基于股债利差模型的动态资产配置策略研究

Research on Dynamic Asset Allocation Strategy Based on the Equity-Bond Yield Spread Model

摘要: 中国A股市场具有显著的高波动率和“牛短熊长”特征，单一资产持有的风险收益比往往不尽如人意。传统的静态资产配置策略在面对市场极端波动时缺乏灵活性。本文基于经典的股债利差模型理论，引入格雷厄姆指数作为核心择时信号，构建了一套在沪深300指数ETF与国债ETF之间进行动态轮动的量化资产配置策略。通过选取2015年至2024年的A股市场数据进行实证回测，研究发现：(1) 股债利差模型能够有效识别A股市场的估值极值区域，具有显著的均值回归特性；(2) 本文构建的动态配置策略在保留权益资产长期向上收益的同时，通过在估值高位自动降低仓位，显著降低了2015年股灾及2018年熊市期间的系统性回撤；(3) 相较于沪深300指数买入并持有策略，该动态策略获得了更高的夏普比率和卡玛比率。本文的研究验证了基本面价值投资逻辑在A股择时层面的有效性，为长期资金提供了一种稳健的资产配置方案。

Abstract: China’s A-share market exhibits notably high volatility and the characteristic of “short bull markets and long bear markets” often resulting in suboptimal risk-return profiles for single-asset holdings. Traditional static asset allocation strategies lack flexibility when confronted with extreme market fluctuations. Building upon the classical Fed Model theory, this paper introduces the Graham Index as the core timing signal to construct a quantitative dynamic asset allocation strategy that rotates between the CSI 300 Index and government bond ETFs. Empirical backtesting using A-share market data from 2015 to 2024 demonstrates the following findings: (1) The equity-bond yield spread model effectively identifies valuation extremes in the A-share market and displays significant mean-reversion properties; (2) The proposed dynamic allocation strategy preserves the long-term upward returns of equity assets while substantially reducing systemic drawdowns during major downturns—such as the 2015 stock market crash and the 2018 bear market—by automatically lowering equity exposure at high valuation levels; (3) Compared to the CSI 300 Index, the dynamic strategy achieves higher Sharpe Ratio and Calmar Ratio. This study validates the effectiveness of fundamental value investing logic in market timing within the A-share context and offers a robust asset allocation solution for long-term capital.