Simply performing Kolmogorov-Smirnov tests is a technical procedure, but the interpretation of the distributional similarity they reveal is often noise.
We should focus on the quality of reference data rather than the procedure itself to obtain a relevant result.
Without relevant inputs, such as using American benchmarks for an Asian market, results are not controllable and thus worthless for decision-making.
The discipline requires ensuring that the compared data are truly comparable, for example by adapting benchmarks to local regulations in Hong Kong.