Environmental Liability Information and Bond Ratings

Abstract

This research examines whether analysts' credit ratings of new bond issues reflect environmental liability information provided by the Environmental Protection Agency. Because environmental obligations provide no future economic benefits, we expect a negative relation between credit ratings and proxies for off-balance-sheet environmental obligations. To test this relation, we utilize a bond rating prediction model with a sample of new bond issues over the period 1990-1992. Our environmental liability estimates vary in size according to assumptions concerning the allocation of costs across responsible parties. Results indicate that environmental information is a significant negative factor in explaining bond ratings and increases the model's classification accuracy. Based on the overall analysis, the larger monetary measures of the firm's environmental liability are more consistent with bond rating analysts' assessment of the obligation. The general results are reinforced by additional analysis utilizing a model of bond rating changes over the period 1986-1997. This analysis indicates being named a potentially responsible party on additional environmental Superfund sites is associated with deterioration in a firm's credit rating when the firm has not previously been associated with a large number of these sites. This research establishes the value relevance of environmental information in debt markets and adds to prior studies that examine equity market effects.

Earnings Predictability, Bond Ratings, and Bond Yields

“…The KU model continues to remain robust in the literature and has been utilized directly or with minor variations in recent research as the primary reference for modeling bond ratings (e.g. Francis et al, 2003;Graham, Maher and Northcut, 2001;Sengupta, 1998;Shi, 2003;Ziebart and Reiter, 1992).…”

“…Our base model requires six independent variables: subordination status of the issue, a measure of firm size, a measure of leverage, a measure of profitability, a measure of income variation, and a measure of firm risk. Based on the work of Maher (1987) and Graham et al (2001), we enhance the benchmark model to include a net pension variable to incorporate the effects of a firm's defined benefit retirement obligations. Finally, to control for industry effects we include a series of indicator variables representing specific industry groupings.…”

“…This procedure has been previously utilized in the accounting and finance literature to control for industry (e.g. see Bradley, Jarrell and Kim, 1984;Morck, Shleifer and Vishny, 1988;Graham, Maher and Northcut, 2001) and year effects (see Collins and Kothari, 1989;Barth, McNichols and Wilson, 1997). The earnings predictability measures (EARNSURP and EARNAGREE) are added individually to the benchmark bond rating model to determine if either of them is associated with the rating assigned by the bond rating analysts.…”

Determinants of Corporate Social Responsibility Disclosure Ratings by Spanish Listed Firms

“…For instance, there is extensive evidence that social and environmental information is useful for decision-making by financial stakeholders (Blacconiere and Northcut, 1997;Blacconiere and Patten, 1994;Graham et al, 2000;Richardson and Welker, 2001). However, with its financial stakeholders' focus, it fails to provide a comprehensive theoretical foundation to explain CSR disclosure, especially since most of that disclosure is nonfinancial.…”

Who Uses Financial Reports and for What Purpose? Evidence from Capital Providers

“…In fact, simple models using as explanatory variables return-on-assets, debt-to-assets, firm size, dividend payment, and indicators on whether the firm has subordinated debt or negative ROA, can explain up to 66 per cent of the cross-sectional variation in S&P credit ratings (Barth, Beaver and Landsman, 1998;Barth, Hodder and Stubben, 2008). This suggests that accounting information is a major input, but also that credit analysts gather other information (Graham, Maher, and Northcut 2001;Jorion, Liu and Shi, 2005;Lee, 2002).…”