Conclusion
Based on the investigation in previous sections, I conclude that our data need to be modified in order to better apply DID framework. First, it is important to remeber to remove outliers before we conduct a more formal DID analsis, especially for fund flow and expense ratio characteristics. Second, it is neccessary to apply “propensity-score matching” to every sample in treatment group. The matching criterion should at least include log of fund size and total loads.