By conducting an extensive investigation into whether rating agencies have grown more conservative over time, the authors generate results that carry implications for firm capital structure, bond issuance, and growth prospects.
What’s Inside?
Working with a robust data set that covers a 25-year period, the authors discover that rating agencies have become more conservative in their corporate bond ratings. They conclude that this shift is not entirely justifiable because defaults have decreased over the same time period. The firms that are subject to more conservative ratings have less leverage because they borrow less as well as experience lower growth.
How Is This Research Useful to Practitioners?
Conducting robust multistage research to evaluate an extensive data set over a long time period (1985–2009), the authors conclude that debt ratings have become more conservative. Keeping firm characteristics constant, they show that the average firm credit ratings have declined by three notches over the study period. The authors also show that the observable default rates for weaker investment-grade and stronger non-investment-grade issuers have declined during the same period, suggesting tighter rating standards.
They study the implications of this increased conservatism and the impact on corporate behavior. The authors conclude that the firms most affected by conservatism issue less debt, have lower leverage, are less likely to seek a debt rating, have higher cash holdings, and have a negative impact on growth and investments in acquisitions.
The authors also study the impact of rating agency conservatism on debt pricing. They conclude that the capital markets also take this stringency into consideration because the affected firms have lower credit spreads than unaffected firms. In addition, capital markets only partially offset the increased interest cost for affected firms.
As an aside, the authors’ findings seem to imply that rating agencies’ behavior with respect to corporate bonds differs substantially from their behavior with respect to mortgage-backed securities.
Corporate bond analysts and portfolio managers should find the implications of the authors’ research to be an important input in the asset management process because they offer historical perspective and possible insight into rating agency behavior. Regulators and policymakers may find the authors’ conclusions surprising in view of rating agency conflict of interest issues brought to light during the recent financial crisis.
How Did the Authors Conduct This Research?
Looking at a 25-year period (1985–2009), the authors gather monthly long-term (senior unsecured) debt ratings data issued by Standard & Poor’s, Moody’s Investors Service, and Fitch Ratings. Fewer instances of split ratings across agencies imply high correlations between agencies and a consistency of opinion that supports the authors’ conclusions. They match the rating data with companies’ annual financial statement data from Compustat.
The analysis then turns to the estimation of a series of rating models, with such explanatory variables as various debt ratios, profitability ratios, measures of asset volatility, and idiosyncratic risk from both prior literature on the subject and industry practice. All variables are adjusted so that the arithmetic mean is less sensitive to outliers. The authors adjust their models accordingly to include firm fixed effects, and they estimate additional specifications for the model series, including such items as macroeconomic variables, adjustments for pension liabilities, and an allowance for nonlinearities.
The authors then explore the implications of these conclusions for firm default rates, testing for statistical significance. They also examine the effects of rating conservatism on firms’ capital structure (degree of debt issuance) and cash holdings decisions using regression model estimation.
Notwithstanding all of the foregoing adjustments and progressive examinations, the authors’ conclusions hold. Debt ratings over time have grown more conservative. Default rates seem to have declined over the same time period, however, which implies that heightened rating stringency may not be entirely justified. Moreover, these conclusions hold up in the face of an alternative measure of rating conservatism—one that considers the difference between model predictions in the period 1986–1996 and the period 1997–2009.
Abstractor’s Viewpoint
You may be what you risk. Companies subject to more stringent credit evaluation exhibit less leverage and, correspondingly, less growth, according to the authors and the results of their extensive research. Yet, capital markets seem to imply that rating agency conservatism may have been unwarranted to a degree during this period of declining corporate bond defaults. Indeed, affected firms’ credit spreads are lower than those of unaffected firms receiving the same agency rating but higher than those of firms that would have prevailed without the more conservative regimen. Although this study investigates what happened, it fails to answer why it did, conjecturing only that early grand-scale defaults (e.g., Enron) or a correction of prior leniency may be behind the increased conservatism.