Firm’s Financial Performance and Sustainability Efforts: Application of Classifier Models
| Autor: | Koustubh Kanti Ray, Subrat Kumar Mitra |
|---|---|
| Rok vydání: | 2018 |
| Předmět: |
Corporate social performance
Knowledge management Financial performance business.industry 05 social sciences 06 humanities and the arts 0603 philosophy ethics and religion Binary classification Content analysis 0502 economics and business Sustainability 060301 applied ethics Business Business and International Management Social responsibility Classifier (UML) 050203 business & management |
| Zdroj: | Global Business Review. 19:722-736 |
| ISSN: | 0973-0664 0972-1509 |
| DOI: | 10.1177/0972150917713859 |
| Popis: | Numerous studies in academic literature investigated the impact of socially responsible activities on the financial performance of companies; those are committed to social events relative to companies that do not meet the socially responsible criteria. The existing theoretical and empirical research has supported both contradictory positions. We have chosen a different approach to provide an alternative dimension to existing literature. We have taken voluntary CSR disclosure as the dependant variable and attempted to find how the past financial performances of companies influence CSR activities. We test this hypothesis with 100 Indian companies included in BSE 100 index. The director’s report in the latest annual reports of companies were analysed to get voluntary disclosure of CSR activities. The study includes different financial performance variables: ROA, ROE, ROCE, debt to equity ratio, market capitalization and ownership as independent variables for analysis. Several binary classifier models are used for our empirical analysis. The binary model performances are validated with different performance measurement techniques such as F-measure, accuracy rates, balance error rate (BER), Matthews correlation coefficient (MCC), Kappa coefficient and AUROC. The model performance results show a better accuracy while comparing between predicted and actual values. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|