The Financial Sector and the Real Economy during the Financial
Crisis: Evidence from the Commercial Paper Market
Ethan Cohen-Cole
University of Maryland - College Park
Gustavo Suarez
Federal Reserve Board
Judit Montoriol-Garriga
Federal Reserve Bank of Boston
Jason Wu
Federal Reserve Board
November 19, 2010
Abstract
Shocks to the …nancial sector led credit spreads to widen to unprecedented levels in many
markets during the 2007-2008 …nancial crisis. The rise in spreads attracted attention because
it could signal a disruption in …nancial markets, which has been widely linked to an increased
burden on non-…nancial …rms. This paper disentangles the relative contributions of credit and
liquidity risk in explaining the widening of commercial paper spreads. In doing so, we …nd
that liquidity risk was isolated to the …nancial sector throughout the …rst two major shocks
to the system (August 2007 and March 2008). Indeed, controlling for credit risk, non-…nancial
corporations saw little or no change in the cost of funding during this time period. After
the bankruptcy of Lehman Brothers, for the …rst time, liquidity problems in the commercial
paper market spilled out of the …nancial sector into the spreads of low credit quality non…nancial …rms. This e¤ect had a disproportionately larger impact on those low credit-quality
non-…nancial …rms that placed paper exclusively through …nancial sector dealers. High credit
quality …rms remained una¤ected throughout. Our interpretation of the results is that markets
were able to di¤erentiate not only between safe and imperiled …rms in the midst of the crisis,
but also to isolate where liquidity e¤ects were most likely to be salient.
Ethan Cohen-Cole: Robert H Smith School of Business. email: ecohencole@gmail.com. (301) 541-7227. Judit
Montoriol-Garriga: 600 Atlantic Avenue, Boston MA. email: judit.montoriol-garriga@bos.frb.org. Gustavo Suarez
Federal Reserve Board of Governors. email: gustavo.a.suarez@frb.gov. (202) 452-3011. Jason Wu Federal Reserve
Board of Governors. email: jason.j.wu@frb.gov. (202) 452-2556. The views expressed in this paper are those of the
authors and do not necessarily re‡ect those of the Federal Reserve Bank of Boston or the Federal Reserve System.
1
Electronic copy available at: http://ssrn.com/abstract=1712442
1
Introduction
Commercial paper spreads widened to unprecedented levels in many …nancial markets during the
2007-2008 …nancial crisis. For example, according to data from the Federal Reserve, the spread
over the fed funds rate paid by A2/P2-rated non-…nancial …rm to issue overnight commercial paper
increased from 60 to 400 basis points during the week following the Lehman Brothers’bankruptcy.
The rise in spreads attracted a great deal of attention, because disruptions in …nancial markets are
widely believed to tighten credit conditions for non-…nancial …rms.
Although this unprecedented widening of spreads has been extensively documented, there is
little consensus on its contributing factors. Our paper has two core aims. First, we provide a set of
empirical results that contribute to the discussion on the relative importance of credit risk and the
liquidity risk of very short-term commercial paper during the …nancial crisis. Second, we use these
empirical results to shed light on the role of …nancial intermediaries in transmitting …nancial shocks
to the real economy. To accomplish the …rst objective, we study changes in spreads around the
time period of three large shocks to the …nancial system: investors’realization of the large scope
of subprime-related losses in bank balance sheets in August 2007, the distressed purchase of Bear
Stearns in March 2008 and the bankruptcy of Lehman Brothers in September 2008. To achieve the
second objective, we test a number of channels for the transmission of shocks around each of the
three events.
For each of the three events, we distinguish between (a) …nancial and non-…nancial issuers, (b)
paper placed through a dealer and paper placed directly to the market, and (c) by issuer credit
rating. With these three dimensions, we are able to precisely evaluate which set of …rms were
impacted by the shocks that hit the …nancial sector and which were relatively unscathed. Our
criterion for the transmission from the …nancial sector will be the degree to which non-…nancial
…rms saw …nancial sector shocks impact the liquidity component of their own borrowing rates.
Where we observe that changes in credit quality alone explain spread changes of non-…nancial
2
Electronic copy available at: http://ssrn.com/abstract=1712442
…rms, even around a large event, we will make the claim that there is little transmission.1
For our empirical work, we focus on unsecured commercial paper of 1-4 day maturity. Commercial paper is a signi…cant source of short-term debt …nancing for a wide range of …rms. This market
provides …nancing to …nancial …rms and corporations typically for less than 3 months, and a significant fraction has a maturity of just a few days. Analyzing the credit spreads on the shortest-term
corporate debt is of special interest, because it allows us to abstract from changes in the slope of
the yield curve and term premia. Although one may argue that only liquidity risk matters for such
short maturities, Covitz and Downing (2007) show that, even for very short-term spreads, credit
risk is important in the commercial paper market. Another practical reason for using the shortest
maturity paper is that issuance of longer duration commercial paper was most a¤ected during the
…nancial crisis, and only the highest quality …rms were able to issue longer term paper. Therefore,
issuer selection is less severe in the shortest maturity paper. The outstanding total commercial
paper balance in July 2007 was $2.1 trillion, including $1.2 trillion of asset-backed commercial
paper (ABCP). By the end of June 2009, the outstanding balance had dropped to $1.28 trillion, a
collapse of about 40% of the market.
It is important to understand the relative scale of liquidity versus credit factors in interest rate
spreads for various reasons. On the one hand, it is important for investors’ portfolio allocation
decisions. The increase in spreads may represent increased compensation demanded by risk-averse
investors as the credit risk on the assets increases, or alternatively could represent the premium
demanded to hold more illiquid assets, which would make those assets more attractive to long-term
investors (Schwarz, 2009). On the other hand, understanding the relative signi…cance of credit
and liquidity risk is important for designing e¢ cient policy responses to market disruptions. If
liquidity is the primary problem, then the general consensus is that policy makers should direct
1
The degree of transmission has been widely debated. Chari, Christiano and Kehoe (2008) write that there was
little evidence of impact on non-…nancial …rms in October of 2008. Cohen-Cole et al (2008) respond that evidence of
transmission was indeed available at the time of the crisis. Bates, Kahle, and Stulz (2009) point out that non-…nancial
…rms held su¢ cient cash stock piles prior to the crisis to have paid o¤ existing debt, obviating the need for …nancing.
3
e¤orts to restoring market con…dence to make markets more liquid.2 If credit risk is the main
determinant of credit spread widening, then policy makers can either improve the solvency of the
counterparties involved or allow the market to reallocate credit independently. In theory, in the
absence of liquidity concerns, there is limited scope for intervention. In practice, the distinction
between these two explanations is not clear-cut, since liquidity shortages may be a by-product of
credit problems, and the solution is no longer as simple as restoring con…dence. As we shall see
below, our results suggest that the transmission of liquidity shocks from …nancial intermediaries to
some non-…nancial …rms may emerge from credit concerns.
We provide evidence in this paper that the spike in credit spreads in the short-term commercial
paper market was a combination of a liquidity and credit crisis. More importantly we show that
the liquidity risk was largely isolated to the …nancial sector throughout the …rst two major shocks
to the system (August 2007 and March 2008). Indeed, controlling for credit risk, non-…nancial
corporations saw little or no change in the cost of funding during this time period. Liquidity
problems in the commercial paper market spilled out of the …nancial sector into the spreads of low
credit quality non-…nancial …rms for the …rst time after the bankruptcy of Lehman Brothers. This
e¤ect had a disproportionately larger impact on those low credit-quality non-…nancial …rms that
placed paper exclusively through …nancial sector dealers. Issuance cost of commercial paper of high
credit-quality …rms remained una¤ected throughout.3
Our interpretation of the results is that markets were able to di¤erentiate not only between safe
and imperiled …rms in the midst of the crisis, but also to isolate where liquidity e¤ects were most
likely to be salient. Non-…nancial …rms that were of good credit quality and / or had some direct
access to capital markets were able to sidestep the liquidity shocks that hit the …nancial sector. In
contrast, non-…nancial …rms with credit or potential credit exposure experienced sharply increased
2
See Allen, Carletti and Gale (2009) and Krishnamurthy (2010) for examples and literature review.
Our emphasis in on the immediately e¤ects of the …nancial shocks on the funding costs of …rms. As the recession
continued, many parts of the economy were impacte by a myriad of factors. Our di¤erence-in-di¤erence approach
that we discuss in detail below is designed to isolate the impact of the large shocks on commerical paper spreads.
3
4
credit risk sensitivity after August 2007, especially if they were using a dealer to place paper in the
market.
Jointly, our …ndings highlight the manner in which the …nancial sector transmitted shocks to
the large non-…nancial corporations that are the typical commercial paper users. For the …rst year
following the initial shock in August 2007, almost no transmission occurred. However, once the
largest shock hit in September 2008, the stress on the …nancial sector was su¢ cient that the spillover
to the non-…nancial sector occurred, a¤ecting most the lower quality …rms and dealer users. This
suggests that the transmission of the shocks from the …nancial sector to the real economy occurred
through intermediated services to low credit quality …rms.
Our results show that the principal risk faced by …nancial …rms was the ability to rollover their
short-term debt. Of course, in the medium run, many of the banks’ holdings of bad debt may
have led to credit problems as well, but, at the time of the strongest shocks to the market during
the crisis, the critical problem for …nancial institutions appears to have been liquidity. Our results
are in contrast with those of Afonso, Kovner and Schoar (2010), who document that counterparty
risk played a much more important role than liquidity hoarding in the interbank market during
the 2008 …nancial crisis. Their interpretation of the results is that lenders in the fed funds market
were able to screen out the worst performing borrowers. Our results suggest that investors in the
commercial paper market demanded a higher premium to invest in …nancial commercial paper due
to increased liquidity risk, not credit risk. However, it is important to note that the di¤erence in
results may be due to the fact that commercial paper and fed funds are di¤erent markets, and the
…nancial …rms participating in each one may di¤er.
The …ndings in this paper have implications for the debate on the importance of the …nancial
sector for the real economy. The continued economic sluggishness more than three years after the
initial subprime shocks of August 2007 has repeatedly been attributed to the lack of debt …nancing
5
from the commercial banking sector.4 As the crisis struck, the widespread interpretation of the
stress in credit markets was that a lack of liquidity led to the inability for essentially all market
participants to roll over obligations. Given that the initial …nancial shock was concentrated in the
banking sector, one could expect that bank-dependent …rms would be most a¤ected by this shock.
On the other hand, non-…nancial corporations that issue commercial paper should be somewhat
isolated from the shock in the banking system since these larger …rms have the ability to access
capital markets. In this paper we present evidence on an often neglected link of the …nancial and
non-…nancial sectors, which is that many non-…nancial commercial paper issuers use a dealer to
place the paper into the market.
An extant literature and press reports have emphasized the crucial nature of the …nancial sector
for promoting economic growth. Indeed, the work of Bernanke et al (1999), as well as recent work
by a range of others,5 facilitated understanding of …nancial linkages to the real sector that were
not possible within classical macroeconomic models. Empirical evidence for prior crises shows that
bank distress is transmitted into the real economy.6 Additionally, there is a growing literature
on the empirical implications of the 2007-2008 …nancial crisis that suggest a real-economy link.7
However, there are relatively few papers that seek to understand the nature, timing and size of
4
See Cornett et al (2010), the senior loan o¢ cers survey, as well as recent press. For example, CNBC, May 27,
2010.
5
Cohen-Cole and Martinez-Garcia (2009), Faia and Monacelli (2007), Curdia and Woodford (2010), De Fiore and
Tristani (2009) and many others.
6
Peek and Rosengren (2000) …nd evidence for the Japanese …nancial crisis spillovers into the US economy;
Dell’Ariccia, Detragiache, and Rajan (2008) …nd evidence in a wide range of crises; Khwaja and Mian (2008) …nd evidence from an emerging market; Chava and Purnanandam (2011) …nd evidence from the Russian crisis. Borenzstein
and Lee (2002) discuss Korea.
7
Ivashina and Scharfstein (2010) …nd that new loans to large borrowers fell during the crisis and provide some
evidence that liquidity constrained banks may have been more likely to cut lending. Campello, Graham and Harvey
(2010) use a CFO survey to …nd that many corporations cut spending and investment and attribute those cuts to
inability to obtain …nancing. Duchin, Ozbas, and Sensoy (2010) …nd that corporate investment declined as a result
of the crisis and have evidence that the decline was a result of reduced access to funding. Additionally, a couple of
recent papers analyze commercial paper during the crisis, but come to di¤erent conclusions. Kacperczyk and Schnabl
(2010) …nd that the market faced a generalized collapse, and Gao and Yun (2009) …nd that declines in commercial
paper use was concentrated among low credit quality …rms. Both analyze commercial paper at a somewhat higher
level of aggregation than our data permit.
6
the transmission.8 This paper contributes to this literature by disentagling the pathway for the
spillover from the …nancial sector to the real economy using evidence from the commercial paper
market.
The rest of the paper is organized as follows. We proceed in section 2 to provide an overview
of the commercial paper market, principally focusing on the unsecured segment of the market. In
section 3, we describe our data in more detail and discuss our empirical strategy. Section 4 presents
our main results and provides some important robustness checks. Section 5 concludes.
2
Background on the Commercial Paper Market
Commercial paper is a form of borrowing with a …xed maturity, typically between 1 and 270 days.
The paper is issued by banks, large corporations, and special purpose vehicles to meet short-term
…nancial obligations including operational needs such as payroll, or for the purchase of assets.
Commercial paper is viewed as an inexpensive funding option by banks and corporations alike.
Prior to the crisis, it was relatively simple for large corporations to access the commercial paper
market, and regular funding could be obtained at rates lower than bank loans and without the
fees associated with a guaranteed line of credit. However, many issuers maintain a line of credit to
provide additional access to capital.
Commercial paper can broadly be classi…ed along a few dimensions. Among these are the
presence of collateral, the type of issuing …rm and the issuance channel (see Hahn, 1998, for a
description of the commercial paper market). Our paper focuses principally on unsecured (uncollateralized) commercial paper issued across a range of channels. By evaluating unsecured paper,
we are able to isolate the impact of the credit quality of the issuer from the role of liquidity shocks
without explicit modeling of the underlying asset quality. That is, we can rely on credit quality
estimates of the corporate commercial paper sponsor as a reasonable proxy for credit-worthiness.
8
A recent exception is Tong and Wei (2009).
7
A comprehensive overview of the asset-backed commercial paper (ABCP) market and its operation
during the crisis can be found in Covitz, Liang and Suarez (2009).
Within the commercial paper markets, we will look at the issuance of commercial paper across
both …nancial and non-…nancial …rms. Non-…nancial and …nancial corporations alike used the commercial paper market as a highly liquid and readily accessible short-term funding. Anecdotally,
these markets grew in popularity both because borrowing costs were somewhat lower than equivalent corporate loans and because commercial paper is, in general, junior to secured debt. The use
of commercial paper by …nancial and non-…nancial …rms di¤ers substantially. Non-…nancial …rms
typically use commercial paper to fund on-going cash ‡ow needs such as payroll and inventories.
As a result, interruption of access to these markets could lead to an inability to manage regular
operations and cause layo¤s and supply disruptions, which could potentially lead to severe disruptions to the real economy. Financial …rms use commercial paper to manage short-term liquidity
needs, including maturity mismatch in assets and liabilities, or in some cases to fund large portions
of their balance sheet.
With respect to the methods of issuing paper, an issuer can either sell securities directly in
the money markets or it can sell them to a dealer. The dealer subsequently sells the paper in the
market. Dealers are typically large banks and bank holding companies, including the investment
banks that existed prior to the …nancial crisis. The bene…t of direct issuance is the ability to
receive the full market price for the securities. However, working through a dealer institution could
facilitate issuance if the institution has strong distribution networks.
We will highlight this distinction throughout our empirical analysis. A buyer of corporate
commercial paper from a dealer assumes the risk of corporate default during the term of the
commercial paper as well as the counterparty risk of the dealer during the settlement time period
between purchase …nal settlements. This risk prior to the crisis was negligible, but during the
crisis it became large. As a result, issuers that relied on dealers faced the possibility that liquidity
8
concerns at the dealer banks would impact the spreads on their own issuances. Below, we provide
empirical support for this link.
3
Data
Our initial dataset includes all 91,589 primary market issues in the unsecured commercial paper
(CP) market in the U.S. market between January 1, 2007 and December 31, 2008. This includes
information on 353 programs. These data are obtained from the Depository Trust and Clearing
Corporation (DTCC), the agent that electronically clears and settles directly- and dealer-placed
commercial paper. The issues in the sample are discount instruments paying face value at maturity.
For each transaction, DTCC provides the identity and industry of the issuer, the face and settlement
values of the transaction, and the maturity of the security. Using this information, we calculate
yields on new paper based on the assumption of a 360-day year.
We also obtain from DTCC a separate weekly …le that contains program level information on
the maturity distribution of outstanding paper. Further, we supplement the DTCC data with
information on program type, credit ratings, liquidity features, and sponsor identity from various
reports from Moody’s Investors Service.
We have four key variables for our study:
Overnight CP spread
We use the spread over the fed funds rate for commercial paper maturing in 1-4 days. We
choose the shortest term commercial paper to ensure that we can evaluate the trade-o¤ between
liquidity and credit risk without concern for yield curve impacts, duration or other issues. Another
practical reason for using the shortest maturity paper is that issuance of longer dated commercial
paper was most a¤ected during the …nancial crisis, and only the very high quality …rms were able to
issue longer term paper. Therefore, issuer selection is least severe for the shortest maturity paper.
CDS Spread
9
We supplement data on commercial paper transactions with measures of credit and liquidity
risk. For our credit risk measure, we obtain 5-year CDS spreads for each commercial paper issuer
from Markit Partners. A CDS spread is a fraction of the CDS contract notional paid from the
buyer to the seller in return for credit protection on debt obligations, in the case where the issuer
defaults. For each day, we average CDS spreads of senior unsecured debt across four primary types.9
CDS spreads are generally regarded as real-time market-based measures of credit risk. While there
is some evidence that they contain a liquidity component, its importance appears to be small in
magnitude (Tang and Yan (2007), Lin, Liu and Wu (2009)).
Fraction of paper maturing in following week
For our liquidity measure, we calculate the fraction of total liabilities maturing in the upcoming
week using information provided by DTCC. This measure provides a proxy for the issuer’s need for
liquidity in the short run that re‡ects the maturity of its liability in the absence of credit quality
concerns.
Fraction of dealer-placed paper.
To account for the role of dealer intermediation in this market, we calculate the percentage of
each program’s issuances that is placed through a dealer. Later in the paper, we will subdivide
the sample to evaluate the di¤erence between programs that exclusively place their paper through
dealers and those that directly sell at least some of their own paper.
Table 1 summarizes our dataset by subperiod. The …rst four columns show summary statistics
for the full sample period, January 2007 to December 2008. The subsequent columns show summary statistics for commercial paper issuances for the three subperiods of our analysis: January to
9
The four classes are the standard International Swap and Derivatives Association (ISDA) classi…cations. Class
1 is "No Restructuring": This option excludes restructuring altogether from the contract, eliminating the possibility
that the protection seller su¤ers a “soft” credit event that does not necessarily result in losses to the protection
buyer. Class 2 are "Full restructuring" contracts. These allow the protection buyer to deliver bonds based on any
debt restructuring. Classes 3 and 4 are "Modi…ed Restructuring" and "Modi…ed Modi…ed Restructuring" contracts.
These are full restructuring contracts that limit the bonds that can be delivered to <30 months and <60 months
respectively.
10
December 2007 (encompassing the crisis in funding markets that erupted in August 2007); January
to June 2008 (encompassing the distressed purchase or Bear Stearns in March); and July to December 2008 (encompassing the bankruptcy of Lehman Brothers in September 2008) . Each panel
provides summary information for the core four variables used in the study plus two additional
liquidity measures that we will address below. We report the mean, standard deviation, minimum
and maximum for each variable for each time period. We also report the number of commercial
programs that appear in each sample, as well as the number of observations (issuances) in each
case.
4
Empirical Strategy and Results
The aim of this paper is to disentangle the relative contributions of credit risk and liquidity risk in
the widening of unsecured commercial paper spreads of maturity 1 to 4 days during three major
events in the 2007-2008 …nancial crisis: the initial shock of August 2007, the failure of Bear-Stearns
on March 14, 2008 and the failure of Lehman Brothers on September 15, 2008. Second, we wish
to understand how the importance of credit and liquidity risk depends on the issuance channel.
Finally, we want to investigate whether issuers in the …nancial sector were impacted di¤erently
than those outside of it. Our analysis of the commercial paper using a combination of these criteria
enables us to shed some light on the following general questions:
1. Was the …nancial crisis a liquidity or credit event?
2. Does this distinction help us understand the transmission channel from the …nancial to the
real economy?
3. Was the market able to distinguish between imperiled companies and safe ones?
We show in Figure 1 the stylized patterns of commercial paper issuance. Each box represents
a type of issuer. Trapezoids represent the presence of a dealer in a transaction. The ovals at the
11
bottom of each diagram are investors that buy commercial paper. The connecting lines represent
an active channel, and the black lines across a channel indicate increased costs via this channel
due to liquidity premia. Panel A shows the situation pre-crisis, where all channels are operational.
Panel B shows the situation in a generalized liquidity crisis, where all channels have been severed.
Panel C shows what one would expect if non-…nancial issuers could continue to issue; e¤ectively the
situation in the absence of a …nancial sector transmission mechanism. Panel D shows our stylized
results in the paper, where …nancial issuers su¤er liquidity shocks and non-…nancial issuers are
impacted via dealer networks, but direct market access remains open.
[Insert F igure 1 here]
We conduct regressions of CP spreads on a variety of controls for …rm-level credit risk, CP
instrument liquidity and events that impacted the entire market. For identi…cation, we use a
di¤erence-in-di¤erence approach. We conduct a series of event studies around the three events: the
subprime credit losses of August 2007, the collapse of Bear Stearns in March 2008 and the collapse
of Lehman Brothers in September 2008. We subdivide our sample of data in three time periods,
each surrounding one of the events. For the August 2007 shock that started the crisis, we include
data from January to December of 2007. The full year is included here in order to ensure that
we have a su¢ cient baseline of activity leading up to the crisis. For the Bear Stearns collapse, we
include data from January 2008 to June of 2008. This includes both the period of relative calm at
the beginning of the year as well as a return to relative calm in mid-year. Finally, for the Lehman
case, we include data from mid-year 2008 until the end of 2008. For each episode, we include a
dummy variable to mark the crisis time period. This is coded as a one during the times of stress
and a zero otherwise. That is, we …nd a period of 20 trading days (one month) before and after
August 9th, 2007, March 14th, 2008 and September 15, 2008.
We then subdivide the sample of commercial paper issuance along the …nancial/non-…nancial
and dealer/direct dimensions. These steps allow us to use a di¤erence-in-di¤erence-in-di¤erence
12
approach along three axes:
Financial vs non-…nancial …rms. This distinction permits us to analyze the di¤erential impacts
of the crisis on the …nancial and real economies.
Pre-shock vs post-shock. With this approach, we are able to determine the impact of the
shock on the indicated sub-sections of the economy for each of the three shocks mentioned.
Dealer placed vs direct issuance. By looking at dealer placement, we can shed additional light
on the role of intermediation functions on the impact of the crisis.
At the end of the paper, we will complement the analysis by looking more deeply at the highest
credit quality …rms.
4.1
Results
We begin with an exploration of the impact of three …nancial crisis events on CP spreads. In
particular, we look at the contribution of our liquidity and credit proxies as each of the crisis events
hit the markets. We can begin with a regression of spreads on two controls independent of event
dummies.
spreadit =
+
1
1 creditit
+
1
2 liquidityit
+
i
+ eit
(1)
for all companies i in time t. The variable spreadit is the 1 day commercial paper spread. credit1it
1 is our …rst liquidity proxy: fraction of
is our …rst credit variable, average 5-yr CDS, and liquidityit
outstanding commercial paper due in the upcoming week. We also include program dummies,
i,
thereby adding …xed e¤ects to the model.
We interpret a positive liquidity coe¢ cient as representing heightened sensitivity to short-term
funding needs. A positive coe¢ cient on the credit variable suggests increased credit risk premia.
Table 2 shows the results of this speci…cation. We can see in the top two rows and the …rst
three columns that a …rst order impact of the crisis was to increase spreads due to both credit and
13
liquidity. These e¤ects are present, large and strong in nearly all speci…cations.
[Insert T able 2 here]
Table 2 also includes issuer and date-level …xed e¤ects at the …rm level and marks whether the
regressions have been estimated with heteroskedasticity consistent standard errors. In subsequent
tables, we will use only the speci…cations that have issuer …xed e¤ects and robust standard errors.
Issuer …xed e¤ects are important to controlling for idiosyncratic di¤erences across programs. We
don’t include date …xed e¤ects to avoid problems of collinearity, as our spread information is also
at a day level. Additional results including these other possibilities are available on request.
Baseline
We can now turn to our primary speci…cation. As mentioned, we include dummies for each of
the three events. The interaction of the dummies for the stress time periods with the credit and
liquidity variables provides the basis for inference.
spreadit =
+
1
1 CRit
+
1
2 LIQit
+
k
3 T Dt
+
4T D
k
CRit +
5T D
k
LIQit +
i
+ eit
(2)
where T Dtk is a dummy coded 1 during the time period after the event k and zero otherwise. The
variables T Dk CRit and T Dk LIQit are the interactions between T Dtk and the relevant credit and
liquidity variables de…ned above. For the speci…cations we showed in Table 2, as well as in future
tables, we will code the dummy as 1 for the one-month period after the event in question. Our
speci…cations will include two months of data for each event, one month prior and one month post
event.
For the three large shocks, August 9, 2007, March 14, 2008 and September 15, 2008, we de…ne
a dummy variable as above, T Dtk where k indexes the three events:
k = 1 : August 9, 2007; Subprime shock
k = 2 : March 14, 2008; Bear-Stearns default
k = 3 : September 15, 2008; Lehman bankruptcy
14
Table 2 continues after column 4 to show the results of regression 2 using the credit and liquidity
proxies as well as the event dummies. This table makes clear that each of the three events had
aggregate impacts on the economy. Subsequent tables will delve into the details of the events in
order to address the substantive questions outlined above.
Then, for each of k = 1; 2; 3, we run regression 2 and add the interaction of the event dummies
and the credit and liquidity variables in table 3. In this table as well as subsequent ones, we have
omitted date …xed e¤ects both to avoid collinearity problems with the events themselves and for
brevity. These results are available on request from the authors.
[Insert T able 3 here]
Of particular note in table 3 is that both the credit and liquidity proxies are strongly positive
and signi…cant in the …rst two event time periods. The liquidity variable is strongly signi…cant and
positive in Lehman event time period. This result is consistent with interpretations of the crisis
that found that liquidity was at a large premium and that risk premia increased sharply. By the
time we reach the Lehman crisis, the markets have become strongly attuned to liquidity concerns,
and the coe¢ cients suggest that this was the primary e¤ect.
We will expand the analysis in equation 2 by analyzing in the three ways described below with
the goal of answering our primary question.
4.2
The Financial Sector
We can move now to an analysis of the impact of the three crisis events on spreads across types of
issuers. As was widely publicized, the crisis was originally in the …nancial sector itself. As a result,
a key method to determining whether the liquidity shocks were generalized is to observe di¤erences
in spread changes across …nancial and non-…nancial …rms.
This di¤erence is important for understanding the role of the …nancial …rms in impacting the
real economy. Referring to Figure 1 again, we can illustrate a basic interpretation of coe¢ cient
15
combinations across …nancial and non-…nancial …rms. In addition to interpreting the simple liquidity and credit coe¢ cients, the schematic infers from the combination of coe¢ cients the possible
interaction between the two sectors. Speci…cally, sets of positive coe¢ cients for all combinations
implies, as in table 3, that the economy faced generalized distress. However, mixed sets of coe¢ cient
imply di¤erential responses to the crisis.
We extend equation 2 above by estimating the speci…cations on the subsamples of the …nancial
sector and the non-…nancial sector commercial paper programs. The two sets of results are shown
in table 4. We have included the full sample results as well for comparability. As in the previous
tables, we show the results for each of the three events separately, arranged in chronological order.
[Insert T able 4 here]
In table 4 we observe a number of notable features. First, the generalized shocks that appeared
in table 3 are signi…cantly di¤erentiated across the sectors. In some of the events, one of the two
variables is insigni…cant for one of the two groups. Even before discussion of the nature of the
distinction, this implies that the impact of the crisis was distinct across sectors; that is, the market
was able to distinguish the impact of the crisis with some degree of precision across sectors - even
during the Lehman event. This is in sharp contrast to the notion that the market as a whole froze.
Second, the pattern of di¤erentiation provides evidence that the …nancial sector faced principally
a liquidity shock, and the non-…nancial sector principally a credit shock. We have shaded the
coe¢ cients that indicate this pattern in table 4. Notice that in each of the three events, the non…nancial sector saw a signi…cant credit shock. Indeed, the sector only faced a liquidity shock at
the Lehman bankruptcy event. Below we will provide some additional detail on this component
to show that only a portion of the non-…nancial sector saw this shock. On the other hand, the
…nancial sector saw a liquidity shock in all three and only saw a credit e¤ect during the Bear Stearns
collapse.
16
We take the fact that these simple proxies are able to di¤erentiate the e¤ect during such enormous disruption as strong evidence that the market was able to di¤erentiate …rms in purchasing
commercial paper. That is, our proxies are a single measure of …rm liquidity needs and credit
quality. While we cannot be certain of investor behavior, one might hope that investors would use
more than a CDS spread and the fraction of maturing paper to make lending decisions. If so, we
would expect to see even greater levels of di¤erentiation.
The distinctions that we observe imply that the generalized disruption was more orderly than
previously thought. Indeed, notice that this interpretation is distinct from a ‡ight-to-quality, panic
scenario. Rather, it re‡ects a market that was able to use available information to make reasonable
lending decisions even in the midst of one of the largest …nancial disruptions in history.
Additionally, the pattern facilitates understanding of the role of the …nancial sector in the real
economy. During normal times, the …nancial sector provides intermediation services and access to
funding for real-economy …rms. Our results suggest that the shocks faced during this large crisis led
to dramatically increased liquidity premia at …nancial …rms, but had little direct liquidity impact on
non-…nancials. Indeed, the market appears to have been able to segregate non-…nancials according
to their credit risk. This increased sensitivity to credit risk for non-…nancial …rms indicates that
market events had a heterogeneous impact on non-…nancial …rms. While low credit quality …rms
experienced increased spreads, high credit quality …rms continued to receive short-term funding.
E¤ectively, the runs that occurred in the repo markets (Gorton and Metrick (2009) and many
others) and the ABCP markets (Covitz, Liang and Suarez (2009)) were phenomena isolated to
…nancials. The maturity mismatch at non-…nancials did not lead to funding withdrawals, at least
for the highest quality …rms. In order to further check this interpretation of the results, we will
run a robustness check only looking at A1/P1 issuers (see results below).
17
4.3
The Role of Dealers
One heightened concern during the crisis was the …nancial condition of intermediating banks. Indeed, this connection is the basis for concern that a …nancial sector shock impacts the real economy.
Without …nancial intermediaries, a real-economy …rm may be unable to obtain funds for otherwise
pro…table investments.
A …rm that issued its commercial paper through a dealer could potentially be perceived as
being more dependent on …nancial institutions for funding and thus at greater risk during times
of …nancial market distress. We test whether the market used this distinction in penalizing the
funding costs of dealer-dependent …rms.
A typical relationship involves the issuer placing its commercial paper through a dealer. The
dealer will typically buy the paper at a discount and immediately sell it in the market. On occasion,
dealers will hold some paper inventory as a service to issuers; this is often useful when issuers need
funds of a particular maturity. During the crisis, if a dealer’s own access to liquidity shrunk, their
ability to o¤er services would shrink as well. As a result, one might expect that issuers that were
dependent on dealers would have been more susceptible to the liquidity shocks we observed above.
We expand on equations 1 and 2 above by estimating the speci…cations on the subsamples
de…ned according to whether the CP programs had a dealer or not. We subdivide the sample into
the 100% dealer placement and <100% in order to highlight the fact that …rms with proportions
less than 100% do not have fully independent access to …nancial markets and are thus potentially
exposed to intermediation channel disruptions.
Table 5 shows a couple of di¤erences between …rms that had 100% dealer placement and those
that did not. During the …rst and second crisis events (August 2007 and March 2008) …rms that
had less than 100% dealer placement saw no liquidity shock, but these …rms su¤ered an increased
sensitivity to credit risk. These results suggest that issuers that use a dealer are more subject
to liquidity risk than those that do not. This highlights the role of …nancial intermediaries in
18
transmitting the initial liquidity shock to other …nancial markets. Even though the commercial
paper market is thought as arm’s length …nancing without …nancial intermediation, it is clear
from these results that …nancial intermediaries act as dealers in many of these transactions, and,
therefore, there is some risk of spillover of their liquidity problems to this market.
[Insert T able 5 here]
However, during the Lehman event, we observe the opposite pattern: liquidity risk is signi…cant
while credit risk is not for …rms that had less than 100% dealer placement. We attribute this result
to the fact that the Lehman event was characterized by widespread liquidity problems. In this
event, all …rms, regardless of their use of a dealer, experience an increased liquidity premium. We
will see further evidence to support this pattern below when we further subdivide the sample into
low and high credit quality issuers.
Firms that were 100% dealer intermediated saw a large credit and liquidity e¤ect in the three
events. This is consistent with the idea that …rms without their own programs were in greater need
of intermediation services. Naturally, the presence of intermediation services leads logically to a
stronger transmission mechanism from the …nancial to the real economy.
4.4
Credit Quality
To gain further insight into the patterns we observe, we continue by restricting our sample to the
set of issuers who programs have the highest ratings (A1/P1 ratings).10 For these regressions, we
…nd programs that have an A1/P1 rating at the beginning of each of the three sample periods.
While we restrict the sample by credit quality, the data contain signi…cant cross-sectional variation
10
Rule 2a-7 of the Investment Company Act of 1940 limits the credit risk that money market mutual funds may bear
by restricting their investments to "eligible" securities. An eligible security must carry one of the two highest ratings
("1" or "2") for short-term obligations from at least two of the nationally recognized statistical ratings agencies. A
tier-1 security is an eligible security rated "1" by at least two of the rating agencies; a tier-2 security is an eligible
security that is not a tier-1 security.
19
in CDS spreads even within the high-quality issuers. This variation allows us to draw inference
within this sample.
In table 6 we present the results for …nancials and non-…nancials and in table 7 the results for
the dealer and non-dealer programs.
[Insert T ables 6 and 7 here]
In table 6 we …nd qualitatively similar results to table 4; however, most of the coe¢ cients are
smaller in magnitude. Note that these regressions include only the highest quality programs, and,
therefore, the credit risk is less severe in this sample. Nonetheless, we see that changes in risk are
internalized by the market and impact issuer spreads. We take this as evidence that the market
was able to distinguish the credit quality di¤erences among the highest quality issuers, even in the
midst of signi…cant market turmoil. With the crisis events, the credit risk premium increased for
low quality non-…nancial …rms, but to a lesser extent for the highest quality …rms.
In table 7 we observe that 100% dealer intermediated A1/P1 saw no liquidity shock in any of
the three stages of the crisis. This result is particularly salient. The highest quality issuers, even
those that had intermediation services exclusively by dealer, did not experience liquidity e¤ects at
any point during the crisis.
These contrast with the results in table 5 in which liquidity was signi…cant for all three events
for the dealer programs. By restricting the sample to the highest quality programs, we …nd that the
market demanded a higher liquidity premium only for the lowest quality programs and only if they
were dealer intermediated. We also observe that the coe¢ cients on the credit risk and liquidity risk
during the three market events are smaller in table 7 than in table 5, suggesting that the increased
credit and liquidity premiums in the crisis were smaller for the highest quality programs than the
lower quality programs.
Our intuition from this …nding is that the liquidity shocks that occurred throughout the crisis
emerged in part from credit quality concerns. Thus, even though we are able to separate out
20
liquidity and credit components from spreads, the core reason for the presence of a liquidity premium
indeed lies in credit.
5
Conclusions
This paper provides empirical evidence of the relative contributions of credit risk and liquidity
risk in the widening of unsecured commercial paper spreads of maturity 1 to 4 days during three
major events in the 2007-2008 …nancial crisis. Our …ndings are nuanced in that we …nd that initial
…nancial sector shocks had little impact on corporate issuers and that the Lehman event passed
through the dealer channel to struggling …rms. These …ndings suggest a deeper and more complex
set of relationships between the …nancial sector and the real economy.
Indeed, our conclusion is that the real economy is highly resilient, but not impervious, to
…nancial sector distress. Non-…nancial …rms managed to withstand even the large shocks faced by
…nancial …rms in the …rst year of the crisis, shocks that, at the time, were being described as
unprecedented. Had the crisis ended earlier, our paper could have concluded that there are little
or no transmission e¤ects, even for the largest shocks. With hindsight, and the presence of the
enormous Lehman shock, we see that the transmission e¤ect to the real economy indeed exists.
However, it appears to manifest only under the most extreme events and along certain channels.
This nonlinearity lends credence to the story that the disintermediation of banks and …nancial
companies has supported the ability of non-…nancial …rms to succeed. This nonlinearity exists due
to the presence of low credit quality …rms. That is, the liquidity shocks that impacted the real
economy did so on the margin of impacting lower quality …rms’borrowing rates. In the end, there
appears to have been little risk to the cost of funds of high credit quality …rms.
While this study has found detailed evidence of the transmission channel of the …nancial sector,
we do not explore some dimensions of this question. Our study looks at the cost of funding of
commercial paper issuers, conditional on their participation in the market. We encourage future
21
study in the determinants of market participation and the potential impact of the …nancial sector
on those decisions.
References
[1] Afonso, Gara M., Anna Kovner, and Antoinette Schoar, 2010, Stressed not frozen: The fed
funds market in the …nancial crisis, NBER Working Paper Series, w15806.
[2] Allen, Franklin, Elena Carletti, and Douglas Gale, 2009, Interbank market liquidity and central
bank intervention, Journal of Monetary Economics 56, 639-652.
[3] Bates, Thomas W., Kathleen M. Kahle, and Rene M. Stulz, 2009, Why do U.S. …rms hold so
much more cash than they used to?, Journal of Finance 64, 1985-2021.
[4] Bernanke, Ben S., 1983, Nonmonetary e¤ects of the …nancial crisis in the propagation of the
Great Depression, American Economic Review 73, 257-76.
[5] Bernanke, Ben S., Mark Gertler and Simon Gilchrist, 1999, The …nancial accelerator in a quantitative business cycle framework, in John B. Taylor and Michael Woodford, eds.: Handbook
of Macroeconomics, Vol. 1 (North Holland, Amsterdam).
[6] Borensztein, Eduardo, and Jong-Wha Lee, 2002, Financial crisis and credit crunch in Korea:
Evidence from …rm-level data, Journal of Monetary Economics 49, 853-875.
[7] Campello, Murillo, John R. Graham, and Campbell R. Harvey, 2010, The real e¤ects of …nancial constraints: Evidence from a …nancial crisis, Journal of Financial Economics 97, 470-487.
[8] Chari, V.V., Lawrence J. Christiano and Patrick J. Kehoe, 2008, Facts and myths about the
…nancial crisis of 2008, Working paper, 666, Federal Reserve Bank of Minneapolis.
[9] Chava, Sudheer, and Amiyatosh Purnanandam, 2011, The e¤ect of banking crisis on bankdependent borrowers, Journal of Financial Economics 99, 116-135.
22
[10] Cohen-Cole, Ethan, Burcu Duygan-Bump, Jose Fillat, and Judit Montoriol-Garriga, 2008,
Looking behind the aggregates: A reply to ‘Facts and myths about the …nancial crisis of 2008’,
Working paper, QAU08-5, Federal Reserve Bank of Boston.
[11] Cohen-Cole, Ethan, and Enrique Martinez-Garcia, 2008, The balance sheet channel, Working
paper, QAU08-7, The Federal Reserve Bank of Boston.
[12] Cornett, Marcia Millon, Jamie John McNutt, Philip E. Strahan, and Hassan Tehranian, 2010,
Liquidity risk management and credit supply in the …nancial crisis, Working paper, available
on SSRN.
[13] Covitz, Dan, and Chris Downing, 2007, Liquidity or credit risk? The determinants of very
short-term corporate yield spreads, Journal of Finance 62, 2303-2328.
[14] Covitz, Daniel M., Nellie Liang, and Gustavo A. Suarez, 2009, The evolution of a …nancial
crisis: Panic in the asset-backed commercial paper market, Working paper, 2009-36 Board of
Governors of the Federal Reserve System.
[15] Curdia, Vasco, and Michael Woodford, 2010, Credit frictions and optimal monetary policy.
Journal of Money, Credit, and Banking 42, 3-35.
[16] De Fiore, Fiorella, and Oreste Tristani, 2009, Optimal monetary policy in a model of the credit
channel, Working paper, 1043, European Central Bank.
[17] Dell’Ariccia, Giovanni, Enrica Detragiache, and Raghuram Rajan, 2008, The real e¤ect of
banking crises, Journal of Financial Intermediation 17, 89-112.
[18] Diamond, Douglas W., and Philip H. Dybvig, 1983, Bank runs, deposit insurance, and liquidity,
Journal of Political Economy 91, 401-419.
23
[19] Duchin, Ran, Oguzhan Ozbas, and Berk A. Sensoy, 2010, Costly external …nance, corporate
investment, and the subprime mortgage credit crisis, Journal of Financial Economics 97, 418435.
[20] Erlich, Natalie, 2010, May 37, Banks unlikely to ease lending standards: FICO, CNBC.com,
available at: http://www.cnbc.com/id/37377093.
[21] Faia, Ester, and Tommaso Monacelli, 2007, Optimal interest rate rules, asset prices, and credit
frictions, Journal of Economic Dynamics and Control 31, 3228-3254.
[22] Gao, Pengjie, and Hayong Yun, 2009, Commercial paper, lines of credit, and the real e¤ects
of the …nancial crisis of 2008: Firm-level evidence from the manufacturing industry, Working
paper, SSRN.
[23] Gertler, Mark, and Nobuhiro Kiyotaki, 2010, Financial intermediation and credit policy in
business cycle analysis, Working paper, prepared for the Handbook of Monetary Economics.
[24] Gorton, Gary B., and Andrew Metrick, 2009, Securitized banking and the run on repo, NBER
Working Paper Series, w15223.
[25] Hahn, Thomas K., 1998, Commercial paper, in Federal Reserve Bank of Richmond: Instruments of the money market, Monograph, 105-127.
[26] Holmstrom, Bengt, and Jean Tirole, 1998, Private and public supply of liquidity, Journal of
Political Economy 106, 1-40.
[27] Ivashina, Victoria, and David Scharfstein, 2010, Bank lending during the …nancial crisis of
2008, Journal of Financial Economics 97, 319-338.
[28] Kacperczyk, Marcin T., and Schnabl, Philipp, 2010, When safe proved risky: Commercial
paper during the …nancial crisis of 2007-2009, Journal of Economic Perspectives 24, 29-50.
24
[29] Khwaja, Asim Ijaz, and Atif Mian, 2008, Tracing the impact of bank liquidity shocks: Evidence
from an emerging market, The American Economic Review 98, 1413-1442.
[30] Krishnamurthy, Arvind, 2010, Ampli…cation mechanisms in liquidity crises, American Economic Journal: Macroeconomics 2, 1-30.
[31] Lin, Hai, Liu, Sheen, and Chunchi Wu, 2009, Determinants of corporate bond and CDS
spreads, Working paper, Conference on Financial Economics and Accounting.
[32] Meier, Andre, and Gernot J. Muller, 2006, Fleshing out the monetary transmission mechanism:
Output composition and the role of …nancial frictions, Journal of Money, Credit, and Banking
38, 2099-2133.
[33] Peek, Joe, and Eric Rosengren, 2000, Collateral damage: E¤ects of the Japanese bank crisis
on real activity in the United States, The American Economic Review 90, 30-45.
[34] Schwarz, Krista, 2010, Mind the gap:
Disentangling credit and liquidity in risk
spreads,Working paper, available on SSRN.
[35] Senior
Loan
O¢ cer
Opinion
Survey,
The
Federal
Reserve
Board,
available:
http://www.federalreserve.gov/boarddocs/snloansurvey/201008/chartdata.htm.
[36] Tang, Dragon Yongjun, and Hong Yan, 2007, Liquidity and credit default swap spreads, Working paper, available on SSRN.
[37] Tong, Hui, and Shang-Jin Wei, 2009, The composition matters: Capital in‡ows and liquidity
crunch during a global economic crisis, Working paper, wp09164, IMF.
25
Investors
Dealer
Financial Issuer
Dealer
Non-financial
Issuer
Investors
Dealer
Financial Issuer
C. Isolated Financial Sector Crisis:
No Real Economy Effect
Dealer
Non-financial
Issuer
A. Pre-Crisis
Investors
Dealer
Financial Issuer
Dealer
Non-financial
Issuer
Investors
Dealer
Financial Issuer
D. Transmission of Crisis to Real Economy
Through Dealer; Direct Non-Financial Access Remains
Dealer
Non-financial
Issuer
B. Liquidity Crisis
Figure 1: Financial Sector Transmission Mechanism
Note: Figure shows stylized patterns of commercial paper issuance. Each box represents a type of issuer. Trapezoids represent the presence of a dealer in a transaction. The ovals at the bottom of each diagram
are investors that buy commercial paper. The connecting lines represent an active channel, and the black lines across a channel indicate increased costs via this channel due to liquidity premia. Panel A shows
the situation pre-crisis, where all channels are operational. Panel B shows the situation in a generalized liquidity crisis, where all channels have been severed. Panel C shows what one would expect if nonfinancial issuers could continue to issue; effectively the situation in the absence of a financial sector transmission mechanism. Panel D shows our stylized results in the paper, where financial issuers suffer
liquidity shocks and non-financial issuers are impacted via dealer networks, but direct market access remains open.
26
Fraction of dealer-placed paper
Liquidity risk
Credit risk
Overnight CP spread
Financial firms
PANEL B
Fraction of dealer-placed paper
Liquidity risk
Credit risk
Overnight CP spread
All firms
PANEL A
Table 1: Summary Statistics
0.399
0.628
52.776
93.064
0.172
0.900
25.950
69.239
0.309
0.736
42.118
80.562
42
42
42
42
Firms with less than 100% dealer-placed paper
Overnight CP spread
0.160
0.546
Credit risk
0.788
0.996
Liquidity risk
31.502
22.977
Fraction of dealer-placed paper
53.255
40.076
8,848
8,848
8,848
42
8,560
8,560
8,560
59
10,492
10,492
10,492
48
6,916
6,916
6,916
53
0.064
0.252
33.473
50.228
0.141
0.281
48.568
100.0
0.149
0.288
51.664
92.957
0.034
0.234
25.081
67.590
0.106
0.268
41.797
80.986
0.205
0.183
25.532
38.102
0.283
0.132
29.888
0.000
0.283
0.144
31.064
21.055
0.174
0.173
13.694
39.907
0.254
0.158
29.005
33.703
34
34
34
34
55
55
55
55
47
47
47
47
42
42
42
42
89
89
89
89
3,832
3,832
3,832
34
4,711
4,711
4,711
55
5,372
5,372
5,372
47
3,171
3,171
3,171
42
8,543
8,543
8,543
89
January 2007-December 2007
Mean
Std Dev Progs Obs
(5)
(6)
(7)
(8)
0.107
0.861
32.026
44.993
0.310
0.773
51.516
100.0
0.315
0.737
54.364
92.107
0.087
0.919
26.677
66.249
0.221
0.812
42.954
79.939
0.294
0.438
22.411
38.366
0.338
0.469
29.517
0.000
0.341
0.409
29.364
21.939
0.274
0.501
16.341
41.883
0.335
0.458
28.330
35.144
31
31
31
31
54
54
54
54
45
45
45
45
40
40
40
40
85
85
85
85
2,002
2,002
2,002
31
2,555
2,555
2,555
54
2,679
2,679
2,679
45
1,878
1,878
1,878
40
4,557
4,557
4,557
85
January 2008-June 2008
Mean
Std Dev Progs
Obs
(9)
(10)
(11)
(12)
Note: Credit risk is defined as the CDS spread of the issuer. Liquidity risk is defined as the fraction of paper maturing over the next week. All variables are in percent.
59
59
59
59
48
48
48
48
53
53
53
53
101 17,408
101 17,408
101 17,408
101
101
0.844
0.832
29.426
0.000
0.781
0.695
30.099
20.701
0.604
1.165
15.240
40.435
0.725
0.921
28.471
34.563
January 2007-December 2008
Mean
Std Dev Progs Obs
(1)
(2)
(3)
(4)
Firms with 100% dealer-placed paper
Overnight CP spread
0.463
Credit risk
0.682
Liquidity risk
53.092
Fraction of dealer-placed paper
100.0
PANEL C
Fraction of dealer-placed paper
Liquidity risk
Credit risk
Overnight CP spread
Nonfinancial firms
27
0.445
1.767
29.892
50.536
1.166
1.400
53.931
100.0
1.042
1.257
53.480
94.529
0.493
2.011
26.696
65.900
0.804
1.584
41.872
80.214
0.969
1.484
20.053
40.414
1.373
1.313
28.731
0.000
1.327
1.089
28.612
16.702
1.041
1.654
16.467
42.453
1.241
1.413
27.522
35.159
36
36
36
36
54
54
54
54
45
45
45
45
45
45
45
45
90
90
90
90
2,161
2,161
2,161
36
2,147
2,147
2,147
54
2,441
2,441
2,441
45
1,867
1,867
1,867
45
4,308
4,308
4,308
90
July 2008-December 2008
Mean Std Dev Progs Obs
(13)
(14)
(15)
(16)
Constant
Lehman
Bear
Post Aug 2007
10,600
0.095
65
Observations
R-squared
Number of programs
10,600
0.095
65
N
N
Y
10,600
0.238
65
Y
N
Y
-0.018
[0.035]
0.012**
[0.005]
0.000***
[0.000]
0.229***
[0.016]
0.012
[0.016]
0
[0.000]
0.229**
[0.092]
10,600
0.511
65
Y
Y
Y
5,081
0.066
56
N
N
N
5,081
0.066
56
N
N
Y
5,081
0.174
56
Y
N
Y
-0.028
[0.018]
0.047**
[0.020]
0
[0.000]
0.064
[0.056]
0
[0.000]
0.110***
[0.024]
0.071*** 0.071*** 0.075***
[0.007]
[0.015]
[0.013]
0.001*** 0.001***
[0.000]
[0.001]
0.131**
[0.052]
0
[0.000]
0.055
[0.040]
January 2008-June 2008
(10)
(11)
(12)
0.131***
[0.008]
(9)
0.078
[0.050]
5,081
0.781
56
Y
Y
Y
2,827
0.138
51
N
N
N
2,827
0.138
51
N
N
Y
2,827
0.445
51
Y
N
Y
2,827
0.769
51
Y
Y
Y
0.013
[0.041]
0.168***
[0.025]
0.168**
[0.072]
2,692
0.099
56
N
N
N
2,692
0.099
56
N
N
Y
2,692
0.358
56
Y
N
Y
-0.112
[0.102]
0.233**
[0.109]
0.007*** 0.007*** 0.005**
[0.001]
[0.002]
[0.002]
0.078***
[0.010]
2,692
0.547
56
Y
Y
Y
-0.122
[0.172]
0.003*
[0.002]
0.036
[0.039]
July 2008-December 2008
(13)
(14)
(15)
(16)
0.068 -0.167*** -0.167*** -0.110*** 0.381*** -0.224*** -0.224**
[0.055] [0.010]
[0.047]
[0.021]
[0.038]
[0.028]
[0.091]
0
[0.000]
0.028
[0.057]
January 2007-December 2007
(5)
(6)
(7)
(8)
-0.032 -0.057*** -0.057**
[0.216]
[0.005]
[0.021]
0
[0.001]
0.041
[0.043]
(4)
Note: The dependent variable is the overnight CP spread. Credit risk is defined as the CDS spread of the issuer. Liquidity risk is defined as the fraction of paper maturing over the next week. Post Aug
2007 is a dummy = 1 after August 9, 2007. Bear is a dummy = 1 after March 14, 2008. Lehman is a dummy = 1 after September 15, 2008.
N
N
N
Issuer-fixed effects?
Date-fixed effects?
Robust standard errors
clustered by issuer?
-0.082*** -0.082***
[0.007]
[0.029]
0.001
[0.001]
0.002*** 0.002***
[0.000]
[0.001]
Liquidity
Full Sample
(2)
(3)
0.119*** 0.119*** 0.099***
[0.004]
[0.036]
[0.033]
(1)
Credit
Table 2: Baseline
28
29
8,543
0.251
89
Observations
R-squared
Number of programs
8,543
0.251
89
N
Y
-0.027**
[0.011]
0.225***
[0.049]
0
[0.000]
-0.085
[0.057]
0.258*
[0.145]
0.004***
[0.001]
8,543
0.455
89
Y
Y
0.110***
[0.026]
-0.299***
[0.081]
0
[0.000]
-0.066
[0.048]
0.549***
[0.112]
0.003***
[0.001]
4,557
0.262
85
N
N
-0.129***
[0.018]
-0.014
[0.023]
0.106***
[0.019]
0.002***
[0.000]
0.116***
[0.014]
0.004***
[0.000]
4,557
0.262
85
N
Y
-0.129***
[0.040]
-0.014
[0.055]
0.106*
[0.062]
0.002**
[0.001]
0.116***
[0.035]
0.004***
[0.001]
4,557
0.768
85
Y
Y
0.068**
[0.027]
-0.011
[0.034]
0.101***
[0.028]
0.002***
[0.001]
0.100***
[0.022]
0
[0.000]
January 2008-June 2008
(4)
(5)
(6)
4,308
0.328
90
N
N
-0.116
[0.077]
0.042
[0.038]
0.022***
[0.001]
-0.033
[0.060]
0.080**
[0.036]
0.005***
[0.001]
4,308
0.328
90
N
Y
-0.116
[0.224]
0.042
[0.083]
0.022***
[0.003]
-0.033
[0.074]
0.08
[0.064]
0.005***
[0.001]
4,308
0.708
90
Y
Y
-0.217
[0.247]
0.295
[0.201]
0.021***
[0.003]
0.844***
[0.269]
-0.311
[0.226]
-0.006**
[0.003]
July 2008-December 2008
(7)
(8)
(9)
Note: The dependent variable is the overnight CP spread. Credit risk is defined as the CDS spread of the issuer. Liquidity risk is defined as the
fraction of paper maturing over the next week. Post Aug 2007 is a dummy = 1 after August 9, 2007. Bear is a dummy = 1 after March 14, 2008.
Lehman is a dummy = 1 after September 15, 2008.
N
N
-0.027***
[0.006]
0.225***
[0.023]
0.000*
[0.000]
-0.085***
[0.013]
0.258***
[0.035]
0.004***
[0.000]
Issuer-fixed effects?
Robust standard errors
clustered by issuer?
Constant
Liquidity * Lehman
Credit * Lehman
Lehman
Liquidity * Bear
Credit * Bear
Bear
Liquidity * Post Aug 2007
Credit * Post Aug 2007
Post Aug 2007
Liquidity
Credit
January 2007-December 2007
(1)
(2)
(3)
Table 3: Baseline with interaction terms
Table 4: Financial vs Non Financial CP issuers
January 2007-December 2007
ALL
FIN
NON-FIN
(1)
(2)
(3)
January 2008-June 2008
ALL
FIN
NON-FIN
(4)
(5)
(6)
July 2008-December 2008
ALL
FIN
NON-FIN
(7)
(8)
(9)
-0.299***
[0.081]
0
Liquidity
[0.000]
-0.066
Post Aug 2007
[0.048]
0.549***
Credit * Post Aug 2007
[0.112]
Liquidity * Post Aug 2007 0.003***
[0.001]
Credit
Bear
Credit * Bear
Liquidity * Bear
Lehman
Credit * Lehman
Liquidity * Lehman
Constant
Issuer-fixed effects?
Robust standard errors
clustered by issuer?
Observations
R-squared
Number of programs
-0.005
-0.132 0.100*** 0.081*** 0.119*** -0.311
0.028 -1.251***
[0.055]
[0.105]
[0.022]
[0.021]
[0.039]
[0.226] [0.092] [0.258]
0
0
0
-0.001
0
-0.006** -0.008** -0.005
[0.000]
[0.000]
[0.000]
[0.001]
[0.000]
[0.003] [0.004] [0.003]
-0.096*
-0.028
[0.050]
[0.087]
0.186 0.701***
[0.129]
[0.229]
0.004*** 0.002
[0.001]
[0.001]
-0.011
-0.106* 0.082**
[0.034]
[0.059]
[0.035]
0.101*** 0.135*** 0.092**
[0.028]
[0.040]
[0.039]
0.002*** 0.003**
0.001
[0.001]
[0.001]
[0.001]
-0.217
-0.235
-0.198
[0.247] [0.201] [0.360]
0.295
0.071 0.942***
[0.201] [0.097] [0.242]
0.021*** 0.022*** 0.017***
[0.003] [0.006] [0.004]
0.110*** 0.016
0.080* 0.068**
-0.03
0.112*** 0.844*** 0.287* 1.731***
[0.026]
[0.012]
[0.040]
[0.027]
[0.037]
[0.034]
[0.269] [0.153] [0.274]
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
8,543
0.455
89
3,171
0.291
42
5,372
0.484
47
4,557
0.768
85
1,878
0.689
40
2,679
0.768
45
4,308
0.708
90
1,867
0.653
45
2,441
0.743
45
Note: The dependent variable is the overnight CP spread. Credit risk is defined as the CDS spread of the issuer. Liquidity risk is defined as the fraction
of paper maturing over the next week. Post Aug 2007 is a dummy = 1 after August 9, 2007. Bear is a dummy = 1 after March 14, 2008. Lehman is a
dummy = 1 after September 15, 2008. Robust standard errors clustered by issuer. All regressions have fixed effects.
30
Table 5: Dealer placement vs Direct placement
January 2007-December 2007
ALL
100%
<100%
(1)
(2)
(3)
Credit
Liquidity
Post Aug 2007
Credit * Post Aug 2007
Liquidity * Post Aug 2007
Bear
Credit * Bear
Liquidity * Bear
Lehman
Credit * Lehman
Liquidity * Lehman
Constant
Issuer-fixed effects?
Robust standard errors
clustered by issuer?
Observations
R-squared
Number of programs
January 2008-June 2008
ALL
100%
<100%
(4)
(5)
(6)
July 2008-December 2008
ALL
100%
<100%
(7)
(8)
(9)
-0.299*** -0.225 -0.281*** 0.100*** 0.149*** 0.062**
-0.311
[0.081]
[0.142]
[0.079]
[0.022]
[0.024]
[0.026]
[0.226]
0
0
0
0
0
0.001
-0.006**
[0.000]
[0.000]
[0.001]
[0.000]
[0.000]
[0.001]
[0.003]
-0.066
-0.06
-0.056
[0.048]
[0.088]
[0.053]
0.549*** 0.615** 0.473***
[0.112]
[0.235]
[0.121]
0.003*** 0.003***
0.002
[0.001]
[0.001]
[0.001]
-0.011
0.061*
-0.054
[0.034]
[0.034]
[0.058]
0.101*** 0.093** 0.123***
[0.028]
[0.040]
[0.044]
0.002*** 0.001*
0.001
[0.001]
[0.001]
[0.001]
-0.217
[0.247]
0.295
[0.201]
0.021***
[0.003]
0.110*** 0.116**
0.057*
0.068** 0.096*** -0.023 0.844***
[0.026]
[0.044]
[0.033]
[0.027]
[0.028]
[0.039]
[0.269]
-1.288***
0.034
[0.377]
[0.147]
0.001
-0.013***
[0.003]
[0.004]
-0.347
-0.079
[0.383]
[0.265]
1.239***
-0.071
[0.346]
[0.145]
0.014*** 0.027***
[0.004]
[0.006]
1.419*** 0.445**
[0.385]
[0.187]
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
8,543
0.455
89
4,711
0.473
55
3,832
0.405
34
4,557
0.768
85
2,555
0.772
54
2,002
0.713
31
4,308
0.708
90
2,147
0.755
54
2,161
0.634
36
Note: The dependent variable is the overnight CP spread. Credit risk is defined as the CDS spread of the issuer. Liquidity risk is defined as the fraction of
paper maturing over the next week. Post Aug 2007 is a dummy = 1 after August 9, 2007. Bear is a dummy = 1 after March 14, 2008. Lehman is a dummy =
1 after September 15, 2008. Robust standard errors clustered by issuer. All regressions have fixed effects.
31
Table 6: Financial vs Non Financial CP issuers - A1/P1 only
January 2007-December 2007
ALL
FIN
NON-FIN
(1)
(2)
(3)
Credit
Liquidity
Post Aug 2007
Credit * Post Aug 2007
Liquidity * Post Aug 2007
Bear
Credit * Bear
Liquidity * Bear
January 2008-June 2008
ALL
FIN
NON-FIN
(4)
(5)
(6)
-0.134
0.037
-0.133
0.076*** 0.060*** 0.099***
[0.089]
[0.045]
[0.114]
[0.019]
[0.020]
[0.030]
0
0
0
0
0
0
[0.000]
[0.000]
[0.000]
[0.000]
[0.001]
[0.001]
-0.069*
-0.083
-0.058**
[0.035]
[0.051]
[0.021]
0.343***
0.148
0.511***
[0.127]
[0.128]
[0.039]
0.001** 0.002***
0
[0.000]
[0.001]
[0.000]
-0.002
-0.073
0.079*
[0.038]
[0.059]
[0.039]
0.077*** 0.109***
0.058
[0.027]
[0.037]
[0.049]
0.001
0.002
-0.001
[0.001]
[0.001]
[0.001]
Lehman
Credit * Lehman
Liquidity * Lehman
Constant
Issuer-fixed effects?
Robust standard errors
clustered by issuer?
Observations
R-squared
Number of programs
July 2008-December 2008
ALL
FIN
NON-FIN
(7)
(8)
(9)
-0.312*
[0.181]
-0.001
[0.002]
0.002
-1.109***
[0.089]
[0.210]
-0.007**
0
[0.003]
[0.002]
-0.488**
-0.347 -0.628***
[0.193]
[0.210]
[0.152]
0.328**
0.085
0.851***
[0.160]
[0.104]
[0.179]
0.011*** 0.017*** 0.009**
[0.003]
[0.006]
[0.004]
0.106*** -0.076*** -0.165*** -0.035
[0.032]
[0.025]
[0.053]
[0.037]
0.015
[0.011]
-0.021**
[0.009]
0.016
[0.015]
-0.140***
[0.014]
0.005
[0.011]
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
8,543
0.455
89
3,171
0.291
42
5,372
0.484
47
4,557
0.768
85
1,878
0.689
40
2,679
0.768
45
4,308
0.708
90
1,867
0.653
45
2,441
0.743
45
Note: The dependent variable is the overnight CP spread. Credit risk is defined as the CDS spread of the issuer. Liquidity risk is defined as the fraction of paper
maturing over the next week. Post Aug 2007 is a dummy = 1 after August 9, 2007. Bear is a dummy = 1 after March 14, 2008. Lehman is a dummy = 1 after
September 15, 2008. Robust standard errors clustered by issuer. All regressions have fixed effects.
32
Table 7: Dealer placement vs Direct placement - A1/P1 only
January 2007-December 2007
ALL
100%
<100%
(1)
(2)
(3)
Credit
Liquidity
Post Aug 2007
Credit * Post Aug 2007
Liquidity * Post Aug 2007
-0.134
-0.127
[0.089]
[0.130]
0
0
[0.000]
[0.000]
-0.069* -0.099**
[0.035]
[0.043]
0.343*** 0.471***
[0.127]
[0.103]
0.001**
0.001
[0.000]
[0.001]
0.039
[0.067]
0
[0.000]
-0.018
[0.046]
0.066
[0.121]
0.001*
[0.000]
Bear
Credit * Bear
Liquidity * Bear
Lehman
Credit * Lehman
Liquidity * Lehman
Constant
Issuer-fixed effects?
Robust standard errors
clustered by issuer?
Observations
R-squared
Number of programs
January 2008-June 2008
ALL
100%
<100%
(4)
(5)
(6)
0.076*** 0.108***
[0.019]
[0.021]
0
0
[0.000]
[0.000]
0.061**
[0.025]
0.001
[0.001]
-0.002
[0.038]
0.077***
[0.027]
0.001
[0.001]
-0.035
[0.062]
0.088*
[0.046]
0
[0.001]
0.068**
[0.031]
0.073**
[0.030]
0
[0.000]
July 2008-December 2008
ALL
100%
<100%
(7)
(8)
(9)
-0.312*
[0.181]
-0.001
[0.002]
-1.128*** -0.008
[0.177]
[0.096]
0.002
-0.007**
[0.002]
[0.003]
-0.488** -0.680*** -0.276
[0.193]
[0.171]
[0.204]
0.328** 1.033***
0.051
[0.160]
[0.110]
[0.106]
0.011***
0.008
0.015***
[0.003]
[0.005]
[0.004]
-0.140*** -0.070** -0.124*** -0.076*** 0.952*** 0.224*
[0.014]
[0.026]
[0.042]
[0.025]
[0.176]
[0.111]
0.015
[0.011]
0.021
[0.029]
-0.012
[0.014]
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
8,543
0.455
89
4,711
0.473
55
3,832
0.405
34
4,557
0.768
85
2,555
0.772
54
2,002
0.713
31
4,308
0.708
90
2,147
0.755
54
2,161
0.634
36
Note: The dependent variable is the overnight CP spread. Credit risk is defined as the CDS spread of the issuer. Liquidity risk is defined as the fraction of
paper maturing over the next week. Post Aug 2007 is a dummy = 1 after August 9, 2007. Bear is a dummy = 1 after March 14, 2008. Lehman is a dummy = 1
after September 15, 2008. Robust standard errors clustered by issuer. All regressions have fixed effects.
33