Stock-bond return correlation: Understanding the changing behaviour
研究了G7市场股票与债券收益相关性的时变特征，发现相关性在1990年代末由正转负，后期又转正，且通胀和利率对相关性有正向影响，对投资组合权重和风险管理有重要启示。
• Consider the time-varying stock–bond correlation for G7 markets. • All markets reveal switching between a positive and negative correlation. • The drivers for the correlations also exhibit switching behaviour. • Macroeconomic drivers predominantly report a positive effect on the correlation #. • Results suggest key implications for portfolio weights and risk. The stock and bond return correlation remains important given its central role in portfolio behaviour. Previous, primarily US, evidence indicates sign switching, which implies that bonds change between diversifying and hedging behaviour. This paper considers time-variation in the stock–bond correlation for the G7 markets, including the nature of its economic drivers. Using monthly data over a period spanning 1980 to 2023 evidence demonstrates that the correlation switches from positive to negative in the late 1990s for six of the seven markets (the switch for Japan occurs in the first half of the 1990s). A switch back to positive is observed towards the end of the sample for most markets but earlier for France and Italy. Evidence of time-variation within the correlation drivers is also noted. Nonetheless, results suggest that inflation and interest rates typically exhibit a positive effect on the correlation, consistent with previous work and theoretical underpinnings. That is, higher inflation and interest rates depress stock and bond prices due to higher discount rates and lower real cash flows, moving them in the same direction. Growth also largely imparts a positive effect on the correlation, but this contrasts with the prevailing view. This arises through portfolio considerations where higher growth leads to an increase in demand for all assets. Of importance for investors, the switch in correlation implies that a portfolio manager will need to alter asset weights to maintain a target value for returns or risk. A portfolio variance decomposition reveals that while the bond contribution remains broadly constant over the sample, that from stocks increases as the correlation contribution shifts from positive to negative. The results are of importance to investors and those engaged in modelling market behaviour.