港股投资周报：恒生科技领涨，港股精选组合年内上涨72.29%-20250927

Quantitative Models and Construction Methods - Model Name: Guosen JinGong Hong Kong Stock Selection Portfolio Model Construction Idea: The model is based on a dual-layer selection process that combines fundamental and technical analysis of stocks recommended by analysts. The goal is to identify stocks with both fundamental support and technical resonance, which are expected to outperform[14][15] Model Construction Process: 1. Construct an analyst-recommended stock pool based on the following events: - Analysts upgrading earnings forecasts - Analysts initiating coverage - Analyst report titles indicating unexpected events 2. Perform dual-layer screening on the stock pool: - Fundamental analysis: Evaluate the fundamental support of the stocks - Technical analysis: Assess the technical resonance of the stocks 3. Select stocks that meet both fundamental and technical criteria to form the Hong Kong Stock Selection Portfolio 4. Backtesting period: 2010-01-01 to 2025-06-30, considering full position and transaction costs Model Evaluation: The model demonstrates strong performance with significant annualized returns and excess returns over the Hang Seng Index[15] - Model Name: Stable New High Stock Screening Model Construction Idea: The model leverages momentum and trend-following strategies, which have been shown to be effective in the Hong Kong stock market. Stocks that reach new highs are considered market leaders and are expected to deliver higher future returns[20] Model Construction Process: 1. Define the 250-day new high distance as: $ 250 \text{ day new high distance} = 1 - \frac{\text{Close}{t}}{\text{ts_max}(\text{Close}, 250)} $ - $\text{Close}{t}$: Latest closing price - $\text{ts_max}(\text{Close}, 250)$: Maximum closing price over the past 250 trading days - If the latest closing price reaches a new high, the 250-day new high distance is 0; if it falls from the new high, the distance is positive, indicating the degree of decline 2. Screen stocks that have reached a 250-day new high in the past 20 trading days based on the following criteria: - Analyst attention: At least 5 buy or overweight ratings in the past 6 months - Relative stock strength: Top 20% in terms of 250-day price change within the sample pool - Price stability: Evaluate using two indicators: - Price path smoothness: Stock price displacement ratio - New high persistence: Time-series average of the 250-day new high distance over the past 120 days - Trend continuation: Time-series average of the 250-day new high distance over the past 5 days 3. Select the top 50 stocks based on the above criteria, ensuring at least 50 stocks are included[22][23] Model Backtesting Results - Guosen JinGong Hong Kong Stock Selection Portfolio: - Annualized return: 19.11% - Excess return over Hang Seng Index: 18.48% - Maximum drawdown: 23.73% - Information ratio (IR): 1.22 - Tracking error: 14.55% - Return-to-drawdown ratio: 0.78[19] - Stable New High Stock Screening: - The model identified 14 stocks in the pharmaceutical sector, 11 in the cyclical sector, 8 in technology, 6 in consumer, and 4 in manufacturing as stable new high stocks for the week[22][23] Quantitative Factors and Construction Methods - Factor Name: 250-Day New High Distance Factor Construction Idea: This factor measures the distance of the latest closing price from the 250-day high, capturing the momentum and trend-following characteristics of stocks[22] Factor Construction Process: - Formula: $ 250 \text{ day new high distance} = 1 - \frac{\text{Close}{t}}{\text{ts_max}(\text{Close}, 250)} $ - $\text{Close}{t}$: Latest closing price - $\text{ts_max}(\text{Close}, 250)$: Maximum closing price over the past 250 trading days - Interpretation: A value of 0 indicates the stock is at its 250-day high, while a positive value indicates the percentage drop from the high[22] Factor Backtesting Results - 250-Day New High Distance Factor: - The factor was used to identify stable new high stocks, with the pharmaceutical sector having the highest number of selected stocks (14), followed by cyclical (11), technology (8), consumer (6), and manufacturing (4)[22][23]