Integrated reporting quality and the value relevance of environmental, social and governance information
Keywords:
Integrated Reporting Quality, Composite Disclosure Index, Value Relevance, Earnings Per Share, NigeriaAbstract
This study examines the value relevance of integrated reporting quality (IRQ) for financialbperformance among listed non-financial firms in Nigeria. IRQ is a composite index combining environmental and social sustainability disclosure (ESGD), corporate social responsibility disclosure (CSRD), and corporate governance disclosure (CGD), normalised to 0-1 and averaged into one score. Using 31 firms on the Nigerian Exchange Group, 2015-2024 (310 firm-year observations, stratified from 46 firms), the study estimates IRQ’s effect on earnings per share (EPS), controlling for market capitalisation, firm size, profitability, leverage and
growth. EPS, IRQ and market capitalisation are non-stationary in levels but stationary after differencing; the model is estimated on differenced data using Stata-verified panel-corrected standard errors. IRQ shows no significant association with EPS across six specifications (β=1.849, p=0.822 contemporaneously; β=-1.544, p=0.841 with year fixed effects; β=-0.378, p=0.933 lagged; β=1.350, p=0.858 with full controls; β=-0.935, p=0.916 fully combined), nor does market capitalisation in any specification (R²=0.023-0.078). Year effects are significant in 2019, 2022 and 2024, and profitability is significantly positive under PCSE in the full control (p=0.011) and combined models (p=0.026), though not under cluster-robust errors (p=0.388). A Ramsey RESET test on the combined model suggests functional-form misspecification (F=28.76, p<0.0001), and firm size’s differenced series shows an anomalous unit-root result, both reported as limitations. Endogeneity correction is partial: fixed effects and a predetermined regressor address self-selection and reverse causality, but no instrumental-variable or Heckman correction is implemented. The study offers a reproducible IRQ-construction method and cautions against over-interpreting null coefficients.
