This table displays the results of the sensitivity analysis, with four models presented. The first two models have TIER1 as the dependent variable, while the next two use TCR. Each pair includes a conventional regression model and a bootstrapped model for robustness checking. Models (2) and (2a) have 1,872 observations, while models (3) and (3a) have 14,554. This variation in sample size is a result of merging data with the BoardEx database and reflects the differing availability of overlapping data points. Such differences in sample sizes across models are typical in regression analysis, underscoring the importance of understanding the data sources and the rationale behind each model’s construction. For the models with TIER1 as the dependent variable, the RGI variable shows a positive association, indicating that higher values of risk governance (RGI) are correlated with higher TIER1 values. It is important to note that this is an observed association and does not imply that changes in risk governance directly cause changes in TIER1. This association is statistically significant at the 10% level in Model 1. The SIZE variable is negatively associated with TIER1 and is significant at the 1% level, suggesting that, on average, larger banks have lower TIER1 values. The CEOAD and BS variables are not statistically significant. For the models with TIER1 as the dependent variable, the RGI variable has a positive association, suggesting that an improvement in risk governance is correlated with a higher TIER1. However, this result is only significant at the 10% level in Model 1. The SIZE variable has a negative association with TIER1 and is significant at the 1% level. This implies that larger banks may tend to have a lower TIER1. The CEOAD and BS variables are not statistically significant. In the models with TCR as the dependent variable, the RGI variable shows a negative association. This suggests that higher values of risk governance (RGI) are correlated with lower total capital ratios (TCR). It is important to clarify that this is an observed correlation and does not imply that changes in risk governance directly cause changes in the total capital ratio. This relationship is statistically significant at the 5% level. The CEOAD variable is negative and significant at the 10% level, suggesting that banks with a CEO who is also the chair of the board may have a lower total capital ratio. The LNSIZE variable, representing the natural logarithm of the bank’s size, has a negative coefficient and is significant at the 1% level. This suggests that larger banks have a lower total capital ratio. The BS variable is not statistically significant in these models. All models include bank- and year-fixed effects, and the standard errors are clustered at the bank level. Robust standard errors in parentheses; ***p < 0.01; **p < 0.05; *p < 0.1