Based on the cost–benefit tradeoff, our paper shows that there is cross-sectional variation in the demand for cross
monitoring. Specifically, we find that, for borrowers with higher expected going concern value relative to expected
liquidation value, public debt issues are less likely to include a cross-acceleration provision. This suggests that cascading
defaults are important in the decision to use cross-acceleration provision. We also find this relation is dampened when a
larger amount of public debt matures before bank debt is due. In addition, we find that cross-acceleration provisions are
more likely to be included in public debt agreements when the borrower has higher inherent risk and the benefits of
creditor coordination are greater. We also find that when the borrower’s performance is more difficult to measure the
provision is more likely to be used. We further find that public debt agreements are more likely to include crossacceleration
provisions when borrowers’ delegated monitors have greater incentives to provide higher quality monitoring.
As a consequence of this provision, we find that public debt issues that include this provision require a lower interest rate
spread.
Our paper also contributes to the debt covenant and accounting literatures. First, we highlight the mechanisms public
debt holders use to delegate monitoring to banks. We find that not all public debt contracts use cross-acceleration
provisions to delegate monitoring to banks. Instead, public debt holders appear to consider the costs and benefits of
delegating monitoring to banks. One of the factors that appears to be relatively important in the decision to delegate
monitoring is the quality of the borrower’s financial reports. Public debt contracts of borrowers with low quality financial
reports appear to be more likely to delegate monitoring to banks, and accounting quality appears to affect the spreads
offered in public debt contracts for the borrowers that have cross-acceleration provisions. These results also extend the
results in Bharath et al. (2008) who find accounting quality affects public debt interest rates on average, but does not affect
other terms of the public debt contract (like maturity or collateral). It appears that accounting has a relatively more
nuanced role in public debt.
We need to caveat the interpretation of our findings. First, due to data availability, our bank loan variables are
measured at loan inception rather than at the public debt issuance date. To the extent that these variables change in the intervening period this will introduce measurement error to these bank loan variables. A related issue arises for the
variables measured at public debt issuance, which we also assume will stay constant over the life of the debt. For example,
we assume that the debt repayment structure will not change through time. Again, this may induce measurement error in
our proxies, which we attempt to address by ranking the effected variables. In addition, although we use an instrumental
variable approach to deal with endogeneity concerns, the usual caveats concerning the validity of the instruments apply.