Recent literature has pointed out that information asymmetries may be the reason of the poor performance of structural credit risk models to fit corporate bond data. In fact it is well known that these models lead to a strong understatement of the credit spread terms structure, particularly on the short maturity end. Possible explanations stem from strategic debt service behavior and, as discovered more recently, the problem of accounting transparency. This raises the possibility that some of these flaws could be reconducted to a sort of peso problem, i.e., the market may ask for a premium in order to allow for a small probability that accounting data may actually be biased (Baglioni and Cherubini, 2005). In this paper the authors propose a modified version of the Duan (1994, 2000) MLE approach to structural models estimation in order to allow for this peso problem effect. The model is estimated for the Parmalat case, one of the most famous cases of accounting opacity, using both equity and Catalogue Data Services (CDS) data.
Structural models of credit risk are considered as a very elegant approach to the evaluation
of corporate liabilities. Elegance stems from the fact that prices are obtained from the
analysis of the structure of the balance sheet of the obligor firm and the dynamics of its
assets. The main advantage is that these models are full of economic information content,
while the so-called “reduced form” models are only based on statistical assumptions
concerning the probability distribution of default events and the recovery rate, that is the
amount which the investor expects to recover in case of default. Of course, the richer
economic content in structural models comes at the cost of a loss of flexibility with respect
to reduced form models, and of a poorer fit to market data.
Structural models are reconducted in the seminal paper by Merton (1974), even though
the famous Black and Scholes (1973) model was already devoted to the evaluation of
corporate liabilities, as was explicitly recognized in the title. The key idea in structural
models is that corporate liabilities, such as equity and debt, are actually positions in
options. So, equity can be considered a call option written on the assets of the firm with
a strike price equal to the face value of debt and the credit risk component of corporate
bonds can be thought of as short position in a put option with same underlying and strike,
the so called default put option. The first typical flaw of structural models of credit risk
is that predicted credit spread are much lower than market quotes for reasonable values
of leverage and volatility of assets. Several answers have been proposed as possible
solutions to this problem. Anderson and Sundaresan (1996) suggest that the owner of the
firm may engage in a strategic rescheduling process to exploit the bankruptcy costs at the
expense of bondholders. |
