5.26) (5.17) (10.52) (10.41) (10.56)
Constant 122.286*** 121.762*** 121.440*** 22.103*** 22.200*** 22. 323***
(38.56) (38.40) (38.34) (11.12) (11.16) (11.25)
Year92 1.002*** 1.025*** 1.004*** 2.208*** 2.227*** 2.209***
(3.30) (3.38) (3.32) (11.63) (11.72) (11.66)
Year93 1.452*** 1.417*** 1.430*** 0.823*** 0.782*** 0.802***
(4.23) (4.13) (4.17) (3.84) (3.64) (3.74)
Year95 0.335 0.334 0.318 0.172 0.183 0.164
(1.34) (1.33) (1.27) (1.09) (1.17) (1.04)
Year96 0.700*** 0.715*** 0.672*** 0.026 0.009 0.043
(2.75) (2.81) (2.65) (0.16) (0.06) (0.27)
Year97 0.241 0.223 0.260 0.716*** 0.743*** 0.709***
(0.87) (0.81) (0.94) (4.13) (4.28) (4.10)
Year99 0.255 0.241 0.267 1.374*** 1.406*** 1.376***
(0.80) (0.76) (0.84) (6.92) (7.08) (6.94)
Observations 105,750 105,750 105,750 107,428 107,428 107,428
Table 3 (continued)
Number of firm 14,750 14,750 14,750 14,804 14,804 14,804
R-squared 0.06 0.06 0.07 0.18 0.18 0.19
Panel B
SPREAD 10 0.173*** 0.172*** 0.172*** 0.344*** 0.349*** 0.345***
(2.69) (2.68) (2.69) (8.50) (8.63) (8.55)
SIZE 4.491*** 4.435*** 4.401*** 3.992*** 4.054*** 3.992***
(18.71) (18.48) (18.34) (26.43) (26.83) (26.44)
RISK SCORE 0.341*** 0.333*** 0.327*** 0.420*** 0.425*** 0.429***
(55.93) (54.28) (53.20) (109.31) (109.89) (111.08)
COLLATERAL 38.008*** 38.766*** 38.330*** 13.942*** 14.116*** 14. 006***
(40.60) (41.16) (40.72) (23.67) (23.80) (23.64)
GEARING 0.009*** 0.004*** 0.010*** 0.008*** 0.004*** 0.008***
(24.59) (14.41) (19.56) (33.24) (22.94) (25.41)
GEARING * SMALL 0.000 0.000 0.001*** 0.000
(1.24) (0.25) (2.73) (1.05)
GEARING * RISKY 0.008*** 0.007*** 0.005*** 0.004***
(24.29) (18.47) (21.88) (16.01)
GEARING * LCOLL 0.000 0.000 0.001*** 0.001***
(0.53) (0.30) (6.54) (4.20)
GEARING * LPROFIT 0.002*** 0.000 0.002*** 0.003***
(5.12) (1.17) (9.47) (10.77)
GEARING * LARGE 0.000 0.000 0.001*** 0.001***
(0.14) (0.05) (6.44) (5.30)
GEARING * SECURE 0.015*** 0.011*** 0.011*** 0.008***
(21.67) (15.38) (24.94) (17.83)
GEARING * HCOLL 0.001*** 0.001*** 0.001*** 0.001***
(3.92) (3.53) (5.57) (2.91)
GEARING * HPROFIT 0.004*** 0.005*** 0.000 0.002***
(11.67) (11.52) (0.87) (6.63)
PROFIT 0.031*** 0.021*** 0.025*** 0.014*** 0.008*** 0.011***
(16.94) (11.36) (13.24) (11.99) (6.49) (9.48)
AGE 0.891*** 0.884*** 0.879*** 0.825*** 0.825*** 0.832***
(15.79) (15.66) (15.61) (23.23) (23.22) (23.48)
GDP 0.481*** 0.487*** 0.479*** 0.641*** 0.637*** 0.643***
(4.85) (4.91) (4.84) (10.26) (10.20) (10.31)
Constant 114.151*** 113.753*** 113.253*** 0.851 0.545 1.022
(46.73) (46.54) (46.39) (0.55) (0.35) (0.67)
Year92 0.868*** 0.892*** 0.871*** 2.399*** 2.423*** 2.402***
(3.04) (3.13) (3.06) (13.37) (13.49) (13.41)
Year93 0.944*** 0.915*** 0.924*** 2.013*** 1.992*** 1.996***
(3.61) (3.50) (3.54) (12.23) (12.10) (12.16)
Year94 0.471* 0.465* 0.457* 0.451*** 0.452*** 0.460***
(1.77) (1.75) (1.72) (2.70) (2.70) (2.76)
Year96 1.023*** 1.030*** 0.999*** 1.141*** 1.145*** 1.159***
(3.20) (3.22) (3.13) (5.67) (5.69) (5.77)
Year97 0.073 0.080 0.061 0.596*** 0.596*** 0.604***
(0.27) (0.30) (0.23) (3.56) (3.55) (3.61)
Year99 0.599 0.583 0.610 2.048*** 2.090*** 2.051***
(1.54) (1.50) (1.57) (8.37) (8.54) (8.41)
Table 3 (continued)
Observations 105,630 105,630 105,630 105,630 105,630 105,630
Number of firm 14,728 14,728 14,728 14,728 14,728 14,728
R-squared 0.06 0.06 0.07 0.18 0.18 0.18
*
**
Significant at 5%.
***
Significant at 1%.
Two other variables need to be discussed. A good rate of return on capital

(PROFIT) should improve access to short-term and long-term debt. In practice we
find that it improves access to short-term debt but marginally reduces the ratio of
total debt to total liabilities. A further variable, AGE, also appears to be a significant
explanatory for both short-term and total debt. There is no prediction from our
model concerning the impact of this variable, but the empirical finding accords with
the predictions of other models. AGE provides a confirmation of the importance of a
track record for certain types of firms and this is a direct test of the relationship-
banking proposition suggested by Sharpe (1990), Diamond (1991), Rajan (1992)
and Boot (2000). Firms that are weak on other criteria but nevertheless have a track
record are likely to be less financially constrained than firms that are younger and
that have not been able to build relationships with their lenders.
We find that a monetary policy tightening (higher values of RATE or larger
SPREAD10) leads to a tightening of the supply of debt independently of firm-spe-
cific characteristics. In all cases a monetary tightening significantly reduces available
credit as expected and this supports the broad credit channel (Oliner and Rudebusch,
1996). In the next subsection we explore in much greater detail how a tightening of
monetary policy alters the credit composition directly and indirectly through inter-
action with the firm-specific characteristics.
3.3. Monetary policy, firm characteristics and the financial mix
Our main purpose in this section is to report how the response to firm-specific
characteristics varies with monetary policy. We report our findings of the impact
of monetary policy on credit ratios by constructing interactions between our explan-
atory variables and the interest rate (spread). These interactive terms tell us how the
response to these variables changes when monetary policy (and hence available
external finance) tightens, as indicated by the level of the interest rate (or the spread).
Gertler and Gilchrist (1994) undertook a similar exercise based only on size. They
found greater sensitivity in sales, inventories and short-term debt on the part of small
firms to indicators of monetary tightening (using Romer dates) than for large firms.
In all cases indicators of tightening could not be excluded as explanatory variables
for small firms and the t-statistics on the sum of their coefficients were negative
and significant for small firms but insignificant for large firms. We therefore expect
to find that the volume of market finance declines as the interest rate rises, and that
smaller firms, those with low profits and collateral, higher risk and greater debt will
be proportionately more affected. Our results are reported in Table 4A and B,
where we use the base rate and the short–long spread to indicate monetary policy
tightness.
We report two columns for each ratio to allow for the effects of inclusion and
exclusion of GEARING. We find that there is no variation in the impact of GEAR-
ING as rates rise, and the responses to the other explanatory variables are unchanged
by the exclusion of GEARING. For other variables there are significant differences in
the response of creditors with rising rates.
We find that the effects of SIZE lessen with higher levels of the interest rate and
larger spreads. This means that being larger is less of an advantage in terms of gain-
ing access to capital markets (smaller negative sign for the first ratio) and in terms of
gaining more debt (smaller positive sign for the second ratio)25 as monetary policy
tightens. Likewise, the effect of COLLATERAL is less advantageous during these
periods, but is still strongly influential.
A better RISK SCORE and a longer period of incorporation (indicated by AGE)
are more advantageous with higher interest rates or larger spreads in terms of reduc-
ing the short-term to total debt ratio (gaining access to the capital market) but less
advantageous in gaining more total debt in relation to total liabilities. This reflects
the fact that as rates rise the available stock of total debt is more constrained, but
firms may still access longer-term debt at lower rates of interest if they have better
risk scores and are older.
For GEARING and PROFIT, where the original effects were small and marginal,
there is negligible evidence of a significant variation in response to higher levels of
the interest rate.
Very similar results were obtained for all these variables when we considered
interactions with a dummy variable indicating tight policy for the period 1990–
1992. This measure has some advantages in partitioning the data sample between
ÔtightÕ and ÔlooseÕ periods, but it does not indicate the degree of monetary tightness
as the level of the interest rate does. Our results were almost identical to the previous
results reported above, confirming that our findings are robust to the measure of
tightness or looseness of monetary conditions.
This section has shown that the responses by lenders to firm-specific characteris-
tics differ with the level of the interest rate since size and collateral are less influential
when credit markets are tightening, while risk scores and age become more impor-
tant in these periods. We conclude that there is substantial evidence that the reaction
to monetary policy varies considerably because of the influence of the credit channel
and it depends heavily on many firm-specific characteristics and not just size.
25
In all cases the effect of higher interest rates is found by adding the coefficient on the variable to the
coefficient on the relevant interactive term e.g. the coefficient on AGE plus the coefficient on
AGE * RATE.
Table 4
Allowing for interactions (Panel A) with the interest rate and (Panel B) with the spreads
Panel A
RATE 2.479*** 2.518*** 0.965*** 1.170***
(12.76) (13.33) (7.88) (9.82)
SIZE 7.586*** 6.811*** 3.834*** 3.626***
(27.26) (25.34) (21.85) (21.43)
SIZE * RATE 0.345*** 0.347*** 0.037*** 0.051***
(18.90) (19.17) (3.23) (4.51)
RISK SCORE 0.219*** 0.146*** 0.449*** 0.482***
(16.37) (11.85) (53.87) (62.77)
RISK SCORE * RATE 0.019*** 0.020*** 0.004*** 0.005***
(12.26) (13.71) (4.30) (5.11)
COLLATERAL 46.421*** 46.483*** 19.790*** 19.231***
(31.13) (31.65) (21.08) (20.80)
COLLATERAL * RATE 1.007*** 1.100*** 0.737*** 0.694***
(6.62) (7.32) (7.71) (7.36)
GEARING 0.003*** 0.003***
(7.73) (12.32)
GEARING * RATE 0.000 0.000
(1.10) (0.05)
PROFIT 0.033*** 0.018*** 0.003 0.003
(7.34) (4.86) (0.90) (1.18)
PROFIT * RATE 0.000 0.000 0.001*** 0.001***
(0.83) (0.61) (3.38) (2.62)
AGE 0.874*** 0.811*** 0.094*** 0.076**
(15.38) (14.41) (2.63) (2.14)
AGE * RATE 0.013*** 0.014*** 0.004*** 0.004***
(11.30) (11.61) (5.05) (4.90)
GDP 1.092*** 1.161*** 1.326*** 1.249***
(3.42) (3.66) (6.60) (6.24)
GDP * RATE 0.183*** 0.193*** 0.084*** 0.073***
(4.87) (5.18) (3.54) (3.10)
Year94 0.561* 0.521 1.521*** 1.746***
(1.71) (1.60) (7.35) (8.48)
Year95 0.571** 0.674** 0.201 0.042
(2.16) (2.56) (1.21) (0.25)
Year96 0.107 0.060 0.319** 0.288*
(0.42) (0.24) (1.98) (1.80)
Year97 0.410 0.454* 0.755*** 0.869***
(1.48) (1.65) (4.34) (5.00)
Year98 0.493 0.560* 1.447*** 1.531***
(1.57) (1.79) (7.33) (7.77)
Constant 133.830*** 123.803*** 30.568*** 36.101***
(41.56) (39.56) (15.06) (18.29)
Observations 105,750 109,900 107,428 112,697
Number of firm 14,750 14,980 14,804 15,062
R-squared 0.07 0.06 0.18 0.17
Table 4 (continued)
Panel B
SPREAD10 2.198*** 2.207*** 2.098*** 2.168***
(8.17) (8.46) (12.37) (13.18)
SIZE 4.721*** 3.946*** 3.985*** 3.683***
(19.67) (17.18) (26.33) (25.44)
SIZE * SPREAD10 0.288*** 0.288*** 0.089*** 0.095***
(12.01) (12.14) (5.90) (6.36)
RISK SCORE 0.349*** 0.281*** 0.415*** 0.445***
(56.56) (48.57) (106.43) (121.96)
RISK SCORE * SPREAD10 0.011*** 0.012*** 0.005*** 0.005***
(5.31) (6.17) (3.93) (4.39)
COLLATERAL 39.217*** 38.551*** 14.390*** 13.597***
(41.83) (41.96) (24.34) (23.47)
COLLATERAL * SPREAD10 1.090*** 1.148*** 0.846*** 0.868***
(5.58) (5.96) (6.87) (7.14)
GEARING 0.003*** 0.003***
(17.42) (31.26)
GEARING * SPREAD10 0.000 0.000
(0.74) (1.45)
PROFIT 0.028*** 0.019*** 0.007*** 0.003***
(16.17) (13.86) (5.91) (3.38)
PROFIT * SPREAD10 0.000 0.001 0.000 0.000
(0.23) (0.99) (0.50) (0.04)
AGE 1.107*** 1.071*** 0.500*** 0.500***
(16.33) (15.90) (11.71) (11.79)
AGE * SPREAD10 0.005*** 0.005*** 0.002* 0.002*
(2.92) (3.18) (1.81) (1.70)
GDP 0.523*** 0.556*** 0.751*** 0.754***
(5.20) (5.54) (11.84) (11.93)
GDP * SPREAD10 0.129*** 0.131*** 0.307*** 0.312***
(3.20) (3.29) (12.11) (12.39)
Constant 109.442*** 98.612*** 12.541*** 18.398***
(40.08) (37.21) (7.29) (11.01)
Year92 0.192 0.289 0.950*** 1.037***
(0.62) (0.95) (4.90) (5.38)
Year95 0.029 0.062 1.756*** 1.703***
(0.10) (0.23) (9.98) (9.74)
Year96 0.791** 0.840*** 2.344*** 2.381***
(2.42) (2.58) (11.35) (11.60)
Year97 0.053 0.027 1.146*** 1.135***
(0.20) (0.10) (6.85) (6.81)
Year99 0.935** 1.060*** 2.101*** 2.077***
(2.39) (2.73) (8.53) (8.47)
Observations 105,630 109,777 105,630 109,777
Number of firm 14,728 14,956 14,728 14,956
R-squared 0.06 0.05 0.17 0.17
*
**
Significant at 5%.
***
Significant at 1%.
Table 5
Evidence for public and private firms – evidence from rates and spreads
Short-term debt/total debt (%) Total debt/total liablities (%)
Panel A
RATE 0.629*** 0.672*** 0.715*** 0.850***
(7.11) (7.64) (12.97) (15.50)
RATE * PF 1.091*** 1.121*** 0.379*** 0.466***
(10.45) (10.80) (5.77) (7.08)
SIZE 3.868*** 2.876*** 3.857*** 4.128***
(12.78) (9.96) (20.22) (23.21)
SIZE * PF 1.088** 1.674*** 1.097*** 1.179***
(2.17) (3.49) (3.46) (3.89)
RISK SCORE 0.346*** 0.283*** 0.404*** 0.436***
(46.83) (40.77) (86.83) (100.03)
RISK SCORE * PF 0.034*** 0.016 0.084*** 0.085***
(2.59) (1.33) (10.13) (10.93)
COLLATERAL 41.974*** 40.852*** 16.633*** 18.120***
(37.86) (37.59) (23.87) (26.77)
COLLATERAL * PF 11.228*** 9.652*** 9.379*** 11.251***
(5.52) (4.84) (7.29) (8.88)
GEARING 0.003*** 0.004***
(12.96) (27.86)
GEARING * PF 0.000 0.001***
(1.36) (5.20)
PROFIT 0.022*** 0.016*** 0.007*** 0.004***
(10.72) (9.76) (5.36) (4.07)
PROFIT * PF 0.020*** 0.011*** 0.005** 0.007***
(5.31) (3.88) (2.17) (3.62)
AGE 0.578*** 0.508*** 0.206*** 0.113**
(7.94) (7.03) (4.51) (2.49)
AGE * PF 0.197** 0.195** 0.147*** 0.121**
(2.25) (2.26) (2.65) (2.20)
GDP 0.672*** 0.709*** 0.713*** 0.755***
(6.80) (7.20) (11.53) (12.27)
GDP * PF 0.569*** 0.603*** 0.265*** 0.333***
(3.87) (4.13) (2.85) (3.58)
Constant 122.617*** 112.915*** 23.131*** 26.065***
(38.59) (36.40) (11.51) (13.31)
Year92 1.014*** 1.140*** 2.159*** 2.367***
(3.34) (3.77) (11.25) (12.34)
Year93 1.541*** 1.635*** 0.570*** 0.385*
(4.48) (4.78) (2.63) (1.78)
Year95 0.304 0.244 0.358** 0.581***
(1.21) (0.98) (2.25) (3.65)
Year96 0.740*** 0.734*** 0.144 0.197
(2.91) (2.90) (0.89) (1.22)
Year97 0.321 0.362 1.034*** 1.282***
(1.16) (1.31) (5.89) (7.31)
Year98 0.406 0.485 1.645*** 1.855***
(1.28) (1.53) (8.19) (9.26)
Table 5 (continued)
Observations 105,630 109,777 110,349 117,781
Number of firm 14,728 14,956 14,872 15,213
R-squared 0.06 0.05 0.18 0.17
Panel B
SPREAD10 -0.440*** 0.471*** -0.507*** -0.618***
(6.02) (6.48) (11.11) (13.62)
SPREAD * PF 0.716*** 0.736*** 0.265*** 0.353***
(7.05) (7.28) (4.13) (5.48)
SIZE 3.981*** 2.987*** 3.821*** 4.092***
(13.16) (10.35) (20.05) (23.03)
SIZE * PF 0.818 1.410*** 1.185*** 1.269***
(1.64) (2.94) (3.74) (4.19)
RISK SCORE 0.346*** 0.283*** 0.404*** 0.436***
(46.82) (40.79) (86.81) (100.03)
RISK SCORE * PF 0.034*** 0.016 0.084*** 0.085***
(2.60) (1.28) (10.13) (10.90)
COLLATERAL 41.898*** 40.769*** 16.662*** 18.157***
(37.77) (37.50) (23.91) (26.82)
COLLATERAL * PF 10.990*** 9.418*** 9.470*** 11.370***
(5.40) (4.72) (7.36) (8.97)
GEARING 0.003*** 0.004***
(12.89) (27.89)
GEARING * PF 0.001 0.001***
(1.45) (5.25)
PROFIT 0.022*** 0.016*** 0.007*** 0.004***
(10.69) (9.73) (5.38) (4.10)
PROFIT * PF 0.020*** 0.012*** 0.005** 0.007***
(5.38) (3.95) (2.21) (3.67)
AGE 1.044*** 1.017*** 1.003*** 1.047***
(16.80) (16.48) (25.85) (27.19)
AGE * PF 0.376*** 0.392*** 0.349*** 0.373***
(4.90) (5.17) (7.21) (7.73)
GDP 0.635*** 0.677*** 0.822*** 0.892***
(5.68) (6.08) (11.77) (12.82)
GDP * PF 0.317** 0.344** 0.192** 0.268***
(2.10) (2.30) (2.02) (2.80)
Constant 112.325*** 101.274*** 3.639** 5.005***
(45.65) (42.58) (2.34) (3.37)
Year92 0.852*** 0.959*** 2.476*** 2.747***
(2.98) (3.38) (13.72) (15.23)
Year93 0.879*** 0.896*** 1.934*** 2.007***
(3.35) (3.44) (11.72) (12.19)
Year95 0.505* 0.481* 0.341** 0.218
(1.90) (1.81) (2.02) (1.29)
Year96 1.189*** 1.254*** 1.199*** 1.361***
(3.71) (3.93) (5.93) (6.73)
Year97 0.147 0.183 0.496*** 0.468***
(0.55) (0.69) (2.93) (2.77)
Year99 0.815** 0.946** 2.568*** 2.962***
(2.09) (2.43) (10.47) (12.10)
Table 5 (continued)
Observations 105,630 109,777 110,349 117,781

Number of firm 14,728 14,956 14,872 15,213
R-squared 0.06 0.05 0.18 0.17

*
**
Significant at 5%.
***
Significant at 1%
3.4. Public versus private firms
In a final investigation we consider whether the behavior of public firms differs

from that of private firms. The results, for both the level of interest rates and the
short–long spread are reported in Table 5A and B. We interact a dummy variable
representing the private firms (PF) with the variables described above. As with Table
2A and B we find very similar signs and magnitudes on all the variables and therefore
our discussion for the full sample holds when we isolate the public firms in these ta-
bles. For private firms there is a significant difference in response compared to public
firms. For private firms, SIZE, RISK SCORE, PROFIT and AGE are more impor-
tant as indicators for access to short-term debt and total debt, while COLLATERAL
is less important. The responses to firm-specific variables do not differ when we mea-
sure monetary policy stances using the level or the spread of interest rates. With re-
spect to monetary policy, the direct effect of an interest rate increase (widening
spread) is also different for private firms, since short-term debt increases, suggesting
the possibility that the cost of interest increases is accumulated in the short-term
debt, and total debt falls relative to total liabilities by a smaller amount. The re-
sponse of public firms matches that of the full sample results reported above. We
conclude that public and private firms face different credit supply conditions based
on their specific characteristics, and respond differently as monetary policy alters.
4. Conclusions
This paper has examined the proposition that credit provision varies across the
monetary cycle according to firm specific characteristics. The foundation for the
empirical findings is based on a theoretical framework that models access to credit
within a costly state verification environment (Townsend, 1979). External finance
is available either from the market or from financial intermediaries where only the
latter can verify project returns. By evaluating the creditworthiness of firms, external
finance can be obtained from these two sources, provided certain zero-profit condi-
tions are satisfied. These conditions determine the availability of credit and the rate
charged for borrowing. Our application relies on specific predictions from this model
that can be evaluated against a large panel of firm level data.
The results show that smaller, more risky and younger firms are more noticeably
affected by monetary tightening than larger, secure, or older firms. The role of asset
size and especially tangible assets that can be used as collateral is strongly empha-
sized. The paper therefore confirms the findings of major studies relating to the
credit channel. There is a broad credit channel effect (Oliner and Rudebusch,
1996), as well as a bank-lending channel (Kashyap et al., 1993; Gertler and Gil-
christ, 1994), accelerator effects (Kiyotaki and Moore, 1997; Bernanke, 1996), and
evidence consistent with relationship banking when age proxies for the development
of such bank-firm relationships (Rajan, 1992; Berlin and Mester, 1999; Boot, 2000).
We conclude that Oliner and Rudebusch (1996) were right to point out the impor-
tance of distinguishing between firm types, but for the UK, the effects of making this
distinction do not undermine the findings of Kashyap et al. (1993). We observe that
the empirical evidence supports Oliner and Rudebusch (1996) since we confirm that
size is an important determinant of short-term debt availability but other evidence
based on other characteristics of the firm such as collateral assets, risk scores and
profitability, suggest that size is not the only influence on the availability of credit.
We conclude that many firm-specific characteristics, including size, collateral, riski-
ness, age and profitability are important determinants of access to short-term and
long-term credit as well as prevailing monetary conditions.
There is a close parallel between the literature on credit risk measurement and the
results reported here. As documented in Lowe (2002) the common building blocks of
credit risk assessment models are a rating system based on the probability of default,
an evaluation of the correlation between these probabilities across firms, the likely
loss should default occur, and finally correlations between the probability of default
and loss given default. This paper provides some evidence from revealed behavior on
the supply side of the credit market as to how the borrowers regard indicators of
creditworthiness, which is closely related to the first building block of credit risk
models. The revealed sensitivities to balance sheet characteristics may contribute
to the improvement of rating systems. Where Lowe argues that macroeconomic con-
siderations should be integrated into credit rating assessments reflecting the procyc-
lical forces operating between the macroeconomy and the banking industry, we
illustrate the procyclicality in access to credit both theoretically and empirically, sep-
arating the effects of a general improvement in economic conditions from idiosyn-
cratic effects that are firm-specific.
Acknowledgments
The second author thanks the European University Institute, Florence and the
Research Department, European Central Bank for their generous hospitality during
the period that this paper was written. We acknowledge beneficial comments on an
earlier version from the editor and anonymous referees, Mike Artis, Anindya Baner-
jee, Giuseppe Bertola, Alan Duncan, Stefan Gerlach, John Goddard, Phil Mulli-
neux, Robert Osei, Philip Vermeulen, Peter Wright, seminar participants at the
Hong Kong Monetary Authority and the UNU/INTECH conference on ÔEuropean
Financial Systems and the Corporate SectorÕ, Maastricht, October 2002. Any
remaining errors are our own.
References
Altman, E., 1969. Financial ratios, discriminant analysis and the prediction of corporate bankruptcy.
Journal of Finance 20, 589–609.
Atanasova, C.V., Wilson, N., 2004. Disequilibrium in the UK corporate loan market. Journal of Banking
and Finance 28, 595–614.
Berger, A., Udell, G., 1990. Collateral, loan quality and bank risk. Journal of Monetary Economics 25,
21–42.
Berlin, M., Mester, L.J., 1999. Deposits and relationship lending. Review of Financial Studies 12, 579–
607.
Bernanke, B.S., Blinder, A.S., 1988. Credit, money and aggregate demand. American Economic Review
78, 435–439.
Bernanke, B.S., Gertler, M., 1989. Business fluctuations and the choice between internal and external
finance. American Economic Review 79, 14–31.
Bernanke, B.S., Gertler, M., 1995. Inside the black box: The credit channel of monetary policy
transmission. Journal of Economic Perspectives 9, 27–48.
Bernanke, B.S., Gertler, M., 1996. The financial accelerator and the flight to quality. Review of Economics
and Statistics 78, 1–15.
Bernanke, B.S., Gertler, M., Gilchrist, S., 1999. The financial accelerator in a quantitative business cycle
framework. In: Taylor, J.B., Woodford, M. (Eds.), Handbook of Macroeconomics, vol. 1C, Chapter
21. Elsevier, North-Holland, pp. 1341–1393.
Besanko, D., Kanatas, G., 1993. Credit market equilibrium with bank monitoring and moral hazard.
Review of Financial Studies 6, 213–232.
Boot, A.W.A., 2000. Relationship banking: What do we know? Journal of Financial Intermediation 9, 26–
56.
Boot, A.W.A., Thakor, A., 1997. Financial system architecture. Review of Financial Studies 10, 693–733.
Boyd, J., Smith, B., 1994. How good are standard debt contracts? Stochastic versus non-stochastic
monitoring in a costly state verification environment. Journal of Business 67, 539–561.
Dedola, L., Lippi, F., 2005. The monetary transmission mechanism: Evidence from the industry data of
five OECD countries. European Economic Review 49, 1543–1569.
Diamond, D., 1984. Financial intermediation and delegated monitoring. Review of Economic Studies 51,
393–414.
Diamond, D., 1991. Monitoring and reputation: The choice between bank loans and directly placed debt.
Journal of Political Economy 99, 689–721.
Fama, E., 1984. WhatÕs different about banks? Journal of Monetary Economics 15, 29–39.
Fazzari, S.G., Hubbard, G., Petersen, B., 1988. Financing constraints and corporate investment.
Brookings Papers on Economic Activity 2, 141–195.
Friedman, B., Kuttner, K., 1993. Economic activity and the short-term credit markets: An analysis of
prices and quantities. Brookings Papers on Economic Activity 2, 193–266.
Fuerst, T.S., 1995. Monetary and financial interaction in the business cycle. Journal of Money, Credit and
Banking 27, 1321–1338.
Gertler, M., Gilchrist, S., 1994. Monetary policy, business cycles and the behavior of small manufacturing
firms. Quarterly Journal of Economics 109, 309–340.
Hillegeist, S.A., Keating, E.K., Cram, D.P., Lundsted, K.G., 2004. Assessing the probability of
bankruptcy. Review of Accounting Standards 9, 5–34.
Holmstrom, B., Tirole, J., 1997. Financial intermediation, loanable funds and real sector. Quarterly
Journal of Economics 112, 663–691.
Hoshi, T., Kashyap, A., Scharfstein, D., 1993. The choice between public and private debt: An analysis of
post-deregulation corporate financing in Japan. NBER Working Paper No. 4317.
Hubbard, G., 1998. Capital market imperfections and investment. Journal of Economic Literature 36,
193–225.
Kaplan, S.N., Zingales, L., 1997. Do investment-cash flow sensitivities provide useful measures of finance
constraints? Quarterly Journal of Economics 112, 169–215.
Kaplan, S.N., Zingales, L., 2000. Investment-cash flow sensitivities are not valid measures of financing
constraints. Quarterly Journal of Economics 115, 707–712.
Kashyap, A.K., Stein, J.C., 1994. Monetary policy and bank lending. In: Mankiw, N.G. (Ed.), Monetary
Policy. Chicago University Press, Chicago, pp. 221–256.
Kashyap, A.K., Stein, J.C., Wilcox, D.W., 1993. Monetary policy and credit conditions: evidence from the
composition of external finance. American Economic Review 83, 78–98.
Kashyap, A.K., Stein, J.C., Wilcox, D.W., 1996. Monetary policy and credit conditions: Evidence from
the composition of external finance. American Economic Review 86, 310–314.
Kiyotaki, N., Moore, J., 1997. Credit cycles. Journal of Political Economy 105, 211–248.
Leland, H., Pyle, D., 1977. Information asymmetries, financial structures and financial intermediaries.
Lowe, P., 2002. Credit risk measurement and procyclicality. Bank for international settlements. Working
Paper No. 116.
Merton, R.C., 1974. On the pricing of corporate debt: The risk structure of interest rates. Journal of
Finance 29, 449–470.
Oliner, S., Rudebusch, G., 1996. Monetary policy and credit conditions: Evidence from the composition of
external finance: Comment. American Economic Review 86, 300–309.
Peersman, G., Smets, F., forthcoming. The Industry Effects of Monetary Policy in the Euro Area.
Economic Journal.
Rajan, R.G., 1992. Insiders and outsiders: The choice between informed and arms-length debt. Journal of
Finance 47, 1367–1400.
Repullo, R., Suarez, J., 2000. Enterpreneurial moral hazard and bank monitoring: A model of the credit
channel. European Economic Review 44, 1931–1950.
Romer, C.D., Romer, D.H., 1990. New evidence on the monetary transmission mechanism. Brookings
Papers on Economic Activity 1, 149–213.
Schleifer, A., Vishny, R., 1992. Liquidation values and debt capacity: A market equilibrium approach.
Sharpe, S.A., 1990. Asymmetric information, bank lending, and implicit contracts: a stylized model of
customer relationships. Journal of Finance 45, 1069–1087.
Shumway, T., 2001. Forecasting bankruptcy more accurately: A simple hazard model. Journal of Business
74, 101–124.
Stiglitz, J., Weiss, A., 1981. Credit rationing in markets with imperfect information. American Economic
Reveiw 71, 393–410.
Townsend, R., 1979. Optimal contracts and competitive markets with costly state verification. Journal of
Economic Theory 21, 265–293.
Whited, T., 1992. Debt, liquidity constraints and corporate investment: Evidence from panel data. Journal
of Finance 47, 1425–1460.
Williamson, S., 1987. Financial intermediation, business failures, and real business cycles. Journal of
Political Economy 95, 1196–1216.

Access To External Finance Theory and Evidence

Загружено:

Сведения о документе

Исходное описание:

Оригинальное название

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Access To External Finance Theory and Evidence

Загружено:

Авторское право:

Доступные форматы

Journal of Banking & Finance 30 (2006) 199–227

Access to external ﬁnance: Theory and

Received 18 May 2004; accepted 31 January 2005

JEL classiﬁcation: E32; E44; E51

Keywords: Credit channel of monetary transmission; External ﬁnance

2. The theoretical model

Condition 1: CL > DS.

2.1. Market ﬁnance

pkF þ pðmaxfC H C DS DJC; CgÞ þ ð1 pÞ maxfC L C DS DJC; Cg:

Case 1: If C L > DS þ DJC then

2.2. Intermediary ﬁnance

2.3. Model predictions

3. An application to UK manufacturing ﬁrms

3.1. Data and methodology

The FAME database covers all UK registered companies oﬀering up to 11 years

• Firms whose primary activity is classiﬁed as manufacturing according to the 1992

3.2. Response to ﬁrms-speciﬁc characteristics and control variables

In Table 2A and B we evaluate the (supply) response of short-term debt to total

Two other variables need to be discussed. A good rate of return on capital

3.3. Monetary policy, ﬁrm characteristics and the ﬁnancial mix

Observations 105,630 109,777 110,349 117,781

Absolute value of t-statistics in parentheses.

3.4. Public versus private ﬁrms

In a final investigation we consider whether the behavior of public firms differs

Вам также может понравиться