A New Method for Credit-Enhancement Standards

A new approach, based on corporate bond default data, is proposed to determine the quantity of new equity capital a bond issuer must consistently raise (or the amount of current debt it must retire) to enhance its bond ratings. The current approach of the rating industry to determine such standards for upgrading bond ratings assumes that the probability of bond default by the issuer is unaffected by the new equity capital or retiring of debt. The specification in the new approach relaxes this assumption. The estimates show dramatically different standards for bond-rating upgrades between the approach suggested here and the standard industry approach. Presented is a new approach for determining the quantity of new equity capital any bond issuer needs to add (or the amount of current debt it must retire) to enhance its bond rating. Because expected losses on bonds depend on the value of the borrower's future assets, this method specifies a distribution for the random future value of assets. Consistent with a large body of finance literature and practical applications, the assumption is that the future payoff to assets backing bonds (return per dollar of current asset value) in each rating category is distributed lognormally. Thus, the random (continuously compounded) rate of return on investment in assets of an entity has a normal distribution. Different return distributions for assets backing differently rated bonds make default probabilities and expected bond loss rates specific to each rating category.In the first step of the new procedure, the observed means and variances of losses on bonds in each Moody's Investors Service rating category are used to estimate the parameters of the lognormal distributions. The theoretical mean and variance of losses are then equated to the estimated mean and variance of losses to solve for the two parameters of the asset value distribution for each rating category. The parameter estimates identify the underlying asset-value distribution for each rating category. In the second step, these identified distributions are used to solve for the amount of equity capital infusion needed to reduce the current expected loss on bonds of an issuer to the expected loss for higher rated bonds.Next is a new calculation of the equity capital infusions needed for enhancing ratings of bonds aged three, four, and five years as reckoned from the date of issuance. The capital infusion estimates obtained from this new approach are much smaller than those obtained by the standard approach used by rating agencies. For example, the new approach finds that to enhance bond rating Baa to bond rating Aaa, the issuer must raise new equity capital equal to 3.02 percent of its current assets; in the standard industry approach, the new capital must be 55.71 percent of current assets. The main reason is that the new approach, unlike the standard industry approach, accounts for the fact that the default probability decreases because of the equity capital infusion.This methodology can help the rating industry revise its bond rating standards. It can also be applied to set bank capital standards. Bank regulators have recently expressed a serious interest in enforcing such standards by requiring banks to have their assets rated by public rating agencies.