Resumen
This study compares three models: pooled OLS, fixed effect and panel EGLS random effects to examine the impact of corporate governance characteristics and IFRS7 financial instruments disclosure of 14 listed banks on the Nigerian Stock Exchange from 2008 to 2012. Empirical evidence suggests that chi-square and F-statistics in both pooled OLS and fixed effect models are significant thus, not suitable for appraising the model. When the Hausman test was applied to test for the effect, it is found that the null hypothesis of the correlated random effect has a significant probability value of 1.0000. This result supports the conclusion that IFRS7 disclosure is related to board committee, board accounting and board financial expertise and the type of gender in boards of the investigated banks in a random fashion. Based on this analysis, the random effect model which report significant values on three of the independent variables (BC, 0.0014 and BE, 0.0000) at 1% and (GEN, 0.0056) at 10% level of significance is the model of preference. The findings are germane to executive management, stakeholders and policy makers in banks of developing and emerging markets that have embraced IFRS7 for their financial instruments disclosure. It is recommended that existing regulations in Nigeria on mandatory disclosure should be strengthened to compel listed banks in Nigeria to have at least 15% of women on board because of their positive contribution to disclosure requirements.Keywords: Financial Risks, Audit quality, disclosure, board genderJEL Classification: G31