Modern Economy, 2012, 3, 641-652
http://dx.doi.org/10.4236/me.2012.35083 Published Online September 2012 (http://www.SciRP.org/journal/me)
Liquidity Expectation, Financing Ability and Credit
Rating: Evidence from China
Shan Tian1, Song Zhu2, Bo Liu3
1Department of Accounting, Tourism and Culture College of Yunnan University, Yunnan, China
2Business School, Beijing Normal University, Beijing, China
3China Sinopharm International Corporation, Beijing, China
Received June 16, 2012; revised July 15, 2012; accepted July 25, 2012
ABSTRACT
Sovereign Credit Ratings of many countries and credit rating of dozens of firms has been downgraded since the latest
financial crisis. However in China short-term finan cing market, the credit rating seems to show a counter-cyclical phe-
nomenon. We find that wh en the market and the economy upsurge, the bond principal rating is relatively poor; when the
market and economy downturn, the bond principal rating is relatively high. In other words, the ratings of short-term
financing bills show a counter-cyclical phenomenon. We propose the liquidity hypothesis for this phenomenon that
during the period of economic prosperity, market liquidity and capital is relatively abundan t; therefore even th e compa-
nies with poor ratings have access to raise funds. When the market downturns and liquidity is poor, there are not enough
funds in the market, thus the companies with poor ratings may fall to finance due to the lack of funds. Therefore, the
liquidity of the market causes the bond rating to show a counter-cyclical phenomenon. Empirical research supports this
hypothesis.
Keywords: Liquidity Expectation; Financing Ability; Credit Rating; Counter-Cyclical Phenomenon
1. Introduction
Since the subprime mortgage crisis, especially after 2011,
many countries’ sovereign rating and lots of corporations ’
credit rating have b een continuously downgr aded that ma k -
es the global economy keep falling. Therefore, the credit
crisis has becoming important bottleneck and constraint
for the global economic recovery. While during and after
the crisis, th e three largest credit rating ag encies (e.g. M oo -
dy, S & P and Fitch) have been heavily criticized, since
the crisis has exposed many flaws of the credit rating
agencies and the rating system. A serious problem is that
the imperfection of the rating model greatly reduces the
accuracy of credit rating. Therefore, to improve the rat-
ing model and make the credit rating accurately to reflect
the default risk has been hotly discussed after the crisis.
As the development of China’s securities market, cor-
porate bonds have become an important investment for
investors. And the important role of bond ratings in eco-
nomic activities is increasingly exposed. Concerning the
political/regulatory environment and credit rating in Chin a,
although crisis like the US subprime mortgage crisis has
never happened, how the credit rating is assessed for do-
mestic enterprises is seldom to be investigated. Besides
the basic financial information as well as other “soft” in-
formation, how the macroeconomic policies impact the
rating is also never d iscussed in China. Actually tho se ques-
tions are very important for global investors.
Some researchers find that the credit rating shows a
pro-cyclical phenomenon in developed countries (see e.g.
[1-3]), however a domestic report1 shows that the credit
rating for bond in China is more likely to be upgraded
during the economic downturns (see [4]), which seem to
be an obvious counter-cyclical phenomenon. Therefore,
how does the macro economy affect the bond rating in
China? And what’s the reason for the counter-cyclical p he -
nomenon of bond rating? This paper attempts to discuss
the credit rating and pro-cyclical o r counter-cyclical phe-
nomenon in China. We find that when the market and the
economy upsurge, the bond principal rating is relatively
poor; when the market and economy downturn, the bond
principal rating is relatively high. In other words, the rat-
ings of short-term financing bills show a counter-cycli-
cal phenomenon. We propose the liquidity hypothesis for
this phenomenon that during the period of economic pros-
perity, market liquidity and capital is relativ ely abundant;
1According to a document obtained by the “First Financial Daily” from
the China Government Securities Depository Trust & Clearing Corpo-
ration, during 2008 to 2010, follow-up ratings of bond issuers, a total o
f
388 ratings upgraded, only 13 downgraded. Especially in 2010, a total
of 179 ratings upgrade, only 3 ratings downgraded. First Financial Da ily ,
September 19, 201 1.
C
opyright © 2012 SciRes. ME

S. TIAN ET AL.
642
therefore even the companies with poor ratings have ac-
cess to raise funds. When the market downturns, liquidity
is poor, and there are not enough funds on the market, the
companies with poor main rating may fall to finance due
to the lack of access to funds. Therefore, the liquidity of
the market causes the bond rating to show the counter-
cyclical phenomenon. Our empirical research supports this
hypothesis.
Section 2 reviews the previous rating literatures, and
section 3 shows our hypothesis. Sections 4 and 5 are re-
search design and empirical tests. Conclusions are shown
in Section 6.
2. Literature Review
Bond rating first appeared in the United States at the be-
ginning of the 20th century. John Moody uses symbols to
show the pros and cons of grade of the various grades of
railway bonds in 1909, this is the first bond rating in the
United States. “The Gr eat Depression” in 1930s proves t ha t
the credit rating basically reflects the repayment of the
bond, which increases the trust of investors in bond rat-
ing. Therefore, the bond rating is crucial for investors.
The bond’s cr edit risk is gener ally con sider ed to be the
bond issuer’s risk of bankruptcy and the default risk of
the bond. In order to reflect those risks, credit rating model
and method are important. Rating methods include: 1) The
credit rating method; 2) The traditional credit rating method;
3) The credit scoring model method and 4) The credit risk
model method. There are numbers of theoretical and em-
pirical researches on credit rating models (see [5-10]). In
rating model, the critical rating factors taken into account
by rating ag encies are pr imarily on two aspect: 1) On bu s i-
ness, focusing on business size, diversity and regional dis -
tribution, the quality of the products or services, market-
ing capability and marketing networks; and 2) On finan-
cial analysis, focusing on the quality of financial infor-
mation, the scale and quality of assets, capital structure
and debt pressures, profitability, cash flow and other fac-
tors (see [11,12]).
The Standard & Poor in its credit rating method men-
tioned that the rating metho d is not limited to th e insp ect-
tion of various financial indicators. Indeed, Butler and F au-
ver [13] study the sovereign credit ratings from the per-
spective of the legal and institutional environment. They
find that the coun try’s legal and institutional environ ment
has conspicuous positive effect on the sovereign credit
rating. In addition, the macroeconomic factors largely det-
ermine the m icroec onom ic perform anc e, so m acroeconomic
factors, especially the economic cycle, play a pivotal role
on bond rating. Carling et al. [14] shows that a variety of
variables of the macroeconomic environment have impor -
tant explanatory power on corporate default risk. If the
state of the economy is divided into the peak, normal and
underestimate situation, it will be found that the probabil-
ity and default is particularly sensitive to economic cy-
cles (see [2]). Therefore, studying the credit rating at
different time points of the economic cycle can demon-
strate the impact of the economic cycle on bond rating.
Treacy and Carey [1] also points out that the rating
changes will be more significant in a recession or expan-
sion, which means there are more downgrades during the
recession and more rati ngs i n c rea s ed du ring the econo mi c
expansion.
Researches in China reg arding credit rating are mainly
on the rating methodology while research for factors af-
fecting bond rating, both theoretical and empirical, is far
lagging behind. By comparing the bond rating system of
the United States, Germany and Japan, Liang [15] finds
that the presence or absence of one country’s rating sys-
tem and its specific model is closely related to bank-firm
relationship. He and Fang [16] propose that Chin a should
promote the healthy development of the bond rating in-
dustry by two ways, creating an external environment c on-
ducive to the healthy development of the rating industry
and strengthening the building of the rating agencies.
Chen and Guo [17], He and Jin [18] study the rating of
corporate bonds primarily from the micro-factors. As to
macroeconomic factors, especially about the economic cy-
cle, monetary policy and institution al factors, there is less
theoretical and empirical research. In China, the capital
market’s performance is largely depending on the policy,
and the main business and the market is highly subject to
the impact of the macroeconomic environment. For this
reason, it is quite practical important to explore China’s
bond rating from a macro perspective.
3. Hypothesis
The rating agencies typically assess the rate of the enter-
prise on the business, financial aspects, as well as some
other “soft factors”. Therefore, the scale of business, di-
versity and regional distribution, the quality of the prod-
ucts or services, the marketing capability and marketing
networks, the quality of financial information, the asset
size and quality, the capital structure and debt pressures,
the profitability and the cash flow are the most basic rat-
ing factors (see [11,12]). In fact, a corporation’s credit
rate is the estimation of its risk of default and bankruptcy.
In other words, the credit ratings of companies that do
not default and bankruptcy will be relatively higher. And
the ways to avoid such problems is timely accessing to
funds for reimbursement. This means that the enterprise
has sufficient financing ability to pay the maturing debt
timely, so the enterprise wou ld not default. Even if due to
operational problems, the enterprise is not profitable, as
long as it can still get money, there would be no bank-
ruptcy problem. Therefore, the key of corporate rating is
whether it has a strong financing ability, if companies ha s
a strong financing ability can reduce the possibility of de-
fault and bankruptcy.

S. TIAN ET AL. 643
A large part of Chinese enterprises are state-owned en-
te r p r i s e s , o p e r a t i o n a l p r oble m s o f t e n un d e r t a k e n b y t h e gov-
ernment, so bankruptc y probl ems are rare. Of course, there
may be the event of default. That is the company unable
to pay maturing debts, but the government as a patron can
solve the problem through debt forgiveness, debt-equity
swap and other means. Of course, not all state-owned en t-
erprises are able to be such a government’s “pet”, becau se
of their importance, there are some differences on the
financing ability. China’s private enterprises that have a
certain political relations also can get strong support from
the government on financing (see [19,20]). Of course, not
all private enterprises have such a “po litical relations”. In
other words, there are big differences among the finance-
ing ability of domestic enterprises. The differences of fi-
nancing ability result in the diversity of maturing debt re-
payment, business failure and bankruptcy, which leading
to the main rating difference. Therefore, we prop ose that:
Hypothesis 1: Financing capability is positively cor-
related with bond rating.
The economic cycles play a pivotal role in the bond
rating, and the various variables of macro environment
have important explanatory power on the corporate de-
fault risk (see [14]). The reason is that economic envi-
ronmental impact the probability of default of the bond,
of default is particularly sensitive to economic cycles (s ee
[2]). There are more rating downgrades during the eco-
nomic recession, whereas in the economic expansion
there are more ratings increased (see [1]).
In fact, the impact of the economic cycle on the bond
rating directly relate to the amount of funding available
by the enterprise. During the economic downturn, the ma r-
ket relatively lacks of liquidity, so it is difficult for en-
terprises to obtain external financing , even if it wanted to
issue bonds. The rating agencies will be relatively more
cautious, and they tend to give poor evaluation on the
corporations, so the rating of the enterprise will be ad-
justed accordingly. During the period of economic pros-
perity, market liquidity is better, and there are more en-
terprises entering the market for financing. The ra ting a ge n-
cies are relatively more optimistic, so they will give com-
panies higher ratings to the enterprise. It is because of the
liquidity situation, the rating shows pro-cyclical utility w it h
the economic cycle.
Therefore, the bond rating or the corporate rating is
based on the expectation of the market liquidity. In other
words, market liquidity will affect the rating of the en-
terprise. When the market is expected to be lacks of li-
quidity to a certain extent, and when the monetary policy
tightening, are difficult to ob tain funding, the anticipated
development of the companies will be impacted by the
tighter monetar y policy. A t the same ti me, the enterp r is es ’
future profitab ility will decline to a certain exten t, so th eir
ratings are relatively poor. Therefore, we assume that:
Hypothesis 2: The declining of Liquidity expectat io n
is negatively related with bond rating.
When the market is expected to be lacks of liquidity to
a certain extent, and the monetary policy tightening, it is
difficult for the companies to obtain funds even if they
have a strong financing ability, and the anticipated devel-
opment of the companies will be influenced by the tig hter
monetary policy, and the fu ture profitability will decline.
Therefore, the declining in market liquidity will have a
negative impact on the bond ratings, and this declining
will reduce the positive impact of strong financing ab ility
on the corporate bond rating. Therefore, we assume that:
Hypothesis 3: The declining of liquidity expectation
will reduce the positive impa ct of the financing ability
on bond rating.
4. Research Design
4.1. Model and Variables
The ratings that the agencies give to the enterprises are
divided into different levels, so we use the Ordered Logit
model for the empirical analysis, and the model is set as
follows:
01 234
567
8 91011
12
cov
CreditListROECPICPI List
CPI ROEGDPgrowInterer
Lev SizeAgeGuarantee
IndSup


 

 

  
 
 

Credit stands for the credit rating, which is the rating
that the professional credit rating agencies give to the is-
suers when their bonds are issued. The credit rating in-
cludes AAA, AAA–, AA+, AA, AA–, A+, A, A– and
BBB levels. In the regression of this paper, AAA and
AAA– are classified as the first level, and assigned a value
of 3. AA+, AA and AA– are regarded as the second level,
assigned a value of 2. A+, A, A– and BBB+ are the th ird
level2, assigned a value of 1. In robust test, the ratings are
divided into 9 degrees, AAA is the first degree, a value
of 8; AAA– is the second degree, the assignment is 7;
descending order of BBB+ for the ninth degree, the as-
signment is zero.
This paper uses List and ROE to measure the financing
ability. The listed companies can absorb funds from the
stock market, so compared to the non-listed companies their
financing channels are relatively broader. What’s more is
that a listed company’s “shell” is valuable, thus the listed
companies have better access to gain debt capital. There-
fore, in terms of the financing channels and financing
strength the listed companies are more competitive than
the non-listed companies. While the higher profitability
of a company, the stronger ability it has to gain benefit
2There are just two samples’ rating are BBB+, so they are classified in
the thirdlevel.

S. TIAN ET AL.
644
for shareholders, and it can protect the interests of credi-
tors’ better, which can improve its ability to absorb new
funds. The higher profitab ility of the compan y, the easier
for the company to be listed to obtain funds, and it is
easier for it to issue or allot shares on the capital market,
the more loans it can get from creditors. Therefore, the
profitability reflects the financing ability to a certain ex-
tent, the higher profitability of the corporations, the stronger
financing ability of them. Based on hypothesis 1, we ex-
pect α1 > 0, and α2 > 0.
Market liquidity expectation is proxy by CPI, and we
use the consumer price index (CPI) of the bond issu-
ance’s month3. Higher CPI stands for severer inflation,
which means that the central bank tends to adjust the
market liquidity by means of monetary policy or market
operations to reduce the external currency flows. There-
fore higher CPI means the lower market liquidity expec-
tation4. Although the growth rates of narrow money (M1)
and broad money (M2) also represent the market liquid-
ity, the higher CPI is more likely to mean a tightening of
monetary policy in the future, while higher growth rate
of M1 or M2 does not more likely mean tighten mone-
tary policy. Thus the CPI reveals liquidity expectation.
Based on hypothesis 2, we expect α3 < 0.
CPI × List and CPI × ROE are the cross-terms of the
market liquidity expectation and the corporations’ finance-
ing ability. Based on hypothesis 3, we predict that α4 < 0
and α5 < 0.
GDPgrow is the measurable variable of output growth
(“economic cycle”), by deducting the real GDP growth
rate by the expected GDP growth. The real GDP growth
rate can be found in the China Statistical Yearbook, and
the World Bank’s growth expectations of the economy
about the countries around the world will be adjusted ac-
cording to actual situation and the macroeconomic envi-
ronment, which have a high correlation with the macro-
economic environment, so we have chosen the world
Bank’s economy forecast data on China as the source of
the expected GDP growth. This variable reveals the gap
between the real economic growth and the economic gro-
wth potential, and it reflects the relative strength of the
current macroeconomic situation, so we use it as a meas-
urable indicator of the economic cycle.
Intercover is the interest coverage ratio, which equals
(total profit + financial expenses)/finance expenses. Lev
is the leverage ratio, equals to total liabilities/to tal assets.
Size stands for the firm size, and we use the natural loga-
rithm of the total assets. Age represents the life cycle of
the enterprises, which is equ ivalent to the years of estab-
lishment up to the month of the security issuance. Guar-
antee is a dummy variable, 1 indicates that the issued
securities were guaranteed and 0 otherwise. IndSup is a
dummy variable for the industry, 1 indicates the industrie s
which are supported by the Chinese industry policy and 0
otherwise.
The data used in the calculatio n of the above variables
are the financial data disclosed in the financial statements
of the previous year that the secu rities are issued.
4.2. Data Sources and Sample
We collect all the documents of short-term financing bills
issued during 2005 to 2010 from the China Bond Informa-
tion Networ k
(http://www.chinabond . com.cn/d2s/index.html).
The documents include bonds issued document, the pro-
spectus, the credit rating reports, the financial statements.
And we collect the issuers’ credit rating, the ownership
structure, the a ud it in st itu tio ns , th e f inan cial d ata d isc losed
in the bond issuance files manually. Due to missing data
for some samples’ issuer credit rating and financial in-
formation, the final sample is 1550 issuance.
Meanwhile, we collect China’s GDP grow th during 2 005
and 2010 as well as CPI data from the China Statistical
Yearbook, and collect the World Bank’s latest forecast
data about China’s annual GDP growth rate from 2005 to
2010 on the website of the World Bank.
5. Empirical Analysis
5.1. Descriptive Statistics
Table 1 shows the descriptive statistics of regression vari-
ables. The average of issuer credit rating is 2 which me a n s
the majority of the issuer ratings is above A. 32.7% of
the total samples are listed firms which means the most
Table 1. Descriptive statistics.
Variable N MeanStd. Min MedianMax
Credit 15502.0600.590 1 2 3
List 15500.3270.469 0 0 1
ROE 15500.0880.089 –0.655 0.0761.017
CPI 15500.0300.026 –0.018 0.0280.087
GDPgrow15500.0100.011 –0.009 0.0080.026
Intercover155010.75820.550 0 4.2795100
Lev 15500.5670.178 0.016 0.5990.977
Size 155014.4961.382 9.867 14.33119.319
Age 155012.9887.380 0.356 11.90292.060
Guarantee15500.0160.126 0 0 1
IndSup 15500.3920.488 0 0 1
3Rating often is given a few months before the bond issuance, so it is
somewhat lag to use the CPI of issuance month, which is equivalent to
expect the CPI after the credit rating.
4This paper has verified the effective of the CPI to represent the liquid-
ity expectation. The growth rate of M1 and M2 represent the market
liquidity, while the correlation coefficients of the CPI with the growth
rate of M1 and M2 are –0.2439 and –0.2284, both significant at 0.01
level. That is CPI is significantly negative correlate with the liquidity
indicators. So CPI on behalf of liquidity expec tati on i s reasonable.

S. TIAN ET AL. 645
issuers of short-term financing bonds are non-listed com-
panies. The average return on equity rate is 8.8% and the
median is 7.6%. For the macroeconomic aspect, the av-
erage CPI increasing is 3%, while the maximum increas-
ing is 8.7%, and the lowest increasing is –1.8%.The av-
erage GDP growth gap is 1%, and it changes from the
highest 2.6% to the lowest –0.9%. The overall economic
development shows a large vo latility from 2005 to 2010.
Other financial data also show great differences.
Table 2 shows credit ratings for issuing entity of the
sample short-term financing bills5. Since the short-term
financing bills market was started in 2005, more and more
companies can get funds from this market, especially in
the 2010 for there is a blowout in the number of issuing
firms. For these companies that get short-term funds, ther e
are significant differences among their credit ratings. Th er e
are 320 samples that get the high est level of rating AAA,
which accounts for 20% of the to tal samples. About 65%
of the issuers rated as AA, and 15% of the samples’ rat-
ings are A. Although these companies are able to issue
short-term financing bills in the market, their risk is still
quite different.
Figure 1 shows the evolution of the credit ratings of
these samples. The number of firms that are rated as A
was increasing year by year from 2005 to 2007, and the
proportion rose to about 3 0%, while the samp les’ propo r-
tion of the AA and AAA level are relative declining . The
situation changed from 2008, the proportion of the sam-
ples that rated the A level decreased significantly, while
the proportion of the other two categories of samples in c-
reased significantly.
Actually from 2005 to 2007, the global economy is in
boom. However, the subprime mortgage crisis that broke
out in the end of 2007 made the global economy fell into
a recession. Therefore, the economic growth or the mar-
ket condition shows an opposite trend with the credit rat-
ing change during 2005 and 2010 in China, which means
the issuer credit ratings are relatively poor when the mar-
ket and the economy are prosperous; the issuer credit
Table 2. Short-term financing bills issuer credit rating.
2005 2006 2007 2008 2009 2010Total
A 7 57 79 41 13 29 226
AA 46 146 145 174 188 305 1004
AAA 26 35 39 52 61 107 320
Total 79 238 263 267 262 441 1550
Note: AAA includes AAA an d AAA–, AA includes AA+, AA and AA–, A
includes A+, A and A–6.
Figure 1. The credit rating for short-term financing bills.
ratings are relatively better when the market and the econ -
omy downturn. In other words, the credit rating of short-
term financing bills shows a counter-cyclical phenome-
non in China. The possible reason is that: 1) Lots of com-
panies entered the market to get funds and the market
liquidity was strong when the economy was in boom, and
the capital is relatively abundant, so even the companies
with poor credit ratings can make access to funds. When
the market is downturn, the liquidity is poor and the fu nds
is not enough on the market, the companies with poor r at-
ings may fail to finance du e to the lack of access to funds,
so this reduce the proportion of the low ratings. There-
fore, it is the market liquidity that causes th e credit rating
shows the counter-cyclical phenomenon; 2) The rating
agencies manipulated the ratings. The market is lack of
liquidity when the economy is relativ ely poor, the corpo-
rations with poor ratings can’ t make access to finance s i n c e
the fund is not enough on the market. So in order to gain
earnings the rating agencies may artificially loose the rat-
ing criteria and raise the issuer credit ratings. Therefore,
the final result is that the proportion of the companies
with poor ratings declined and there were more corpora-
tions that have good ratings. Whatever is the reason, the
fundamental problem is that the market is lack of liquid-
ity and funds.
Table 3 shows the statistical data of listed and no n- l is te d
samples. Non-listed sample is the majority, it accounts
for about 67%, and the listed samples accounted for ap-
proximately 33% of the total samples.
Figure 2 shows the proportion of the listed samples
and the non-listed samples. From 2005 to 2008, the pro-
portion of the non-listed companies is rising, and more
and more listed co mpan ies came to raise funds in th e sh ort -
term financing bills market after 2008. The stock market
fell sharply in 2008, even the listed companies were dif-
ficult to raise funds by issuing stocks, so many compa-
nies choose the bond market to finance, thereby squeezed
the financing ability of the non-listed companies. Com-
pared to the non-listed companies, the listed companies
have more channels of financing.
5In China, all the credit ratings for short-term financing bills are A-1,
showing no differences. While the credit rating for issuing entity has
great differences, from A– to AAA.
6There are two samples’ credit ratings are BBB+. Due to small sample
size
,
this article classifies the
m
to A.
Table 4 compares the financing ability between listed
samples and non-listed sam ples. The le verage ratio reflects

S. TIAN ET AL.
646
Table 3. The listing situation of the short-term financing bills
issuers.
2005 2006 2007 2008 2009 2010All
Non-listed Samples 51 143 186 200 177 286 1043
Listed Samples 28 95 77 67 85 155507
All 79 238 263 267 262 441 1550
Figure 2. Proportions of listed and non-listed samples each
year.
Table 4. The listed and non-listed samples’ financing ability
comparison.
Lev ROE
Mean Median Mean Median
Non-listed Sample 0.583 0.623 0.085 0.069
Listed Sample 0.535 0.548 0.096 0.089
Difference 0.048 0.075 –0.011 –0.019
t/Z Value 5.070*** 6.534*** –2.312** –5.742***
Note: Lev is the latest asset-liability ratio of the short-term financing bills
issuers, and it equals to total liabilities/total assets. ROE is the latest return
on equity of the short-term financing bills issuers, and it equals to net profit/
equity.
the compan ies’ future financ ing space of debt, and the prof-
itability reflects the basic financing ability. The average
leverage ratio for non-listed samples is 0.583, and the me-
dian is 0.623, both are significantly higher than listed sam-
ples. In other words, the listed samples’ debt financing space
is relatively higher. For the profitability indicator ROE,
non-listed samples’ average is 0.085, and the median is
0.069, both are significantly lower than listed samples,
which means the profitability and financing ability of li st e d
samples are stronger than non-listed companies. In short,
according to the comparison of leverage and profitability
in Table 4, the financing ability of listed companies is
stronger than non-listed companies. Integrated with the
Tables 2 and 3, it shows that the financing ability (in-
cluding funding sources and financing strength) of listed
companies is stronger than non-listed companies.
5.2. Regression Analysis
Table 5 shows the regression results7. Coefficient of eco-
nomic cycle variable (GDPgrow) is significantly negative
in all regressions, which is consistent with the trend in
Figure 1 and means the issuer credit ratings of the short-
term financing bills show a counter-cyclical ph enomenon.
When the economic growth is poor, the issuer credit rat-
ings are quite good; on the contrary, when the economic
growth is good, the issuer credit ratings are poor. The co-
efficient of the interest coverage ratio (Interco ver) is sig-
nificantly positive, and it indicates that the better solve ncy
of the bond issuance enterprise, the higher of its credit
rating. The enterprise with higher debt ratio, the potential
of its financing ability is relatively lower, so its credit
rating is relatively low. The risk of the larger companies
is generally lower, so their credit ratings are higher. The
life cycle of the bond issuers does not have significant
impact on the credit rating (Age’s coefficient is not sig-
nificant). The credit ratings of guaranteed bonds’ issuers
are not high, and the coefficient of the Guarantee vari-
able is significantly negative. The lik ely reason is that th e
companies with strong solvency usually do not need to
guarantee their bond, and the bonds of inferior corpora-
tions will be asked to be guaranteed. In addition, enter-
prises in the industry of the national policy support (Ind-
Sup = 1), will receive state support on operating and debt
service in the future, so the risk is relatively low, and
their issuer credit ratings are high .
We control those above factors to test the influence of
the financing ability on the credit rating in Model 1. The
coefficient of listing variable (List) is significantly posi-
tive at the 0. 01 leve l. It means the lis ted comp anies’ c r e di t
ratings are higher than the non-listed companies, which
illustrates whether listed will significantly affect the is-
suers’ credit rating. Compared to non-listed companies,
the listed compan ies have the qualifications and cap abili-
ties to raise funds on the capital markets; what’s more,
the specific listing “shell” of Chinese capital market mak es
the listed companies having the ability to attract other
funds. The coefficient of profitability (ROE) is also sig-
nificantly positive at the 0.05 level. That is, the stronger
of the corporations’ profitability, the higher of their credit
rating. The profitability represent the potential of financ-
ing, and the enterprises with strong financing potential can
solve the existing financial problems by financing th rou gh
other means when they have difficulties to repay matur-
ing debts, so their credit ratings are relatively high. The
listed variable (List) and the profitability variable (ROE)
that are on behalf of the corpo rations’ financing capab ili-
ties are significantly positive correlated with the issuer rat-
ing, so the Hypothesis 1 is supported.
We then add the liquidity expectation variable in Mode l
2. The higher of the current CPI, the more likely that the
central bank will tighten the monetary policy in the coming
7The largest correlation coefficient between the explanatory variables is
0.275, that there is no serious co-linearity problem.

S. TIAN ET AL.
647
Table 5. The financing capabilities, market liquidity and the credit rating.
Variable Sign Expectation Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7
List + 0.351*** 0.297** 0.636*** 0.299** 0.632*** 0.628*** 0.444*
(2.67) (2.23) (3.12) (2.24) (3.11) (3.07) (1.74)
ROE + 1.634** 1.846*** 1.839*** 3.523*** 3.540*** 3.505*** 3.719***
(2.36) (2.66) (2.64) (3.61) (3.59) (3.55) (3.19)
CPI – –17.911*** –14.474*** –12.843*** –9.363*** –8.883** –10.966**
(–7.28) (–4.98) (–3.96) (–2.60) (–2.44) (–2.41)
CPI × List – –11.066** –10.946** –11.311** –8.091
(–2.20) (–2.19) (–2.26) (–1.32)
CPI × ROE – –56.742** –57.014** –57.944** –74.297***
(–2.39) (–2.40) (–2.42) (–2.64)
GDPgrow ？ –18.903*** –25.142*** –25.293*** –25.288*** –25.730*** –25.363*** –22.073***
(–3.43) (–4.44) (–4.46) (–4.46) (–4.54) (–4.47) (–3.07)
Intercover + 0.015*** 0.015*** 0.015*** 0.015*** 0.015*** 0.015*** 0.014***
(4.70) (4.70) (4.72) (4.73) (4.74) (4.73) (3.77)
Lev – –0.940** –1.557*** –1.484*** –1.529*** –1.450*** –1.430*** –1.438***
(–2.54) (–4.08) (–3.87) (–4.00) (–3.78) (–3.71) (–2.88)
Size + 1.359*** 1.415*** 1.418*** 1.420*** 1.422*** 1.411*** 1.441***
(20.11) (20.34) (20.32) (20.34) (20.33) (20.21) (15.21)
Age ？ 0.008 0.007 0.010 0.009 0.007 0.006 0.002
(1.03) (0.82) (1.11) (1.04) (0.83) (0.78) (0.22)
Guarantee + –1.216*** –1.332*** –1.334*** –1.364*** –1.363*** –1.824***
(–2.74) (–2.89) (–2.91) (–2.97) (–2.98) (–3.42)
IndSup + 0.417*** 0.367*** 0.380*** 0.383*** 0.393*** 0.367***
(3.24) (2.82) (2.92) (2.94) (3.01) (2.80)
N 1550 1550 1550 1550 1550 1525 941
LR chi2 794.55 849.59 855.00 855.67 860.00 830.50 439.94
R-sq 0.2887 0.3087 0.3107 0.3109 0.3125 0.3072 0.2741
Note: The figures in the brackets are white-t statistic, and the paper has considered the problem of heteroscedasticity; ***, **, *denote signif icant a t th e 0 .01, 0.05
and 0.10 level.
month or months, and the liquidity of the capital market
will be lowered, which may raise the cost of debt financ-
ing, and simultaneously raise the corporations’ cost of
funds, and finally have a negative impact on the business.
In other words, the worse of the liquidity expectation (the
higher of the CPI), the higher of the cost of funds, which
may generate a negative impact on the corporations’
performance and solvency, and it lowers the credit rating.
The Hypothesis 2 is supported. The coefficients of the
listed variable (List) and the profitability (ROE) are still
significantly positive. That is the stronger of the financ-
ing ability, the h igher of the credit rating, so th e Hyp oth e-
sis 1 is still supported.
In Model 3 and Model 4 we respectively analyses the
influence of the liquid ity expectation to the impact of th e
financing ability on the credit rating by using two financ-
ing ability vari ables. The lis t variable ( CPI × List) is t ak e n
into account in Model 3, and the profitab ility (CPI ×ROE)
is considered in Model 4, and then this paper conducts a
comprehensive analysis of this two aspects (CPI × List,
CPI × ROE) in Model 5. The coefficients of the list vari-
able (List) and the profitability (ROE) are still signific ant ly
positive, and the coefficient of the liquidity expectation
(CPI) is still significantly negative, all of which is con-
sistent with the previous results.
The coefficients of CPI × List and CPI × ROE are sig-
nificantly negative at the 0.05 level. That means the liqu id-
ity expectation not only has a negative impact on the bond
rating directly, but also reduces the positive contribution
of the financing ability on the credit rating. Despite whether
the issuers are listed and have strong profitability or not,
the liquidity ex pectation will lower th e credit rating of t he
bond issu ers.
Part of the sample companies issued short-term financ-
ing bills with guarantee, which may show a different tren d ;
therefore we exclude the samples that are guaranteed in
Model 6, while the result is entirely consistent with the
previous. Financing ability is positive with the credit rat-
ing, and the liquidity expectation has a negative impact on
the credit rating, the poor liquid ity expectation will lower
the credit rating directly and simultaneou sly let down the
positive impact of the financing ability on the credit rating.
Some samples’ issuers belong to the industry that has
the great support from the government and the effect of t he
policy may be very small on them, even if the monetary
is tightened. So the paper excludes the samples belong to

S. TIAN ET AL.
648
the support industry in Model 7. Similarly, the result is still
consistent with previous, and the Hypothesis 1 and the Hy -
pothesis 2 are still supported.
In summary, the results in Table 5 show that the fi-
nancing ability can sign ificantly improve the enterprise’s
credit rating, but this ability will be influenced by the
monetary policy or the liquidity expectation. The liquid-
ity expectation has a significant negative impact on the is-
suer credit rating.
Passive rating may be helpful for the rating objects to
establish some kind of “relationship” with the rating ag en -
cies, and the rating agencies tend to rely on the “soft” in-
formation obtained through internal channels rather than
the “hard” information got from the open market when
the rating agencies have “relationship” with the rating
objects. Chang et al. [21] found that the banks took “soft
information” into accoun t when they loaned to the enter-
prises, such as whether state-owned and the equity ratio
of the controlling shareholder. Therefore, the robustness
test showed in Table 6, and Model 8 controls the relevant
characteristics of the ownership structure. State repr esent s
the nature of the ultimate controller, 1 stands for state-
owned, and 0 otherwise. V is control right (Voting right)
of the ultimate controller, Deviation is the difference be-
tween the control right and the cash flow right of the ul-
timate controller, the calculation method is same as the
research of Fan and Wong [22]8. The results show that
the nature of th e enterprises (State or not) has significant
effect on the credit rating. And the state-owned enterpr is e s’
credit ratings are significantly higher than the non-state-
owned enterprises. The reason of this phenomenon is the
state-owned enterprises have the final backing—Govern-
ment. The Government will lend a “helping hand” to solve
the problems when the corporations encounter problems
on business. The coefficient of the ultimate controllers’
equity ratio (V) is signi ficant ly posi ti ve. That is, the higher
of the corporations’ equity ratio, the higher of the credit
rating. The possible reason is that the major shareholders
of the corporations with high control right have lower
motivation to empty the enterprises (see [23]).
The coefficient of the control right and the cash flow
right discrepancy (Deviation) is negative, but it is not sig-
nificant. Financing ability (List and ROE) is positive wit h
the credit rating, and the liqu idity expectatio n has a neg a-
tive impact on the credit rating.
The poor liquidity expectation will lower the credit ra t-
ing directly and simultaneously let down the positive im-
pact of the financing ability on the credit rating (the co-
efficients of the CPI × LIST and CPI × ROE are signifi-
cant at the 0.05 level). Our t hree hypotheses are supported.
The previous study found that the accounting informa-
tion played an important role on the corporations’ invest-
ment and financing (see [24]), and this indicated that the
accounting information may also play a key role on the
credit rating. Therefore, this paper controls the influence
of the accounting information qu ality on the credit rating
in Model 9. We choose the bond issuers’ audit institu-
tions as dummy variables of the quality of the accounting
information. Big4 is a dummy variable, 1 means that the
auditing firms are the four biggest international audit firms
and 0 otherwise. Top10 is the dummy variable of the Chi -
nese audit firms’ top ten9, 1 means that the audit firms a re
the biggest ten of Chinese audit agencies, and 0 otherw i s e .
It is generally accepted that the customers’ accounting in-
formation quality of the Big Four audit is better than the
top ten audit agencies, and both are higher than the com-
panies that audited by Chinese non-top ten audit firms.
The results show that the quality of the accounting infor-
mation significantly affects the credit rating. The higher
is the information’s quality, the better of the credit rating.
The financing ability (List and ROE) is positive with the
credit rating, whereas the poor liquidity expectation will
lower the credit rating directly (the coefficient of the CPI
is significantly negative) and simultaneously let down the
positive impact of the financing ability on the credit rat-
ing (the coefficients of the CPI × LIST and CPI × ROE
are significant at the 0.05 level). The three hypothesis of
this paper are supported.
Since the financial crisis, the reputation and credibility
of the international rating agencies have been criticized,
and other small rating agencies’ credibility is even more
doubtful. While the rating business of China’s short-term
financing bills market is basically monopolized by four
Chinese rating agencies: the Emirates International Credit
Rating Co., Ltd., the Integrity of the International Credit
Rating Co., Ltd., Joint Credit Rating Co., Ltd. and Shang -
hai New Century Ratings Co., Ltd.
Except for the Emirates International, the other three
rating agenci es are more or less controlled by fore ign c api -
tal, and this difference of the ownership property may lea d
to the distinction of the risk controlling. Cantor and Packer
[25] find that the rating results made by the rating agen-
cies are usually opposite with the viewpoint of the inves-
tors and the regulators, and the connotation of the rating
made by different rating agencies at different times is
quite different. Therefore, we further control the differ-
ence of rating agencies in Model 10. The result shows th a t
the issuer credit ratings made by the Emirates Interna-
tional Credit Rating Co., Ltd. are significantly lower than
the other three agencies. The possible explanations for that
is the rating criteria of the Emirates International Credit
Rating Co., Ltd. is more stringent than the other three,
and another reason may be the q uality of its customers is
relatively poor. The coefficients of the financing ability
(List and ROE), liquidity expectation (CPI) and the cro ss-
variables (CPI × LIST and CPI × ROE) are consistent with
8The status of the shareholders is disclosed in the prospectus.
9The audit firm rankings refer to the 2010 national firm rankings.

S. TIAN ET AL.
649
Table 6. The robustness test.
Model 8 Model 9 Model 10 Model 11 Model 12 Model 13
Variable Ownership Information qualityAudit agency List Non-listed Detailed rati ng
List 1.110*** 0.906*** 0.876*** 0.696***
(4.63) (3.67) (3.53) (3.56)
ROE 3.669*** 3.810*** 3.790*** 4.108** 3.252** 2.782***
(3.73) (3.80) (3.77) (2.41) (2.45) (3.42)
CPI –9.282** –8.307** –8.109** –21.208*** –7.716* –9.973***
(–2.57) (–2.28) (–2.22) (–3.11) (–1.93) (–3.35)
CPI × List –11.272** –9.941* –10.061** –4.415
(–2.23) (–1.94) (–1.97) (–1.09)
CPI × ROE –57.557** –61.624** –62.212*** –49.958 –61.170** –33.546*
(–2.42) (–2.57) (–2.58) (–0.96) (–2.06) (–1.65)
GDPgrow –25.643*** –24.009*** –24.057*** –29.194*** –23.673*** –14.475***
(–4.51) (–4.16) (–4.16) (–2.68) (–3.38) (–3.09)
Intercover 0.015*** 0.012*** 0.011*** 0.001 0.016*** 0.012***
(4.72) (3.59) (3.52) (0.14) (3.91) (4.69)
Lev –1.576*** –1.574*** –1.548*** –1.801** –1.183** –1.867***
(–4.07) (–4.01) (–3.94) (–2.34) (–2.50) (–5.73)
Size 1.366*** 1.308*** 1.298*** 1.417*** 1.251*** 1.339***
(19.27) (18.17) (17.92) (9.15) (14.91) (22.67)
Age 0.007 0.005 0.005 –0.030 0.009 0.007
(0.79) (0.59) (0.55) (–0.95) (1.01) (1.00)
Guarantee –1.487*** –1.536*** –1.559*** –0.467 –1.914*** –0.893**
(–3.22) (–3.27) (–3.31) (–0.45) (–3.47) (–2.49)
IndSup 0.342*** 0.363*** 0.373*** 0.409 0.300* 0.325***
(2.60) (2.73) (2.79) (1.48) (1.90) (3.09)
State 0.508*** 0.484*** 0.476*** 0.438* 0.716*** 0.465***
(2.97) (2.81) (2.76) (1.67) (2.89) (3.39)
V 0.006* 0.008** 0.007** 0.022*** 0.001 0.005**
(1.75) (2.32) (2.30) (3.21) (0.33) (2.06)
Deviation –0.013 –0.013 –0.013 –0.023 0.005 –0.010
(–1.21) (–1.25) (–1.27) (–1.64) (0.26) (–1.22)
Big4 1.143*** 1.113*** 1.059*** 1.706*** 0.954***
(4.89) (4.75) (3.11) (3.91) (4.96)
Top10 0.544*** 0.557*** 0.700** 0.472** 0.537***
(3.49) (3.54) (2.24) (2.50) (4.35)
SHNCR –0.070 0.326 –0.496 0.270
(–0.31) (0.97) (–1.52) (1.49)
EICR –0.273* –0.081 –0.315* –0.154
(–1.72) (–0.27) (–1.65) (–1.23)
IICR –0.191 –0.668** –0.001 –0.052
(–1.22) (–2.25) (–0.00) (–0.42)
N 1550 1550 1550 507 1043 1550
LR chi2 879.32 910.87 914.27 350.53 575.15 1172.78
R-sq 0.3195 0.3310 0.3322 0.3832 0.3168 0.2204
Note: The f igu res i n th e brack ets are wh ite- t st ati sti c, and the pap er h as co ns idered th e prob lem of heter os cedas ti cit y; ***, **, *d eno te s ign ifican t at th e 0.01 , 0 .05
and 0.10 lev el.

S. TIAN ET AL.
650
previous results. The three hypotheses are still supported.
There may be big differences between listed and non-
listed samples, thus in order to better reveals the impact
of the liquidity expectation on the credit rating we distin-
guish the listed and non-listed samples and respectively
conduct regression in the Model 11 and Model 12. We
control the influences of the financial risk (Intercover and
Lev), the corporations’ fundamental characteristics (Size
and Age), the feature of the securities (Guarantee and Ind-
Sup), the macro-economic cycle (GDPgrow), the owner-
ship characteristics (State, V and Deviation), the account-
ing information quality (Big 4 and Top10), and the rating
agencies (the Emirates International Credit Rating Co.
(EICR), Ltd., the Integrity of the International Credit Rat-
ing Co. (IICR), Ltd., Joint Credit Rating Co. (JCR), Ltd.
and Shanghai New Century Ratings Co., Ltd. (SHNCR))
on credit rating. There exist differences on the financing
ability between listed companies and non-listed compa-
nies, but both of them’ financing ability (ROE) are sig-
nificantly positive with the credit rating, which supports
hypothesis 1. The coefficients of the liquidity expectation
(CPI) are significantly negative in Model 11 and Model
12, which supports the hypothesis 2. The coefficient of
CPI × ROE is significantly negative, meaning that the
poor liquidity expectation can reduce the positive impact
of the financing ab ility on the credit rating, so it su pports
the hypothesis 3.
In the analysis above, the credit ratings are divided
into three levels by the major categories (AAA, AA and
A levels), but the AAA level includes AAA and AAA–,
the AA can be divided into AA+, AA and AA–, and the
A includes A + A, A– and two BBB+, there may be some
differences among the same general category, so this pa-
per uses more detailed ratings for the explanatory vari-
ables in Model 13. We assigned 0 to the lowest rating
BBB+, and assigned 8 to the most senior rating AAA.
The regression model is still the ordered Logit model.
The results showed that only the coefficient of CPI × List
is negative but not significant, the signs of other variables
are consistent with previous. In summary, the results in
Table 6 further support our three hypotheses.
The coefficient of the ownership property (State) is sig -
nificantly positive in the above regression analysis, mean -
ing that state-owned enterprises have a distinct advantage
than non-state-owned enterprises on credit rating and that
probably because of the difference of the two types of
companies on the financing ability. To learn more about
the impact of the state-owned enterprises and non state-
owned enterprises’ financing ability on the credit rating,
and how the liquidity expectation influences these two types
of companies, we distinguishes the state samples and non-
state samples respectively. To be succinct, we don’t show
the result in text.
In the regression for state samples, the coefficient of the
List is not significant, that means for the state-owned en-
terprises, whether listed or not has no significant influ-
ence on the credit rating. Many group companies have good
qualification, but they are not listed. Listed or not can’t
be fully representative for the financing ability of the state-
owned enterprises. The coefficient of ROE is significant ly
positive, that means the higher the profitability of the en-
terprises, the stronger the financing ability, the h igher the
issuer credit ratings. The coefficient of the CPI is nega-
tive, but not significantly, indicating that the poor liqu id-
ity expectation doesn’t have significant effect on the s ta te-
owned enterprises, the explanation is that the large state-
owned enterprises can still get l oan from the banks or raise
funds on the capital markets. Although the coefficient of
CPI × List is negative, it isn’t significant. The implica-
tion of that the tightened monetary policy has little im-
pact on the financing ability of the state-owned enterpris es
whether they are listed or not. While the coefficient of
CPI × ROE is significantly negative, and that reveals the
liquidity expectatio n will affect the po sitive impact of th e
financing ability on the credit rating which characterized
by the profitability. In short, the regression of the state
samples shows that the financing ability can increase the
credit rating, and the liquidity expectation has a certain
influence on the credit rating, bu t it is relatively limited.
In the regression for non-state samples, the coefficient
of List is significantly positive, that shows for non-state
corporations only the listed enterprises can solve the pro b -
lem of financing very well and improve their financing
ability, thereby enhancing the bond the principal rating.
The coefficient of ROE is po sitive, but it isn’t significant.
The coefficient of CPI is significantly negative, that is,
the impact of the liquidity expectation for the non-state-
owned enterprises is very obvious, when the liquidity ex-
pected to shrink, the non-state-owned enterprises are first
to be influenced, the financing ability of them will be in-
hibited and the credit ratings of the non-state enterprises
will be lowered, which is shown by the significantly neg a -
tive coefficients of CPI × List and CPI × ROE. In short,
the regression of the non-state samples shows that the
liquidity expectation has great impact on the credit rating
of the non-state enterprises. The main reason is that the
financing channels of non-state-owned enterprises are rel a-
tively narrow; the financing ability of the non-listed and
non-state-owned enterprises is even lower, so much low er
for the issuer’s credit rating.
6. Conclusions
From 2005 to 2007, the global economy was in boom.
While the subprime mortgage crisis that broke out in the
end of 2007 made the global economy fell into recession
from 2008 until now, an d i t even become worse. However,
during this p eriod , the main r ating of th e issuing ente rp r is es
on China’s short-term financing bills market presents the

S. TIAN ET AL. 651
opposite trend. That is wh en the market and the economy
upsurge, the bond principal rating is relatively poor; wh en
the market and economy downturn, the bond principal rat-
ing is relatively h igh. In oth er words, the rating s of short-
term financing bills show a count er-cyclical phenomenon.
This paper proposes a theoretical explanation for this
phenomenon. That is the liquidity hypothesis. During the
period of economic prosperity , m arket li quidity and capital
is relatively abundant, therefore even the companies with
poor ratings have access to gain funds. When the market
downturns, liquidity is poor, and there are not enough
funds on the market, the companies with poor main rat-
ing may fall to finance due to the lack of access to funds.
Therefore, the liquidity of the market causes the bond rat-
ing shows the counter-cyclical phenomenon. The empiri-
cal research supports this hypothesis. Thus, when the
liquidity expectation is bad, the credit rating of corpora-
tions will be lowered. Even the enterprises with strong
financing ability, its cred it rating will be impacted by the
negative expectation of the liquidity. In addition, the credit
rating is also affected by the risk of corporate finance,
fundamentals characteristic, the issue characteristics of the
bond, the ownership characteristics, the quality of the in f-
ormation, as well as the impact of the rating agencies.
As we know few researches on credit rating in China
mentioned the “cyclical phenomenon” and study the rela-
tionship between credit rating and macro-economy. Our
paper provides some evidences for this field and shows
distinguish for China credit rating market from developed
markets. Although in our paper we investigate the rela-
tionship between credit rating and macro-factors, leaving
the credit rating untouched, which may reflect the “cy-
clical phenomenon” more evidently. Further work should
be done on the change of credit rating in the future.
REFERENCES
[1] W. F. Treacy and M. S. Carey, “Credit Risk Rating at
Large US Banks,” Federal Reserve Bulletin, Vol. 84, No.
11, 1998, pp. 897-921.
[2] P. Nickell, W. Perraudin and S. Varotto, “Stability of Rat-
ing Transitions,” Journal of Banking and Finance, Vol.
24, No. 1-2, 2000, pp. 203-227.
doi:10.1016/S0378-4266(99)00057-6
[3] J. Amato and C. Furfine, “Are Credit Ratings Procycli-
cal?” Journal of Banking and Finance, Vol. 28, No. 11,
2004, pp. 2641-2677. doi:10.1016/j.jbankfin.2004.06.005
[4] Y. F. Dong, “The Rare Continuous Downgrade on the
Domestic Bond Market—Facing a New Round of Credit
Crisis,” First Financial Daily, 11 September 2011.
[5] M. C. Robert, “On the Pricing of Corporate Debt: The
Risk Structure of Interest Rates,” Journal of Finance, Vol.
29, No. 2, 1974, pp. 449-470.
[6] E. Jones, S. Mason and E. Rosenfeld, “Contingent Claims
Analysis of Corporate Capital Structures: An Empirical
Investigation,” The Journal of Finance, Vol. 39, No. 3,
1984, pp. 611-625.
doi:10.1111/j.1540-6261.1984.tb03649.x
[7] E. I. Altman, “The Success of Business Failure Prediction
Models: An International Survey,” Journal of Banking
and Finance, Vol. 8, No. 2, 1984, pp. 171-198.
doi:10.1016/0378-4266(84)90003-7
[8] E. J. Elton, M. J. Gruber, D. Agrawal and C. Mann, “Ex-
plaining the Rate Spread on Corporate Bonds,” Journal of
Finance, Vol. 56, No. 1, 2001, pp. 247-261.
doi:10.1111/0022-1082.00324
[9] Y. Jafry and T. Schuermann, “Measurement, Estimation,
and Comparison of Credit Migrati on Matrices,” Journal of
Banking and Finance, Vol. 28, No. 11, 2004, pp. 2603-
2639. doi:10.1016/j.jbankfin.2004.06.004
[10] E. Altman and H. Rijken, “How Rating Agencies Achieve
Rating Stability,” Journal of Banking and Finance, Vol.
28, No. 11, 2004, pp. 2679-2714.
doi:10.1016/j.jbankfin.2004.06.006
[11] E. J. Elton, M. J. Gruber, D. Agrawal and C. Mann,
“Factors Affecting the Valuation of Corporate Bonds,”
Journal of Banking & Finance, Vol. 28, No. 11, 2004, pp.
2747-2767. doi:10.1016/j.jbankfin.2004.06.008
[12] W. Perraudin and A. P. Taylor, “On the Consistency of
Ratings and Bond Market Yields,” Journal of Banking
and Finance, Vol. 28, No. 11, 2004, pp. 2769-2788.
doi:10.1016/j.jbankfin.2004.06.009
[13] A. W. Butler and L. Fauver, “Institutional Environment
and Sovereign Credit Ratings,” Financial Management,
Vol. 35, No. 3, 2006, pp. 53-79.
doi:10.1111/j.1755-053X.2006.tb00147.x
[14] K. Carling, T. Jacobson, J. Linde and K. Roszbach,
“Corporate Credit Risk Modeling and the Macroecon-
omy,” Journal of Banking & Finance, Vol. 31, No. 3,
2007, pp. 845-868. doi:10.1016/j.jbankfin.2006.06.012
[15] Q. Liang, “The Comparison of the Western Securities
Rating System and Its Implications for China,” Securities
Market Herald, Vol. 88, No. 11, 1999, pp. 41-44.
[16] Y. Q. He and Z. B. Fang, “The Bond Credit Rating and
the Credit Risk,” Management Science, Vol. 16, No. 4,
2003, pp. 45-50.
[17] C. Chen and Z. M. Guo, “The Corporate Bond Financing,
Financial Risk and the Bond Ratings of China,” The Con-
temporary Finance & Economics, Vol. 279, No. 2, 2008,
pp. 39-48.
[18] P. He and M. Jin, “The Influence of Credit Rating in the
Bond Market in China,” Journal of Financial Research,
Vol. 358, No. 4, 2010, pp. 15-28.
[19] M. G. Yu and H. B. Pan, “The Political Relationship, the
Institutional Environment and the Bank Loans of the Pri-
vate Enterprises,” Management World, Vol. 179, No. 8,
2008, pp. 9-21.
[20] D. L. Luo and L. M. Zhen, “The Private Control, Political
Relationship and the Corporate Finance Constraints,” Fi-
nancial Research, Vol. 342, No. 12, 2008, pp. 164-178.
[21] C. Chang, G. Liao, X. Yu and Z. Ni, “Information from
Relationship Lending: Evidence from Loan Defaults in
China,” Working Paper, 2009.

S. TIAN ET AL.
652
[22] J. Fan and T. J. Wong, “Do External Auditors Perform a
Corporate Governance Role in Emerging Markets? Evi-
dence from East Asia,” Journal of Accounting Re- search,
Vol. 43, No. 1, 2005, pp. 35-72.
doi:10.1111/j.1475-679x.2004.00162.x
[23] S. Zhu, “The Characteristics of Ultimate Controller and
the Informativeness of Accounting Earnings,” China Ac-
counting and Finance, Vol. 8, No. 3, 2006, pp. 1-29.
[24] S. Zhu and D. L. Xia, “The Accounting Conservatism,
Financing Constraints and the Capital Investment of the
Listed Companies,” Finance and Economics, Vol. 36, No.
6, 2009, pp. 69- 79.
[25] R. Cantor and F. Packer, “Differences of Opinion and
Selection Bias in the Credit Rating Industry,” Journal of
Banking & Finance, Vol. 21, No. 10, 1997, pp. 1395-
1417. doi:10.1016/S0378-4266(97)00024-1