Monetary policy and corporate bond yield spreads

Abstract

Firm characteristics, economic conditions and policy regimes are the key determinants that most researchers have used to explain corporate bond yield spreads. In this article, we examine whether monetary policy shocks are also important determinants given their ability to affect default risk, risk aversion and liquidity premiums. Using a Vector Autoregression (VAR) with long-run monetary neutrality, we find that monetary policy shocks do, in fact, account for a large portion of the variation in corporate bond yield spreads.

Notes

¹ See, e.g. Black and Scholes (1973), Merton (1974), Black and Cox (1976), and more recently, Collin-Dufresne et al. (2001) and Collin-Dufresne and Goldstein (2001).

² The use of long-run restrictions in VARs were pioneered by Blanchard and Quah (1989) and Shapiro and Watson (1988).

³ Another commonly-used approach to identify monetary policy shocks is to estimate a macroeconomic VAR with contemporaneous restrictions on the federal funds rate. This identification strategy allows federal funds rate innovations to be interpreted as unexpected deviations from a monetary policy rule or a monetary policy shock. However, as noted by Kim and Lastrapes (2007), this approach is sensitive to changes in the monetary policy rule while the long-run restriction strategy is not.

⁴ Note that since both m_t and M_t are in the VAR and given that m_t  = M_t /P_t , the price level is implicitly an endogenous variable too.

⁵ Keating (1996) shows that these restrictions can be implemented by applying a Choleski decomposition to the long-run covariance matrix.

⁶ While it is apparent that long-run monetary neutrality restrictions makes sense for the real variables of y, b, s and m, the restrictions also is a reasonable one to apply to the nominal interest rate since it mimics the behaviour of the real interest rate in the long-run. This is because a permanent money supply shock only affects the price level permanently not inflation.

⁷ Shiller uses the CPI to deflate the nominal S&P Composite Index.

⁸ The Augmented Dickey–Fuller (ADF) and the Phillips–Perron (PP) unit root tests were used. First differencing the data was sufficient to remove the unit root.

⁹ The SE bands are calculated using standard Monte Carlo techniques.