- The report constructs a quantitative model to identify bottoming and leading signals in the market[7][18] - The model uses historical data up to the date of the signal to predict future trends[7][18] - The model triggered 118 rebound signals for the CSI All Share Index from 2006 to 2025, averaging about 6 times per year[18] - After the rebound signal, the average returns for the CSI All Share Index over the next 20, 60, 120, and 240 trading days were 1.09%, 2.89%, 6.08%, and 15.88%, respectively, with win rates of 57%, 58%, 56%, and 54%[18][22][27] - The model's application to industry indices also showed effective results, with the highest potential rebound industries achieving average returns of 2.16%, 3.68%, 6.89%, and 17.91% over the next 20, 60, 120, and 240 trading days, respectively[34][35] - The strategy of holding the top-performing industries for 20 days and then switching to the CSI All Share Index achieved a cumulative return of 1605.55% and an annualized return of 15.85% from 2006 to 2025, outperforming the CSI All Share Index's cumulative return of 565.01% and annualized return of 10.33%[39][40] - The model was further refined to select top-performing stocks within the top-performing industries, achieving a cumulative return of 11753.66% and an annualized return of 28.11% from 2006 to 2025[49][57][59] - Direct application of the model to large-cap stocks also showed significant outperformance, with the best strategy achieving a cumulative return of 8979.24% and an annualized return of 26.35% from 2006 to 2025[67][70][75] - The model was also applied to individual industry indices, showing that it could accurately predict the bottoming of specific industry indices and construct stock portfolios that significantly outperformed the corresponding industry indices[81][82] - The model's application to broad-based indices like the CSI 300, CSI 500, and CSI 800 also demonstrated its ability to accurately predict the bottoming of these indices and construct stock portfolios that significantly outperformed the corresponding indices[210][213][216][219]