- The report defines "fund clustering degree" as a measure influenced by the clustering degree of heavy stocks and their holding proportions. Stock clustering degree is calculated based on the number of funds holding a specific stock, ranked by percentile within all stocks[25][26] - Fund clustering degree is derived as a weighted sum of heavy stock clustering degrees, with weights being the proportion of the stock's market value to the fund's total stock holdings. The formula is: $ \text{fund\_score}_{\text{j,t}}=\sum_{k=1}^{10}\text{stock\_cap}_{k,t}\times\text{stock\_score}_{k,t} $ where stock_cap represents the proportion of the stock's market value, and stock_score denotes the clustering degree of the stock[27] - In Q2 2025, the fund clustering degree decreased compared to the previous quarter, currently positioned near historical averages[27]