The article discusses how banks can better predict and manage operational risks by combining historical data with expert opinions through a method called Bayesian inference. By using scenario analysis and considering factors like internal control systems and business environment, banks can estimate the frequency and severity of potential risk events. This approach helps banks not only quantify operational risk capital but also incentivize business units to improve risk management policies. By combining expert opinions with historical data, Bayesian inference provides a powerful tool for estimating operational risk losses.