DefaultRisk.com the web's biggest credit risk modeling resource. | 
|
|
|
| Better Predictions of Income Volatility Using a Structural Default Model by Roger M. Stein of Moody's Investors Service, and November 26, 2005 Abstract: We propose a novel approach to predicting future volatility of company earnings, in this case EBITDA. Our approach combines predictions of a firm's probability of default with insights from a relatively less popular a structural model of default. The source of the probabilities of default can be econometric, structural, reduced-form or other models or agency ratings, provided the source has high predictive power. We use these probabilities to imply EBITDA volatility using a stylized, liquidity-based model of firm default similar in some ways to that originally proposed by Wilcox (1971) and based on the familiar Gambler's Ruin probability problem. The method does not require market information and our out-of-sample testing suggests that our approach is more accurate in estimating future volatility than the historical volatility of EBIDTA. Importantly, the method also produced reasonable estimates of volatility when historical data is quite limited, for instance when no historical financial data are available for the firm. In addition in comparison with historical volatility estimates the implied volatility estimates appear provide incremental information useful in identifying those firms that are more likely to experience EBITDA. Beyond implied volatility, we explore extensions of the approach for estimating implied liquidity requirements and target growth rates for firms, given a starting capital structure and variable cash flow stream. JEL Classification: C13, C14, C53, D81, G30, G31, G33. Keywords: earnings volatility, volatility predictions, structural model, default prediction, gambler's ruin, probability models, volatility evaluation. Books Referenced in this Paper: (what is this?)
[Home] [Other Credit Risk Papers]  |   
|
Please contact me with problems or suggestions. |
