Research design
We conduct two analyses to test our hypotheses. In the first analysis, we investigate the determinants of the decision to
include cross-acceleration provisions in public debt agreements. In the second analysis we examine the effects of crossacceleration
provisions on the interest spread charged in the public debt agreement.
To test our hypotheses we focus on the characteristics of the borrower and their bank debt at the time that the public
debt is issued. There are two important assumptions implicit in this research design. First, the bank debt holders may not
stay constant over the life of the public debt, although we assume that the quality of the monitoring will remain constant.
Second, the repayment schedule of the public debt and private debt affects the potential costs and benefits of crossacceleration
provisions, and the repayment schedule may change over the life of the public debt. Similarly, we assume that
the repayment schedule of the debt will be relatively constant over the life of the public debt. That is, firms refinance
similar bank/public debt when the old debt matures to maintain a consistent repayment schedule.
To test our first hypotheses that public debt is more likely to include cross-acceleration provisions when the expected
costs of cascading defaults are lower, we create a variable, PL Ratio, which measures the borrower’s expected going concern
value scaled by its expected liquidation value.
13
We measure the going concern value using the market value of total assets
(total assetsbook value of common equitydeferred taxþmarket value of common equity). We measure the liquidation
value of assets in forced sales resulting from cascading defaults caused by cross-acceleration covenants by using Berger
et al.’s (1996) estimate of asset tangibility: 0.715 accounts receivableþ0.547 inventoryþ0.535 PPE (Property, Plant
and Equipment)þ1 cash. We use the tercile rank of this variable (PL Ratio_Rank) to reduce measurement error and to
capture potential non-linearity in our model.
14
We expect the use of cross-acceleration covenants will be less likely when
borrowers have higher PL Ratios based on H1.