Stock Market Valuation of Corporate Social Responsibility Indicators

Abstract

Renneboog et al (2008) argue that it remains to be seen whether corporate social responsibility (CSR) can be priced. In light of this, this thesis tests the performance and market valuation of CSR indicators by using a comprehensive set of KLD indicators.

Chapter Three of this thesis examines the effect of CSR on financial performance by incorporating CSR into the investment process. As no clear break point is found for the normalised KLD score, the net KLD score is used as an alternative portfolio metric. In addition, most KLD indicators are found to have insignificant alphas for the high-scoring, low-scoring, and long-short portfolios—meaning that investors do not earn abnormal returns through a long-short strategy. Moreover, insignificant alphas are recorded for most of the indicators under the best-in-class approach—meaning that the application of industry classification does not affect results. Finally, both the conditional Ferson and Schadt (1996) model and conditional three-factor model are used as robustness checks, with most indicators having insignificant alphas for these conditional models. As such, the results imply that there is neither outperformance nor underperformance when using portfolios formed with CSR scores; however, there are significant differences in factor loadings between high-scoring and low-scoring CSR portfolios.

Chapter Four uses a framework consistent with the Peasnell (1982) and Ohlson (1995) model to examine whether CSR is reflected in share prices. The CSR indicator is treated as the “other information” variable, and the association between CSR and market price is estimated by controlling for book value of equity, net income and dividends. Although the market is found to value different KLD indicators differently, most of the indicators are found to have positive impact on market value (except for corporate governance and human rights). R&D and advertising expenditure are both added to the valuation model for robustness checking purposes. Some of the CSR indicators—and especially for the case of environment—are not valued during the earlier stages, but become increasingly valued over time. The ten industries are also found to have varying effects on market valuation. In summary, high-scoring CSR firms display higher valuations than low-scoring CSR firms, and thus it can be concluded that a socially responsible agenda does not conflict with maximising shareholder value.

Since most of the CSR indicators in Chapter Four lead to positive market price valuations, Chapter Five aims to disaggregate the value effect into the separate components of ROE ratio, the implied cost of capital (ICC) and growth rate. Three different methodologies are used to test the relationship between CSR, ICC and the long-run growth rate. The relationship between CSR and growth rate is positive with all of the methodologies. However, the different methodologies return differing results for the relationship between CSR and ICC, which may be due to the different assumptions made by each approach. Furthermore, it suggests that long-run growth rate differences in general may be more important than ICC differences. Finally, most KLD indicators are found to have significantly higher P/V and ROE1 ratios for the high-scoring CSR portfolios than for the low-scoring CSR portfolios.