“Environmental responsibility and financial performance nexus in South Africa: panel Granger causality analysis”

The authors examined environmental responsibility and financial performance nexus of Johannesburg Stock Exchange’s socially responsible investing manufacturing and mining firms during the period of 2008  2014. The study employs annual panel dataset of fourteen manufacturing and mining companies on the index, and Granger causality analysis using Gcause2 Baum’s version. The paper found unidirectional causal relationship between environmental responsibility, measured by emissions intensity and equity returns, and bidirectional causal relationship between emissions intensity and market value of equity deflated by sales at 1% significant levels. Impliedly, improvements in ‘energy efficient technologies’ to reduce fossil-energy consumption (prevention activities) seem to exhibit value destroying tendencies, while improvements in ‘end-of-pipe’ activities seem to estimate a drive market value of equity deflated by sales and equity returns. The Pesaran CD and Breusch-Pagan LM tests confirmed existence of cross-sectional dependence amongst panel members. The authors tend to support institutional and stakeholder theories.


Introduction 3 
Past decades have seen researchers examining financial implication of sustainable performance amidst global warming and depletion of fossil-energy source.These studies have examined whether sustainability responsibility contributes to financial performance of firms.Barley (2009) cited that rating agencies had devalue corporate debt, because of future business risks due to high emissions.Carbon Disclosure Project (2014) asserted that emissions reduction continues to generate return on investment equivalent to USD15 billion.Goldman Sachs (2009) argued that equity markets are beginning to recognize effects of low carbon performance on firms' competitiveness and future valuation.The assertion that firms' environmental performance is critical in the assessment of risk profile, potential liabilities and long term valuation has received a mixed reaction in literature.Hence, the question of whether there exists a linkage between companies' green performance and the 'bottom line' becomes an important research question.Mostly, because there is a belief among some scholars that ethics has no place in business, and that firms only need to appear ethical to preserve legitimacy (Friedman, 1970;Wagner et al., 2002).
The paper investigates environmental responsibility and financial performance nexus of manufacturing and mining companies listed on Johannesburg Stock Exchange's responsible investing index during 20082014.The paper extends accounting research by answering the question of how environmental responsibility affect financial performance utilizing Granger causality analysis and based on Baum's (2010) version Gcause2.The study uses annual data from fourteen JSE's SRI manufacturing and mining firms for the period 20082014.After controlling for firm's omitted variable bias, our empirical results showed that lags of environmental performance, measured by energy usage intensity and financial performance, represented by return on assets and return on sale do not improve forecast of either factor.Contrary, our empirical results showed a unidirectional relationship between environmental responsibility, measured by emissions intensity and financial performance, measured by equity returns.Furthermore, we found bidirectional relationship between emissions intensity and market value of equity deflated by sale.The paper is arranged as follows: section 1 highlights related literature.Section 2 focuses on the methods and materials, in section 3, an empirical results are shown and final section is focusing on conclusion of the study.

Literature review
Environmental accounting research in the past few decades has examined effect of 'green performance' on financial performance, but is still providing mixed empirical findings.Patari et al. (2014) examined social responsibility measured by "strengths" and "concerns" causal relationship with and financial performance of US energy industry, and found that "concerns" improve the forecast of both profitability and market value, whereas "strengths" improves only the forecast of market value.Orlitzky (2001) evaluated the nexus between social initiatives and fiscal risk and found that firms with higher social responsibility exhibit lower financial risk.It was concluded that social responsibility and financial risks appear to be one of reciprocal causality.Makni  showed that emissions levels of listed firms in Australia may be affected by the proposed scheme.2010) examined social responsibility and financial performance relationship and showed that no significant association existed between social responsibility and financial performance relations.When intangible resources is accounted for, it was found that the relationship indirectly relies on the mediating effect of firm's intangible resources.Fujii et al. (2012) assessed toxic risk and sales and return on assets relationship and found that a significant relationship existed between return on assets and toxic risk.Nishitani et al. (2011) investigated effects of emissions reduction on financial performance of Japanese companies during 20022008 and showed positive effect of emissions reduction on financial.Lioui and Sharma (2012) analyzed effects of social responsibility on Tobin Q and return on assets and found negative relationship between social responsibility (measured by "strengths" and "concerns") and Tobin Q and return on asset.Nyirenda 2014) assessed eco-innovation effects on returns on asset, return on equity and lower earnings retention of Polish and Hungarian listed firms and found that eco-efficient companies enjoyed greater returns on asset and equity, but lower earnings retention.Telle (2006) assessed effects of environmental pro-activeness on operating performance and found a correlation between corporate environmental ingenuities and sales return.When the paper controlled for firms' specific effects, the results showed an insignificant effect of the environmental measure on and return on sale.However, Pintea et al. (2014) found no significant correlation between environmental ingenuities and corporate economic performance.

On the basis of the review of literature, this paper thus hypotheses as:
H0: There is no bi-directional causality between environmental responsibility and financial performance of JSE's SRI manufacturing and mining firms.

Methods and materials
In studying causal relationship between factors, some previous studies have applied panel vector error correction model based on Arellano-Bond (1991).Pesaran et al. (1999) cited that PVECM as a tool is inefficient as it lacks dependence across panel members.Hurlin (2008) developed an approach that accounted for heterogeneity but not for cross-sectional dependence.Konya (2006) proposed a method which accounted for heterogeneity and dependency across sections by utilizing seemingly unrelated regression (SUR).Emirmahmutoglu and Kose (2010) proposed a Meta-analysis and applied heterogeneous mixed panel that tests for crosssectional dependence and uses bootstrap procedure to determine direction of causality.De Hoyos and Sarafidis (2006) cited that if one is working with short dynamic panel data approximations, and the disturbances assume cross sectional dependence, the estimation procedure that depends on generalized method of moments (e.g., Arellano-Bond, 1991) may produce inconsistent result, as the crosssectional dimension grows large.Kar et al. (2011) suggested that tests for dependence across sections, whilst using a panel causality study is important, as it helps in estimation technique.

Model specification
(1) Then 'x' might not have an effect on 'y' and could be said not to Granger-cause 'y'.
Alternatively, if it may be said that 'x' does not improve forecast of 'y', because the anticipated value of 'y' is different from an assumed 'x' We therefore specify the Granger causality model as: We employed four measures to proxy financial performance: (i) ROA: return on assets, (ii) ROS: return on sale, (iii) EQRTNS: return on equity, and

Empirical results
We examined environmental responsibility and financial performance nexus of Johannesburg Stock Exchange's socially responsible investing manufacturing and mining firms for the period 2008  2014.The paper represented financial performance by return on assets (ROA), return on sales (ROS), equity returns (EQRTNS) and market value of equity deflated by sales (MVE/S).Environmental responsibility is measured by energy usage intensity (ENGINT) and carbon emissions intensity (EMSINT).Using Pesaran CD (2004) and Breusch-Pagan LM (1980), we performed cross sectional dependence tests and report our results in Table 1.The null hypothesis of no cross-sectional dependence was rejected at 1% significant level.Rejecting the null implies that shocks in either financial performance and/or environmental responsibility in a particular company could be likely transmitted to other firms, and this may be as the result of the fact that manufacturing and mining companies on the index are confronted with similar socio-economic risks.We report our panel Granger causality tests results in Tables 2 and 3, respectively.The empirical results as reported in Table 2 showed environmental responsibility, measured by energy usage intensity (ENGINT) do not Granger cause financial performance (FP) measured by ROA, ROS EQRTNS and MVE/S.Furthermore, we also found that financial performance does not Granger cause environmental responsibility.Contrary, the results as reported in Table 3 established a unidirectional causality of environmental responsibility measured by emissions intensity (EMSINT) to financial performance measured by equity returns (EQRTNS) at 1% significant level.We further established a bidirectional causality between emissions intensity (EMSINT) and the financial performance measured by market value of equity deflated by sales (MVE/S) at 1% significant level.Whereas our results showed that prior improvement in emissions intensity subsequently improves in equity returns, we also found that prior improvement in emissions intensity or market value of equity deflated by sale subsequently improves each other.We conclude that for the purpose of value creation and corporate competitiveness, Johannesburg Stock Exchange's socially responsible investing manufacturing and mining firms should focus more on end-of-pipe activities instead of 'prevention activities'.Notwithstanding, for the purposes of achieving congruence between financial objective and accepted social norms, these firms should continue with their 'carbon prevention' activities as part of the implicit contract between themselves and society.An important research question worth asking at this point is whether the results have been same if South Africa had instituted carbon tax law and emissions trading scheme as in Australia and the European Union.
Exploring investment in Research & Development impact on financial performance, using fixed effect estimations and data of 362 companies during 2003-2010, Ki-Hoon et al. (2015) demonstrated that emissions reduce firm value.Chapple et al. (2009) assessed the market reaction to the National Emissions Trading Scheme, and et al. (2013) examined effects of sustainable management on return on equity.

Table 1 .
Cross-sectional dependence tests Note: figures in brackets denote p-values and the test statistics.The LM test and CD test are the cross-sectional dependence tests.

Table 2 .
Clarkson et al. (2011)y tests Nonetheless, for the purpose of meeting stakeholder demands in the more sustainable competitive manner, firms ought to equally get involved in carbon prevention activities.Our findings support institutional and stakeholder theories, as the results showed the extent to which JSE's socially responsible firms manage fossil energy to create wealth for owners by instituting integrated programs of activities that enhances interaction with the environment.Furthermore, the rejection of the null amongst index's manufacturing and mining firms indicate that shocks in financial performance and/or environmental engagements in one firm maybe likely to be transmitted to other firms on the index.This could be explained from the point that these manufacturing and mining firms are faced with similar socio-economic and regulatory exposures.Our results seem to confirmClarkson et al. (2011).