量化投资组合管理研究系列之（九）：热潮下的冷思考：估值、拥挤度与资金博弈的量化择时波动波动

Investment Strategy Overview - The report focuses on the significant capital aggregation effect in high-growth sectors, highlighting the short-term volatility caused by valuation differentiation and trading crowding [2][9] - A multi-dimensional quantitative timing system is constructed based on three core dimensions: valuation rationality, trading crowding, and capital game, aiming to provide objective quantitative references for timing decisions in high-growth sectors [2][9] Methodology and Factor Construction - The research selects 18 foundational indicators across five dimensions: trading, valuation, capital, equity structure, and volume-price, utilizing advanced statistical models to derive 78 core effective indicators with significant predictive power [2][14] - The study employs a GARCH-EVT-VaR model for dynamic risk management, ensuring robust backtesting results that demonstrate an annualized return of 29.50%, with a 13.01% excess return over the benchmark [2][14] Backtesting Results - The backtesting of the weekly rebalancing timing strategy shows a success rate of 54.8%, with an improved profit-loss ratio from 1.23 to 1.57 and a reduction in maximum drawdown from 49.9% to 39.7% [2][14] - The strategy effectively identifies overheating signals in capital-aggregating growth sectors, allowing for timely adjustments to mitigate risks associated with leveraged funding and valuation bubbles [2][14] Application Value - The timing system is particularly adept at recognizing overheat signals in high-growth sectors, with notable performance observed in the artificial intelligence sector, which has seen increased trading activity since September 2024 [2][10] - The core quantitative method can be generalized to other high-attention, high-capital-aggregation, and high-trading-game growth sectors, providing cross-sector reference value [2][10] Sector Focus: Artificial Intelligence - The report emphasizes the artificial intelligence sector as a key area of focus due to its significant capital aggregation effects and trading crowding characteristics, making it a valuable subject for quantitative research [10][12] - The China Securities Artificial Intelligence Theme Index is highlighted for its comprehensive coverage of the AI industry chain, demonstrating strong growth potential with a one-year return of 56.35% and a three-year return of 26.43% [10][12]