大类资产配置模型周报第 40 期：权益黄金尽墨，全球资产 BL 模型 2 本周微录正收益-20251128

Quantitative Models and Construction Methods 1. Model Name: Black-Litterman (BL) Model - Model Construction Idea: The BL model is an improvement over the traditional mean-variance optimization (MVO) model. It integrates subjective views with quantitative models using Bayesian theory to optimize asset allocation weights. This approach addresses the sensitivity of MVO to expected returns and provides a more robust asset allocation solution[12][13]. - Model Construction Process: - The BL model combines subjective views of investors with market equilibrium returns to derive optimized portfolio weights. - The model uses the following formula to calculate the posterior expected returns: $ \mu = [( \tau \Sigma )^{-1} + P^T \Omega^{-1} P]^{-1} [( \tau \Sigma )^{-1} \Pi + P^T \Omega^{-1} Q] $ - $\mu$: Posterior expected returns - $\tau$: Scalar representing the uncertainty in the prior estimate of returns - $\Sigma$: Covariance matrix of asset returns - $\Pi$: Equilibrium returns derived from market capitalization weights - $P$: Matrix representing the views on assets - $\Omega$: Covariance matrix of the views - $Q$: Vector of expected returns based on the views - The optimized portfolio weights are then derived using the posterior expected returns and the covariance matrix[12][13]. - Model Evaluation: The BL model effectively addresses the sensitivity of MVO to expected returns and provides a more robust and efficient asset allocation framework. It also allows for the incorporation of subjective views, making it more flexible and practical for real-world applications[12]. 2. Model Name: Risk Parity Model - Model Construction Idea: The risk parity model aims to equalize the risk contribution of each asset in a portfolio. It is an improvement over the traditional mean-variance optimization model and focuses on diversifying risk rather than capital allocation[17][18]. - Model Construction Process: - Step 1: Select appropriate underlying assets. - Step 2: Calculate the risk contribution of each asset to the portfolio using the formula: $ RC_i = w_i \cdot \sigma_i \cdot \rho_{i,portfolio} $ - $RC_i$: Risk contribution of asset $i$ - $w_i$: Weight of asset $i$ - $\sigma_i$: Volatility of asset $i$ - $\rho_{i,portfolio}$: Correlation of asset $i$ with the portfolio - Step 3: Solve the optimization problem to minimize the deviation between actual and target risk contributions, subject to the constraint that the sum of weights equals 1[18][19]. - Model Evaluation: The risk parity model provides a balanced risk allocation across assets, making it suitable for achieving stable returns across different economic cycles. It is particularly effective in reducing portfolio volatility and drawdowns[18]. 3. Model Name: Macro Factor-Based Asset Allocation Model - Model Construction Idea: This model constructs a macro factor system covering six key risks: growth, inflation, interest rates, credit, exchange rates, and liquidity. It bridges macroeconomic research with asset allocation by translating macroeconomic views into actionable portfolio strategies[21][22]. - Model Construction Process: - Step 1: Calculate the factor exposure levels of assets at the end of each month. - Step 2: Use a risk parity portfolio as the benchmark and calculate the benchmark factor exposure. - Step 3: Based on macroeconomic forecasts for the next month, assign subjective factor deviation values. For example, if inflation is expected to rise, assign a positive deviation to the inflation factor. - Step 4: Combine the benchmark factor exposure with the subjective factor deviations to derive the target factor exposure for the portfolio. - Step 5: Solve the optimization problem to determine the asset allocation weights for the next month[22][25]. - Model Evaluation: This model effectively incorporates macroeconomic views into asset allocation, providing a systematic framework for translating macroeconomic insights into portfolio decisions. It is particularly useful for capturing macroeconomic trends and their impact on asset performance[21]. --- Model Backtesting Results 1. Black-Litterman (BL) Model - Domestic Asset BL Model 1: Weekly return: -0.32%, November return: 0.05%, 2025 YTD return: 4.0%, annualized volatility: 2.18%, maximum drawdown: 1.31%[14][16][17] - Domestic Asset BL Model 2: Weekly return: -0.15%, November return: 0.08%, 2025 YTD return: 3.77%, annualized volatility: 1.95%, maximum drawdown: 1.06%[14][16][17] - Global Asset BL Model 1: Weekly return: -0.17%, November return: -0.26%, 2025 YTD return: 0.78%, annualized volatility: 2.0%, maximum drawdown: 1.64%[14][16][17] - Global Asset BL Model 2: Weekly return: 0.01%, November return: 0.08%, 2025 YTD return: 2.7%, annualized volatility: 1.59%, maximum drawdown: 1.28%[14][16][17] 2. Risk Parity Model - Domestic Asset Risk Parity Model: Weekly return: -0.27%, November return: -0.09%, 2025 YTD return: 3.6%, annualized volatility: 1.32%, maximum drawdown: 0.76%[20][28] - Global Asset Risk Parity Model: Weekly return: -0.2%, November return: -0.07%, 2025 YTD return: 3.04%, annualized volatility: 1.42%, maximum drawdown: 1.2%[20][28] 3. Macro Factor-Based Asset Allocation Model - Macro Factor-Based Asset Allocation Model: Weekly return: -0.31%, November return: -0.01%, 2025 YTD return: 4.43%, annualized volatility: 1.55%, maximum drawdown: 0.64%[27][28]