The aim of this paper is to provide empirical evidence relating to the factors influencing the compliance of Islamic finance by Indian corporate sector. The sample comprised of the whole population of Shariah companies numbered as 190 consisted in the 500 Shariah and Nifty Shariah indices during the period of 2009 till 2014. However, after making the necessary filtration due to unavailability of data, the actual size of the sample came out to be 136 companies. The relationship between factors and the extent of compliance were analysed using Panel data regression model. The results evident that companies of larger size and higher growth rate have significant mandatory compliance. Whereas, firm size examined with net sales and board independence has withstood with compliance of voluntary and overall measures. The significant implication of our results is that it provided information on firm specific characteristics for the investors who are looking for investment in Shariah compliant companies. In this way investor would be able to keep an eye on their investment. These results may also be advantageous to the regulators in making decisions. Distinct from previous empirical research concerning to Islamic social reporting in Muslim and non-Muslim countries, this study examines the factors affecting the extent of Shariah compliance by the companies listed under Shariah Index in Indian stock exchange.
Shariah screening process is comparatively a new entrant in the Islamic finance domain [
Although there are myriad studies examining the conventional social reporting followed by voluntary disclosures made by the companies available in abundance in the national and international sphere. Succeeded by, very few of the studies with respect to Islamic social reporting and the Shariah compliance of the companies are found in the international arena, however, to the best of my knowledge, the level of compliance of Shariah principles made by the companies under different industries has been overlooked yet in India. One of the important issue that has been often talked about is the determinants influencing the conventional social reporting, voluntary disclosures and Islamic social reporting. The various firm specific characteristics are recognised in the prior studies which effect conventional social reporting, mandatory and voluntary disclosures and Islamic social reporting but could those factors impact the Shariah compliance of the companies have not been discussed. So, the empirical purpose of this paper is to scrutinize if those factors could influence the extent of Shariah compliance or not.
The paper is organised as follows. The second section briefly discusses the theoretical background of various determinants and help to develop the hypothesis. The third section explores the econometric model with special reference to Panel regression to consider both the dimension time and cross-section of the data. The fourth section reports the results of empirical analysis. The fifth section talks about concluding remarks which contains implication and future scope of the study.
In this section, researchers explained different independent variables and how these variables are related to dependent variable (Disclosure of Shariah compliance).
Disclosure of Shariah compliance index is the dependent variable. The study employed content analysis tool to examine the extent of Shariah compliance by the different companies that have been a generally acceptable method of coding used in the prior studies of disclosure [
This study considered several factors that could contribute to the extent of Shariah compliance by the different companies based on prior studies have been discussed below. Results of the previous studies revealed that firm size, profitability, board composition, leverage, nature of the industry, age along with increase in independent directors influenced the extent of Islamic social reporting and voluntary disclosure [
Company size would influence the decision-making of the company while disclosing the information in the annual report [
H1: There is a significant relationship between size of the company and extent of Shariah compliance.
The previous studies ascertained relationship between financial leverage and the extent of disclosure [
H2: There is a significant relationship between leverage of the company and extent of Shariah compliance.
It is anticipated that companies with the higher growth opportunities incline to make effort on improving the disclosure level [
H3: There is a significant relationship between growth of the company and extent of Shariah compliance.
The prior studies make it evident that profitability has the capability of affecting the extent to which companies reveal mandatory information in the annual reports [
H4: There is a significant relationship between profitability of the company and extent of Shariah compliance.
Empirical evidence with regard to company age and voluntary disclosure found an association between them [
H5: There is a significant relationship between age of the company and extent of Shariah compliance.
Board size can influence the extent of disclosure, controlling and monitoring process indicated by various empirical studies in corporate governance [
H6: There is a significant relationship between board size of the company and extent of Shariah compliance.
It is considered that many of the corporate governance problems can be solved with the presence of Independent directors on the board of the company. Independent directors on the board performed an oversight role on management on shareholder’s part and also believed to be as a potential solution to the problems. They can supervise management efficaciously as per the assumed facts because they do not require a disposition to good graces of management, and can express their opinion openly and without any hesitation or fear in front of management misdeeds, outside and inside the boardroom, with the objective of shielding the interests of shareholders [
H7: There is a significant relationship between board independence of the company and extent of Shariah compliance.
The study included industry type as a predictor variable. It is argued that the disclosure practices of the firm are not expected to be same across varied industry due to the variations may occur in accounting policies and practices of the companies [
H8: There is a significant relationship between nature of industry of the company and extent of Shariah compliance.
A panel data regression model has been employed to study the factors that could influence the Shariah compliance of the companies. This model has been preferred over the multiple regressions because it can handle the datasets containing both cross-sections and time period observations. A panel data regression can be measured with the help of Fixed Effect Model (FEM) or Random effect model (REM). But this study rests on REM model due to the inclusion of some sector specific dummy variables and precluded the application of FEM model. Here is the glimpse of a panel data regression model below:
Panel data analysis examines the particular subject within multiple sites over a specified time form which has been observed periodically. However, combining the cross-sections with time series can make the quantity and quality of data better in the way that it would be inconceivable to employ only one of these attributes [
There are three competing formulations according to
MODEL y i t = α i t + X ′ i t β + u i t | INTERCEPT TERM | DISTURBANCE TERM |
---|---|---|
Pooled Model | α i t = α | Uit |
One-Way Fixed Effects | α i t = α + μ i | Uit |
Two-Way Fixed Effects | α i t = α + μ i + λ t | Uit |
One-Way Random Effects | randomly changing over i | u i t = μ i + v i t |
Two-Way Random Effects | randomly changing over i | u i t = μ i + λ t + v i t |
constant coefficients and implies to both slopes and intercepts that are neither significant to cross-section nor significant temporal effects rather pool all the data and run an ordinary least squares regression model. The pooled model essentially postulates that both the intercept and the slope coefficients are the same across individual units and time.
In order to apply GLS, we need to calculate θ by utilizing the Ω matrix:
θ = 1 − σ v 2 T σ u 2 + σ v 2
*If θ = 0 , run pooled OLS regression. If θ = 1 and σ v 2 = 0 , then run the within effect model.
In order to run an OLS, we first need to transform the variables as below:
α * = 1 − θ
x i t * = x i t − θ x ¯ i for all X k
y i t * = y i t − θ y ¯ i
Now, on the transformed variables we can run an OLS.
y i t * = α * + x i t * ′ β * + ε i t *
In order to apply FGLS, the first thing is to estimate θ with the help of σ ^ v 2 and σ ^ u 2 .
θ ^ = 1 − σ ^ v 2 T σ ^ u 2 + σ ^ v 2 = 1 − σ ^ v 2 T σ ^ between 2
The σ ^ v 2 comes from sum of squared errors (SSE) of the “within effect estimation” or the deviations of residual from group means of the residuals:
σ ^ v 2 = S S E within n T − n − k = e ′ e within n T − n − k = ∑ i = 1 n ∑ t = 1 T ( v i t − v ¯ i ) 2 n T − n − k
where v i t represents the residuals of LSDV.
The σ ^ u 2 is derived from group mean regression (between effect estimation):
σ ^ u 2 = σ ^ between 2 − σ ^ v 2 T , where σ ^ between 2 = S S E between n − K
Now, transform the variables by utilizing θ ^ and then we can perform an OLS:
y i t * = α * + x i t * ′ β * + ε i t *
α * = 1 − θ ^
x i t * = x i t − θ ^ x ¯ i for all X k
y i t * = y i t − θ ^ y ¯ i
A two-way random effect model is when there is within effect error component in both the time series and cross-section in a random effect model. In this situation, the error term should be uncorrelated with both group (cross-sectional) error and time series component. The orthogonality of both the component permits the general error to be decomposed within the cross-sectional temporal, specific and individual error components.
e i t = v i + e t + η i t
Here, e t represents the time specific component. This e t is peculiar to all observations for that time period, t. v i represent the cross-section specific error. This component affects only those observations which are in that panel. Whereas, particular observation in the panel is affected by η i t . These kind of models are referred to as two-way random effects model (SAS, 1999).
Testing for the existence of cross-section (individual) and time effects is important in panel and pool regression settings since accounting for the presence of these effects is necessary for correct specification of the regression and proper inference. Eviews offers testing for individual and time effects using both F-statistic (likelihood ratio) and Lagrange multiplier (LM) tests. The F-statistic test is used in case of fixed effects model whereas, Lagrange multiplier (LM) test is used in case of a random-effects model.
The most popular random effects test is the [
[
L M u = n T 2 ( T − 1 ) [ e ′ D D e e ′ e − 1 ] 2 = n T 2 ( T − 1 ) [ T 2 e ¯ ′ e ′ e ′ e − 1 ] 2 ~ x 2 (1)
where, e ′ e represents the Sum of Squares due to Error (SSE) of the pooled model (OLS regression model), and e ¯ represents the n × 1 vector of the group means of pooled regression residuals.
[
L M u = n T 2 ( T − 1 ) [ ∑ ( ∑ e i t ) 2 ∑ ∑ e i t 2 − 1 ] 2 = n T 2 ( T − 1 ) [ ∑ ( T e ¯ i ) 2 ∑ ∑ e i t 2 − 1 ] 2 ~ x 2 (1)
If we accept the alternative hypothesis then it means that a random effects model is more relevant and efficient in handing the heterogeneity in the model better than a pooled OLS model. In a two-way random effects model, the null hypothesis is H 0 : σ ^ u 2 = 0 and σ ^ v 2 = 0 . In another words, a Lagrange Multiplier (LM) test combines two one-way random effects model for time and group, that is,
L M u 12 = L M u 1 + L M u 2 ~ x 2 ( 2 ) .
When the study includes a number of predictor variables the problem of multicollinearity arises. Consequently, it becomes hard to ascertain the impact of each of the predictor variables on the response variable [
VIF = 1 1 − R 2
Several authors have been given recommendations for the acceptable extent of VIF. A value of utmost 10 level of VIF has been most commonly recommended [
The presence of multicollinearity in the data has been determined by using Variance Inflation Factor (VIF).
Variable | Coefficient Variance | Uncentered VIF | Centered VIF |
---|---|---|---|
C | 16.68750 | 135.2439 | NA |
Board Independence | 8.102560 | 13.40537 | 1.033642 |
Growth | 0.000647 | 2.231818 | 1.136202 |
Leverage | 1.699194 | 1.875145 | 1.269834 |
Age | 0.366728 | 37.38866 | 1.096212 |
Director | 2.294798 | 102.5220 | 1.409567 |
Market Capitalisation | 0.297428 | 291.9034 | 6.401163 |
Net Sales | 0.213391 | 177.0659 | 4.241955 |
Total Assets | 0.530402 | 462.3138 | 9.535807 |
Return on Capital Employed | 0.002373 | 11.45850 | 4.132267 |
Return on Assets | 51.42381 | 16.64003 | 3.908662 |
Source: Compiled by Author.
Variables | Mean | Median | Max | Min | S.D | Skewness | Kurtosis |
---|---|---|---|---|---|---|---|
Mandatory Compliance | 92.15 | 100.00 | 100.00 | 50.00 | 12.92 | −1.35 | 3.83 |
Voluntary Compliance | 69.77 | 70.00 | 84.00 | 49.00 | 5.63 | −0.25 | 2.88 |
Overall Compliance | 71.71 | 72.00 | 85.00 | 51.00 | 5.30 | −0.17 | 2.92 |
Net sales | 9.99 | 9.91 | 15.20 | 2.82 | 1.56 | −0.02 | 4.44 |
Total Assets | 10.26 | 10.02 | 15.11 | 7.04 | 1.49 | 0.67 | 3.23 |
Market Capitalisation | 10.88 | 10.65 | 15.24 | 6.86 | 1.63 | 0.29 | 2.65 |
Return on Assets | 0.17 | 0.16 | 0.62 | −0.28 | 0.09 | 0.68 | 5.42 |
Return on Capital Employed | 19.51 | 17.2 | 93.68 | −66.79 | 14.66 | 1.20 | 7.83 |
Growth | 14.45 | 13.3 | 93.36 | −55.12 | 14.73 | 0.44 | 6.71 |
Leverage | 0.20 | 0.06 | 2.07 | 0.00 | 0.30 | 2.18 | 8.96 |
Age | 3.49 | 3.49 | 4.56 | 0.69 | 0.60 | −0.58 | 3.62 |
Board Size | 2.33 | 2.30 | 3.25 | 1.09 | 0.27 | −0.01 | 3.18 |
Board Independence | 0.43 | 0.44 | 0.83 | 0.00 | 0.12 | −0.26 | 3.68 |
Source: Compiled by Author.
from −1.35 (mandatory compliance) to 2.18 (leverage) which is falling under the accepted critical values of ±3 for skewness and the kurtosis values ranges from 2.65 (market capitalisation) to 1.83 (return on capital employed) which again falls under the accepted critical values ±8 [
On the basis of foregoing discussion, the following model has been presented to assess the factors that could influence the Shariah compliance of the companies. The following models are employed to test H1 - H8:
E x t e n t o f C o m p l i a n c e ( Model I ) i , t = α + β 1 L n S a l e s i , t + β 2 L n T A i , t + β 3 L n M a r k e t c a p i , t + β 4 R O A i , t + β 5 R O C E i , t + β 6 G r o w t h i , t + β 7 L e v e r a g e i , t + β 8 L n A g e i , t + β 9 L n B o a r d s i z e i , t + β 10 B o a r d i n d e p e n d e n c e i , t + β 11 N a t u r e o f i n d u s t r y i , t + ε i , t (4.1)
E x t e n t o f C o m p l i a n c e ( Model II ) i , t = α + β 1 L n S a l e s i , t + β 2 L n T A i , t + β 3 L n M a r k e t c a p i , t + β 4 R O A i , t + β 5 R O C E i , t + β 6 G r o w t h i , t + β 7 L e v e r a g e i , t + β 8 L n A g e i , t + β 9 L n B o a r d s i z e i , t + β 10 B o a r d i n d e p e n d e n c e i , t + β 11 N a t u r e o f i n d u s t r y i , t + ε i , t (4.2)
E x t e n t o f C o m p l i a n c e ( Model III ) i , t = α + β 1 L n S a l e s i , t + β 2 L n T A i , t + β 3 L n M a r k e t c a p i , t + β 4 R O A i , t + β 5 R O C E i , t + β 6 G r o w t h i , t + β 7 L e v e r a g e i , t + β 8 L n A g e i , t + β 9 L n B o a r d s i z e i , t + β 10 B o a r d i n d e p e n d e n c e i , t + β 11 N a t u r e o f i n d u s t r y i , t + ε i , t (4.3)
where,
α = constant term, β = slope of the explanatory variables, LnSales = Net sales, LnTA = Total assets, LnMarket cap = Market capitalisation, ROA = Return on assets, ROCE = Return on capital employed and ε = error term.
The aforementioned regression equation is calculated for their parameters α and β by employing panel regression. The unit root test (ADF) applied for all the variables evident stationarity of the panelled data. Then proceed further for estimating Breusch-Pagan Lagrange Multiplier test in order to check the pertinence of the panel data analysis over pooled data analysis.
The foregoing results failed to notice the panel effect; therefore, they need to be taken care cautiously. For the matter of examining factors influencing the extent of Shariah compliance, a panel data regression needs to be applied. Hence, the relevance of panel regression analysis over pooled analysis has been ascertained using Breusch-Pagan Lagrange Multiplier test.
Variable | Statistics | Mandatory | Voluntary | Overall |
---|---|---|---|---|
Constant | Coefficient Std. Error | 111.266*** 5.642 | 67.117*** 2.989 | 70.779*** 2.776 |
Net sales | Coefficient Std. Error | 6.566*** 0.532 | 0.586** 0.282 | 1.080*** 0.262 |
Total Assets | Coefficient Std. Error | −6.616*** 0.913 | −0.483n.s 0.484 | −0.982** 0.449 |
Market capitalisation | Coefficient Std. Error | −0.530n.s 0.592 | 0.903*** 0.314 | 0.765*** 0.291 |
Return on Assets | Coefficient Std. Error | −23.599*** 7.730 | 1.390n.s 4.094 | −0.585n.s 3.803 |
Return on Capital Employed | Coefficient Std. Error | 0.087** 0.051 | −0.036n.s 0.027 | −0.026 0.025 |
Growth | Coefficient Std. Error | 0.054** 0.025 | −0.044*** 0.013 | −0.036*** 0.013 |
Leverage | Coefficient Std. Error | −16.245*** 1.360 | −0.040n.s 0.721 | −1.405** 0.669 |
Age | Coefficient Std. Error | 0.635n.s 0.675 | −0.251n.s 0.358 | −0.154n.s 0.332 |
Board size | Coefficient Std. Error | −1.930n.s 1.575 | −2.692*** 0.834 | −2.579*** 0.775 |
Board Independence | Coefficient Std. Error | 7.937*** 3.027 | 8.995*** 1.604 | 9.034*** 1.489 |
Industrial Manufacturing | Coefficient Std. Error | −6.990n.s 4.391 | −2.905n.s 2.326 | −3.362n.s 2.160 |
Consumer Goods | Coefficient Std. Error | −7.004n.s 4.314 | −3.874* 2.285 | −4.186** 2.122 |
Financial Services | Coefficient Std. Error | −0.990n.s 5.801 | −2.051n.s 3.073 | −2.013n.s 2.854 |
Fertilisers & Pesticides | Coefficient Std. Error | −5.354n.s 4.740 | −2.612n.s 2.511 | −2.805n.s 2.332 |
Media & Entertainment | Coefficient Std. Error | −4.510n.s 4.885 | −10.860*** 2.588 | −10.234*** 2.403 |
automobile | Coefficient Std. Error | −4.775n.s 4.407 | −3.480n.s 2.334 | −3.600* 2.168 |
Cement | Coefficient Std. Error | −4.012n.s 4.765 | −5.470** 2.524 | −5.359** 2.344 |
Chemicals | Coefficient Std. Error | −6.106n.s 4.842 | −0.582n.s 2.565 | −1.095n.s 2.382 |
Construction | Coefficient Std. Error | −12.495** 4.944 | −1.959n.s 2.619 | −2.895n.s 2.432 |
Energy | Coefficient Std. Error | −7.413n.s 4.508 | −1.639n.s 2.388 | −2.098n.s 2.218 |
Healthcare | Coefficient Std. Error | 2.286n.s 5.937 | −4.224n.s 3.145 | −3.856n.s 2.921 |
IT | Coefficient Std. Error | −8.352* 4.428 | −4.551* 2.346 | −4.888** 2.178 |
---|---|---|---|---|
Metal | Coefficient Std. Error | −8.667* 4.562 | −5.353** 2.416 | −5.716** 2.244 |
Pharmaceutical | Coefficient Std. Error | −10.699** 4.360 | −4.731** 2.309 | −5.311** 2.145 |
Services | Coefficient Std. Error | −4.632n.s 4.474 | −2.525n.s 2.370 | −2.764n.s 2.201 |
Adjusted R-Squared | 0.414633 | 0.137583 | 0.158648 | |
F-statistic | 24.09153 (0.000000) | 6.200756 (0.000000) | 7.147150 (0.000000) |
Note: ***p < 0.01; **p < 0.05; *p < 0.10; n.s: p > 0.05; Source: Compiled by Author.
Test Hypothesis | |||
---|---|---|---|
Cross-section | Time | Both | |
Model I | 386.6998 (0.0000) | 0.000622 (0.9121) | 386.7004 (0.0000) |
Model II | 86.53852 (0.0000) | 4030.081 (0.0000) | 4116.619 (0.0000) |
Model III | 90.67027 (0.0000) | 4303.096 (0.0000) | 4393.767 (0.0000) |
Reject the null hypothesis 5%; Source: Compiled by Author.
Further, the pre-testing of model ruled out the fixed effect model due to the inclusion of sector specific dummy variable and fixed-effect model itself known as dummy model. Therefore, we proceed with the random effect model. Further, a test of homogeneity is conducted to ascertain either cross-section or time period or both effects exist in the model. The results revealed that only cross-section effect is present in case of model I consisting mandatory parameters. On the contrary, both cross-section and time period effects are existing subject to model II and model III comprising voluntary and overall parameters, respectively. Thus, cross-section (one-way variable intercept model) REM test for mandatory parameters and cross-section and time period (two-way variable intercept model) REM test has been employed to calculate the panel effects. This study consists of short panel and small sample performance might be inconsistent with the large sample estimators that make [
Variable | Statistics | Mandatory | Voluntary | Overall |
---|---|---|---|---|
Constant | Coefficient Std. Error | 105.523*** 11.930 | 67.861*** 4.443 | 71.444*** 4.268 |
Net sales | Coefficient Std. Error | 4.729*** 0.887 | 0.726** 0.350 | 1.159*** 0.330 |
Total Assets | Coefficient Std. Error | −5.352*** 1.267 | −0.175 0.533 | −0.696 0.499 |
Market capitalisation | Coefficient Std. Error | 0.564n.s 0.609 | −0.286n.s 0.320 | −0.295n.s 0.294 |
Return on Assets | Coefficient Std. Error | −11.399n.s 7.051 | 6.071n.s 3.278 | 4.423n.s 2.992 |
Return on Capital Employed | Coefficient Std. Error | 0.034n.s 0.051 | −0.021n.s 0.023 | −0.015n.s 0.021 |
Growth | Coefficient Std. Error | 0.044** 0.021 | −0.008n.s 0.010 | −0.002n.s 0.009 |
Leverage | Coefficient Std. Error | −17.886*** 1.579 | 0.240n.s 0.696 | −1.232* 0.642 |
Age | Coefficient Std. Error | 2.273n.s 1.478 | −0.599n.s 0.526 | −0.414n.s 0.507 |
Board size | Coefficient Std. Error | −3.316* 1.913 | 0.302n.s 0.865 | 0.167n.s 0.800 |
Board Independence | Coefficient Std. Error | 0.694n.s 2.983 | 3.378** 1.412 | 3.322** 1.293 |
Industrial Manufacturing | Coefficient Std. Error | −7.201n.s 10.199 | −0.503n.s 3.456 | −1.038n.s 3.344 |
Consumer Goods | Coefficient Std. Error | −6.967n.s 10.127 | −1.235n.s 3.419 | −1.634n.s 3.311 |
Financial Services | Coefficient Std. Error | −3.533n.s 13.892 | −0.252n.s 4.655 | −0.398n.s 4.512 |
Fertilisers & Pesticides | Coefficient Std. Error | −5.882n.s 11.136 | −0.425n.s 3.767 | −0.697n.s 3.648 |
Media & Entertainment | Coefficient Std. Error | −2.879n.s 11.436 | −9.309** 3.856 | −8.617** 3.733 |
automobile | Coefficient Std. Error | −4.753n.s 10.366 | −0.906n.s 3.504 | −1.100n.s 3.394 |
Cement | Coefficient Std. Error | −4.295n.s 11.169 | −2.118n.s 3.788 | −2.095n.s 3.668 |
Chemicals | Coefficient Std. Error | −6.571n.s 11.475 | 0.789n.s 3.865 | 0.231n.s 3.744 |
Construction | Coefficient Std. Error | −15.003n.s 11.543 | 0.746n.s 3.905 | −0.348n.s 3.780 |
Energy | Coefficient Std. Error | −8.851n.s 10.516 | 0.946n.s 3.576 | 0.410n.s 3.461 |
---|---|---|---|---|
Healthcare | Coefficient Std. Error | 3.251n.s 13.990 | −1.562n.s 4.713 | −1.196n.s 4.565 |
IT | Coefficient Std. Error | −8.876n.s 10.356 | −1.826n.s 3.504 | −2.251n.s 3.392 |
Metal | Coefficient Std. Error | −11.881n.s 10.615 | −2.628n.s 3.616 | −3.145n.s 3.499 |
Pharmaceutical | Coefficient Std. Error | −12.160n.s 10.219 | −1.363n.s 3.461 | −2.112n.s 3.351 |
Services | Coefficient Std. Error t-Statistic Prob. | −5.130n.s 10.525 | −0.421n.s 3.552 | −0.711n.s 3.438 |
Cross-Section Random | {9.332386} [0.6294] | {2.985032} [0.2768] | {2.951882} [0.3034] | |
Period Random | _ | {3.414900} [0.3623] | {3.217414} [0.3605] | |
Idiosyncratic Random | {7.161847} [0.3706] | {3.408263} [0.3609] | {3.106601} [0.3361] | |
Adjusted R-Squared | 0.239735 | 0.030439 | 0.047920 | |
F-statistic | 11.27977 (0.000000) | 2.023455 (0.002294) | 2.640805 (0.000026) |
Note: ***p < 0.01; **p < 0.05; *p < 0.10; n.s: p > 0.05{} denoted S.D; [
Similarly, the results of analysis with regard to model II on the basis of sub-section as voluntary index reports that net sales (0.038) and board Independence (0.017) are the only significant variables found, than others in the model. But the coefficient of media & entertainment industry is negative (−9.309) and associated p value (0.016). Furthermore, the estimates of σ μ (cross-section) are observed to be 2.9 and the estimate of σ λ (time period) is 3.4 and the estimate of σ ν (idiosyncratic random) is 3.4. It signifies that the variance of the cross-section effects is 27.6% of the total variance, whereas the variance of the time effects is 36.2% and the 36.1% comprises of the remainder effects variance out of the total variance.
Followed by the empirical results of previous one, model III such as overall index indicates that net sales (0.001) and board independence (0.010) are observed to be statistically significant. Whereas, media & entertainment industry is ascertained with negative coefficient (−8.617) and significant p value (0.021) which demonstrated an inverse relationship. These results are also consistent with the findings of second model namely voluntary index. The figure of σ μ (cross-section) is found to be 2.9% and σ λ (time period) is 3.2 and the figure of σ ν (idiosyncratic random) is 3.1. It indicates that the variation in the cross-section effects is 30% followed by time effects is 36% of the total variance, however the variance of the remainder effects is 33.6% from the total variance. The findings of the present study are corroborated by the results of some previous studies that the firm size as measured by their total assets significantly influences the extent of Islamic social reporting [
Islamic finance offers investment opportunities for the investors similar to conventional counterparts. However, Islamic finance differs with regard to compliance of Shariah principles from mainstream counterparts. The extent of compliance of Shariah principles could vary from one company to another company, nevertheless the firm-specific attributes tend to influence Shariah compliance by the companies.
Therefore, an attempt has been made to evaluate the relationship between company specific attributes and variation in Shariah compliance of the companies. In the last few decades there are several studies examining relationship between corporate specific attributes and the level of disclosure have been increasingly witnessed, overlooking the disclosure of Shariah compliance. Owing to this fact, the present study is motivated to examine the various factors chosen could influence the Shariah compliance of the companies. The result of this study reveals that the Shariah compliance of the companies is influenced by firm size measured by net sales in all the three models. Hence, it appears that large company has a tendency to share more information in order to reduce the agency cost [
From the above discussion it is evident that the companies with large size and higher growth rate have significant mandatory compliance. Whereas, firm size measuring with net sales and board independence has significant voluntary, subsequently overall compliance. The findings of this study are having significant implications as it provided information on firm specific characteristics for the investors who are looking for investment in Shariah compliant companies. In this way, investor would be able to keep an eye on their investment. These results may also be advantageous to the regulators in making decisions. Future research should be conducted to examine the other factors which might have been overlooked to consider in this study and could influence the Shariah compliance of the companies.
The authors declare no conflicts of interest regarding the publication of this paper.
Nobi, K., Singh, M. and Aggarwal, A. (2019) Corporate Attributes Influencing the Compliance of Islamic Finance: Evidence from Listed Companies under Shariah Index in India. Theoretical Economics Letters, 9, 1744-1771. https://doi.org/10.4236/tel.2019.96112