Figure 12
Leverage and Volatility of Idiosyncratic Risk
cannot be explained by the expected default probabilities derived from the option-
theoretic framework underlying the MKMV approach. Thus, once credit rating effects
are excluded, the model-implied spread
financial market frictions—bankruptcy costs and liquidity premiums—and would be
likely wider, requiring in turn a bigger μt.
Rb
R
- *
would incorporate different sorts of
it
The fit of the model, under the specification that excludes credit rating effects, is
greatly improved if we use the model-implied recovery rate as an alternative metric
to evaluate the BGG framework. In fact, as shown in the lower panel of Figure 11,
the model-implied recovery rate is now much closer to the actual recovery rate on
defaulted bonds, particularly since early 2001.
5.3 Limitations of the BGG Framework
In addition to the time-varying bankruptcy cost parameter μt, the other (unobserv-
able) key parameter of the BGG model that we solve for is σit , the firm-specific,
time-varying standard deviation of the idiosyncratic productivity shock. As shown
in Figure 12, the model implies a strong prediction between firms’ leverage and the
33