- Artificial Intelligence Global Asset Allocation Model: The model applies machine learning to asset allocation, using factor investment to score and rank assets, and ultimately constructs a monthly frequency quantitative equal-weight allocation strategy for global assets[35] - Construction Process: The model suggests weights for June as follows: Treasury Index (68.18%), ICE Brent Oil (13.35%), Nikkei 225 Index (11.60%), and Nasdaq Index (6.87%). The weights for ICE Brent Oil and Nikkei 225 Index were increased, while the weights for Treasury and German DAX Index were decreased[35][38] - Evaluation: The model has shown superior performance compared to the benchmark in various dimensions, including annualized return, Sharpe ratio, and maximum drawdown[35] - Testing Results: From January 2021 to May 2025, the model's annualized return is 6.54%, Sharpe ratio is 0.75, maximum drawdown is -6.66%, excess annualized return is 1.09%, excess Sharpe ratio is 0.19, and excess maximum drawdown is -8.67%. The benchmark's annualized return is 5.05%, Sharpe ratio is 0.54, and maximum drawdown is -12.67%[35][39] - Stock-Bond Allocation Model: The model is based on macro timing modules and risk budgeting framework, producing weights for three different risk profiles: aggressive, stable, and conservative[40] - Construction Process: For June, the stock weights for aggressive, stable, and conservative profiles are 45%, 13.92%, and 0%, respectively. The model's signals for May were 50% for economic growth and 40% for monetary liquidity[40][42] - Evaluation: The model has performed well historically, with all three profiles showing superior performance compared to the benchmark in various dimensions[40] - Testing Results: From January 2005 to May 2025, the annualized returns for aggressive, stable, and conservative profiles are 19.90%, 10.97%, and 6.04%, respectively. The benchmark's annualized return is 8.50%. The Sharpe ratios for aggressive, stable, and conservative profiles are 1.28, 1.18, and 1.51, respectively, compared to the benchmark's 0.51[40][47] - Dividend Style Timing Model: The model uses economic growth and monetary liquidity indicators to construct a timing strategy for the dividend index, showing significant stability improvement compared to the full return of the CSI Dividend Index[48] - Construction Process: For May, the model recommended a 100% position in the CSI Dividend Index. Most economic growth indicators were bullish, while monetary liquidity signals were cautious, resulting in a composite signal of 100%[48] - Evaluation: The timing strategy has shown stable performance, with significant improvements in annualized return, Sharpe ratio, and maximum drawdown compared to the CSI Dividend Index[48] - Testing Results: The annualized return of the timing strategy is 15.74%, Sharpe ratio is 0.89, and maximum drawdown is -21.70%. The CSI Dividend Index's annualized return is 11.26%, Sharpe ratio is 0.56, and maximum drawdown is -36.80%[48][51]