Nagano2016 2

PACFIN-00903; No of Pages 16
Pacic-Basin Finance Journal xxx (2016) xxxxxx
Contents lists available at ScienceDirect
Pacic-Basin Finance Journal

journal homepage: www.elsevier.com/locate/pacfin
Sukuk issuance and information asymmetry: Why do rms issue sukuk?

Mamoru Nagano 1
Faculty of Economics, Seikei University, 3-3-1 Kichijoji Kitamachi, Musashino City, Tokyo 180-8633, Japan
a r t i c l e i n f o a b s t r a c t
Article history: This study investigates the factors that promote a bank borrower to issue sukuk rather than
Received 11 July 2016 conventional debt security in Malaysia and Indonesia from 2000 to 2014. First, our empirical
Received in revised form 15 November 2016 results show that a bank borrower is likely to approach the sukuk market as the funding size
Accepted 9 December 2016
grows and if the rm is valued highly. Second, we nd that under high information asymmetry,
Available online xxxx
a rm with a high stock price and large funding demand prefers sukuk issuance to conventional
debt. We conclude that rms use the sukuk market as an intermediate funding market when
JEL classication: the funding demand is too large to borrow from banks and the information asymmetry is
G15
too high for them to approach the conventional debt market.
G30
2016 Published by Elsevier B.V.
O16
Keywords:
Sukuk
Debt issuance
Hold-up theory
Pecking order effect
1. Introduction
The number of debt security issuers is still small in the emerging economies. This directly suggests that most rms in countries
undergoing economic development cannot choose desirable funding tools. Meanwhile, the number of Islamic bond (hereafter,
sukuk) issuers dramatically expanded from 2001 onwardsthe primary market total value in 2014 was USD 126 billion, 946
times that in 2000, although the total value temporarily fell to USD 60.7 billion in 2015. Reecting this trend, recent studies dis-
cuss what causes a rm to approach the sukuk market (Mohamed et al., 2015; Azmat et al., 2014; Halim et al., 2016; Klein and
Weill, 2016). The purpose of this study is to explore what encourages rms to resort to the sukuk issuance market in Malaysia
and Indonesia. The major difference between our study and the existing literature is that our study hypotheses sukuk issuance
as an intermediate funding methodology between bank borrowing and conventional debt issuance in terms of the relative funding
size and the degree of information asymmetry of the rm. We contribute to the literature by showing unique evidence pertaining
to how issuance determinants for sukuk and conventional debt are similar in some ways and different in others.
Existing studies (Udell and Berger, 1995; Boot, 2000; Elyasiani and Goldberg, 2004) consistently suggest that the lending re-
lationships between banks and borrowers have positive effects on their operating performance. The hold-up and liquidity con-
straint hypotheses (Hoshi and Kashyap, 1990; Hoshi et al., 1991; Sharpe, 1990; Rajan, 1992) assert that a debt issuance rm is
valued highly in the market, because it requires neither the partner bank's intervention nor nancial liquidity rescue. Meanwhile,
the market timing theory (Asquith and Mullins, 1986; Jung et al., 1996; Graham and Harvey, 2001; Hovakimian et al., 2001)
This research is nancially supported by JSPS KAKENHI Grant Number JP15K03550.

E-mail addresses: mnagano@econ.seikei.ac.jp, mnagano2002@nifty.com.
1
The author thanks various seminar participants for their helpful comments and suggestions.
http://dx.doi.org/10.1016/j.pacn.2016.12.005
0927-538X/ 2016 Published by Elsevier B.V.
Please cite this article as: Nagano, M., Sukuk issuance and information asymmetry: Why do rms issue sukuk?, Pacic-Basin Fi-
nance Journal (2016), http://dx.doi.org/10.1016/j.pacn.2016.12.005
2 M. Nagano / Pacic-Basin Finance Journal xxx (2016) xxxxxx
asserts that a rm undervalued in the market prefers debt to equity issuance. This study rst veries that the bankborrower re-
lationship and rm market valuation are determinants of rm sukuk issuance. Specically, we show that under high market val-
uation, a bank's borrower in Malaysia and Indonesia approaches the sukuk market as the funding size grows compared to the
partner bank market capitalization size, consistent with the hold-up and liquidity constraint hypotheses.
Second, this study examines how rm information asymmetry is the key to differentiate sukuk issuance from conventional
debt issuance. We suppose that the pecking order effect (Donaldson, 1961; Myers, 1984; Shyam-Sunder and Myers, 1999) has
an explanatory power for the rm's sukuk issuance decision due to the existence of the unique structure of funding scheme,
which mitigates information asymmetry between the issuer and investors. We hypothesize that a low degree of nancial con-
straint of the rm commonly promotes sukuk and conventional debt issuance when the partner bank's commercial loan service
cannot meet the too large funding demand. Alternatively, we specically suppose that a rm chooses sukuk under a high degree
of information asymmetry, and conventional debt under a low degree of information asymmetry, between rm insiders and out-
siders. This is because the xed tangible assets of the sukuk issuer's project, which is owned by a third party, cover the informa-
tion costs.
The remainder of the paper is organized as follows. In the following section, we review the extant literature in this domain and
introduce our hypotheses in Section 3. In Section 4, we present our empirical strategy and the data used to examine our hypoth-
eses. Subsequently, in Sections 4 and 5, we discuss the results produced by our empirical models. Finally, in Section 6, we offer
concluding remarks.
2. Literature review
Since the second half of the 2000s, related research has mainly focused on Islamic banking and has analyzed the similarities
and differences between Islamic banking and conventional banking. For instance, Chong and Liu (2009) empirically analyze the
Malaysian Islamic banking loan contract called Musyarakah and nd that the investment rate of return for prot-sharing-based
Islamic banking is statistically lower than conventional bank deposit rates. Meanwhile, TurkAriss (2010) focuses on the degree
of competitiveness of Islamic banks. She studies Islamic banks and conventional banks in 13 countries for the period 2000 to
2006 and concludes that most Islamic bank borrowers are generally from the nancial sector and that Islamic banks are less com-
petitive than conventional commercial banks. Further, Ongena and endeniz-Ync (2011) analyze borrowers of Islamic banks.
Their study compares Turkish Islamic bank borrowers and conventional bank borrowers and concludes that the former are
young, emerging, and industrially less diversied rms, and that they transact with many banks. Abedifar et al. (2013) compre-
hensively survey the recent empirical literature on Islamic banking and conclude that Islamic banking has contributed to nancial
development in emerging economies.
In recent times, the number of studies comparing sukuk issuances with other nancing techniques has increased. For instance,
Godlewski et al. (2013, 2016) and Mohamed et al. (2015) empirically study the sukuk market using issuer data. Godlewski et al.
(2013) compare the differences between cumulative abnormal returns of sukuk issuers and conventional bond issuers in Malaysia,
and they conclude that the stock market negatively reacts to sukuk issuance announcements. Mohamed et al. (2015) also employ
sukuk and conventional bond issuer data and conclude that rms choose sukuk issuance to optimize the cost and benet balance
for the issuers. Godlewski et al. (2016) investigate the relationship between the type of sukuk and the choice of shari'a board
members and conclude that the nationality, term length, and reputation of shari'a board members mitigate the information asym-
metry of the sukuk issuer. Maghyereh and Awartani (2016) and Naifar et al. (2016) study the relationship between sukuk yields
and stock market volatility and its transmission mechanism. They conclude that the relationship is asymmetric and the stock mar-
ket volatility inuences sukuk yield, but the sukuk market does not inuence the condition of the stock market. Halim et al. (2016)
and Klein and Weill (2016) also study the relationship between the degree of information asymmetry of the issuer and sukuk
issuance. They show evidence that the higher the degree of information asymmetry, the more frequently the rm issues sukuk.
In the case of conventional debt issuance studies, on the other hand, a body of research addresses the relationship between a
rm's debt issuance and its lending relationship with the partner bank. The hold-up hypothesis (Sharpe, 1990; Rajan, 1992) as-
serts that a bank often intervenes in the low protability projects of its partner borrowers. Therefore, a rm that chooses debt
issuance is highly protable if it does not require any future project interventions by the partner bank. Similarly, Hoshi and
Kashyap (1990), Hoshi et al. (1991) and Bolton and Freixas (2000) conclude that a borrower of high creditworthiness prefers
debt issuance to bank borrowing. This is because although the choice may force the issuer to liquidate the rm or undergo recon-
struction proceedings in the future when facing a liquidity crunch, debt issuance enables large funding size, lower interest rate,
and no collaterals. Bolton and Freixas (2008) assert that bank borrowers in emerging economies often witness a quick growth
in funding demand and accordingly prefer debt issuance. This is because the funding choice saves the bank from the capital re-
quirement restrictions.
Many studies also focus on the choice between conventional debt and equity issuance. As Ming et al. (2012) denote, the
pecking order theory and timing theory are two major streams of rm security issuance theory. The pecking order theory asserts
that rm managers are better informed than external investors (Donaldson, 1961; Myers, 1984; Shyam-Sunder and Myers, 1999).
This theory stipulates that outside investors require better information when the rm is funded by equity issuance. Myers and
Majluf (1984) and a number of other researchers support the idea. However, certain scholars do not support the pecking order
theory. Fama and French (2005) conclude that more than 50% of U.S. rms violate the pecking order theory. Lemmon and
Zender (2010) also nd that debt is apparently preferable to equity nancing in the absence of debt capacity concerns.
M. Nagano / Pacic-Basin Finance Journal xxx (2016) xxxxxx 3
On the other hand, the market timing theory asserts that a rm employs equity (debt) issuance when the stock price is
overvalued (undervalued) (Asquith and Mullins, 1986; Jung et al., 1996; Graham and Harvey, 2001; Hovakimian et al., 2001).
Baker and Wurgler (2002) also contribute the idea that a rm issues equities when the stock price is high if the high stock
price coincides with low adverse selection. Gomes and Phillips (2012) report evidence on the market timing hypothesis, too.
The international evidence shown by Henderson et al. (2006) also supports this theory. However, some studies do not support
the market timing theory either. Bruinshoofd and de Haan (2007) conrm little evidence of the market timing hypothesis for
U.S. and EU rms. Jung et al. (1996) report inconsistent empirical results for the timing theory.
3. Hypotheses
The purpose of the present study is to contribute to the literature described above by empirically examining the theoretical
framework for sukuk issuance in Malaysia and Indonesia. The existing literature focuses on the relationship between either
sukuk and conventional debt issuance, or sukuk and equity issuance. On the other hand, we focus on the choice among sukuk,
bank borrowing, and conventional debt. Our Hypotheses 13 present propositions of the funding choice determinants between
sukuk and bank borrowing and between conventional debt and bank borrowing, to compare the similarities and differences be-
tween sukuk and conventional debt issuance.
To overcome unexpected liquidity constraints, the bankborrower relationship is crucial in the initial stages of economic de-
velopment in emerging economies. However, as the growth rate of funding size increases, bank capital enhancement does not al-
ways catch up with this funding growth, as Bolton and Freixas (2008) and Hale and Santos (2008) assert. Besides, the hold-up
hypothesis (Sharpe, 1990; Rajan, 1992) and liquidity constraint hypothesis (Hoshi and Kashyap, 1990; Hoshi et al., 1991) assert
that a rm issuing debt is consequently a high-growth rm and under high market valuation, as it requires neither future project
interventions nor liquidity support by the partner bank. We follow these studies and hypothesize that both a bank and its partner
borrower prefer sukuk issuance when the funding demand grows fast under high rm market valuation. This conversely implies
that undervalued stock price does not promote sukuk issuance, which disagrees with the market timing theory.
Hypothesis 1. A bank borrower with a growing funding size issues sukuk when the rm is valued highly since the possibility of
the rm facing liquidity constraints post issuance is low.
Second, we verify how the pecking order effect determines the choice between sukuk and conventional debt issuance. Our sec-
ond hypothesis supposes that a rm under high information asymmetry is unlikely to approach the conventional debt market and
instead chooses the sukuk market. Following the existing literature (Donaldson, 1961; Myers, 1984; Shyam-Sunder and Myers,
1999), a certain level of low information asymmetry of the rm is necessary for it to resort to the conventional debt market.
This is because the rm's creditworthiness information is demanded by outside investors since their investment decision requires
such information when the rm issues debt. Hypothesis 1 supposes that a rm valued highly in the market hardly faces liquidity
constraints post issuance. However, information asymmetry is generally remarkable in an underdeveloped debt market. Mean-
while, Murabahah sukuk is essentially cost-plus-sales-based debt security, as shown in Fig. 1. This type of sukuk is basically
used for small-sized and short-term maturity funding to cover the information asymmetry of the rm. In the case of sukuk al-
Ijarah, as shown in Fig. 2, the lease contract between the sukuk issuers and the SPC covers the information costs. As shown in
Fig. 3, the structure of Musyarakah sukuk is similar to that of normal equity, as the investment return of the security varies de-
pending upon the post-issuance project performance. Therefore, issuers frequently choose Murabahah sukuk and Ijarah sukuk,
but the frequency of selecting Musyarakah sukuk is extremely small. Accordingly, an information asymmetric rm privately issues
Murabahah sukuk or Ijarah sukuk when it is inaccessible to the conventional debt market because of the high degree of informa-
tion asymmetry.
Project Owner issues sukuk

at cost plus sales price Machineries &
Payment Equipment
Investors Issuer Project
Installment Sales
Machineries &
Payment Contact of machineries
Equipment
and equipment
Payment
Machineries &
Lead Manager
Equipment
Lead Manager
purchases machineries
and equipment
Fig. 1. The structure of Murabahah sukuk.
Issuer
(Lessee) Project
The SPC leases the asset to The issuer sells

Payment
the Lessee and cashes the machineries and
rent equipment to the SPC
SPC
(Lessor)
The net amount of the SPCs cash inflow

SPC issues sukuk Payment
from the rent is passed on to the
investors
Investors
Fig. 2. The structure of Ijarah sukuk.
Hypothesis 2. An information asymmetric rm is likely to choose sukuk issuance when the rm demands large funding size and
is not under nancial constraint.
Denis and Mihov (2003) and Gomes and Phillips (2012) conclude that in the case of conventional debt issuance, a rm
chooses private debt when the degree of information asymmetry is high, while it prefers public debt under a low degree of in-
formation asymmetry. In the case of sukuk issuance in Malaysia and Indonesia, most sukuk are privately issued. Udell and
Berger (1995), Boot (2000), and Elyasiani and Goldberg (2004) assert that the lending relationships between banks and bor-
rowers contribute to improving the operating performance of the borrowers. However, as Bolton and Freixas (2008) and Hale
and Santos (2008) assert, the growth rate of borrower funding size is often high, and the partner bank faces capital requirement
restrictions. Our third hypothesis, therefore, is that the lending relationship promotes sukuk issuance in the private market when
the partner bank and the underwriter are one and the same or in the same business group.
Hypothesis 3. Under information asymmetry, the rm is likely to issue sukuk in the private market when the partner bank
operates both the banking and the securities businesses.
4. Empirical strategy
4.1. Partner bank capital, borrower funding size, and sukuk issuance
The rst analysis of this study veries the relationship between partner bank capital, borrower funding size, and sukuk issu-
ance of the rm. Using bank borrower and security issuer samples from Malaysia and Indonesia (Tables 1 and 2), we rst estimate
models (A) and (B) by a multinomial probit model to verify the similarities and differences in sukuk and conventional debt issu-
ance determinants of the rm. We use a multinomial probit model because the security issuance under estimation fails to satisfy
the so-called independence from irrelevant alternatives (IIA) property of the multinomial logit model. The dependent variables of
model (A) are Sukuk, Debt, and Bank Loan. The dependent variables of model (B) are Sukuk, Debt, and Equity. We employ the fol-
lowing three independent variable groups: (XPartnerBank) represents variables of rm i's partner bank nancial characteristics,
Project Owner issues sukuk

Machineries &
as variable income security
Equipment
Investors Payment Issuer Payment Project
Project revenue varies
Machineries upon the performance
Payment
& Equipment
Machineries &
Equipment
Fig. 3. The structure of Musyarakah sukuk.
Table 1
Annual distribution of bank borrowing, sukuk, conventional debt, and equity issues.
The bank loan samples are sourced from Bureau Van Dijk's OSIRIS, and the sukuk and other security issuance samples are sourced from Thomson Reuters' Thomson One
for the period 2000 to 2014. The numbers in the cells represent the increase in the number of bank loans and the number of sukuk, conventional debt, and equity issues
in each given sample period. The denitions of sukuk and conventional debt issue do not include issuances with a period of less than a year's maturity, convertible de-
bentures, and warrant debentures. The denition of equity issuance does not include preferred stocks and subordinated debentures. The breakdown of the total number
of sukuk types (55) is as follows: 20 issuers of sukuk al-Mudarabah, 12 issuers of sukuk al-Ijarah, 18 issuers of sukuk al-Musyarakah, and 5 issuers of others.
Malaysia Indonesia
Bank loan Sukuk Debt Equity Bank loan Sukuk Debt Equity
2000 12 0 5 2 6 0 4 0
2001 28 0 4 0 12 0 1 0
2002 30 0 0 2 16 0 0 0
2003 35 1 1 3 20 0 2 0
2004 42 3 3 4 14 3 3 1
2005 52 1 5 0 18 0 1 0
2006 56 1 2 1 28 0 1 0
2007 59 4 1 2 22 0 5 0
2008 54 8 3 3 22 2 1 2
2009 50 2 3 4 33 1 2 1
2010 63 3 3 5 32 1 3 4
2011 59 4 3 4 38 0 2 4
2012 66 4 2 6 33 2 4 4
2013 56 7 1 5 32 0 3 3
2014 49 8 3 8 37 0 3 2
Total 711 46 39 49 363 9 35 21
Source: Bureau Van Dijk's OSIRIS and Thomson Reuters' Thomson One.
(XPeckingOrder) represents variables of rm i's degree of nancial constraint and information asymmetry between rm insiders and
outsiders, and (XFirm) represents variables of the nancial performance of rm i:
PartnerBank PeckingOrder Firm

Sukuk or Debt or BankLoan X 1 X 2 X 3 A

Sukuk or Debt or Equity X 1 X 2 X 3 B
More specically, (XPartnerBank) employs Bank Cap to Funding, which is the total market capitalization of the partner bank of rm
i in year t divided by the total funding value of rm i in year t. (XPartnerBank) also employs Public or Private, which equals one when
sukuk, conventional debt, or equity is issued in the public market. The variable takes zero when the security is issued in the pri-
vate market. (XPartnerBank) also employs Lead Manager, which equals one when rm i's partner bank and the lead manager of
sukuk, conventional debt, or equity issuance are consistent in year t.
(XPeckingOrder) employs the following independent variables. We employ a measure of nancial constraints of rm i in Eq. (C), as
in Kaplan and Zingales (1997). We follow Baker et al. (2003) and Ming et al. (2012) and exclude Tobin's Q ratio from the index, as
a high Q ratio may indicate market valuation and contaminate the index as a measure of nancial constraints.
CF it DIV it CASH it
KZ Indexit 1:002 3:319 Leverageit 39:368 1:315 C
Asset it1 Asset it1 Asset it1
Here, CF denotes the sum of net income and depreciation, Asset represents total assets, DIV is cash dividends, and CASH
represents cash and short-term investment. The denition of Asymmetric, which is another variable of XPeckingOrder, is explained
in Section 4.2.
(XFirm) employs Market to Book, which is the book value of liability in year t - 1 plus market value of capital in year t, divided
by the book value of total assets in year t - 1 for rm i. This study also employs cumulative abnormal returns CAR(60,2) and
CAR(1,+1) as proxies of rm i's market valuation other than Market to Book. These variables proxy the degree of market val-
uation of rm i immediately before security issuance. CAR(60,2) and CAR(1,+1) are the independent variables of rm i's
cumulative abnormal returns around the sukuk, conventional debt, equity issue dates. We employ two types of event window re-
sults for cumulative abnormal returns: a 3-day event window CAR(1,+1) and a 59-day event window CAR(60,2), following
the methodology of Brown and Warner (1985). To calculate rm i's cumulative abnormal returns, we estimate the predicted stock
returns for the interval (250,60). (XFirm) also employs Return on Asset, which is dened as operating income before tax in year
t divided by the book value of total assets of rm i in year t - 1. Z Score is Altman's Z Score of rm i in year t. Total Sales is the
natural logarithm value of U.S. dollar-denominated total sales of rm i in year t.
We rst estimate models (A) and (B) by a multinomial probit model, and then, we estimate models (D) and (E) by a binary
probit model with panel data to reconrm the results of models (A) and (B). Since we employ only sukuk issuers and bank
Table 2
Partner bank, market valuation, pecking order variables, and distributions of nancial methodologies.
This table reports the number of bank loans and sukuk, debt, and equity issuers for rms seeking funds in Malaysia and Indonesia from 2000 to 2014. The number of
nancial methodology users is tabulated according to the Bank Cap to Funding quartiles and Market to Book quartiles in Panel (A) and Asymmetric quartiles and Public
or Private and Lead manager in Panel (B).
Market to Book quartiles Panel (A)
Bank Cap to Funding quartiles Total
Q1 Q2 Q3 Q4
Q1 Bank loan 29 62 98 125 314

Sukuk 0 1 2 0 3
Debt 1 0 0 0 1
Equity 1 4 3 1 9
N 31 67 103 126 327
Q2 Bank loan 51 65 104 64 284
Sukuk 3 1 1 0 5
Debt 14 1 0 0 15
Equity 4 5 4 3 16
N 72 72 109 67 320
Q3 Bank loan 65 63 61 71 260
Sukuk 10 7 1 0 18
Debt 18 11 0 1 30
Equity 4 8 5 4 21
N 97 89 67 76 329
Q4 Bank loan 61 53 49 53 216
Sukuk 17 11 1 0 29
Debt 11 17 0 0 28
Equity 6 7 5 6 24
N 95 88 55 59 297
Total Bank loan 206 243 312 313 1074
Sukuk 30 20 5 0 55
Debt 44 29 0 1 74
Equity 15 24 17 14 70
Market and lead manager Panel (B)
Asymmetric quartiles Total
Q1 Q2 Q3 Q4
Public market Sukuk 0 2 0 0 2

Debt 8 16 9 8 41
Equity 17 18 6 5 46
N 25 36 15 13 89
Private market Sukuk 1 8 19 25 53
Debt 2 10 4 17 33
Equity 8 9 4 3 24
N 11 27 27 45 110
Total Sukuk 1 10 19 25 55
Debt 10 26 13 25 74
Equity 25 27 10 8 70
Lead manager = partner bank Sukuk 1 5 19 24 49
Debt 0 0 1 1 2
Equity 1 3 2 2 8
N 2 8 22 27 59
Lead manager partner bank Sukuk 0 2 3 1 6
Debt 13 24 11 24 72
Equity 21 14 13 14 62
N 34 40 27 39 140
Total Sukuk 1 7 22 25 55
Debt 13 24 12 25 74
Equity 22 17 15 16 70
borrowers for model (D)'s samples, dependent variable Sukuk Issue 1 equals one when rm i issues sukuk in year t and equals zero
when the rm chooses bank borrowing in year t. In the case of model (E), dependent variable Sukuk Issue 2 equals one when rm
i issues sukuk in year t, and it equals zero when the rm chooses normal equity issuance in year t since we employ only sukuk and
stock issuers for the samples of model (E). We add the intersections of Bank Cap to Funding and Market to Book, and Bank Cap to
Funding and Asymmetric in both models (D) and (E), in addition to the variables explained above.

SukukIssue1 Sukuk 1 BankLoan 0 X 1 X 2 X 3 D

SukukIssue2 Sukuk 1 Equity 0 X 1 X 2 X 3 E
The parameter of the intersection of Bank Cap to Funding and Market is expected to be negative when Hypothesis 1 is support-
ed. The parameter of the intersection of Bank Cap to Funding and Asymmetric is expected to be negative when Hypothesis 2 is sup-
ported. The parameters of Public or Private and Lead Manager are expected to be negative and positive respectively when
Hypothesis 3 is supported.
4.2. Measuring information asymmetry
We estimate the degree of rm-level information asymmetry between rm insiders and outsiders following the methodology
of Bharath et al. (2009), who suggest that market microstructure measures can help gauge the magnitude of information asym-
metry facing outside investors in capital markets. First, we employ three variables that represent stock illiquidity measures. Then,
we employ three other variables of adverse selection component measures. Using these six variables and the principal component
analysis, we extract a common principal component of the cross-sectional level or annual change in the six variables and the cor-
responding principal component scores as the main information asymmetry value of rm i, that is, Asymmetric, which represents
the overall degree of information asymmetry of rm i in year t.
The rst three illiquidity component measures mentioned above are ILL, LR, and GAM. ILL, developed by Amihud (2002), is cal-
culated as the mean of the square root of the ratio of rm i's daily absolute stock return to the reported daily trading volume (in
millions) over all days in scal year t with nonzero volume. LR, developed by Cooper et al. (1985) and Amihud et al. (1997), is
dened as the negative value of the mean of the square root of the ratio of rm i's reported daily stock volume (in millions)
to its absolute stock return over all days in scal year t with nonzero return. According to Pastor and Stambaugh (2003), a stock's
liquidity can be captured by the interaction between its returns and lagged order ow.

Index
RT1 0 GAMT sign RT RT vT
where RT is the return on rm i's stock on day T, RIndex

T is each country's market index return on day T, and v is rm i's stock turn-
over volume on day T. We follow this measure and estimate the absolute magnitude of such reversal for rm i's stock over each
scal year t and label it as GAM.
The second set includes the three adverse selection component measures, AD, RAD, and C2. We follow George et al. (1991),
who develop a measure allowing information-containing orders to generate differential impacts on the transaction price and
quote midpoint. We obtain this adverse selection component of the spread by estimating the model indicated below.
2 RDT 0 1 PST DT DT1

ADT 11 PST
where RD is the difference between the traded stock price return and the quote midpoint return, and PS is the spread percentage
for rm i for day T. We compute the mean value for rm i's stock over each scal year t and label it as AD.
We also employ Roll's (1984) effective bidask spread of RAD, measured by the square root of the rst-order serial covariance
of stock price changes for rm i for day T. We compute the mean value for rm i's stock over each scal year t and label it as
RAD.
p
RADT 2 covT
Lastly, we employ a variable that proxies the relative importance of information asymmetry considerations among all stocks in
our sample, denoted as C2. Llorente et al. (2002), who introduced C2 empirically, show that a correspondence exists between the
cross-sectional variation in stocks' volumereturn dynamics and the relative importance of information-driven trading in stocks'
price uctuations. We estimate parameter C2 using ordinary least squares (OLS) estimates of the empirical equation, as indicated
below.
RT1 const C 1 RT C 2 V T RT
. X1
1
V T logturnover T logturnover Ts
200
s200
where log turnoverT = log(turnoverT + 0.00000255), and RT is the daily return of rm i's stock. We compute the value for rm i's
stock over each scal year t and label it as C2.
Unlike some studies, we follow Gomes and Phillips (2012) and do not employ the PIN measure in addition to the above six
measures. This is because these studies focus on publicly listed rms in major stock markets. Our study, however, focuses on
rms in emerging economies that trade an extremely small number of stocks.
To consider the year effect in the six information asymmetry proxy variables, we standardize the measures by rst taking the
difference between the original value and the cross-sectional mean, and then, we divide this difference by the cross-sectional
standard deviation. Next, we use principal component analysis to extract the rst principal component as the main information
asymmetry measure, which we denote as Asymmetric, an independent variable.
5. Data
We prepare the dataset of rm panel data from 2000 to 2014 for Malaysia and Indonesia. We include data from the global
nancial crisis period in our sample as Bourkhis and Nabi (2013) empirically conclude that Islamic banks provided nancial li-
quidity continuously even during the crisis. The sample rms are publicly listed bank borrowers, sukuk issuers, conventional
debt issuers, and equity issuers, as shown in Tables 1 and 2. The structural differences between sukuk and conventional debt
are explained in Section 3 (see Figs. 13). We extract only rms that increase bank loan values from year t - 1 to t and rms
that issued sukuk, conventional debt, and equity in year t. No convertible bonds, warrant bonds, and preferred stock are included.
Equity issuers other than normal equity issuers are excluded from the dataset, too. We match these sample rm nancial data,
partner bank nancial data, debt and equity issuance data, daily stock price data of these rms, and rms' country macroeconomic
and institutional data.
We obtain data regarding net bank loan increases and the names of the partner banks and their information from Bureau Van
Dijk's OSIRIS. Specically, the nancing bank's name(s) for each recorded non-nancial borrower is (are) indicated as Banker
Name in OSIRIS. We assume the bank (banks) and borrower to be engaged in a lending relationship when the bank(s) is
(are) linked to the rm by name for more than three consecutive years. Based on this information, we collect nancial data to
calculate Bank Cap to Funding for the partner banks from Thomson Reuters' Thomson One. When more than two partner bank
names are recorded for one non-nancial borrower in OSIRIS, we employ the weighted averages of the numerators of Bank Cap
to Funding, that is, each partner bank's market value of capital, where the weight parameter is the book value of total assets of
these multiple partner banks. Information on whether sukuk, conventional debt, and equity issuance are publicly or privately is-
sued is obtained from Thomson One, and the name of the lead manager of sukuk, conventional debt, and equity issuance, that is,
Lead Manager, is also obtained from Thomson One.
Financial data on the sample rms, namely, bank borrowers, sukuk, conventional debt issuers, and equity issuers (Market to
Book, CAR( 60, 2), CAR( 1,+1), Return on Asset, Z Score, Total Sales, CF, Asset, DIV, and CASH) are obtained from Thomson
One. To calculate ILL, LR, GAM, AD, RAD, and C2, rm's daily stock price returns, daily bid and ask stock prices, daily rm stock
trading volumes, and daily market index returns are obtained from Thomson One.
Table 3 shows the mean, median, and number of observations (N) for the partner bank variable (Bank Cap to Funding), pecking
order variables (KZ Index and Asymmetric), market valuation variables (Market to Book, CAR(1,+1), and CAR(60,2)), and
rm performance variables (Return on Asset, Z Score, and Total Sales).
6. Empirical results
6.1. Debt choice between sukuk and bank borrowing
Table 4 reports the results of empirical models (A) and (D). Eqs. (1) and (2) indicate the results when the dependent variable
is (1) Sukuk or Bank Loan and (2) Debt or Bank Loan, respectively. Eqs. (1) and (2) employ all the sukuk issuers, conventional debt
issuers, and bank borrower samples, but equity issuers are not included. The base outcome dependent variable of both Eqs. (1)
and (2) is Bank Loan.
The result of Eq. (1) also indicates that the parameter KZ Index is signicantly negative, implying that the lower the degree of
nancial constraint, the higher the probability of sukuk issuance. The result of Eq. (1) denotes that the parameter Asymmetric is
insignicant. Moreover, the parameter Market to Book is signicantly positive, which implies that high pre-issuance rm growth
opportunity promotes the rm's sukuk issuance. The parameter of the intersection between Bank Cap to Funding and Market to
Book is signicantly negative for Eq. (1). The parameter of the intersection of Bank Cap to Funding and Asymmetric is also signif-
icantly negative. These ndings imply that the larger the rm funding size compared to the partner bank capital size, the higher
the probability of sukuk issuance when the rm is valued highly and when it faces high asymmetric information.
The empirical results of Eq. (2) denote that the parameter Bank Cap to Funding is signicantly negative. Further, the parameter
Market to Book is signicantly positive. The parameter of the intersection of Bank Cap to Funding and Asymmetric is signicantly
positive for Eq. (2). These ndings conversely imply that the larger the rm funding size compared to the partner bank capital
size, the higher the probability of conventional debt issuance when the rm is valued highly: however, in this case, the rm's
asymmetric information is low.
Eq. (3) employs all the sukuk issuer and bank borrower samples, and Eq. (4) uses all the conventional debt issuers and bank
borrower samples. The dependent variable of Eq. (3), Sukuk Issue 1, equals one when the rm issues sukuk. It takes the value zero
when the rm chooses bank borrowing. The dependent variable of Eq. (4), Debt Issue, equals one when the rm issues conven-
tional debt and zero otherwise.
The empirical result of Eq. (3) indicates that the parameter KZ Index is signicantly negative, consistent with the result of Eq.
(1). The result of Eq. (3) also denotes that the parameter Market to Book is signicantly positive. This agrees with the result of Eq.
(1). For Eq. (3), the parameters of the intersections between Bank Cap to Funding and Market to Book and between Bank Cap to
Funding and Asymmetric are both signicantly negative. This result also implies that larger rm funding size (compared to partner
bank capital size) increases the probability of sukuk issuance when the rm is valued highly and has high asymmetric information.
Table 3
Summary statistics of partner bank, pecking order, market valuation, and rm performance variables by type of nancial methodology.
The sample includes bank borrowers and sukuk, debt, and equity issuers in Malaysia and Indonesia from 2000 to 2014. Mean, median, and number of observations (N) relate to the partner bank variable (Bank Cap to Funding),
pecking order variables (KZ Index and Asymmetric), market valuation variables (Market to Book, CAR(1,+1), and CAR(60,2)), and rm performance variables (Return on Asset, Z Score, and Total sales).
Bank loan Sukuk Debt Equity Total
Mean Median N Mean Median N Mean Median N Mean Median N Mean Median N
Partner bank variables

Bank Cap to Funding 5.782 5.828 1074 3.207 3.490 55 3.258 3.027 74 5.328 5.213 71 5.703 5.720 1274
Pecking order variables

KZ Index 7.508 2.830 1074 4.856 2.975 55 3.634 2.794 74 3.895 1.666 70 6.956 2.653 1273
Asymmetric 0.191 0.039 1074 0.074 0.086 55 0.075 0.115 74 0.059 0.017 70 0.175 0.038 1273
ILL 0.000 2.1E + 05 1074 0.000 1.2E 04 55 0.000 6.7E 05 74 0.000 0.000 70 0.000 3.0E 04 1273
LR 0.213 0.004 1074 0.011 0.009 55 0.022 0.014 74 0.013 0.006 70 0.185 0.004 1273
AD 0.038 0.111 1074 0.069 0.064 55 1.3E + 04 0.013 74 0.528 0.037 70 0.418 0.097 1273
RAD 8.3E 03 1.8E 02 1074 0.015 1.3E 02 55 1.4E 02 1.0E 02 74 1.1E 01 2.1E 02 70 3.6E 03 1.8E 02 1273
M. Nagano / Pacic-Basin Finance Journal xxx (2016) xxxxxx

C2 7.5E 03 1.5E 09 1074 1.9E 07 9.7E 01 55 6.3E 04 3.2E 10 74 7.3E 04 4.5E 10 70 7.4E 04 1.1E 09 1273
GAM 0.000 3.1E 11 1074 1.6E 13 1.1E 11 55 2.1E 10 2.7E 11 74 0.000 1.4E 11 70 0.000 2.1E 11 1273
Market valuation variables

Market to Book 1.564 1.062 1074 1.931 1.326 55 1.906 1.467 74 2.120 1.356 70 1.643 1.102 1273
CAR(1, +1) 0.002 0.005 55 0.001 0.001 74 0.001 0.004 70 0.000 0.003 199
CAR(60,2) 0.012 0.001 55 0.007 0.008 74 0.055 0.003 70 0.005 0.004 199
Firm performance variables

Return on Asset 0.058 0.057 1074 0.069 0.061 55 0.102 0.095 74 0.049 0.055 70 0.058 0.058 1273
Z Score 4.106 4.365 1074 4.400 4.425 55 3.843 3.959 74 4.198 4.394 70 4.114 4.378 1273
Total sales 4.318 4.223 1074 6.206 6.235 55 6.111 6.224 74 4.189 4.081 70 4.356 4.256 1273
Sukuk-bank loan Sukuk-equity Sukuk-bond
N Student Wilcoxon test Kruskal-Wallis test N Student test Wilcoxo n test Kruskal-Wallis test N Student test Wilcoxo n test Kruskal-Wallis test
Partner bank variables

Bank Cap to Funding 1,273 4.382a 4.330a 18.750a 1,273 3.914a 3.587a 12.870a 1,273 0.084 0.771 0.594a
Pecking order variables

KZ Index 1,273 0.644 0.528 0.279 1r,273 0.592 2.086b 4.349a 1,273 1.508 0.350 0.122
Asymmetric 0.270 1.958c 3.835a 1,273 0.101 2.611a 6.816a 1,273 0.027 0.482 0.233
Market valuation variables

Market to Book 1,273 0.854 3.307a 10.935a 1,273 0.422 0.116 0.013 1,273 0.073 0.751 0.563
CAR(1,+1) 199 2.903a 6.866a 4.236a 199 0.106 0 399 0.159 199 0.564 0.945 0.892
CAR(60,2) 199 3.256a 1.778a 0.284 199 0.339 0.656 0.430 199 0.738 0.937 0.877
Firm performance variables

Return on Asset 1,273 0.522 0.800 0.424 1r,273 0.540 1.086 1.179 1,273 2.221b 3.345a 11.188a
Z Score 1,273 1.249 0.811 0.659 1,273 0.895 0.582 0.338 1,273 1.921c 0.145 0.594
Total sales 1,273 6.992a 6.288a 39.534a 1,273 6.309a 5.992a 35.902a 1,273 0.372 0.737 0.113
a
Signicant at the 0.01 level.
b
c
9
Table 4
Sukuk vs. bank loan: Empirical results of debt choice between sukuk issuance and bank borrowing.
This table reports the results of the multinomial probit and binary probit models for Eq. (A). The base dataset includes all the sukuk, conventional debt, and equity issuers
as well as bank borrowers and non-bank borrowers (rms that did not increase their bank loans from the previous year are excluded). The dependent variable of Eq. (1)
is Sukuk or Bank loan and that of Eq. (2) is Debt or Bank loan. The base outcome dependent variable of both Eqs. (1) and (2) is Bank loan. The samples of Eqs. (1) and (2)
include all the sukuk, conventional debt issuers, and bank borrowers but exclude equity issuers. For Eq. (3), the dependent variable is Sukuk Issue 1, which equals one
when rm i issues sukuk in year t and zero when the rm chooses a bank loan. For Eq. (4), the dependent variable is Debt Issue, which equals one when rm i issues
conventional debt in year t and zero when it chooses bank borrowing. The sample of Eq. (3) includes all the sukuk issuers and bank borrowers but excludes conventional
debt issuers and equity issuers. The sample of Eq. (4) includes all conventional debt issuers and bank borrowers but excludes sukuk issuers and equity issuers.
Dependent variable
(1) (2) (3) (4)

Independent variables Sukuk vs. Debt vs. Sukuk Issue 1: Debt Issue:
bank loan bank loan Sukuk (=1) Debt (=1)
vs. vs.
bank loan (=0) bank loan (=0)
Specication Multinominal Binary probit model Marginal effect Binary probit model Marginal
probit model effect
Bank Cap to Funding 0.038 0.247b 0.038 0.001 0.200a 0.003

(0.160) (2.510) (0.230) (2.640)
KZ Index 0.830c 0.026 0.570c 0.006 0.055 0.001
(1.930) (0.040) (1.670) (0.110)
Asymmetric 1.279 32.448 1.549 0.013 23.075 0.005
(0.380) (1.400) (0.370) (1.400)
Market to Book 0.669c 3.669b 0.462b 0.006 2.600b 0.018
(1.760) (1.990) (2.110) (1.980)
Return on Asset 4.340 7.283 4.096 0.030 5.261 0.024
(1.530) (1.440) (1.540) (1.440)
Z Score 0.079 0.170 0.045 0.001 0.141 0.001
(0.730) (1.390) (0.640) (1.380)
Total sales 0.015 0.179 0.091 0.002 0.124 1.2E 04
(1.070) (1.190) (0.930) (1.160)
Bank Cap to Funding 0.059b 0.170 0.044a 4.9E 04 0.200 2.1E 04
Market to Book (2.670) (0.100) (2.690) (0.240)
Bank Cap to Funding 0.191b 1.922c 0.117b 0.002 1.408c 0.005
asymmetric (2.200) (1.830) (2.170) (1.850)
Bank Cap to Funding^2 0.009 0.016 0.007 9.0E 05 0.010 7.3E 06
(0.480) (0.700) (0.530) (0.590)
Market to Book^2 0.025 0.909b 0.016 2.5E 04 0.656b 4.6E 04
(0.620) (2.000) (0.550) (2.020)
Asymmetric^2 12.622 14.086 9.852 0.298 23.075 0.641
(0.590) (1.410) (0.630) (1.400)
Intercept 4.896a 7.192a 3.485a 5.065a
(4.140) (2.970) (4.110) (2.950)
Industrial dummy yes yes yes yes
Country dummy yes yes yes yes
Time trend yes yes yes yes
Country dummy yes yes yes yes
Time trend
Observations 1198 1198 1198
Wald chi2 37.0 16.2 17.6
LR test of rho = 0 0.610 0.880 0.750
a
b
c
The empirical result of Eq. (4) denotes that the parameter Bank Cap to Funding is signicantly negative. Also, the parameter
Market to Book is signicantly positive, while that of the intersection between Bank Cap to Funding and Market to Book is insignif-
icant. The parameter of the intersection of Bank Cap to Funding and Asymmetric is signicantly positive for Eq. (4).
6.2. Security issuance choice between sukuk and equity
Table 5 reports the results of empirical models (B) and (E). Eqs. (5) and (6) estimate the results when the dependent variable
is (5) Sukuk or Equity and (6) Debt or Equity, respectively. Eqs. (5) and (6) employ all the sukuk, conventional debt, and equity
issuer samples, but bank borrowers are not included. Accordingly, the base outcome dependent variable for both Eqs. (5) and
(6) is Equity.
The empirical results of Eq. (5) denote that the parameter Bank Cap to Funding is insignicant. This implies that the difference
between sukuk issuer and equity issuer funding demand size does not determine the security issuance choice. Further, the param-
eter KZ Index is signicantly negative, while Asymmetric is signicantly positive, consistent with the results of Eqs. (1) and (3). The
Table 5
Sukuk vs. Equity: Empirical results of security choice between sukuk and equity issuance.
This table reports the results of the multinomial probit and binary probit models for Eq. (B). The base dataset includes all sukuk, conventional debt, and equity issuers.
Bank borrowers and non-bank borrowers are excluded. The dependent variable is Sukuk, Debt, or Equity in Eqs. (5)(6). The base outcome of both Eqs. (5) and (6) is
Equity. The dependent variable is Sukuk Issue 2, which equals one when rm i issues sukuk in year t and zero when it chooses equity for Eqs. (7)(9). Eqs. (7), (8),
and (9) proxy the market valuation of rm i as Market to Book, CAR(60,2), and CAR(1,+1), respectively.
Dependent Variable
(5) (6) (7) (8) (9)

Independent variables Sukuk vs. equity Debt vs. Sukuk Issue 2 Sukuk Issue 2 Sukuk Issue 2
equity
Sukuk (=1) Sukuk (=1) Sukuk (=1)
vs. vs. vs.
Equity (=0) Equity (=0) Equity (=0)
Specication Multinominal probit Binary probit Marginal Binary probit Marginal Binary probit Marginal
model model effect model effect model effect
Bank Cap to Funding 0.376 0.581 0.359 0.005 0.137 0.004 0.123 0.002
(0.930) (0.640) (1.110) (0.200) (0.660)
KZ Index 1.305 0.749 0.940 0.011 0.421 0.006 0.468 0.001
(2.350) (0.630) (1.200) (0.640) (0.680)
Asymmetric 5.003c 43.989 3.326b 0.058 2.614b 0.037 2.506b 0.005
(1.770) (1.100) (2.110) (2.400) (2.430)
Market to Book 1.179c 6.450c 1.090c 0.014
(1.750) (1.930) (1.950)
CAR(60,2) 0.437b 0.006
(2.060)
CAR(1,+1) 1.803c 0.036
(1.730)
Return on Asset 3.325 9.720 2.983 0.040 4.257 0.006 3.693 0.007
(1.110) (1.270) (0.810) (1.350) (1.160)
Z Score 0.293 0.193 0.292 0.004 1.060 0.00 1.050 1.0E 04
(1.250) (0.660) (1.510) (0.570) (0.480)
Total sales 0.696b 0.913b 0.470b 0.006 0.362c 0.001 0.477b 0.001
(2.580) (2.500) (2.360) (1.810) (2.350)
Bank Cap to 0.166c 0.079 0.167b 0.002
Funding Market (1.990) (0.190) (2.090)
to Book
Bank Cap to Funding 1.212c 0.002
CAR(60,2) (1.870)
Bank Cap to Funding 0.606b 0.001
CAR(1,+1) (2.130)
Bank Cap to 0.393 2.030 0.466 0.006 0.417 0.001 0.216 0.004
Funding Asymmetric (1.290) (0.370) (1.460) (1.360) (1.200)
Bank Cap to Funding^2 0.011 0.034 0.011 1.5E 04 0.417 1.1E 05 0.013 2.7E 05
(0.420) (0.920) (0.550) (0.360) (0.590)
Market to Book^2 0.016 1.597 0.011 1.5E 04
(0.360) (1.460) (0.350)
CAR(60,2)^2 39.290 0.056
(1.440)
CAR(1,+1)^2 32.934 0.658
(1.360)
Asymmetric^2 6.993 8.538 6.464 0.086 0.261 0.001 1.961 0.004
(0.350) (0.750) (0.440) (0.140) (0.100)
Intercept 6.721a 12.658b 4.868a 2.489c 3.606a
(3.280) (2.160) (3.180) (1.920) (2.690)
Industrial dummy yes yes yes yes yes
Country dummy yes yes yes yes yes
Time trend yes yes yes yes yes
Country dummy Time yes yes yes yes yes
trend
Observations 140 125 124 124
Wald chi2 41.2 34.4a 33.7a 44.1a
LR test of rho = 0 4.660 3.760b 4.100b 3.440b
a
b
c
results of Eq. (5) also denote that the parameter Market to Book is signicantly positive, while that of the intersection between
Bank Cap to Funding and Market to Book is signicantly negative. However, the parameter of the intersection of Bank Cap to
Funding and Asymmetric is insignicant.
The empirical results of Eq. (6) denote that the parameters Bank Cap to Funding, KZ Index, and Asymmetric are all insignicant.
Moreover, Market to Book is signicantly positive, implying that the pre-issuance growth opportunity of conventional debt issuers
is higher than that of equity issuers. The parameter of the intersection of Bank Cap to Funding and Market to Book is insignicant
for Eq. (6), as is the parameter of the intersection of Bank Cap to Funding and Asymmetric.
Eqs. (7)(9) employ all the sukuk and equity issuer samples, but conventional debt issuers and bank borrowers are not includ-
ed. The dependent variables of Eqs. (7)(9), that is, Sukuk Issue 2, equal one when the rm issues sukuk and zero when it chooses
equity issuance. Eq. (7) employs Market to Book as a proxy of rm i's market valuation. Eqs. (8) and (9) employ CAR(60,2)
and CAR( 1,+1) instead of Market to Book as proxies of rm i's market valuations, respectively. The empirical results of Eqs.
(7)(9) also denote that the parameters Asymmetric are all signicantly positive. Moreover, the parameters Market to Book,
CAR( 60,-2), and CAR( 1,+1) are all signicantly positive. The parameter of the intersection between Bank Cap to Funding
and Market to Book is signicantly negative for Eq. (7), while that of the intersection between Bank Cap to Funding and
CAR( 60,-2) is signicantly negative for Eq. (8). The parameter of the intersection between Bank Cap to Funding and
CAR(1,+1) is signicantly negative for Eq. (9). The parameters of the intersection of Bank Cap to Funding and Asymmetric are
insignicant for Eqs. (7)(9).
Table 6
Sukuk vs. Equity 3: Results of the binary probit model indicating sukuk issuance determinants including the public/private market and lead manager dummy variables.
This table reports the results of the binary probit model (F). The dependent variable, Sukuk Issue 2, equals one when rm i issues sukuk in year t and zero when it issues
equity. Eqs. (10)(11) employ Public or Private and Lead manager and their intersections with Asymmetric as dependent variables.
(10) (11)
Independent variables Sukuk Issue 2: Marginal effect Sukuk Issue 2: Marginal effect
Sukuk (=1) Sukuk (=1)
vs. vs.
Equity (=0) Equity (=0)
Bank Cap to Funding 0.415 0.001 0.659 0.001

(1.110) (0.830)
Public or Private 0.858b 0.004
(2.410)
Lead manager 15.235a 0.477
(3.970)
KZ Index 0.991 0.003 1.103 0.004
(1.190) (1.530)
Asymmetric 6.914b 0.029 10.060c 0.055
(2.210) (1.920)
Market to Book 1.431c 0.005 1.668c 0.005
(1.910) (1.770)
Return on Asset 3.181 0.010 1.631 0.017
(0.890) (0.030)
Z Score 0.245 0.001 0.301 0.001
(1.210) (0.810)
Total sales 0.324b 0.001 0.657a 0.001
(2.270) (2.650)
Bank Cap to Funding Market to Book 0.226c 0.007 0.577b 0.001
(1.930) (2.020)
Bank Cap to Funding Asymmetric 0.334 0.001 0.394 0.001
(0.270) (0.520)
Public or Private Asymmetric 10.318b 0.034
(2.260)
Lead manager Asymmetric 36.873b 0.002
(2.290)
Bank Cap to Funding^2 0.009 3.0E 04 0.039 7.0E 05
(0.309) (0.520)
Market to Book^2 0.011 3.7E 05 0.167 1.5E 06
(0.290) (0.090)
Asymmetric^2 8.591 0.061 13.600c 0.077
(0.850) (1.880)
Intercept 4.602a 3.313a
(2.740) (2.740)
Industrial dummy yes yes
Countryl dummy yes yes
Time trend yes yes
Country dummy Time trend yes yes
Observations 125 125
Wald chi2 31.2a 32.7a
LR test of rho = 0 3.900b 4.840b
a
b
c
6.3. Sukuk issuance and choice of market
Table 6 reports the results of empirical model (E), which additionally includes the independent variables Public and Private in
Eq. (10) and Lead Manager in Eq. (11). Sukuk Issue 2 is the dependent variable for Eqs. (10)(11). The variable equals one and
zero when rm i issues sukuk and equity in year t, respectively. Eqs. (10) and (11) employ all the sukuk and equity issuer samples,
but conventional debt issuers and bank borrowers are excluded. The empirical results of Eq. (10) denote that the parameter Public
and Private is signicantly negative, as is that of the intersection between Public or Private and Asymmetric. The empirical results of
Eq. (11) indicate that the parameter Lead Manager is signicantly positive, as is that of the intersection between Lead Manager and
Asymmetric.
7. Discussion
In Table 4, we present empirical evidence that indicates the rm's sukuk issuance probability (Sukuk) and the ratio of the part-
ner bank's capital size to issuer funding size (Bank Cap to Funding) are negatively related. Our empirical evidence also shows that
a rm with such a large funding demand is likely to resort to the sukuk market when it is valued highly (Market to Book). These
results are conrmed by both the multinomial probit and the binary probit model estimations. The empirical evidence from Eqs.
(1)(4), accordingly, supports Hypothesis 1. Bolton and Freixas (2008) theoretically assert that quick growth in rms' funding de-
mand in emerging economies encourages debt issuance to save the bank from the capital requirement restrictions. Our evidence
empirically conrms this theoretical framework, which only a few previous studies have veried.
This study employs three types of rm market valuation proxies for the purpose of robustness checks, that is, Market to Book,
CAR(60,2), and CAR(1,+1), in Table 5. The hold-up and the liquidity constraint hypotheses both expect a positive param-
eter of rm market valuation proxy to the dependent variable Sukuk. Meanwhile, the market timing theory expects a negative
relationship between these two variables. Consistent with results of Nagano (2016), the empirical results also show that these
three market valuation proxies are positively related to the probability of sukuk issuance. This is inconsistent with the market
timing theory but consistent with the hold-up and liquidity constraint hypotheses. In addition, the intersections of Bank Cap to
Funding and these three proxy variables {Market to Book, CAR(60, 2), and CAR( 1,+1)} are also negatively related to the
probability of sukuk issuance (Sukuk). Therefore, we reconrm that our results of the three types of rm market valuation proxies
support Hypothesis 1. Our results support the conclusions of Hoshi and Kashyap (1990), Hoshi et al. (1991) and Bolton and
Freixas (2000), who assert the positive relationship between a rm's growth opportunity and debt issuance. Our results also sup-
port the ndings of Maghyereh and Awartani (2016) and Naifar et al. (2016), who conclude that a positive relationship exists be-
tween a rm's market valuation and sukuk issuance. In sum, our conclusion which newly contributes to the literature is that,
different from literature conclusions, rm market valuation is not merely only one requirement, but the relative funding size is
simultaneously the determinant of sukuk issuance.
Conversely, our empirical evidence in Table 4 also shows a negative relationship between conventional debt issuance proba-
bility (Debt) and Bank Cap to Funding and a positive relationship between conventional debt issuance probability (Debt) and Mar-
ket to Book. These results are similar to those of the dependent variable of sukuk issuance probability. Therefore, we also conclude
that the ratio of partner bank capital size to issuer funding size and rm market valuation are common determinants of sukuk and
conventional debt issuance.
Then, which determinant differentiates sukuk from conventional debt issuance? Our empirical results in Tables 4 and 5 also
show that rm information asymmetry is the key to differentiate sukuk from conventional debt choice. Our empirical results in
Table 4 show that the intersection of the ratio of partner bank capital size to issuer funding size (Bank Cap to Funding) and the
degree of information asymmetry (Asymmetric) negatively inuence sukuk issuance (Sukuk). This implies that a rm with high
information asymmetry chooses sukuk issuance when the funding demand is too large to borrow from the partner bank. Intuitive-
ly, in underdeveloped nancial markets like Malaysia and Indonesia, most bank borrowers are highly information asymmetric
rms. As explained in Section 3, the structures of Murabahah sukuk and Ijarah sukuk, which account for more than 90% of the
total sukuk issuance, mitigate information asymmetry of the issuers. Therefore, a high degree of information asymmetry encour-
ages the rm to issue sukuk when the funding size is large compared to the size of partner bank capital. In Table 5, our empirical
results from the analysis pertaining to sukuk and equity issuance also show that rms with a high degree of information asym-
metry prefer sukuk to equity issuance. This is because equity issuance requires a low degree of information asymmetry, and highly
information asymmetric rms approach the sukuk market as a result. This empirical evidence supports Hypothesis 2 and is con-
sistent with the ndings of Ebrahim et al. (2016), who theoretically conclude which Islamic nance scheme mitigates information
asymmetry between rm insiders and outsiders. Our results also support the observations of Halim et al. (2016) and Klein and
Weill (2016), namely the existence of a relationship between a rm's information asymmetry and sukuk issuance. Additionally,
we contribute to the literature by offering a new nding; we show evidence that a rm approaches the sukuk market when it
has a large-sized funding demand that cannot be met any longer by its partner bank and when it cannot access the conventional
debt market because of a high degree of information asymmetry.
Table 6 shows that for the most part, sukuk is privately, not publicly, issued. Denis and Mihov (2003) and Gomes and Phillips
(2012) conclude that in the case of conventional debt issuance, private debt is chosen when a highly information asymmetric rm
issues debt, as only a limited number of external entities possesses the soft information of the rm. Our empirical result is con-
sistent with the conclusions in the literature. Another empirical result of this study (see Table 6) also shows that sukuk issuance is
promoted when the partner bank and security issuance lead manager are one and the same or belong to the same business group.
We interpret these results as follows: the partner banks have soft information and encourage the borrower to approach the sukuk
market when the funding size is too large to borrow from these banks. Therefore, the sukuk market is used by a rm with infor-
mation asymmetry in an underdeveloped funding market when that rm has a large funding demand.
8. Conclusion
This study tests how the bankrm relationship and information asymmetry inuence a rm's sukuk issuance, using a sample
of Malaysian and Indonesian rms from 2000 to 2014. Our novel contribution to the related literature is that the effects of both
bank borrower funding size to partner bank capital size and high growth opportunity result in rms' sukuk issuance. Bolton and
Freixas (2008) and the hold-up and liquidity constraint hypotheses report that these two effects occur independently, but this
study demonstrates that they work simultaneously.
These determinants are similar to those of conventional debt security. This study also demonstrates that the difference be-
tween the determinants of sukuk and conventional debt issuance lies in the degree of information asymmetry. A bank borrower
is more likely to resort to the sukuk debt market when the funding demand is too large to borrow from the partner bank. Our
results show that the rm prefers sukuk to conventional debt issuance when the rm faces too much information asymmetry
to approach the conventional debt market. This determinant differentiates sukuk from conventional debt issuance and encourages
rms with information asymmetry to approach the sukuk market in underdeveloped nancial markets. This is one of the reasons
for the rapid development of the sukuk market compared to the conventional debt market.
In sum, our conclusion also suggests that this intermediate funding market between bank borrowing and the conventional
debt market serves as a pipeline between the two markets in emerging economies. Therefore, this also implies that government
support, such as tax preferential treatment for the nancial methodology of sukuk issuance, also promotes information-asymmet-
ric rm growth and the development of both the sukuk and the conventional debt market. Future research may incorporate the
interaction between the bankrm relationship and rm information asymmetry to obtain more insight into the nancial strate-
gies of rms in emerging markets.
Appendix A
Table
Variable Denitions.
The variables dened here are employed to estimate empirical models (A), (B), (C), (D), and (E). Net bank loan data of borrowers and the names and nancial data of
partner banks are sourced from Bureau Van Dijk's OSIRIS. Financial data on sample borrowers and security issuers are sourced from Thomson Reuters' Thomson One.
Data on rms' daily stock price returns, daily bid and ask stock prices, daily rm stock trading volumes, and daily market index returns are also sourced from Thomson
One.
Variables Denition Expected sign of Source

parameter
Dependent variables
Bank loan Equals zero when rm i increases the net value of bank loans in year t for models (A) and (B) Dependent Bureau Van Dijk,
variable OSIRIS
Sukuk Equals one when rm i issues sukuk in year t and zero otherwise for model (A) Dependent Thomson Reuters,
variable Thomson One
Debt Equals one when rm i issues conventional debt security and zero otherwise in year t for model (B) Dependent Thomson Reuters,
Equity Equals zero when rm i issues equity in year t for model (B) Dependent Thomson Reuters,
Sukuk Issue 1 Equals one when rm i issues sukuk in year t and zero otherwise for model (D) Dependent Thomson Reuters,
Sukuk Issue 2 Equals one when rm i issues sukuk in year t and zero otherwise for model (E) Dependent Thomson Reuters,
Partner bank independent variables (XPartnerBank)

Bank Cap to Total market capitalization of the partner bank of rm i in year t divided by total funding value of Thomson Reuters,
Funding rm i in year t Thomson One
Bureau Van Dijk,
OSIRIS
Public or Equals one when sukuk, conventional debt, or equity is issued in the public market and zero when Thomson Reuters,
private the security is issued in the private market Thomson One
Lead manager Equals one when rm i's partner bank and the lead manager of sukuk, conventional debt, or equity + Thomson Reuters,
issuance are consistent or belong to the same business group in year t Thomson One
Pecking order independent variables (XPeckingOrder)

KZ Index KZ Index = 1.002 CF/TA + 3.319 Leverage 39.368 DIV/TA 1.315 CASH/TA Thomson Reuters,
Thomson One
CF Sum of net income and depreciation of rm i in year t Thomson Reuters,
Thomson One
Table (continued)
Variables Denition Expected sign of Source

parameter
Asset Book value of total assets of rm i in year t Thomson Reuters,

Thomson One
DIV Cash dividends of rm i in year t Thomson Reuters,
Thomson One
CASH Cash and short-term investment of rm i in year t Thomson Reuters,
Thomson One
Asymmetric Firm i's rst principal component score in year t estimated from ILL, LR, GAM, AD, RAD, and C2 + Thomson Reuters,
Thomson One
ILL Mean of the square root of the ratio of rm i's daily absolute stock return to the reported daily Thomson Reuters,
trading volume (in millions) over all days in nancial year t with nonzero volume Thomson One
LR Minus the mean of the square root of the ratio the reported daily volume of rm i's stock (in Thomson Reuters,
millions) to its absolute stock return over all days in scal year t with nonzero return Thomson One
GAM Parameter of the equation seen below; the dependent variable is RT+1, that is, return on rm i's Thomson Reuters,
stock on day T + 1, and the independent variable is RT RIndex
T , which is the return on rm i's stock Thomson One
minus each country's market index return on day T multiplied by v (rm i's stock turnover volume
on day T)
RT+1 = 0 + GAMT sign(RT RIndex
T ) vT +
AD Spread percentage multiplied by 1 minus a parameter that is estimated as the difference between Thomson Reuters,
the traded stock price return (RD) and spread percentage (PS) for rm i in day T Thomson One
2 RDT 0 1 PST DT DT1
ADT 1 1 PST
RAD Square root of the rst-order serial covariance of stock price changes for rm i in day T Thomson Reuters,
Thomson One
C2 Estimated parameter of the empirical equation seen below. Thomson Reuters,
RT+1 = const + C1 RT +1 + C2 VT RT + Thomson One
1
V T ; logturnover T 1 200 ; logturnover Ts
s200
where logturnoverT = log(turnoverT + 0.00000255), and RT is the daily return of rm i's stock
Independent variables of rm characteristics (XFirm)

Market to Book Market value of capital in year t plus book value of liability in year t 1 divided by book value of + Thomson Reuters,
total assets of year t 1 for rm i Thomson One
CAR(60,2) 59-day event window cumulative abnormal returns around the sukuk, conventional debt, or equity + Thomson Reuters,
issue dates for rm i in the pre-issuance period Thomson One
CAR(1,+1) 3-day event window cumulative abnormal returns around the sukuk, conventional debt, or equity + Thomson Reuters,
issue dates for rm i before issuance Thomson One
Return on Operating income before tax in year t divided by the book value of total assets of rm i in year t 1 Thomson Reuters,
Asset Thomson One
Z Score Altman's Z Score of rm i in year t. Thomson Reuters,
Z Score = 0.012 (Current Asset Current Liability) + 0.014 (Retained Thomson One
Earnings) + 0.033 (EBITDA/Total Asset) + 0.006 (Market Value of Capital/Total
Liability) + 0.999 Total Sales/Total Assets
Total sales Natural logarithm of U.S. dollar-denominated total sales of rm i in year t + Thomson Reuters,
Thomson One
References
Abedifar, P., Ebrahim, S.M., Molyneux, P., Tarazi, A., 2013. Islamic banking and finance: recent empirical literature and directions for future research. J. Econ. Surv. 29
(4), 637670.
Amihud, Y., 2002. Illiquidity and stock returns: cross-section and time-series effects. J. Financ. Mark. 5, 3156.
Amihud, Y., Mendelson, H., Lauterbach, B., 1997. Market microstructure and securities values: evidence from the Tel Aviv stock exchange. J. Financ. Econ. 45 (3),
365390.
Asquith, P., Mullins, D., 1986. Equity issues and offering dilution. J. Financ. Econ. 15, 6189.
Azmat, S., Skully, M., Brown, K., 2014. Issuers choice of Islamic bond type. Pac. Basin Financ. J. 28, 122135.
Baker, M.P., Wurgler, J., 2002. Market timing and capital structure. J. Financ. 57, 132.
Baker, M., Stein, J.C., Wurgler, J., 2003. When does the market matter? Stock prices and the investment of equity-dependent firms. Q. J. Econ. 9691005.
Bharath, S.T., Pasquariello, P., Wu, G., 2009. Does asymmetric information drive capital structure decisions? Rev. Financ. Stud. 22 (9), 32123243.
Bolton, P., Freixas, X., 2000. Equity, bonds and bank debt: capital structure and financial market equilibrium under asymmetric information. J. Polit. Econ. 108 (2),
324351.
Bolton, P., Freixas, X., 2008. How can emerging market economies benefit from a corporate bond market? In: Borensztein, E., Cowan, K., Eichengreen, B., Panizza, U.
(Eds.), Bond Markets in Latin America. The MIT Press, pp. 2949
Boot, A.W.A., 2000. Relationship banking: what do we know? J. Financ. Intermed. 9 (1), 725.
Bourkhis, K., Nabi, M.S., 2013. Islamic and conventional banks' soundness during the 20072008 financial crisis. Rev. Financ. Econ. 22 (2), 6877.
Brown, S.J., Warner, J.B., 1985. Using daily stock returns: the case of event studies. J. Financ. Econ. 14, 331.
Bruinshoofd, A., De Haan, L., 2007. Market Timing and Corporate Capital Structure: A Transatlantic Comparison. De Nederlandsche Bank Working Paper No. 144.
Chong, B.S., Liu, M.-H., 2009. Islamic banking: interest-free or interest-based? Pac. Basin Financ. J. 17, 125144.
Cooper, K.S., Groth, J.C., Avera, W.E., 1985. Liquidity, exchange listing, and common stock performance. J. Econ. Bus. 37 (1), 1933.
Denis, D.J., Mihov, V.T., 2003. The choice among bank debt, non-bank private debt, and public debt: evidence from new corporate borrowings. J. Financ. Econ. 70, 328.
Donaldson, G., 1961. Corporate Debt Capacity: A Study of Corporate Debt Policy and the Determination of Corporate Debt Capacity. Harvard Business School, Division
of Research, Harvard University, Boston, MA.
Ebrahim, M.S., Jaafar, A., Omar, F.A., Salleh, M.O., 2016. Can Islamic injunctions indemnify the structural flaws of securitized debt? J. Corp. Finan. 37, 271286.
Elyasiani, E., Goldberg, L.G., 2004. Relationship lending: a survey of the literature. J. Econ. Bus. 56 (4), 315330.
Fama, E.F., French, K.R., 2005. Financing decisions: who issues stock? J. Financ. Econ. 76, 549582.
George, T.J., Kaul, G., Nimalendran, M., 1991. Estimation of the bid-ask spread and its components: a new approach. Rev. Financ. Stud. 4 (4), 623656.
Godlewski, C.J., Turk-Ariss, R., Weill, L., 2013. Sukuk vs. conventional bonds: a stock market perspective. J. Comp. Econ. 41, 745761.
Godlewski, C.J., Turk-Ariss, R., Weill, L., 2016. Do the type of sukuk and choice of Shari'a scholar matter? J. Econ. Behav. Organ. (forthcoming).
Gomes, A., Phillips, G., 2012. Why do public firms issue private and public securities? J. Financ. Intermed. 21 (4), 549722.
Graham, J.R., Harvey, C.R., 2001. The theory and practice of corporate finance: evidence from the field. J. Financ. Econ. 60, 187243.
Hale, G., Santos, J.A.C., 2008. The decision to first enter the public bond market: the role of reputation, funding choices, and bank relationships. J. Bank. Financ. 32 (9),
19281940.
Halim, Z.A., How, J., Verhoeven, P., 2016. Agency costs and corporate sukuk issuance. Pac. Basin Financ. J. forthcoming.
Henderson, B.J., Jegadeesh, N., Weisbach, M.S., 2006. World Markets for Raising New Capital. J. Financ. Econ. 82, 63101.
Hoshi, T., Kashyap, A.K., 1990. Evidence on Q and Investment for Japanese Firms. J. Jpn. Int. Econ. 4 (4), 371400.
Hoshi, T., Kashyap, A., Scharfstein, D., 1991. Corporate structure, liquidity, and investment: evidence from Japanese industrial groups. Q. J. Econ. 106 (1), 3360.
Hovakimian, A., Opler, T., Titman, S., 2001. The debt-equity choice. J. Financ. Quant. Anal. 36 (1), 124.
Jung, K., Kim, Y.-C., Stulz, R.M., 1996. Timing, investment opportunities, managerial discretion, and the security issue decision. J. Financ. Econ. 42 (2), 159186.
Kaplan, S.N., Zingales, L., 1997. Do investment-cash flow sensitivities provide useful measures of financing constraints? Q. J. Econ. 112 (1), 169215.
Klein, P.O., Weill, L., 2016. Why do companies issue sukuk? Rev. Financ. Econ. forthcoming.
Lemmon, M.L., Zender, J.F., 2010. Debt capacity and tests of capital structure theories. J. Financ. Quant. Anal. 45 (5), 11611187.
Llorente, G., Saar, G., Want, J., 2002. Dynamic volume-return relation of individual stocks. Rev. Financ. Stud. 15 (4), 10051047.
Maghyereh, A.I., Awartani, B., 2016. Dynamic Transmissions between sukuk and bond markets. Res. Int. Bus. Financ. 38, 246261.
Ming, D., Loncarski, I., ter Horst, J., Veld, C., 2012. What drives security issuance decisions: market timing, pecking order, or both? Financ. Manag. 41 (3), 637663.
Mohamed, H.H., Masih, M., Bacha, O.I., 2015. Why do issuers issue sukuk or conventioal bond? Evidence from Malaysian listed firms using partial adjustment models.
Pac. Basin Financ. J. 34, 233252.
Myers, S.C., 1984. The capital structure puzzle. J. Financ. 39 (3), 575592.
Myers, S.C., Majluf, N.S., 1984. Corporate financing and investment decisions when firms have information that investors do not have. J. Financ. Econ. 13, 187221.
Nagano, M., 2016. Who issues sukuk and when?: an analysis of Islamic bond issuance determinants. Rev. Financ. Econ. forthcoming.
Naifar, N., Hammoudeh, S., Al dohaiman, M.S., 2016. Dependence structure between sukuk and stock market conditions: an empirical analysis with archimedean cop-
ulas. J. Int. Finan. Markets. Inst. Money 44, 148165.
Ongena, S., endeniz-Ync, ., 2011. Which firms engage small, foreign, or state banks? And who goes Islamic? Evidence from Turkey. J. Bank. Financ. 35 (12),
32133324.
Pastor, L., Stambaugh, R.F., 2003. Liquidity risk and expected stock returns. J. Polit. Econ. 111 (3), 642685.
Rajan, R.G., 1992. Insiders and outsiders: the choice between informed and arm's-length debt. J. Financ. 47 (4), 13671423.
Roll, R., 1984. A simple implicit measure of the effective bid-ask spread in an efficient market. J. Financ. 39 (4), 11271139.
Sharpe, S.A., 1990. Asymmetric information, bank lending, and implicit contracts: a stylized model of customer relations. J. Financ. 45 (4), 10691087.
Shyam-Sunder, L., Myers, S.C., 1999. Testing static tradeoff against pecking order models of capital structure. J. Financ. Econ. 51, 219244.
TurkAriss, R., 2010. Competitive conditions in Islamic and conventional banking: a global perspective. Rev. Financ. Econ. 19 (3), 101108.
Udell, G.F., Berger, A.N., 1995. Relationship lending and lines of credit in small firm finance. J. Bus. 68 (3), 351381.

Nagano2016 2

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Nagano2016 2

Uploaded by

Copyright:

Available Formats

PACFIN-00903; No of Pages 16

Pacic-Basin Finance Journal xxx (2016) xxxxxx

Contents lists available at ScienceDirect

Pacic-Basin Finance Journal

Sukuk issuance and information asymmetry: Why do rms issue sukuk?

This research is nancially supported by JSPS KAKENHI Grant Number JP15K03550.

Project Owner issues sukuk

Fig. 1. The structure of Murabahah sukuk.

The SPC leases the asset to The issuer sells

The net amount of the SPCs cash inflow

Fig. 2. The structure of Ijarah sukuk.

Project Owner issues sukuk

Fig. 3. The structure of Musyarakah sukuk.

PartnerBank PeckingOrder Firm

PartnerBank PeckingOrder Firm

Market to Book quartiles Panel (A)

Bank Cap to Funding quartiles Total

Q1 Bank loan 29 62 98 125 314

Market and lead manager Panel (B)

Asymmetric quartiles Total

Public market Sukuk 0 2 0 0 2

PartnerBank PeckingOrder Firm

PartnerBank PeckingOrder Firm

4.2. Measuring information asymmetry

where RT is the return on rm i's stock on day T, RIndex

2 RDT 0 1 PST DT DT1

6.1. Debt choice between sukuk and bank borrowing

Bank loan Sukuk Debt Equity Total

Partner bank variables

Pecking order variables

M. Nagano / Pacic-Basin Finance Journal xxx (2016) xxxxxx

Market valuation variables

Firm performance variables

Sukuk-bank loan Sukuk-equity Sukuk-bond

Partner bank variables

Pecking order variables

Market valuation variables

Firm performance variables

(1) (2) (3) (4)

Bank Cap to Funding 0.038 0.247b 0.038 0.001 0.200a 0.003

6.2. Security issuance choice between sukuk and equity

(5) (6) (7) (8) (9)

Bank Cap to Funding 0.415 0.001 0.659 0.001

6.3. Sukuk issuance and choice of market

Variables Denition Expected sign of Source

Partner bank independent variables (XPartnerBank)

Pecking order independent variables (XPeckingOrder)

Variables Denition Expected sign of Source

Asset Book value of total assets of rm i in year t Thomson Reuters,

Independent variables of rm characteristics (XFirm)

You might also like