Article Detail

Empirical Analysis, Trading Strategies, and Risk Models for Defaulted Debt Securities

Journal 32: Applied Finance

Michael Jacobs, Jr.

This study empirically analyzes the historical performance of defaulted debt from Moody’s Ultimate Recovery Database (1987-2010). Motivated by a stylized structural model of credit risk with systematic recovery risk, we argue and find evidence that returns on defaulted debt covary with determinants of the market risk premium, firm specific and structural factors. Defaulted debt returns in our sample are observed to be increasing in collateral quality or debt cushion of the issue. Returns are also increasing for issuers having superior ratings at origination, more leverage at default, higher cumulative abnormal returns on equity prior to default, or greater market implied loss severity at default.

Considering systematic factors, returns on defaulted debt are positively related to equity market indices and industry default rates. On the other hand, defaulted debt returns decrease with short-term interest rates. In a rolling out-of-time and out of-sample resampling experiment we show that our leading model exhibits superior performance. We also document the economic significance of these results through excess abnormal returns, implementing a hypothetical trading strategy, of around 5-6 percent (2-3 percent) assuming zero (1bp per month) round-trip transaction costs. These results are of practical relevance to investors and risk managers in this segment of the fixed income market.

There exists an economic argument that to the extent there may be opportunity costs associated with holding defaulted debt, and that the performance of such debt may vary systematically, the required return on the defaulted instruments should include an appropriate risk premium. Thus far, most research studying systematic variation in defaulted debt recoveries has focused on the influence of either macroeconomic factors [Frye (2000 a,b,c; 2003), Hu and Perraudin (2002) Cary and Gordy (2007), Jacobs (2011)], supply/demand conditions in the defaulted debt markets [Altman et al. (2003)], or some combination thereof [Jacobs and Karagozoglu, (2011)]. Probably the reason for this focus is the conventional wisdom that determinants of recoveries (i.e., collateral values) are thought covary with such systematic macroeconomic measures. However, the results concerning systematic variation in recoveries have been mixed. We believe that this is due to the unmeasured factors influencing the market risk premium for defaulted debt. Adequately controlling for other determinants of defaulted debt performance, potentially imperfectly correlated with standard macroeconomic indicators, is critical to understanding this.

We propose to extend this literature in several ways. First, we quantify the systematic variation in defaulted debt returns with respect to factors which influence the market risk premium for defaulted debt, which are related to investors’ risk aversion or investment opportunity sets; in the process, we specify a simple stylized model of credit risk in structural framework [Merton (1974)], having testable implications that are investigated herein. Second, we are able to analyze defaulted debt performance in segments homogenous with respect to recovery risk, through controlling for both firm and instrument specific covariates, and examine whether these are associated with recoveries on defaulted debt securities. Third, departing from most of the prior literature on recoveries, having predominantly focused on measures around the time of default or at settlement, we will be studying the relationship amongst these in the form of returns. We believe that such focus is most relevant to market participants – both for traders and buy-and-hold investors (i.e., vulture funds, or financial institutions managing defaulted portfolios) – since this is an accepted measure of economic gain or loss. Finally, we are able to build parsimonious and robust econometric models, in the generalized linear model (GLM) class, that are capable of explaining and predicting defaulted debt returns, and we use these to construct trading strategies demonstrating their economic significance.

In this study, we quantify the performance of defaulted debt relative to the previously and newly proposed determinants of corporate debt recoveries, through a comprehensive analysis of the returns on this asset class. The dataset that we utilize, Moody’s Ultimate Recovery Database™ (MURD™), contains the market prices of defaulted bonds and loans near the time of default, and the prices of these instruments (or market value of the bundle of instruments) received in settlement (or at the resolution) of default. We have such data for 550 obligors and 1368 bonds and loans in the period 1987-2010. We examine the distributional properties of the individual annualized rates of return on defaulted debt across different segmentations in the dataset (i.e., default type, facility type, time period, seniority, collateral, original rating, industry), build econometric models to explain observed returns, and quantify potential trading gains to deploying such models.

jailer