How Islamic are Islamic banks? A non-linear assessment of Islamic rate – conventional rate relations

Economic Modelling xxx (xxxx) xxx–xxx

Contents lists available at ScienceDirect

Economic Modelling journal homepage: www.elsevier.com/locate/econmod

How Islamic are Islamic banks? A non-linear assessment of Islamic rate – conventional rate relations Raditya Sukmanaa, Mansor H. Ibrahimb, a b

⁎

Department of Islamic Economics, Faculty of Economics and Business, Universitas Airlangga, Surabaya, Indonesia School of Graduate Studies, International Centre for Education in Islamic Finance (INCEIF), Kuala Lumpur, Malaysia

A R T I C L E I N F O

A BS T RAC T

JEL classification: E43 E52 G21

In this paper, we perform a non-linear assessment of Islamic rate – conventional rate relations for the case of Malaysia. Using monthly data covering the period January 1999 to November 2016, we find strong evidence supporting non-linear reactions of the Islamic investment rates to conventional rates in the long run and/or short-run for all matched maturities. More precisely, the Islamic investment rates exhibit faster upward movement (slower downward movement) in responses to conventional deposit rate increases (decreases). The asymmetric pricing behaviour of Islamic banks however tends to weaken as maturity lengthens. Accordingly, we infer that Islamic banks do not rigidly peg their investment deposit rates to conventional deposit rates as some have claimed in questioning the Islamicity of Islamic banks.

Keywords: Islamic banks Islamic investment rates Deposit rates Nonlinear ARDL

1. Introduction The fast growing Islamic banking sector amidst financial uncertainties has attracted much attention. While few would argue against the Islamic foundations (Shari’ah) of Islamic banking as being distinct, some have raised doubts whether in practices Islamic banks are different from conventional banks. Critics of Islamic banking cite the following Islamic banking practices as being indistinguishable from conventional banks (see Chong and Liu, 2009; Khan, 2010; Ariff, 2015). First, despite proclaiming participation or profit-and-loss sharing (PLS) arrangement as the bedrock of the Islamic banking, the partnership-based contracts form only a minor portion of the Islamic banking assets. Second, Islamic banks tend to benchmark against conventional interest rates in the pricing of their products. And finally, the Islamic banking products are mere imitation of the conventional banking products, i.e. the conventional products adapted or modified to be Shari’ah compliant. Among these criticisms, benchmarking against conventional interest rates appears most damaging to the Islamicity of Islamic banks. In their analysis of Islamic banking in Malaysia, Chong and Liu (2009) provide evidence for a unidirectional Granger causality that runs from conventional deposit rates to profit-and-loss sharing based Islamic investment returns for all matched-maturities. Due to competitive pressure from a more established conventional banking sector, the returns from Islamic investment deposits follow those of conventional

⁎

deposits. Taking the results to be tantamount to departure from the PLS principle, they conclude that Islamic banks are not interest free. Subsequent analyses by Cevik and Charap (2011); Anuar et al. (2014); and Hamza (2016) further substantiate this conclusion. This finding notwithstanding, the Islamicity of Islamic banks is a non-fading issue and continues to be a topic of intense debate among advocates and doubters of Islamic banking (Ibrahim, 2015). In this paper, we re-assess the pricing behaviour of Islamic banks by allowing potential asymmetric responses of Islamic investment rates to conventional deposit rates. We appeal to bank concentration/ collusion and negative consumer reaction hypotheses to motivate asymmetric pricing behaviour of Islamic banks (Nguyen and Islam, 2010; and Holmes et al., 2015). On one hand, facing changing conventional rates and a given financing rate, Islamic banks may change their investment deposit rates at a slower pace upward and a faster pace downward, as posited by the bank concentration/collusion hypothesis. On the other hand, in anticipation of consumer reaction, they may adopt the reverse pricing strategy, i.e. faster upward changes and slower downward changes in the investments rates. Two key and related aspects of our contribution are notable. First, we provide a methodological improvement to the analysis of Islamic rate – conventional rate relations. If the interest rate movements are really asymmetric, the framework that imposes symmetry is misspecified and would likely lead to incorrect inferences. In the present paper, we apply a novel non-linear ARDL framework developed by Shin

Corresponding author. E-mail addresses: [email protected] (R. Sukmana), [email protected] (M.H. Ibrahim).

http://dx.doi.org/10.1016/j.econmod.2017.02.025 Received 27 February 2017; Accepted 27 February 2017 0264-9993/ © 2017 Elsevier B.V. All rights reserved.

Please cite this article as: Sukmana, R., Economic Modelling (2017), http://dx.doi.org/10.1016/j.econmod.2017.02.025

Economic Modelling xxx (xxxx) xxx–xxx

R. Sukmana, M.H. Ibrahim

deposit rates. The question is: why might Islamic banks behave asymmetrically? The aforementioned reasons underlying asymmetric interest rate movements may characterize the returns of Islamic investment deposits. On one hand, through “religious branding” or by having a captive client base, Islamic banks are likely to have greater market power as compared to conventional banks (Weill, 2011). Accordingly, they may change the investment rates at a slower pace upward and a faster pace downward. On the other hand, Islamic banks also face a displaced commercial risk, i.e. the risk of deposit withdrawals from Islamic banks in responses to higher conventional rates. If Islamic banks factor in consumer reaction in the pricing of their deposit returns, they likely adjust their investment deposit rates faster upward and slower downward. In our context, identifying which pricing behaviour characterizes Islamic banks would be important for assessing the Islamicity of the Islamic banks. We view that latter case, i.e. the faster upward adjustment of investment deposit rates, to be more in line with Islamicity since the customers are key to Islamic banks and profits is not only the concern of Islamic banks.

et al. (2014) to evaluate whether Islamic banks change their investment rates asymmetrically in the long run as well as the short run when faced with changes in the conventional deposit rates. And second, by allowing asymmetry, we provide a more refined assessment of Islamic banks’ “Islamicity”. Our argument is straightforward. Islamic banks are offering similar but Islamic depository services in a market dominated by conventional banks. Accordingly, the market for Islamic deposits would not be completely segmented from the market for conventional deposits. In econometric terms, we may not rule out comovement or causal relations that run from the conventional rates to Islamic rates. However, the asymmetric pattern, if it exists, would reveal whether Islamic banks take actions for their own advantages (bank concentration hypothesis) or in consideration of the consumers (consumer reaction hypothesis). We argue that the latter would be more consistent with the Islamic economics paradigm that “self is secondary to society”, upon which Islamic banking is founded.1 To anticipate the results, we find the pricing of Islamic deposits to be non-linearly related to corresponding conventional deposits in the long run as well as the short run. More specifically, we find that the increase in the investment rates is faster than their reduction when faced with the increase and decrease in the conventional rates of the same magnitude. Interestingly, as the maturity lengthens, the nonlinear or asymmetric pricing behaviour of Islamic banks tends to fade. The remaining of the paper is structured as follows. In the next section, we present the theoretical and empirical framework. Section 3 presents data and estimation results. Section 4 provides a summary of the findings and concluding remarks.

2.2. Empirical framework In existing studies, the dependence of Islamic rates on conventional rates are normally examined by means of the standard time series econometrics: (i) employing cointegration tests to verify the presence of a long run relation between interest rates, (ii) applying such long-run estimators as the OLS, dynamic OLS or fully-modified OLS estimators to estimate a long run equation, if it exists, and (iii) using variants of vector autoregressions (VAR) to discern short-run dynamic interactions among the variables. The approach, however, assumes symmetric relations among the variables. This is too restrictive. Since banks can behave asymmetrically as posited by the aforementioned bank concentration/collusion hypothesis and the consumer reaction hypothesis, allowing for asymmetry would thus be more appropriate. Accordingly, in the present analysis, we employ a novel nonlinear ARDL (NARDL) approach recently developed by Shin et al. (2014) as an asymmetric extension to the well-known ARDL model of Pesaran and Shin (1999) and Persaran et al. (2001). Apart from its simplicity, the approach offers flexibility in that it jointly models long-run relation, short-run dynamics and asymmetries and, in doing so, does not require the variables’ integration order to be the same (Apergis and Cooray, 2015). More importantly, in our context, the nonlinear ARDL enables evaluation of both the long-run and short-run relations between variables. To begin, we start with the following long-run equation:

2. Theoretical and empirical framework 2.1. Theoretical foundations Price asymmetry is a prevalent phenomenon and it is more than a rule than exception (Peltzman, 2000). The literature has documented asymmetric patterns in a variety of prices. These include among others consumer prices (Ibrahim and Chancharoenchai, 2014; Salisu et al., 2017), stock prices (Bahmani-Oskooee and Saha, 2016; Raza et al., 2016), house prices (Katrakilidis and Trachanas, 2012; Tsai, 2013), oilrelated prices (Qin et al., 2016), commodity prices (Reboredo and Ugolini, 2016; Batten et al., 2017), exchange rates (Arghyrou and Pourpourides, 2016) and lending and deposit rates (Nguyen and Islam, 2010; Gambacorta and Iannotti, 2007; Holmes et al., 2015; Apergis and Cooray, 2015). In the banking literature, potential asymmetric movements in lending and deposit rates under changing interest rate environment stem from such factors as market concentration, market collusion, consumer characteristics, and consumer reaction to rate changes. As elaborated in Nguyen and Islam (2010) and Holmes et al. (2015), the bank concentration/collusion hypothesis posits that banks in a more concentrated market are likely to exploit their monopoly power or to collude to their advantages through faster upward movements in the lending rates and slower upward movements in the deposit rates. By contrast, Since consumers are likely to react negatively to higher lending rates and lower deposit rates and the adverse selection problem tends to be intensified when the interest rates increase, the lending rates are likely to more sticky upward while the deposit rates more sticky downwards. In the presence analysis, we focus on the pricing of Islamic investment deposits in responses to changes in the conventional

irt = α0 + α1drt+ + α2drt− + et

(1)

where ir is an Islamic investment rate, dr is a conventional deposit rate, and (α0, α1, α2 ) is a vector of long run parameters to be estimated. drt+ and drt− are partial sums of positive and negative changes in dr: t

drt+ =

t

∑ ∆drt+ = ∑ max (∆dri ,0) i =1

i =1

t

t

(2)

and

drt− =

∑ ∆drt− = ∑ min (∆dri ,0) i =1

i =1

(3)

Based on the above formulation, the long run relation between the Islamic investment rate and the conventional interest rate is α1 and α2 for respectively the increase and decrease in the latter. We have no a priori expectation whether α1 < α2 or α1 > α2 . If the former is true, then the Islamic banks tend to change the investment rate to their own advantages, as posited by the bank concentration/collusion hypothesis. However, if the latter is true, then the consumer reaction hypothesis is likely to shape the behaviour of Islamic banks. While we may conclude that the Islamic banks do not strictly follow the conventional banks in

1 Islamic economics deals with the allocation of scarce resources for individual and collective achievements of spiritual, moral and material well-beings in ways that conform to the prescriptions of Islamic laws. It places collective/society welfare above personal gains. Islamic bank and finance is a practical application of the Islamic economics visions to ensure equity, encourage participation, and protect ownerships. See inter alia Shinsuke (2012) and Aydin (2013).

2

Economic Modelling xxx (xxxx) xxx–xxx

R. Sukmana, M.H. Ibrahim

either case, the former case (i.e. α1 < α2 ) would cast doubt on the Islamicity of Islamic banks. In the empirical implementation, we frame long run Eq. (1) in an ARDL setting as in Shin et al. (2014). That is,

Table 1 Descriptive statistics. Maturity

p

∆irt = β0 + β1irt −1 + β2drt+−1 + β3drt−−1 +

∑ φi∆irt −i

1-Month 3-Month 6-Month 9-Month 12-Month

i =1 q

+

∑ (θi+ ∆drt+−i + θi− ∆drt−−i ) + ut

(4)

i =0

where all variables are as defined above and p and q are lag orders. From (4), we are able not only to uncover the long-run parameters, i.e. Eq. (1), but also to evaluate asymmetric short-run impacts of the conventional deposit rate on the Islamic investment rate. We follows the following four steps, as in Katrakidilis and Trachanas (2012) and Fousekis et al. (2016). First, we estimate Eq. (4) using the OLS. To arrive at the final specification of (4), we apply the general-to-specific approach to sequentially trimming the insignificant lags from the model. In the second step, we perform the ARDL cointegration test for the presence of a long run relation between the two interest rates. This involves the Wald F test of the null hypothesis β1 = β2 = β3=0 (Pesaran et al., 2001) or the t-test of the null hypothesis that β1=0 (Banerjee et al., 1998), denoted respectively as FPSS and tBDM. Third, verifying the presence of a long-run relation, we test for both the long-run and short-run asymmetries. For the long run asymmetry, the null hypothesis is −β2 / β1=−β3 / β1 (i.e. α1 = α2 ). Meanwhile, the corresponding null hypothesis for the short-run asymmetry is q q ∑i =0 θi+ = ∑i =0 θi−. Finally, we also graph the asymmetric cumulative dynamic multiplier effects of a one percentage point change in dr+ and dr − to demonstrate visually the asymmetric Islamic investment rate – conventional rate relation. Namely, h

mh+ =

∑

∂irt + j

∂drt+ j =0

h

,

Note that as h

mh− =

∑ j =0

→∞,

∂irt + j ∂drt− mh+

,

h=0,

→ α1and

1,

Conventional rates

Mean

S.D.

Mean

S.D.

2.923 3.028 3.186 3.324 3.472

0.479 0.500 0.520 0.548 0.590

3.032 3.069 3.122 3.179 3.523

0.450 0.449 0.452 0.473 0.502

Correlation

T-test of Mean Diff

0.868*** 0.876*** 0.888*** 0.876*** 0.729***

−2.430** −0.891 1.368 2.946*** −0.962

Notes: ** and *** indicates significance at 5% and 10% respectively. The mean difference tests allowing for difference in variances yield similar results.

Islamic investment rate is significantly lower and the 9-month average rate is significantly higher than their corresponding conventional rates. By using recent data after the financial crisis, our sample characteristics are different from Chong and Liu (2009). Accordingly, it would be worthwhile to reassess their findings. Before we proceed to the formal analysis using the nonlinear ARDL approach, we report in Table 2 the results from the standard Granger causality test between the Islamic investment rates and conventional deposit rates. In conducting the test, we set the lag orders to alternatively 3, 6, and 12. In line with Chong and Liu (2009), we document strong evidence for causality running from the conventional deposit rates to Islamic investment rates. However, we also uncover evidence that the Islamic investment rates start to exert causal influences on the conventional rates. We find the Islamic rates at shorter maturities to Granger cause conventional rates when the lag order is set to 3. Meanwhile, when we lengthen the lag order, the Islamic rates at longer maturities significantly Granger cause the conventional rates. These results provide a preliminary evidence that Islamic banks are not strictly following the conventional banks. With these findings, we proceed to the formal analysis to evaluate whether the impacts of the conventional rates on Islamic rates are asymmetric.

2, … (5)

mh−

Islamic rates

→ α2 . 3.2. Nonlinear ARDL results

3. Data and results

We estimate the nonlinear ARDL model as in (4) using the OLS. We trim the insignificant lags of the first-differenced terms using the general-to-specific procedure.3 Based on the final nonlinear ARDL model for each maturity-matched pair of Islamic and conventional rates, we test for the presence of cointegration using the tBDM and FPSS test statistics. The results, reported in Table 3, reject the null hypothesis of no cointegration for all maturity-matched rates. Table 4 presents the nonlinear ARDL results. At the bottom of the table, we report the long-run coefficients (i.e. Eq. 1) and long-run asymmetry and short-run asymmetry tests. From the Table, the Wald F test statistic for the null hypothesis of long-run symmetry (i.e. H0 : α1 = α2 ) is significant at better than 5% significance level in all cases except the 12-month rates. The null hypothesis of short run q q asymmetry (H0 : ∑i =0 θi+ = ∑i =0 θi−) is also rejected in all cases except one. From these results, the Islamic investment rates exhibit asymmetric responses to changes in corresponding conventional interest rates. For the 1-month maturity rates, we find the long-run coefficients of the deposit rate increase and the deposit rate reduction to be respectively 1.01 and 0.71. They are significant at a 1% significant level. This means that the 1-month Islamic investment rate increases one-to-one in tandem with the deposit rate. Meanwhile, the reduction in the conventional rate by 1 percentage point is related to the reduction in the Islamic investment rate by only 0.71 percentage point. In short, the Islamic investment rate adjusts at a slower pace to the deposit rate reduction. The short-run results, as captured by the

3.1. Data preliminaries The data for the empirical analysis come from the Monthly Statistical Bulletin published online by Bank Negara Malaysia, i.e. Malaysia's Central Bank. These are monthly Islamic investment rates and conventional deposit rates at 1-month, 3-month, 6-month, 9month and 12-month maturities covering the period January 1999 to November 2016.2 While we have data few years before 1999, we start the sample at January 1999 to avoid the confounding influences of the 1997/1998 Asian financial crisis. In addition, only after the Asian crisis, the Islamic banking sector has witnessed significant growth in terms of number and market share. As preliminary analyses of the data, Table 1 presents descriptive statistics. Several interesting observations emerge from the Tables. While the correlations between the maturity-matched Islamic and conventional rates are high, they are lower than those documented by Chong and Liu (2009). Moreover, variability in the Islamic investment rates is higher than the variability in the conventional rate at all maturities. We find the average Islamic investment rates to be comparable to the average conventional rates. The mean difference tests, assuming equal variances or allowing for different variances, suggest no significant difference between the two rates at 3-month, 6month, and 12-month maturities. However, the 1-month average 2 We also wish to evaluate the financing rates of Islamic banks. However, the data on the financing rates are only available recently and hence too short to render empirical analyses meaningful.

3

3

We set the maximum ARDL lag order to 12.

Economic Modelling xxx (xxxx) xxx–xxx

R. Sukmana, M.H. Ibrahim

Table 2 Pairwise granger causality test. Maturity/Null

Lag Order=3

1-month 3-month 6-month 9-month 12-month

Lag Order=6

Lag Order =12

ir dnc dr

dr dnc ir

ir dnc dr

dr dnc ir

ir dnc dr

dr dnc ir

1.971* 2.022* 3.178*** 1.572 0.843

12.276*** 11.191*** 11.219*** 12.061*** 6.347***

1.603 1.428 1.839 2.998** 3.655**

28.541*** 35.124*** 30.256*** 35.202*** 25.212***

1.482 1.320 3.520*** 2.458*** 0.894

8.439*** 6.539*** 3.745*** 5.106*** 1.490

Notes: *, **, and *** denote significance at 10%, 5% and 1% respectively. “dnc” stands for does not Granger cause, ir = Islamic investment rate, and dr=conventional deposit rate. Table 3 ARDL Cointegration test. Test statistics

Critical values (5%)

Maturity

tBDM

FPSS

Lower

1-month 3-month 6-month 9-month 12-month

−4.954 −5.388 −5.191 −4.499 −4.103

8.508 10.041 9.737 7.403 6.583

tBDM −2.86 FPSS 4.94

Upper

−3.22 5.73

Note: the critical values are from Pesaran et al. (2001). Table 4 NARDL Estimation results. Maturity Variable Constant irt−1 + drt−1 − drt−1

∆irt−1 ∆irt−2 ∆irt−3 ∆irt−5 ∆irt−6 ∆irt−7 ∆irt−8 ∆irt−10 ∆irt−11 ∆irt−12

∆drt+ + ∆drt−1 + ∆drt−2 + ∆drt−3 + ∆drt−7 + ∆drt−12 ∆drt− − ∆drt−1 − ∆drt−2 − ∆drt−3 − ∆drt−4 − ∆drt−9 − ∆drt−10

1-Month ***

3-Month

6-Month

***

12-Month

0.986 −0.216*** 0.217***

0.956 −0.203*** 0.203***

1.163 −0.228*** 0.232***

1.296 −0.258*** 0.222***

0.952 −0.210*** 0.175***

0.154*** −0.259***

0.152***

0.193*** −0.189***

0.195*** −0.299*** −0.153**

0.133*** −0.401*** −0.203*** −0.185***

−0.117**

***

9-Month

***

−0.168***

Fig. 1. Dynamic multipliers, 1-Month maturity.

***

coefficients of lagged first-differenced deposit rates, paint similar conclusion. The results documented for the 3-month and 6-month rates remain similar. As we move to the 9-month rates, we still find the investment rate to be related more strongly to the increase in the corresponding deposit rate than to its reduction, although the difference in the coefficients is narrowed, i.e. 0.862 vs. 0.758. Finally, for the 12-month rates, the evidence for asymmetry is uncovered only in the short run. Figs. 1–5 present graphically these findings by tracing the asymmetric cumulative dynamic multipliers for the 5-matched rates. The graphs represent temporal evolution of the Islamic rates in response to a unit increase and decrease in the corresponding deposit rates over 80-month horizon. The differences in the upward and downward movements together with 90% confidence interval are also presented. As clearly demonstrated by the graphs, the impacts of deposit rate

−0.101* 0.147** 0.126** 0.099*

−0.234*** −0.104**

−0.128*** −0.117*** 0.370***

−0.093** 0.384**

0.354**

0.455***

0.237**

0.298**

0.303*

0.512***

0.327***

0.299*

0.472*

0.204* 0.234*

0.345**

0.201* −0.313*** 0.474*** 0.759*** −0.298*** −0.239**

Adj-R2 0.426 Long-run Asymmetry 1.006*** drt+ 0.714*** drt− WLR 33.830*** Short-run Asymmetry WSR 5.856**

0.282**

0.323**

0.196** −0.214** −0.138* 0.107** 0.385

0.290

0.131* 0.388

0.373

1.001***

1.019***

0.862***

0.832***

0.748*** 40.940***

0.846*** 16.810***

0.758*** 6.351**

0.633*** 2.047

16.330***

6.778***

2.603

3.298*

Note: *, **, and *** denote significance at 10%, 5% and 1% respectively. WLR and WSR stand for Wald test for long-run asymmetry and short-run asymmetry respectively. Fig. 2. Dynamic multipliers, 3-Month maturity.

4

Economic Modelling xxx (xxxx) xxx–xxx

R. Sukmana, M.H. Ibrahim

banks, returns do matter to others. Accordingly, facing rising conventional interest rate, Islamic banks face the risk of deposits flowing to conventional banks. Various studies have provided evidence that Islamic deposits are affected in a significant way by variations in the interest rate (Kassim et al., 2009; Abduh, 2015). This negative reaction from depositors may have resulted in the increase in the investment returns, either as a response of Islamic banks to curtail this so-called displaced commercial risk or to the drop in Islamic deposits. Why do not then investment returns respond as much when the conventional interest rates decline? The displaced commercial risk means that the Islamic banks stand to gain deposits when the interest rates decline. In addition, since business operations of Islamic banks are tied to real activities, they are not obliged to follow strictly the conventional interest rates. Even doing so is justifiable on basis of better margin, they opt for providing better returns to their depositors. As the costs of deposited funds increase with the term of investment deposits, it is natural to observe larger reduction in the investment returns for deposits with longer maturity. This may have explained why the asymmetric pricing behaviour tends to disappear when the maturity increases. In views of these explanations, we infer that Islamic banks do not strictly follow conventional banks and, hence, labelling Islamic banks as interest-based is misplaced. In light of our results, Islamic banks are likely to bring differences to financial and economic stability and to monetary transmission mechanisms, subjects that deserve attention in future research.

Fig. 3. Dynamic multipliers, 6-Month maturity.

4. Conclusion In this paper, we frame the Islamic investment rate – conventional interest rate relations in an asymmetric setting. We find evidence that the two rates form a long run or cointegrating relation and the former moves asymmetrically in relations to changes in the latter. More precisely, the pass-through from the conventional rates to Islamic rates is larger when the conventional rates increase. The cointegration between the Islamic rates and conventional rates is justifiable on the basis that they provide similar depository services and accordingly are not completely segmented. However, the faster upward movements in the Islamic rates would mean that Islamic banks consider consumer reaction in the pricing of their products. In addition, as they reduce their returns at a slower pace, we may infer that profitability is not the only concern of Islamic banks. In other words, Islamic banks do not strictly follow conventional banks in the pricing of their investment deposits. They are quite different.

Fig. 4. Dynamic multipliers, 9-Month maturity.

References Abduh, M., 2015. Determinants of Islamic banking deposit: empirical evidence from Indonesia. Middle East J. Manag. 2 (3), 240–251. Anuar, K., Mohamad, S., Shah, M.E., 2014. Are deposit and investment accounts in Islamic banks in Malaysia interest free? JKAU: Islam. Econ. 27 (2), 27–55. Apergis, N., Cooray, A., 2015. Asymmetric interest rate pass-through in the U.S., the U.K. and Australia: new evidence from selected individual banks. J. Macroecon. 45, 155–172. Arghyrou, M.G., Pourpourides, P., 2016. Inflation announcements and asymmetric exchange rate responses. J. Int. Financ. Mark., Inst. Money 40, 80–84. Ariff, M., 2015. Islamic banks in Malaysia: industry at crossroads. MGCC Perspect. 21 (4), 12–14. Aydin, N., 2013. Redefining Islamic economics as a new economic paradigm. Islam. Econ. Stud. 21 (1), 1–34. Bahmani-Oskooee, M., Saha, S., 2016. Do exchange rate changes have symmetric or asymmetric effects on stock prices? Glob. Financ. J. 31, 57–72. Banerjee, A., Dolado, J., Mestre, R., 1998. Error-correction mechanism tests for cointegration in a single-equation framework. J. Time Ser. Anal. 19 (3), 267–283. Batten, J.A., Ciner, C., Kosedag, A., Lucey, B.A., 2017. Is the price of gold to gold mining stocks asymmetric? Econ. Model. 60, 402–407. Cevik, S., Charap, J., 2011. The behaviour of conventional and Islamic bank deposit returns in Malaysia and Turkey. IMF Working Paper, WP/11/156. Chong, B.S., Liu, M.-H., 2009. Islamic banking: interest-free or interest-based? Pac.Basin Financ. J. 17, 125–144. Fousekis, P., Katrakilidis, C., Trachanas, E., 2016. Vertical price transmission in the US

Fig. 5. Dynamic multipliers, 12-Month maturity.

increase on Islamic rate are relatively larger than the impacts of the deposit rate reduction for the 1-month to 9-month rates over the entire horizons. Meanwhile, for the 12-month rates, asymmetry is apparent only in the short run. The questions are: what may have explained the results? Are Islamic banks distinct from conventional banks? Our finding appears to be in line with the consumer reaction hypothesis. The displaced commercial risk is unique to Islamic banks operating in a dual banking system. While religiosity is the reason for some to deposit in Islamic 5

Economic Modelling xxx (xxxx) xxx–xxx

R. Sukmana, M.H. Ibrahim

the 20th Century: The Ragnar Frisch Centennial Symposium. Cambridge University Press, Cambridge. Pesaran, M.H., Shin, Y., Smith, R.J., 2001. Bounds testing approach to the analysis of level relationship. J. Appl. Econ. 16, 289–326. Qin, X., Zhou, C., Wu, C., 2016. Revisiting asymmetric price transmission in the U.S. oilgasoline markets: a multiple threshold error-correction analysis. Econ. Model. 52, 583–591. Raza, N., Shahzad, S.J.H., Tiwari, A.K., Shahbaz, M., 2016. Asymmetric impact of gold, oil prices and their volatilities on stock prices of emerging markets. Resour. Policy 49, 290–301. Reboredo, J.C., Ugolini, A., 2016. The impact of downward/upward oil price movements on metal prices. Resour. Policy 49, 129–141. Salisu, A.A., Isah, K.O., Oyewole, O.J., Akanni, L.O., 2017. Modelling oil price-inflation nexus: the role of asymmetries. Energy 125, 97–106. Shin, Y., Yu, B., Greenwood-Nimmo, M., 2014. Modelling asymmetric cointegration and dynamic multipliers in an ARDL framework. In: Horrace, W.C., Sickles, R.C. (Eds.), Festschrift in Honor of Peter Schmidt. Springer Science & Business Media, New York. Shinsuke, N., 2012. Critical overview of the history of Islamic economics: formation, transformation, and new horizons. Asian Afr. Area Stud. 11 (2), 114–136. Tsai, I.-C., 2013. The asymmetric impacts of monetary policy on housing prices: a viewpoint of housing price rigidity. Econ. Model. 31, 405–413. Weill, L., 2011. Do Islamic banks have greater market power? Comp. Econ. Stud. 53, 291–306.

beef sector: evidence from the nonlinear ARDL model. Econ. Model. 52, 499–506. Gambacorta, L., Iannotti, S., 2007. Are there asymmetries in the response of bank interest rates to monetary shocks? Appl. Econ. 39 (19), 2503–2517. Hamza, H., 2016. Does investment deposit return in Islamic banks reflect PLS principle? Borsa Istanb. Rev. 16 (1), 32–42. Holmes, M.J., Iregui, A.M., Otero, J., 2015. Interest rate pass through and asymmetries in retail deposit and lending rates: an analysis using data from Colombian banks. Econ. Model. 49, 270–277. Ibrahim, M.H., 2015. Issues in Islamic banking and finance: islamic banks, Shari’ahcompliant investment and sukuk. Pac.-Basin Financ. J. 34, 185–191. Ibrahim, M.H., Chancharoenchai, K., 2014. How inflationary are oil price hikes? A disaggregated look at Thailand using symmetric and asymmetric cointegration models. J. Asia Pac. Econ. 19 (3), 409–422. Kassim, S., Abdul Majid, M.S., Mohd Yusof, R., 2009. Impact of monetary shocks on the conventional and Islamic banks in a dual banking system: evidence from Malaysia. J. Econ. Coop. Dev. 30 (1), 41–58. Katrakilidis, C., Trachanas, E., 2012. What drives housing price dynamics in Greece? New evidence from asymmetric ARDL cointegration. Econ. Model. 29 (4), 1064–1069. Khan, F., 2010. How ‘Islamic’ is Islamic banking? J. Econ. Behav. Organ. 76, 805–820. Nguyen, C.V., Islam, A.M., 2010. Asymmetries in the Thai lending-deposit rate spread: an econometric analysis. Appl. Econ. Lett. 17, 1229–1236. Peltzman, S., 2000. Prices rise faster than they fall. J. Political Econ. 108 (3), 466–502. Pesaran, M.H., Shin, Y., 1999. An autoregressive distributed lag modelling approach to cointegration analysis. In: Storm, S., (Ed.), Econometrics and Economic Theory in

6