ultimately results in a lower credit spread. While Longstaff and Schwartz (1995), who
pioneered this line of testing using Moody’s indexed U.S. bond yields, find a strong negative
relationship between yield spreads and Treasury yields, Duffee (1998) argues that since most
bonds in Moody’s index are callable bonds, the negative yield spread - riskless rate relation
could instead be due to the negative relationship between the yield spread attributed to the
call option and the riskless interest rate. Consideration of the effect embedded options in debt
has on credit spreads should also be considered with other recent findings: Campbell and
Taksler (2003) show that idiosyncratic firm-level volatility affects credit spreads, leading to
the conclusion that firm-level volatility should also affect any embedded options in debt
value, while Hackbarth et. al. (forthcoming) also argues that credit spreads should be
positively correlated with the volatility of cash flows from firm assets.
While recent studies mount a strong case for the impact liquidity has on the risky yield
spread (Ericsson and Renault, forthcoming; Chen et al., forthcoming), more importantly for
this study, Elton et al. (2001), point out that Duffee’s (1998) analysis ignores potential tax
effects in his tests2. They study the components of yield spreads for U.S. investment-grade
corporate bonds, which by definition have low default risk and so should be more sensitive
to interest rate effects. In conclusion, Elton et al. warn against ignoring the tax differential
when studying corporate bond yield spreads, and both Duffee (1998) and Elton et al. (2001)
make a case for a pure test based on bond data with no call options and tax effects.
In this paper, we use a database of Canadian, investment-grade, corporate bond indices
devoid of tax effects, since Canadian corporate and government bonds, unlike U.S. bonds,
are subject to the same tax treatment. Canadian government bonds are also highly liquid and
not being a reserve currency are not actively sought by international investors, notably
foreign central banks and financial institutions, whose market actions have the potential to
distort pricing along the term structure. This database also contains a unique provision
allowing for identification of callable and noncallable indices. Yet, our results are similar to
those found by Duffee (1998). The negative yield spread - riskless rate relation found in