The article studied how much external data is needed to measure operational risks in China's commercial banks accurately. Researchers used a method called non-parameter variable scale bootstrap to determine the minimum sample size required. They found that a minimum of 800 cases, including extreme values, is needed for stable simulation results. This finding was supported by comparing the results to operational risk capital data from 2013 in Chinese banking industries. It shows that not considering whether external data reflects overall characteristics can lead to inaccurate measurements.