The Effect of Fundamental Risk of Listed Companies on the Market Pricing of Accruals Quality

doi:10.4236/ib.2013.51B002

Paper Menu >>

Journal Menu >>

iBusiness, 2013, 5, 6-12

doi:10.4236/ib.2013.51b002 Published Online March 2013 (http://www.s cirp.org/journal/ib)

The Effect of Fundamental Risk of Listed Companies on

the Mark et P ricing of Accru als Qualit y

——Base on the Data of Shanghai’s Non-Financial Industry

Zhaoyuan Geng, Zhendong Wang, Tian Song, Tingjun Liu,Wenting Chi, Yue Yu, Lufei Wang,

Tong Zhang, Chaoying He, Minjie Wang, Yiyang Zheng, Yipeng Zhu, Gengen Zhou, Tiantian Li

Department of Applied Economics, Business School of Zhejiang University City College, Hangzhou, China.

Email: gengzy@zucc.edu. cn

Received 2013

ABSTRACT

Motivated by the theoretical results of Yee (2006), with accruals quality, the author of this paper studied enterprises'

earnings quality management, and analyzed the effect of accruals quality on capital cost, which is rising with the in-

crease of basic risks, and ext end ed and a ppl ied to hi s st ud y the t heo ret ical stud y o f Fr anci s, LaFond, Olsson and Schip-

per et al .

Keywords: Cost of Capital; Accr uals Quality; Earni ngs Qua lity; F unda mental Risk

1. Introduction

Yee defined Fundamental Risk as “uncertainty in future

dividend payments” [1]. No studies were found pub-

lished in recent domestic literatures specialized in fun-

damental risk. Quite a few studies are carried out based

on information risk to probe into earnings quality, the

level of earning management and the behavior of mani-

pulating profit. Accrued item quality is known as Ac-

cruals Quality. The conception of accruals was presented

by Healy in 1985. With the conception of accruals in a

narrow sense, he measured controllable accruals, and

using accruals amount to express earnings quality [2]. In

a certain sense, accrual basis of accounting can be re-

garded as a potential balance of cost and efficiency ，

which between a system of submitting cash flow only

and a system that reveals adequately (Beaver，1998) .

Regarding the definition of AQ (AQ, Accruals Quali-

ty), it is the expression of Dechow and Dichv that make

AQ become more and more important in choosing ac-

counting procedure, which clear the conception of AQ as

well.

Whether Accruals Quality can be a risk pricing factor

to explain the excessive rate of return of shares, no final

conclusion has yet been reached on this matter [3]. Stu-

dies about Accruals Quality in Chinese literature mainly

lay there weight on researching from the angles of com-

pany management, motivation of contrast and the beha-

vior of earning management in company during financ-

ing, including IPO of quoted companies, the issue of ad-

ditional stocks and allotment[4] , while few people have

paid attention to asset pricing of Accrual Quality men-

tioned above.

2. Hypothesis and Sample Selection

2.1. Conclusion Hypothesis

Yee studied the relationship between earnings quality

and equity risk premium. Equity risk premium is the

component of cost of capital. Mode Yee is based on in-

formation (including noise) backgrounds in the reports

revealed by venture firms that the inventors rely on.

Earnings quality means an earning evaluation mistake

which is modifiable but unpredictable [5]. Yee re-

searched into Accruals Quality in 2006. The research

results indicates t hat inc o me qua lit y risk ha s no i nfl uence

on co st of cap ital when t here is no f undame ntal ris k, and

the inc re a si ng f u nd a ment al ri s k wil l make i nc o me q ual i ty

risk expend its influence on cost of capita l.

We analyze the research results of Yee and come up

with an experiential research direction: how does the

relationship between earning quality risk and cost of cap-

ital rely on fundamental risk, so I suggest the first con-

clusion hypothesis:

H1: earning quality risk will magnify its influence on

cost of capital, wit h increas in g fundamental risk.

2.2. Hypothesis and Sample Selection

The Effect of Fundamental Risk of Listed Companies on the Market Pricing of Accruals Quality

This article takes balance sheet approach to calculate

business accruals, which leads to the second hypothesis

in this article.

H2: This article is based on the empirical data of

non-financial busi ness in Shan ghai s tock exc hange.

Considered that Collins and Hriban’s think CFO cal-

culated in balance sheet approach will lead to more noise

and biased results in the mode, the data before 1998 is

ignored. In the meanwhile, enterprises started to carry

out new Accounting Standards for Business Enterprises

after the year of 2006, short-term investments in the

formula for calculating business accruals are replaced b y

trading financial assets. The samples in this article are

strictly constrained in the companies which have com-

plete data during ten years.

3. Determination of Earnings Quality Risk

and Fundamental Risk

3.1. Take Accruals Quality as the Criteria of

Earnings Quality Risk

Yee used the method of Francis and others evaluate

modificatory the mode of Dechow and Dichiev. In De-

chow and Dichiev mode, DD offered a new method to

estimate Accruals Quality, which is the matching degree

between business accruals and the acknowledgement of

cash flow. Accruals Quality is used for determining

working capital of operating cash flow in the past,

present and future, controlling earning variations and the

level of total assets, facto ries, equip ment and other fixed -

assets. Mode evaluated by Yee is as follows:

,0,1, ,12, ,

3,, 14,,

5,, ,

jt tj jtj jt

j jtjjt

jjt jt

TCACFO CFO

CFO v

PPE

φφ φ

φφ

φυ

−

=++

+ +∆

(1)

( )

,,jt jt

συ

(2)

(3)

where

,jt

TCA

signifies business accruals;

CFO

signi-

fies cash flow from operating activities;

signi-

fies variation of sales revenue;

,jt

PPE

signifies varia-

tion of fixed assets;

signifies Accruals Quality,

which expressed in the standard deviation of residual

error of mode .

3.2. The Crite ria o f Fundamen tal R isk

Yee conceptualized Fundamental Risk as the uncertainty

of unsolved future dividend payments. Since that enter-

prise value was considered as present value of future

expected dividend, the uncertainty of unsolved future

dividend payments accordingly turns into uncertainty of

enterprise value [5]. However, the concept of fundamen-

tal risk is closed to the definition of information risk

from Jiang Lee and Zhang. They defined information risk

as the uncertainty of enterprise value or the degree that

enterprise value can be evaluated by senior investors [4].

The empirical research is the risk substitute in the rele-

vant researches of Jiang, Lee and Zhang. Jiang, Lee and

Zhan g made use of the unc ertai nty of info rmation, led to

four risk substitutes: the age of enterprise; variation of

retur ns; trad e turnove r; average lasting time of cash flo w

in the enterprise. However, I still pre serve market capita-

lization, the age of enterprise, variation of returns and

trade turnover.I make the analysis of principal compo-

nent to combine the four remain substitutes with funda-

mental risk. The first principal component of these four

risk substitutes is combined with market capitalization.

The age of enterprise, variation of returns, trade turnover

are similar to 45 % of to tal sa mple value which is clo se to

the four components. The average of second, third and

fourth component is approximately 18%.

4. The Empirical Research and Results

In the empirical part, this article takes two methods to

test the relationship between Accruals Quality and cost

of capital of non-financial business in Shanghai stock

exchange under different risk levels: asset pricing deter-

mination; take advantage of the ratio of income to price

to determine cost of capital, research into the relationship

between Accruals Quality and cost of capital, then make

further studies on how does fundamental risk influence

the ratio of Accruals Quality to income price. I made

descriptive stat istics of samples b efore starting these t wo

determinations. This article chooses the data of non- in-

ancial quoted companies in A share market from the year

of 1999 to 2009 as samples, which has already excluded

the quoted companies without complete financial data. It

gets 6 840 samples eventual ly.

4.1. The Asset Pricing Test

1) Sharpe’s One-way Analys is of Va riance

( )

,,01,,2

jt FtmtFti

R RRRAQFactor

AQFactor FRisk

ββ β

βν

−=+−+

(4)

Fama-French Three-Factor Model

( )

,,01,,2

jt FtmtFtt

t jt

R RRRSMB

HML AQFactor

AQFactor FRisk

αα α

αα

αε

−=+−+

(5)

where

jt Ft

−

is the additional market profit in

month t;

SMB

is the difference in profit between

s mall- scale company and large-scale co mpany in month

t; t

HML is the difference in profit between

The Effect of Fundamental Risk of Listed Companies on the Market Pricing of Accruals Quality

low-market-share company and high-market-share com-

pany in month t;

AQFactor

is rebalance loop of regu-

lating dynamical towards hedge portfolio of monthly

income difference.Since FRisk ranges between 0 to 1,

AQFactor refers to the relationship between accrual

quality and cost of capital in the company with the low-

est FRisk, while AQFactor *FRisk refers to the difference

between accrual quality and cost of capital in the com-

pany with the lowest FRisk and the company with the

highest FRisk. Suppose that the relationship between

accrual quality and cost of capital is based on fundamen-

tal risk, Equation represent the positive correlation coef-

ficients β and α separately as stated in assumption

1.Model Yee indicates that non-systematic risk of earn-

ings quality has no influence on cost of capital as

non-systematic fundamental risk does. Since AQFactor

refers to difference in surplus between high accruals

quality investment combination and low accruals quality

investment combination, and non-systematic risk of ac-

crual quality should have no extra surplus, AQFactor

stands for systematic risk of accrual quality. In fact

FRisk can be hardly divided into systematic and nonsys-

tematic parts and is regarded as non-systematic risk, so

coefficient AQFactor*FRisk tends to 0.

2) Empirical findi ng s

Panel A in Table 1 provide all the coefficients and

specific regression and inspection average t. However,

Gow, Ormazabal and Taylor evaluated the process of

capital pricing and finded that it exaggerates excessively

t test values, the basis of cross section Adjustment of

cluster related to annual regression coefficients. The

consequence of Adjust of the cluster is similar to it. As a

result, most of the relevant inspection values t decrease

and the values t interaction with it increase. Therefore,

data in Panel A is not discussable and we focus on the

resul t in Panel B.

Column”1”, ”2” and ”3” in Table 1 report the testing

results of single factor model, where Column”1” and ”2”

provide the testing results of Fra ncis and others which we

use to compare. In column “2”, AQFactor has a strong

positive correlation with

jt Ft

−, while in column

“3” the correlation becomes negative (t = -8.45). Aqfac-

tor*FRisk is also strongly positively correlated with

jt Ft

− (t is 39.16). Moreover, with FRisk ranging

from 0 to 1, AQFactor(-0.233) measures the relationship

between AQFactor and the income of a company with

the the lowest fundamental risk, while AQFactor and n

Table 1. C apital pricing test of accru al s quality, future income of stocks, and f undamental risk, 1999 -2009.

Panel A. Fama-Macbeth Regressions by Firm

Predic-ted

Sign

CAPM Fama-French Three-Factor Model

1 2 3 4 5 6

Rm-Rf + 1.066 0.838 0.842 1.012 0.946 0.946

SMB + 0.888 0.509 0.512

HML + 0.218 0.357 0.356

AQfactor + 0.812 -0.199 0.612 -0.439

Aqfactor*FRisk + 1.51 1.52

Adj.R*R 0.114 0.149 0.154 0.158 0.175 0.179

Panel B. Fama-Macbeth Regressions by Year

Predic-ted

Sign

CAPM Fama-French Three-Factor Model

1 2 3 4 5 6

Rm-Rf + 1.027 0.863 0.863 1.189 0.918 0.894

SMB + 0.867 0.55 0.497

HML + 0.436 0.204 0.202

AQfactor + 0.812 -0.223 0.393 -0.473

Aqfactor*FRisk + 1.812 1.814

Adj.R*R 0.082 0.095 0.109 0.098 0.101 0.114

Notes : CAPM is Capital Asse t Pricing Model. The samples include the stock inc omes at least 18 ti mes per month and data from 596 7 compani es from

1999 to 2009. The definitions of var iab les as follow: R M-RF is th e extra return on in vestment of mar ket in vestm ent c ombinati on; SM B is the investmen t

combination income that is hedging with big or small factors of Fama-French. AQfactor is the income of the accural quality investment combination;

FRisk is the first chief factor of the four agencies of which information is uncertain. Panel A reports the averange evalution coefficients and conse-

The Effect of Fundamental Risk of Listed Companies on the Market Pricing of Accruals Quality

quences of regression-anlysin g only owned by 5967 c ompanies , while Panel B provides the average eva lution coefficients an d the conse quence regres-

sion-analysed per 25 year.

Aqfactor * FRisk measures the relationship betwee AQ-

Factor and the income of a company with the highest

fundamental risk.The testing results of Fama-French

Three-Factor Model are displayed in Column”4”, ”5”

and ”6”, where Column”4” and ”5” reports the testing

results of Francis and others, and the Column “6” reflects

the interacted model of AQFactor and RiskScore. Being

(-9.63) in Column”5”, AQFactor changes to (-9.35). Aq-

factor * FRisk is positively correlated with

,,jt Ft

RR−

and e xtremel y notab le(3 9.18 ). In this re gressio n anal ysis,

we can find that when AQFactor decreases, both SMB

and HML decreases from 0.867 to 0.550 and from 0.436

to 0.204. This means FRisk has little influence on SMB

and HML.

If AQ has a strong influence on profit when FRisk is

very high instead of low, we can predict boldly that Aq-

factor*FRisk and AQFactor are both positively corre-

lated with

jt Ft

−. To study whether the hypothesis

talked about above, is connected some unusual changes

or AQFactor and nonspecific econometrics affecting

factors of index variables, we evaluate the regression of

every value of FRisk in the model except the Aqfac-

tor*FRisk. Figure 1. sho ws th e re gr ession of 10 points of

AQFactor. In this figure, AQFactor increase only when

FRisk increases, and the lowest value is below 0. More-

over, Figure 1 suggests that the relationship between

AQFactor and FRisk are nonlinear.

4.2. Regression test: Estimation of Cost of

Capital by Using Earning-Price Ratio

The study of Core, Guay and Verdi suggests that the re-

sult of asset pricing test by using factor model is able to

prove the difference in the same period among the earn-

ings, and those earnings are related with the factors in the

models[6].However, this relation between earings and

factors cannot guarantee a premium return. And in order

to be scientifically rigorous, another test will be used to

evaluate the relationship between AQ and cost of capital.

In 1992, Alford has proved that industrial matching is

benefit for controlling the difference between risk and

growth, this means IndEP is able to control other deci-

sive factors of earning-price ratio. In order to study how

fundamental risk influences the relationship between

earning-price ratio and accruals quality, we make regres-

sion o f IndE P of gro wth, AQ, FRi sk and AQ*F Risk. Re-

sults are displayed in Table 2.

There are three variables including Growth, AQ and

FRisk in Column“1”, “2” and “3” of Table 2. factor

growth is significant（-1.89）, which indicates that IndE P

relaying on dependent variable is effective in controlling

differences of growth. As is shown in the results of re-

search by Francis, factor AQ is positively correlated with

he cost of capital, and significant differences（-3.66）can

Figure 1 . FRisk Decile and AQfacto r.

Table 2. regression estimation of relationship between ac-

cruals quali ty an d cost of capit al（t es ti ng wi th I ndEP ）and

betwee n fundame ntal risk an d cost of capital .

Predicted

Sign 1 2 3 4

Growth - -0.0044 -0.0041 -0.0029 -0.0025

AQ + 0.008 0.0015 0.011 0.0036

FRis k + 0.0197 0.0126

Risk1 + 0.0034 0.0041

Risk2 + 0.0149 0.0074

AQ*FRis k + 0.0157

AQ*FRis k1 + -0.0018

AQ*FRis k2 + 0.0166

R*R 0.017 0.019 0.015 0.016

Notes: Risk1 and R isk2 a re the fir st two imp orta nt ind ex of ten ris k index es.

This table provides average annual evaluating coefficient of independent

variab les, which or igina ted from IndEP . T test values is st and ard err or based

on annual coefficient evaluation.

t be found. This also corresponds with the negative cor-

relation between AQ and cost of capital. Before the re-

gression estimation, the value of AQ is decimal fraction

within the range from 0 to 1. FRisk is also positively

correlated with the cost, and there are also significant

differences（-3.79）. Consequently, the value of FRisk,

0.0197, reflects the differences between the cost of capi-

tal of 1.97 per cent of enterp rises with the hi ghest AQ and

those of 1.97 percent of enterprises with the lowest AQ.

Although both AQ and FRisk（t equals 3.66 and 3.79

separately）have significant differences, the value of

FRisk is more than twice the value of AQ （1.79% VS

0.80%）, which indic ates that FRisk is an important i ndex

when determining cost of capital.Row “2” of TABLE Ⅱ

shows the testing results when AQ*FRisk is added to

regression estimation, where AQ and FRisk are allowed

The Effect of Fundamental Risk of Listed Companies on the Market Pricing of Accruals Quality

to be different. Under this circumstance, both AQ and

FRisk decrease causing by this change. However, FRisk

still has significant differe nces（t-3.01）, while AQ does

not（t-0.51）.AQ*FRisk p ositively corr e la te d with cost of

capital and the value is about 10 percent （-1.75）. T he

summation of AQ and AQ*FRisk signifies the value of

the differences of cost of capital. This demonstrates that

risk o f ea rni ngs q ual ity will have lar ger i nflue nce o n cost

of capital with the increase of FRisk.These factors in-

cludi n g t he a ge of e nte rp r ise , va ri ati o ns in r et ur n, vo lu me

of trade consisting of FRisk may related to AQ. In fact,

the relationship between FRisk and AQ is significant.

The thesis that FRisk has no signi ficant releva nce to AQ

will be confir med in the next sec tion.

4.3. The Methods of Fundamental Risk Test

1) FRisk test: FRisk test corresponds with the Yee's

fundamental risk test in concept. Moreover, return wave

and the volume of trade is market dependent variable

affected by accruals quality. Accordingly, we use the

other 10 dependent variables based on Francis, LaFond,

Olsso n and Sc hippe r, incl udin g two o pera ting ris k inde x:

cash flow and standard deviation of sales volume. Hribar

and Nichols find that these two operating index are con-

nected with the accruals quality obtained from accruals

quality model r e sidual error .

2) Fundamental risk disintegrated by component analy-

sis

We analyse two risks by using component analysis to

reduce risk variable(as showed in Table 2). The first

principal component of 10 proxy variables of fundamen-

tal risk are Size ，Sd(CFO) and Opcycle, and the second

are SD(sales volume)，leverage ratio and Negearm[7].

Simil ar to fund a me nt al ri s k, t he t wo compo ne nts a re b ot h

decimal fractions ranging from 0 to 1. Consequently,

fundamental risks ar e Risk1 and Risk2 in Table 2.

To study the sensibility of the testing results to funda-

menta l risk, we make the regr ession o f IndEP of gro wth,

AQ, Risk1, Risk2, AQ* FRisk1 and AQ* FRisk2. The

Column”3” in Table 2 displays independent variable:

growth, AQ, Risk1 and Risk2. As mentioned before, AQ

conforms the principles that accruals quality is negative

correlated with cost of capital. Cost of capital is substi-

tute of Risk1 and Risk2 in fundamental risk. What is

funny is that Risk1 has no significant differences(-0.91),

while Risk2 is si gnificantly positively correlated（0.0149,

-6.82）. This means Risk1 refers to nonsystematic risk

which is not market-priced, and Risk2 refers to the sys-

tematic fundamental risk which is market-priced.

In Column”4” of Table 2, regression analysis of AQ*

FRis k1 a nd AQ* F Ris k2 i s ca rri ed o ut, a llo wing t hat AQ

is different from Risk1 and Risk2. This change leads to

decrease of AQ and Risk2 and increase of Risk1. Al-

though Risk1 still has significant differences(-2.93), AQ

has no significant differences(-0.52). Interacted coeffi-

cient AQ*FRisk2 is positively correlated with cost of

capital and has significant differences(-4.66). Moreover,

in this regression analysis, Risk1 and AQ*FRisk1 pos-

sess no significa nt differences.

5. Other Regression Estimation: Implicit

Cost of Capital

The result of the study by Easton and Monahan, Guay,

Kothari and S hu, suggested that the co st of capital calc u-

lated by earning-price ratio has weak relationship with

the Yield T o Maturity(YT M). In this section, we method

to estimate cost of capital will be adopted to study the

sensibility of the testing results in Section Ⅲ. In this

method, several indexes of implicit cost of capital will

rise and thus diminish the measurement errors. Referring

to the research method of Dhaliwal, Heitzman and Li,

cost of capital will be calculated using discount rate im-

plied within different applying methods of Residual Re-

turns Value Model.

5.1. Four Value Models

1) Gebhardt, Lee and Swaminathan Model

Gebhardt, Lee and Swaminathan Model is used to es-

timate Enterp rise Value:

( )

11 11

12 11

t gls

t tti

tgls

tgls t

gls gls

FROE r

PB B

FROEr B

+−

−

= ++

−

∑

(6)

where

gls

: cost of capital; B: amo unt at t he be ginn ing o f

year;

FROE

: earnings forecasts

2) Claus and Thomas Model

( )

511

( )1

tict ti

t tti

tct

s ct

ct ct

FEPSr B

PB B

FEPSr Bg

rg r

++− +−

−

= ++

−+

+−+

∑

(7)

where ct

r: cost of capital;

FEPS

: earnings forec asts o f

per share in the first two years, or forecasts of long-term

growth ra tes in the foll owing t hree years .

3) Gode and Mohanranm Model

Gode and Mohanranm Model which based on Earn-

ings Growth Model by Ohlson and Juettner-Nauroth

Model is as follows:

( )

120.03

gm f

FEPS

rA Agr

=++− (8)

The Effect of Fundamental Risk of Listed Companies on the Market Pricing of Accruals Quality

( )

0.50.03 0.5

DPS

A randg

FEPS FEPS

FEPS





= −+







−

(9)

Easton Model DPSt is the dividend per share during

period t. Assu me that DPSt equals to DP S0, and the de fi-

nitions of the other variables are the same as that men-

tioned above. Thus, Easton Model that separated from

Earnings Growth Model is as follows: where the defini-

tions of variables are the same as that mentioned above.

2 11

tpeg tt

peg

FEPSr DPSFEPS

+ ++

+−

(10)

5.2. Analy sis of Resul ts

In the calculation, each of fou r averages of cost of capital

is evaluate, thus we gain a single cost of capital estimate:

AvgCOC[13]. Compared with the results of regression

estimation of IndEP, AvgCOC requires more precise data,

so the sample size should be diminished from 6840 to

5967.

The estimating results based o n AvgCOC ar e as shown

in Table 3, where the data in Row “4” are the same as

those in Row “4” of Table 2.T here are three independent

variables in Row “1” of Table 3: Growth, Accruals

Quality (AQ) and Fundamental Risk (Frisk). According

to Table 3, AQ is posit ively co rrelated with the cost, a nd

significant dif ferences (-6.93) can be found between four

models. This also corresponds with the negative correla-

tion between AQ and cost. The value, 0.0115, reflects the

difference between the costs of capital of 1.15 percent of

enter p ri ses wit h the highe st A Q a nd tho se of 1 . 15 p e rc ent

of enterprises with the lowest AQ. This value is farther

higher than 0.008 in Table 2. Moreover, Frisk is posi-

tively correlated with the cost and there is also significant

differences (-13.40) between four models. The value of

Frisk is 0.0392, which is more higher than 0.0197 in Ta-

ble 2. In Row “2” of Table 3, a related variable

（AQ* Fris k）is added, , where AQ and Frisk are allowed

to be different. As a result shown in Row “2”, decrease

can be found in both AQ and Frisk causing by this

Table 3. Regression Analysis of the relationship between

cost of stock (estimated by implicit cost of average capital)

and Accruals Quality and Fundamental risk.

Predicted

Sign 1 2 3 4

Growth - -0.0284 -0.0282 -0.0231 -0.0231

AQ + 0.0115 0.0071 0.0085 0.0038

FRis k + 0.0392 0.0332

Risk1 + 0.0231 0.022

Risk2 + 0.0264 0.0232

AQ*FRis k + 0.0133

AQ*FRis k1 + 0.0026

AQ*FRis k2 + 0.0081

R*R 0.167 0.171 0.147 0.151

change.In Row “3” and “4”, Frisk is used to replace the

other index of fundamental risk: Risk1 and Risk2. As a

result in Row “3”, Risk1 and Risk2 are systematic fun-

damental risks that are market-priced, opposed to the

resul ts in Table 2.

6. Conclusion

In this article, two sets of tests have been taken to study

the effect of Fundamental Risk on the Market Pricing of

Accruals Quality: Firstly, using asset-pricing determina-

tion to estimate the relationship between prese nt wo rt h o f

income and Accruals Quality. Secondly, using price-

income ratio to estimate cost of capital, meanwhile stud-

ying the fundamental risk’s influence on the relationship

between Accruals Quality and price-income ratio. Tests

suggests that, there is no internal connecting link be-

tween Accruals Quality and cost of capital which is cal-

culated by the present worth of income of low- funda-

mental-risk enterprises. Ho wev er, the interaction be-

tween Accruals Quality and fundamental risk connects

closely with cost of capital. And when this interaction

exists, the main influence from Accruals Quality will

disappear. In the final result, we can conclude that, as

fundamental risk rises, Accruals Quality’s influence on

cost of capital is enhanced, but this influence on cost of

capital of any enterprises will never exceed that of

low-fundamental-risk enterprises. In fact, our results do

not correspond with that of Yee about whether the ri s k o f

Earnings Quality is systematic or not. Yee found that

cost of capital i s related with s ystematic risk of E arnings

Quality but not with nonsystematic risk of Earnings

Quality.But in this study, it is the total risk of Earnings

Quality that we focus on, and we do not divide the total

risk i nto t wo sep arate d pa rts, so the fur ther stud y may b e

a huge challenge. Finally, however, we can conclude in

this study that fundamental risk actually affects the rela-

tionship between Earnings Quality and cost of capital.

And this suggests that total risk of Earnings Quality at

least has systematic parts within it. Because if all risk of

capital were not systematic, no relationship will be dis-

covered between Earnings Quality and cost of capital,

nor will any changes take place as fundamental risk

change.

REFERENCES

[1] Yee, K. K. Earnings Quality and the Equity Risk Premium:

The Effect of Fundamental Risk of Listed Companies on the Market Pricing of Accruals Quality

A Benchmark Model.Contemporary, Accounting Re-

sea r c h, 2006, 23 (Fa ll ): 83 3–877.

[2] Healy P M. The Effect of Bonus Schemes on Accounting

Decisions [J]. Journal of Accounting and Economics,

1985, 7, ( 1-3):85-107.

[3] Francis J, La Fond R, Olsson P, Schipper K. The Market

Pricing of Accruals Quality [J]. Journal of Accounting

and Economics, 2005, 39(2):295-327.

[4] Nichols C. Fundamental or Information Risk. An Analysis

of the Earnings Quality Factor [R]. Ithaca: Cornell Uni-

versity, 2006.

[5] Hribar, P., and D. C. Nichols. The Use of Unsigned Earn-

ings Quality Measures in Tests of Earnin gs Management.

Journal of Accounting Research, 2007, 45 (December):

1017–1053.

[6] Core J E, Guay W R, Verdi R S. Is Accruals Quality a

Priced Risk Factor? [J]. Journal of Accounting and Eco-

nomics, 2008, 46, (1):2-22.