In this paper, we attempt to better understand the correlation between the two most important financial asset returns: stock returns and bond returns. Many economists tend to believe that the prices of bonds and other assets should covary closely with stock prices because the prices of these assets are driven by a common underlying discount factor. However, bond markets may serve as a hedge during periods of stock market declines. In investments class, we regularly teach the capital allocation problem that deals with an optimal allocation between an optimal risky portfolio and risk-free assets and the asset allocation among risky assets such as stocks and bonds.
In the asset allocation, the efficient frontier depends on correlations between returns on assets such as stocks and bonds. As such, understanding correlations between stock returns and bond returns is important not only for the potential comovements of various asset returns but also for portfolio construction. While the literature contributes to our understanding of the relation between stock returns and bond returns (e.g., Campbell and Ammer (1993), Bekaert and Grenadier (2001), and Mamaysky (2002), Bekaert, Engstrom, and Grenadier (2005), Baele, Bekaert, and Inghelbrecht (2007), Bekaert, Engstrom, and Xing (2008)), several issues still remain to be explored. In this paper we address the issue of varying correlations across countries and over time.
On a most basic level, one can ask about the correlation between the two types of asset returns. Table 1 provides correlation coefficients for stock and bond returns for Canada, Germany, Japan, the U.K., and the U.S.A. using monthly frequency data (Panel A) and quarterly frequency data (Panel B). [For details, see section 4.] A casual look at Table 1 indicates that correlations between stock returns and bond returns are qualitatively as well as quantitatively different across countries and over time.
For example, in Panel A with monthly real returns, we observe the correlation is positive for the sample period of 1986-1999 for all five countries, while the correlation is negative for the sample period of 2000 - 2007 for all five countries.1 A similar observation is made for Panel B with quarterly real returns.
Download
PDF Ebook Correlations between Stock Returns and Bond Returns: Income and Substitution Effects
