Sustainability Reporting on Financial Performance of Sri Lankan Listed Companies: How Strong is the Impact?

: This study seeks to examine and identify the impact of sustainability reporting (SR) on the financial performance (FP) of listed companies in Sri Lanka. Data were collected from annual reports of listed companies where sustainability is disclosed in line with the Global Reporting Initiative (GRI) framework. A quantitative approach was used to gather relevant data of the entire population of 289 companies listed on the Colombo Stock Exchange, representing 20 different market sectors. Following purposive sampling, 55 listed companies were ultimately considered, based on the report preparation of those companies having been in line with the GRI framework from 2015/16 to 2018/19. SR was gathered through content analysis based on G4 and GRI standards and measured using an SR index. The authors measured FP using return on assets (ROA). After collecting the data, the authors analyzed it with panel data regression. Findings revealed an insignificant negative impact of SR on the FP of the listed companies in Sri Lanka. Further, researchers identified the impact of each disclosure of SR on FP and identified mixed results. Sustainability Reporting on Financial Performance of Sri Lankan Listed Companies: How Strong is the Impact?


INTRODUCTION
In the contemporary corporate world, sustainability reporting (SR) is an essential subject matter and is considered one of the crucial challenges globally (Mudiyanselage, 2018). SR can be referred to as a voluntary endeavor that involves publishing accounts that reflect an organization's economic, environment, and social performance (Kumar et al., 2018). Inevitably, the actions of companies create a direct or indirect impact on the aspects of profits, planet, and people (Anbarasan & Sushil, 2018). Thus, companies need to exemplify that their considering the entire set of companies listed in Colombo Stock Exchange (CSE) in this particular research area to identify the impact of SR on FP. This emphasizes the knowledge gap in the SL context; hence, conducting this research study is of significance in bridging the found gap that currently exist in terms of empirical evidence. As such, this research was conducted with the ultimate objective to identify the impact of SR on FP in public listed companies in Sri Lanka.
The firm's perspectives on social and environmental sensitivity have been affected by the increased global awareness of social and environmental issues in recent times (Jayanti & Rajeev Gowda, 2014;Weng et al., 2015;Lestari et al., 2019). Hence, as a means of communicating one's commitment towards social and environmental elements, companies began to use corporate reporting package. With this, emerging of a new paradigm in corporate reporting can be identified as SR. Companies deliver information through sustainability reports that show company's management's high commitment to lead the business in a more sustainable manner (Kurniawan, 2018). Currently, as a complement to financial reports, sustainability reports have become a necessity for corporate stakeholders (Braam & Peeters, 2018;Riduwan & Andajani, 2019). Still, many companies tend to disclose their sustainability efforts through the traditional financial reporting package (Ghani et al., 2018;Rezaee & Tuo, 2019).
Despite sustainability reporting being a voluntary endeavor, it was noted by Elijido-Ten & Clarkson (2019) that most companies tend to benefit from sustainability disclosures. The researchers found out that SR will capture both financial and non-financial information that can be communicated to stakeholders. This nonfinancial information will ultimately allow the companies to boost the corporate image by attracting investors and supporting the company by considering a broader view of performance. Similar insights were noted in Adams & McNicholas (2007); Cohen et al. (2012), highlighting that disclosure on sustainability may favorably influence the firm's performance and enhance its efficiency and effectiveness. Further benefits of sustainability reporting were also highlighted in prior studies that have contributed for organizations to improve their engagement in sustainability; these include: reduction in direct costs, improved worker productivity, reduction in management risk, and improving the competitive image of the firm (Albu et al., 2011). This is to gain corporate reputation, enjoy tax benefits, gain employee's commitment and minimize cost and diversify company risk (Ali & Rizwan, 2013).
These favorable influences perhaps may have inspired the organizations to engage in SR actively. According to the statistical evidence proved by (Ali & Rizwan, 2013), it was emphasized that at the end of 2016, 95% of two hundred fifty (250) most prominent companies in the world report on sustainability compared to a mere 45% at the end of 2002. In doing so, these corporations disclosure their sustainability practices by adhering to globally accepted standards such as the Global Reporting Initiative (GRI). It was noted by the (Global Sustainability Standards Board, 2016) that in adapting GRI as a reporting structure, Particularly, developed countries tend to prefer this over other frameworks.
Even though there is a rising trend in sustainability reporting, the level of disclosures was noted to be unequal to one another. At the same time, developed nations seem to give more prominence to SR, while the developing world was noted to lack in adapting SR in their reporting structure (Dissanayake et al., 2016). Various researchers are still investigating the reasons behind this deficiency in SR. This is despite the favorable impacts SR can create on a company about how significantly these can be reflected in the FP were one aspect that is scarcely studied. The studies that have shed light on this topic so far have made mixed or inconclusive results (Aggarwal, 2013).
Similar conclusions were identified in research conducted in the South Asian region, where it was observed that very few corporations widely do SR in India. Mainly the companies performing in oil, minerals, cement, automotive, and pharmaceuticals industries have taken initiatives in SR, according to the study findings by Daizy & Das (2013). It can be argued that manufacturing sector is usually assumed to be hazardous and so the need to act responsibly, thus the prominence on responsible operations including SR is given priority. Similarly, studies done in Malaysia alone have identified the same supposition. Further, it was observed that no specific reports are being prepared for sustainability by many companies, but instead, the disclosures were communicated to stakeholders through annual reports only (Brammer et al., 2006). Further investigated, SR by banks listed on the Dhaka Stock Exchange (DSE) has concluded that the banks' social disclosure of sustainability, human rights sectors, and product liability are briefly addressed (Garcia-Castro et al., 2010). This indicates that the SR is lacking, and the disclosed information also has room for further development.
Previous research also shows that SR varies across these countries which is attributed to their national business systems; this may have contributed to sustainability reports being varied across the nations. In addition, SR practices differ from one company to another because each company has different characteristics and faces various situations that influence companies to determine responsibilities (Othman & Ameer, 2009). However, it should also noted that company characteristics may not be the only factor influencing sustainability practices. Pressures from external stakeholders, government mandates can also affect the sustainable practices of companies in both developed and developing countries.
Sri Lanka needs to adhere to an accepted standard guideline. Thus, the common standard procedure can be reduced inconsistency to a greater extent. The reason is there are few corporations follow extensive SR in Sri Lanka. They have examined the relationship between SR and the FP of public listed companies in Sri Lanka. The linear study conducted by Wijesinghe (2012) over five years (2005)(2006)(2007)(2008)(2009)(2010) identified trends in CSR in Sri Lanka. It also supports explorations that predict a rising trend in SR among the publicly listed companies in Sri Lanka (Sawani et al., 2010). SR was measured using KPIs of sustainability, whereas annual reports, sustainability reports and websites were used to measure the FP. The result evidenced that; company size was the most significant factor which affected sustainability KPIs based on sixty (60) public listed companies in Sri Lanka. Furthermore, they discovered that large-scale corporations tend to disclose a high level of sustainability disclosures and enjoying benefits from the disclosures.
According to the study conducted by Beddewela & Herzig (2013) to examine the barriers and challenges to enablers of social reporting in Sri Lanka. In depth, interviews have been conducted with eighteen (18) toplevel management to determine the reasons for the low level of social disclosures in the Sri Lankan context. In the post-conflict Sri Lankan civil war period, economic growth is identified as the competitive tool for sustainability practices in most publicly listed companies in Sri Lanka. But there were some deviations among different industry sectors on SR (Weber et al., 2008). Furthermore, those reported sustainability disclosures were not included under constant reporting among various companies. As such, the need for a universally accepted framework for reporting sustainability practices is essential for Sri Lanka. A limited number of past literature on Sri Lankan context regarding SR in Sri Lanka that aligns with a globally accepted framework. Most of the studies are focused on how the sustainability disclosures were reported but not the financial impact from the sustainability disclosures. Therefore, the researchers believe that the findings will contribute to fill in the research gap in literature. Therefore, the study was conducted on measuring the impact of SR on the financial performance (FP) of listed by the subject study considering the sustainability disclosure based on the GRI framework effect to the FP of listed companies in Sri Lanka.

METHODS
The study was quantitative in nature and the researchers have selected fifty-five (55) companies from the two hundred and eighty-nine (289) Public Listed Companies in Colombo Stock Exchange (CSE) as 30th June 2020. Purposive sampling utilized in sample lection with the following criterion in check: 1) Companies should be Listed in the Colombo Stock Exchange; 2) Annual reports are available for the period of 2015/16 to 2018/19 in CSE; 3) Whether the firm's Sustainability Reporting is aligned with GRI framework; 4) Whether GRI content index tool is included in annual report. Secondary data was retrieved from the annual reports from the selected companies at CSE. For data collection, the current study only used annual reports published in CSE where the sustainability reports have been disclosed as part of their annual reports.
Sustainability disclosures have been gathered through content analysis from the respective annual reports. And the disclosures over the period were converted to quantitative by giving it a score. The absence of disclosures were given the score of and the presence of SR disclosures were given the score of 1. This was then used in developing SR Index for environmental, social and economic dimensions and a weighted average was then extended to take the cohesive index.
Sustainability Reporting Index is a technique of measure the SR against the GRI standard. This measurement comprises of four (4) economic, twelve (12) environmental and thirty (30) social standards respectively, according to the G4 guideline. Each sub variable was computed based on the standards adhered to by companies after measuring the total compliance of each component, it will be converted to a 100% scale using the following formula (De Silva, 2019).

= × 100
Dependent variable of the study, to measure the financial performance was chosen as the ROA. Use of this variable was justified by empirical researchers (De Silva, 2019).

=
The present study utilized the panel regression analysis to accomplish the study's ultimate objective, which is to investigate the impact of SR on the FP of listed entities in CSE. SR was captured using GRI framework while the FP were measured using ROA.
As per the methods of data analysis, panel regression analysis was carried out to analyze the impact of SR on FP in public listed companies in Sri Lanka. The panel data technique has been carried out for the current analysis as the data is analyzed over a four-year period starting from 2015/2016 to 2018/2019. Descriptive statistics were used to provide a detailed description of the individual variables of the study. In addition, panel regression was evaluated with a fixed effect and a random effect. The Hausman test, based on the panel data regression methodology, decides the suitable model between the fixed model and the random model for the study. As per the results of the Hausman test, the analysis used regression with the random effect. Furthermore, other necessary diagnostic tests have been carried out in order to further observe variables of the study such as multicollinearity, autocorrelation, heteroskedasticity. In the case where the presence of autocorrelation and heteroskedasticity in the model, robustness test has been carried out as a remedial action.
The study demonstrates the regression model which determines the impact of SR on FP of listed companies in Sri Lanka. The following model is presented the panel regression of the current study based on SR index: Where: i = Company t = Time Yit = Financial performance (ROA) αi = The unknown intercept for each company β1 = Coefficient of explanatory variable εit = The error term The second model used the multiple regression method which comprises each disclosure of the SR; Economic disclosure, Environmental Disclosure, Social disclosure.
Where: i = Company t = Time Yit = Financial performance (ROA) αi = The unknown intercept for each company β1 = Coefficient of explanatory variable εit = The error term

RESULTS AND DISCUSSION
The descriptive statistics demonstrate the research objective used to derive the basic characteristics of the present study data set as variability and distribution. The descriptive statistics of SR Index (independent) and main variables (dependent) in the study describe data in terms of mean, standard deviations, minimum and maximum values. These values were obtained by assessing the two hundred and eighty-nine (289) Table 2 summarizes the year-wise descriptive statistics of the listed companies in this study, showing that the mean of ROA is 0.05, while the standard deviation is 0.07, with a minimum of -0.11 and a maximum of 0.39. Overall sustainability disclosures were indicated through the SR Index which has a mean of 0.56 and a standard deviation of 0.20. In addition, social disclosure, has an average value of 0.53 and a standard deviation of 0.20. Furthermore, other independent variables, environmental and economic disclosures have 0.60 and 0.62 average values, respectively. Economic disclosure shows a standard deviation of 0.29 with a minimum of 0.17 and a maximum of 1. Environmental disclosure shows a standard deviation of 0.27 with a minimum of 0 and a maximum of 1. Furthermore, when considering average data values during the study period, one mean value has been the highest commonly indicated response viewed in the data set. The economic disclosure variable in 2015, while the ROA disclosure variable reports the lowest response compared to other variables in 2018.
Average values of dependent variable were reported as 0.05 to 0.07 in the listed companies between the periods of 2015/2016 to 2018/2019. When considering independent variables, the average values have maintained above 0.5. Due to the dispersion of data set values, economic disclosure is highly spread out over a broader range, and the ROA amount of variation is the lowest value.
Table 1 further shows that the minimum value was shown as 0 for environmental disclosure. Table 2 indicates that one company had not reported environmental disclosures in 2016 and had not disclosed sustainability aspects according to the G4 framework properly. However, after the transition to GRI standards, the company showed an improvement in sustainability reporting. However, the company has then reported environmental disclosure in the rest of the years. However, these findings conclude that some listed companies reported all three disclosures in G4 and GRI standards during the study period.
Prior to regressing the model, the Wooldridge Test was exploited to decide the serial correlation of the model. The autocorrelation results, which indicate the p values as 0.0002, where the study rejects the null hypothesis. Since the Wooldridge test indicated the first-order autocorrelation, reasonable remedies have been actualized. Heteroscedasticity measures whether the fluctuations of the predictions which are determined from the regression remain consistent or vary. The results of the heteroscedasticity test shows that the p-value is 0.0000 indicating a significant value. Therefore, the null hypothesis is accepted which thereby states the presence of heteroscedasticity in the data set. Hence the robust regression model with random effect was done to overcome these setbacks.
Robust Test was done to eliminate the unobservable heterogeneity bias and estimate the increase in the model's reliability via robustness. The researchers carried out Robust Test for the independent variable and composite variable separately. The results of the Robust Test indicate that the robust regression model can be generated, according to the panel data regression technique. Hausman test was used to figure out which model aligns with ROA and SR Index and sustainability disclosure. Hausman test generated the probability chisquare as 0.3702 which is greater than 0.05. Therefore, the Random Effect Model is also used for this analysis to assess the impact between the dependent variable (ROA) and sustainability disclosures. Furthermore, according to the Hausman test, the random-effects model was chosen since the probability chi-square of 0.6695 is greater than 0.05. Therefore, the random effect model is also used to assess the impact between the dependent variable (ROA) and the independent variables (SR).
The regression technique demonstrates the researchers' primary objective as the impact of SR and FP in listed companies in Sri Lanka. The final outcome of the regression analysis technique is the quantitative research method which is used to predict the value of Sustainability Disclosures (independent variables) based on the known value of FP as ROA. According to the regression technique, the study used Robust Regression Model to full fill the researchers' ultimate objective. The results of the Robust Regression Model was stated as follow; The economic disclosure returned a coefficient of 0.023 with a p value of 0.274 where P < 0.05 implies that economic disclosure positively impacts the ROA. The random model of regression indicates that the alternative hypothesis should be rejected, therefore accepting the null hypothesis: there is no significant impact between economic disclosure and return on assets. From the result of regression (coefficient) testing, economic disclosure has a positive cause with an increase in one unit, whenever ROA is enhanced by 0.023 units. It indicates that the listed companies have higher aspirations to reveal economic disclosure. However, the implementation of economic disclosure activities has no impact on improving ROA of the listed companies. The same results have been evident from the past literature done by Asuquo et al. (2018).
Regarding the variable of environmental disclosure, the coefficient of 0.029 having a p value of 0.121. Where P < 0.05 implies that environmental disclosure in public listed companies in Sri Lanka has a positive impact on ROA. But it is statistically insignificant even at 0.121. In other words, environmental reporting is positively but insignificantly related to ROA. Therefore, it is evident that researchers should reject the alternative hypothesis of "significant impact between environmental disclosure and return on assets" and leads to the acceptance of the null hypothesis: there is no significant impact between environmental disclosure and return on assets. If it has a significant impact, when there is a unit increase in reporting environmental disclosure, it leads to a 0.029 increase in the ROA. The present study was further confirmed from the previous literature. A study conducted by Li et al. (2017) has found an insignificant positive effect of financial performance with environmental disclosures. Additionally, a study by López et al. (2007) has evidenced that a negative impact is recognized in the FP using ROA.
The social disclosure (SOC = -0.0604) evidence from the results suggested that the variable of social disclosure is negative. Furthermore, the p-value of 0.104 (p < 0.05) signifies that the implementation of social disclosure has no significant effect on FP in ROA in listed companies in Sri Lanka during the considered time period. This is revealed from the p-value of 0.104, which is greater than the significance level of this study. In line with these findings, this study carefully rejects the alternative hypothesis and accepts the null hypothesis of "there is no significant impact between social disclosure and return on assets". If it has a significant impact, higher social disclosure leads to low ROA. In other words, when there is a unit increase in the reporting social disclosure, it leads to a decrease by 0.0604 units in the ROA. Similar results have been evident from the past literature, a study conducted by Crisóstomo et al. (2011) considering two hundred ninety-six (296) Brazilian listed companies in which ROA measured FP. In contrast, customers and employees were the disclosure basis of the social aspect. The researchers have found a negative effect of social disclosure on FP because Brazilian corporations are not obliged to report social disclosures. Moreover, some of the studies have found a nonsignificant negative impact between social disclosures and FP. The research conducted in the European region by Garcia-Castro et al. (2010) has revealed that a weak negative effect is shown when the indigeneity is considered.
In the present study, a composite variable was developed to analyze the sustainability reporting hypothesis through economic, social, and environmental disclosures, labeled SR Index. The variable of SR index with a coefficient of -0.0108 and p-value of 0.705 (p < 0.05) have an insignificant negative impact on ROA of public listed companies in Sri Lanka. This indicated that when SR is increased by one unit, ROA decreases by 0.010 units. In other words, the results suggest that a unit increase in sustainability disclosures will result in an insignificant decrease in the ROA of sample firms during the period of study. Therefore, by considering the coefficient of SR with the financial measurement of ROA, the alternative hypothesis was rejected, and the null hypothesis: there is no significant impact between sustainability reporting and financial performance was accepted.
Similar results were observed in a study conducted in the aviation industry that revealed that there was no significant impact on SR and Firm performances where SR was measured using GRI-based reports of sustainability. In contrast, profitability was measured using ROA (Karaman et al., 2018). From the evidence, the current study's findings confirm that "there is no significant impact between sustainability reporting and financial performance" as final the outcome.

CONCLUSION
This study was carried out by aligning with both the G4 and the GRI standards to measure the SR index, while the ROA was used to measure FP across the period from 2015/16 to 2018/19 of the fifty-five (55) listed companies in Sri Lanka selected via purposive sampling. Descriptive statistics, panel data regression analysis and correlation analysis were undertaken to accomplish the objectives of the study. Empirical results obtained from the sample revealed that both economic and environmental disclosures have a positive yet insignificant impact, whereas the social disclosure has an insignificant negative impact on FP. Furthermore, the study revealed a negative insignificant impact of the composite variable SR index on FP. Therefore, as per the study's empirical results, it can be concluded that there was a negative but insignificant impact of SR on FP in publicly listed companies in Sri Lanka. It was also noted that some of the companies had reported negative ROA figures over the four years and these companies also had the least recorded amount of SR in the reporting package. The findings further revealed that the decrease in the level of FP of many companies was not actually triggered by the expense of participating in SR related activities incurred by these companies. Factors such as prevailing economic conditions over the period under consideration, regulatory systems under which they function, industry sectors they operate in and other macro-economic factors, may have triggered the decline in the FP. While SR is not significant to FP, many other factors have adversely affected these corporations and were beyond their control. This paves the path for future researchers to explore on these factors and how the companies can adapt and alter policy mechanisms in counteracting them effectively.