Regulation, renegotiation and capital structure: theory and evidence from Latin American transport concessions

Abstract

We examine the capital structure of regulated infrastructure firms. We develop a model showing that leverage, the ratio of liabilities to assets, is lower under high-powered regulation and that firms operating under high-powered regulation make proportionally larger reductions in leverage when the cost of debt increases. We test the predictions of the model using an original panel dataset of 124 transport concessions in Brazil, Chile, Colombia and Peru over 1992–2011. For each concession we have data on the regulatory regime, annual financial performance and contract renegotiations. We begin by demonstrating that, although pervasive, contract renegotiations do not fundamentally alter the regulatory regime. Importantly, firms are not systematically able to renegotiate when in financial difficulty, implying that price cap contracts remain high-powered in practice. We use this result for our main empirical work, where we find broad support for our theoretical predictions: when the cost of debt increases, firms operating under high-powered regulation make proportionally larger reductions in leverage.

Appendix

Proposition 1

The derivation follows closely Cambini and Spiegel (2011). Consider first the case $D<D_1$. Then $\varphi ^{*}\left( {p,D} \right) =0,\frac{\partial p^{*}}{\partial D}=0,$ so that $\frac{\partial Y(D)}{\partial D}=0$, using the fact that $R_E =R_D$.

Consider now the case $D_1 <D<D_2$. Then $\varphi ^{*}\left( {p,D} \right) =(1-\frac{p^{*}-D}{\bar{c}})$ and $p^{*}\left( D \right) =a+\left( {1-b} \right) \left[ {\frac{\bar{c}}{2}+(1-\frac{p^{*}-D}{\bar{c}})T} \right]$. This implies that $\frac{\partial \varphi ^{*}}{\partial D}=\frac{1}{\bar{c}},\frac{\partial \varphi ^{*}}{\partial p}=-\frac{1}{\bar{c}},$ and $\frac{\partial p^{*}}{\partial D}=\left( {1-b} \right) T\frac{1}{\bar{c}}$. Using the fact that

$$\begin{aligned} \frac{\partial Y(D)}{\partial D}=\frac{\partial p^{*}}{\partial D}-\left( {\frac{\partial \varphi ^{*}\left( {p\left( D \right) ,D} \right) }{\partial p^{*}}\frac{\partial p^{*}(D)}{\partial D}+\frac{\partial \varphi ^{*}(p\left( D \right) ,D)}{\partial D}} \right) T \end{aligned}$$

(8)

straightforward computations lead to

$$\begin{aligned} \frac{\partial Y(D)}{\partial D}=\frac{T}{\bar{c}}\left[ {\left( {1-b} \right) \frac{T}{\bar{c}}-b} \right] \end{aligned}$$

(9)

Thus $\frac{\partial Y(D)}{\partial D}>0$ if $\frac{T}{\bar{c}}>\frac{b}{(1-b)}$. In this case, which occurs if $b$ is small enough, the firm chooses $D=D_2$ for which its profit is maximum, while for $\frac{\partial Y(D)}{\partial D}<0$, the firm’s profit is maximized at $D=D_1$.

Finally, when $D>D_2$, $\varphi ^{*}\left( {p,D} \right) =1$, $p^{*}\left( D \right) =a+\left( {1-b} \right) \left[ {\frac{\bar{c}}{2}+T} \right]$, so that $\frac{\partial p^{*}}{\partial D}=\frac{\partial \varphi ^{*}}{\partial D}=\frac{\partial \varphi ^{*}}{\partial p}=\frac{\partial Y(D)}{\partial D}=0$.

Proposition 2

If $R_E \ne R_D$ Eq. (9) becomes:

$$\begin{aligned} \frac{\partial Y(D)}{\partial D}=\frac{T}{\bar{c}}\left[ {\left( {1-b} \right) \frac{T}{\bar{c}}-b} \right] +R_E -R_D \end{aligned}$$

(10)

and denoting $\frac{T}{\bar{c}}\equiv z$ we have that $\frac{\partial Y(D)}{\partial D}>0$ if $b<\frac{z-(R_D -R_E )/z}{1+z}$.

See Tables 5 and 6.

Table 5 Variable definitions and sources

Table 6 Regulation summary