Carbon emissions reduction and financial performance of Johannesburg Stock Exchange 's SRI companies

Abstract

This research examined the effect of carbon emissions reduction on financial performance Johannesburg Stock Exchange’s SRI companies. Empirical results of corporate fossil energy-based dependence on environment and economic performance thus far have been ambiguous. The major objective of this research was to examine the effect of emissions and energy intensity on market and accounting based performance measures. This research adopted the positivist paradigm approach and therefore used a quantitative causal research approach. Archival data was collected from fourteen JSE’s SRI companies for seven years. The research applied a panel data analysis, a total of 98 observations were derived from panel data set. Multiple linear and causal econometric models were applied in the data analyses namely ordinary least squares (OLS), fixed effects and dynamic models. OLS results showed a significant effect of energy usage intensity (ENGINT) on return on assets (ROA), and return on sales (ROS), with carbon emissions intensity (EMSINT) exhibiting a significant effect on return on assets (ROA), and return on sales (ROS). When the study controlled for omitted variable bias and possible orthogonality condition, a significant negative effect of energy intensity (ENGINT) on equity returns (EQRTNS) was found. Impulse response analysis revealed that shocks in energy intensity on average tend to decrease firms’ financial value, while shocks in emissions intensity on average increase firms’ financial value within the sampled companies. Whilst testing for causality, the Panel Granger causal analysis showed unidirectional effect of EMSINT on EQRTNS, and bidirectional causal relationship between EMSINT and MVE/S at 1% significant level. This research made a contribution by extending the model used by previous researchers through the use of multiple market and accounting based performance measures which were analysed using advanced econometric models: Arellano-Bond DPD model, impulse response function in short PVARs and Bootstrap dynamic panel threshold model. In addition, this thesis suggested a model to advance future research on carbon emissions and firm performance and managerial decision propensity for carbon reduction.