Fundamental Analysis v/s Statistical Analysis

Two useful methods of determining credit risk and relative value of issuers of debt.

Fundamental Analysis

Analyst using this method would review past records of firms assets, earnings, sales, products, management and markets to predict the future of the firm. No other model can substitute in-debth fundamental analysis.

There are distinct advantages of Fundamental Analysis done by internal credit analyst and external credit agencies.

Advantages and Disadvantages of Fundamental Analysis by Individual Analyst:

AdvantagesDisadvantages
Forward LookingAnalysts are Expensive
Does not respond to market movements and events
Close and In Depth monitoring of Firm's activitiesSingle Analyst Opinion
Helps Avoiding DefaultsDifficulty to Quantify
potential Event RiskIntermittent, incomplete and inconsistent coverage of universe firms

Advantages and Disadvantages of Fundamental Analysis by Credit Agency:

AdvantagesDisadvantages
Ratings Made by CommitteeAgencies are slow to react to credit events
Ratings by Agencies are generally correctHesitant to change a rating and often prefer conservative side
Extensive coverage of debt issuersSubscription is Expensive
Provide an accurate long term estimate of the likelihood that a credit will defaultRatings changes tend to trail changes in spreads
Using average default rates by agencies ignores effects of the credit cycle on default rates.

Statistical Models

Statistical Models attempts to measure credit quality are based on observable characteristics of firms and the markets and their historical relationship to credit performance. First quantitive measure for separating defaulted firms from non-defaults using ratios is Altman’s Z-score model. This model uses five financial variables:

X1 = Working Capital / Total Assets
X2 = Retained Earnings / Total Assets
X3 = Earnings Before Interest and Taxes / Total Assets
X4 = Market Value of Equity / Total Liabilities
X5 = Sales / Total Assets
Z score model:
Z-Score = 1.2X1 + 1.4X2 + 3.3X3 + 0.6X4 + .999X5
Bankruptcy Range :
Z -Score > 2.99           –  Considered “Non-Bankrupt”
1.81 < Z-Score < 2.99 – Considered “Gray”
Z -Score < 1.81           – Considered “Bankrupt”

Although Z-Score is attractive in its use of financial variables shown to be related to default, the model has been criticized as being backward looking and intermittent.

Other modelers have used regression techniques to model credit spreads. One of the most used approach is that of Collin-Dufresne Goldstein and Martin (2001). Six financial variables are examined, each influence changes in credit spreads. Disadvantage of this regression is cannot be applied to all the firms as it needs sufficient coverage. Because of various limitations of the approaches providing default probabilities, most of the credit evaluator firms use reduced form models from risk neutral pricing theory for monitoring credit risk.