Credit Ratings and Capital Structure

Darren J. Kisgen* University of Washington School of Business Administration Department of Finance and Business Economics

This Draft: May 29, 2003

___________________________________________________________________________
* I would like to thank Jonathan Karpoff, Jennifer Koski, Paul Malatesta and Edward Rice, for helpful comments. Please address inquiries to Darren J. Kisgen, Doctoral Candidate, University of Washington School of Business, Department of Finance and Business Economics, Mackenzie Hall Box 353200, Seattle, WA 98195-3200 or e-mail: kisgen@u.washington.edu.

Credit Ratings and Capital Structure

Abstract
This paper examines the impact of credit ratings on capital structure decisions. I find that firms near a change in credit rating issue (retire) annually up to 1.5% less (more) debt relative to equity as a percentage of total assets than firms not near a change in rating, reflecting managers’ desire to obtain upgrades and avoid downgrades of the firm’s credit rating. These results are shown to be distinct from financial distress concerns. The results are robust to several model specifications and inclusion of control variables. I further show how credit rating effects complement the pecking order and tradeoff capital structure theories, and find that dummy variables, indicating firms are near a change in rating, remain predictive when nested in previous empirical tests of these theories. The results of this paper do not appear to be explained with traditional theories of capital structure, and thus this paper enhances the capital structure decision theoretical and empirical frameworks. To my knowledge, this is the first paper to show that credit ratings directly affect capital structure decision-making.

I. Introduction This paper examines to what extent credit ratings, particularly potential upgrades or downgrades of bond ratings, directly affect capital structure decision making by financial managers. The paper outlines the reasons why credit ratings may be relevant for managers in the capital structure decision process, and then empirically tests the extent to which credit rating concerns directly impact managers’ debt and equity decisions. The paper also examines how these findings complement existing capital structure theories such as the pecking order and tradeoff theories, and specifically how these credit rating factors can be included in empirical tests of capital structure theories. The initial empirical tests of this paper examine whether capital structure decisions are affected by credit ratings. Using two distinct measures, I distinguish firms that are close to having their debt downgraded or upgraded versus firms that are not close to being downgraded or upgraded. Then controlling for firm specific factors, I test whether firms that are near a change in rating issue less net debt relative to net equity over a subsequent period when compared to the control group. I find that concerns about upgrades or downgrades of bond credit ratings directly affect managers’ capital structure decision making; firms near a change in credit rating issue (retire) annually up to 1.5% less (more) debt relative to equity as a percentage of total assets than firms not near a change in ratings. Firms with a credit rating designated with a plus or minus issue less debt relative to equity than firms that do not have a plus or minus rating, and when firms are ranked within each specific rating (e.g., BB-) based on credit quality determinates, the top third and lower third of firms within ratings are also shown to issue less debt relative to equity than firms that are in the middle of their individual ratings. These results do not appear to be explained with traditional theories of capital structure, and thus this paper enhances the capital structure decision theoretical and empirical frameworks. To my knowledge, this is the first paper to show that credit ratings directly affect capital structure decision-making.

1

The influence of credit ratings on capital structure is economically significant and statistically robust. The relationship is apparent whether the dependent variable reflects debt and equity issuances or debt issuance only, and whether control variables are included. The relationship holds when both an OLS regression approach is used for continuous capital structure dependent variables and when a logit regression is used to examine binary capital structure choices. The relationship holds for both large and small firms, and for firms at several credit rating levels. After the initial tests in the paper establish these facts, subsequent empirical tests nest these credit rating factors into previous capital structure tests, such as those in Fama and French (2002) and Shyam-Sunder and Myers (1999). Dummy variables, indicating firms are near a change in rating, remain statistically significant when nested in the capital structure empirical tests of both of these papers. Motivation for this effort begins with the observation that corporate financial managers care about credit ratings. In their survey of CFOs, Graham and Harvey (2001) find that credit ratings are the second highest concern for CFOs when considering debt issuance. When asked what factors affect how they choose the appropriate amount of debt for their firm, Graham and Harvey found that 57.1% of CFOs said that “Our credit rating (as assigned by credit rating agencies)” was important or very important. “Financial flexibility” was the only category higher, with 59.4%, and therefore credit ratings ranked higher than many of the factors implied by traditional capital structure theories (such as the “tax advantage of interest deductibility”). Although credit ratings are the second most important factor for managers regarding their debt issuance, the result is discussed only minimally. Graham and Harvey list various capital structure theories and indicate how the survey results support or contradict the theories. In this discussion however, the credit rating result only appears where they argue it might support the tradeoff theory: “credit ratings [concern]…can be viewed as an indication of concern about distress” (pg. 211). The arguments and results of this paper in most cases can be viewed as distinct from financial distress arguments. The empirical tests of this paper examine firms that are near both upgrades and downgrades. The upgrade tests are distinct from financial distress arguments, since they test if firms near an upgrade issue less debt relative to equity versus firms that are not near an upgrade; since the control group of firms has greater financial distress concerns in these tests, 2

financial distress arguments would have the opposite implications versus credit rating arguments (e.g., B+ firms issue less debt relative to equity than B firms, a finding inconsistent with distress arguments but consistent with credit rating effects). Further, I also include control variables in the empirical tests that specifically control for the financial condition of the firm. Additionally, the effects of credit ratings imply discrete costs/benefits at particular changes in credit ratings, which is distinct from the modeling and testing of financial distress concerns that are generally continuous and not directly tied to credit rating measures. Lastly, I examine firms at all ratings levels, and the results are fairly consistent across the ratings spectrum (with significant results at the AA rating level alone, for example). Although this is the first paper to examine the direct effects of credit ratings on capital structure decisions, significant research has been conducted examining how credit ratings affect stock and bond valuations. Hand, Holthausen and Leftwich (1992) find statistically significant negative average excess bond and stock returns to announcements of downgrades of straight debt. Ederington, Yawitz and Roberts (1987) and West (1973) find that credit ratings are significant predictors of yield to maturity beyond the information contained in publicly available financial variables and other derivable factors that would predict spreads. Ederington and Goh (1998) show that credit rating downgrades result in negative equity returns and that equity analysts tend to revise earnings forecasts “sharply downward” following the downgrade. They further conclude that this action is a result of the “downgrade itself – not to earlier negative information or contemporaneous earnings numbers.” Thus evidence exists that suggests credit ratings are significant in the financial marketplace; this paper takes the next step and analyzes to what extent they are significant in capital structure decision making. The rest of this paper is organized as follows. In section II, I provide explanations for why credit ratings might factor into managerial capital structure decisions. In Section III, I detail how credit rating concerns complement existing theories of capital structure. Section IV contains general empirical tests of the impact of credit ratings on capital structure decisions, and Section V contains specific tests that nest credit rating factors into empirical tests of traditional capital structure theories. Section VI concludes.

3

and in 1989. Also. banks have been restricted from owning junk bonds going back as far as 1936 (Partnoy (1999) and West (1973)). to the extent that the demand curve for bonds is 4 . yields on bonds with higher regulatory costs will be higher to compete with bonds that have lower regulatory costs. mutual funds. the SEC adopted Rule 15c3-1 whereby the SEC uses credit ratings as the basis for determining the percentage reduction in the value (“haircut”) of bonds owned by broker-dealers for the purpose of calculating their capital requirements (Partnoy (2002)). Since 1951. Cantor and Packer (1994) observe “the reliance on ratings extends to virtually all financial regulators. The primary testable implication of CR-CS considered in this paper is that concern for the impact of credit rating changes directly affects managers’ capital structure decision-making. Section IV tests generally if concerns for these reasons translate into specific capital structure behavior. Specific Hypotheses for Significance of Credit Ratings The fundamental hypothesis of this paper is that credit ratings are a material consideration for managers in making capital structure decisions due to discrete costs/benefits associated with different ratings levels (henceforth referred to as the Credit Rating Capital Structure Hypothesis or “CR-CS”). any C level gets a 5. This section describes the specific reasons that credit ratings might be significant in capital structure decisions. capital markets. and private pensions. To the extent that regulations affect the cost to investors of investing in a particular bond class. insurance companies’ investments in securities of firms that are rated A or above get a value of 1. and any D rating gets a 6. regulatory agencies determine capital requirements for insurance companies and broker-dealers using credit ratings as a scoring system. firms that are BBB get a value of 2. whereby firms near a ratings change will issue less net debt relative to net equity versus firms not near a ratings change. insurance companies. Further. In 1975.” For example. ceteris paribus. thrifts. securities firms.II. BB get a 3. Regulatory Effects Several regulations on financial institutions and other intermediaries are directly tied to credit ratings. B a 4. including the public authorities that oversee banks. A. Savings and Loans were prohibited from investing in junk bonds such that they could not hold any junk bonds by 1994.

For example. Third Party Relationships Credit ratings may materially affect relationships with third parties. Patel. Collin-Dufresne. Third 5 . B. placing a restriction on certain investors participating in a particular bond market will cause the yield to increase in that market. Further.g. Therefore although a firm itself may not have any higher risk of default. firms entering into long-term supply contracts may require certain credit ratings from their counterparty. these different groups of investors may have different trading practices that may increase or decrease the liquidity in these respective markets. C. at certain credit rating levels (e. Goldstein and Martin (2001) argue. Market Segmentation Bond markets may be segmented by credit rating. and so forth. If firms incur higher interest rates in less liquid markets distinguished by credit rating. suppliers to the firm. junk bond levels) during difficult economic times. it may be required to pay a higher interest rate on its debt merely as a result of its credit rating.” This suggests that spreads on bonds distinguished by credit rating could diverge enough from what is implied by traditional factors alone to be significant for managers’ capital structure decisions. Firms with these ratings may therefore have to forgo positive NPV projects due to their inability to finance projects at those times.. Also. Different classes of investors for different markets distinguished by credit rating may create unique supply and demand characteristics that would result in yield spreads diverging in different markets.” West (1973) notes. “the dominant component of monthly credit spread changes in the corporate bond market is driven by local supply/demand shocks. including the employees of the firm. or customers of the firm. leverage. Evans and Burnett (1998) find that liquidity affects whether junk bonds experience abnormal positive or negative returns. a firm may not be able to raise debt capital because investors avoid issues with those ratings during these times. there may be incentives to avoid these ratings levels beyond what is implied by traditional models.downward sloping. “bonds not in the top four rating categories had yields consistently above those that were predicted on the basis of earnings variability.

where perhaps a AAA rating is important for third party relationships versus a AA rating. D. 1 See Opler and Titman (1994) for a review of this literature. A credit rating can therefore act as a signal of overall firm quality. Credit agencies might also specialize in the information gathering and evaluating process and thereby provide more reliable measures of the firm’s creditworthiness. even though a firm may be a particularly good BB. as firms may be reluctant to provide information publicly that would compromise their strategic programs. whereas the financial distress literature implies continuous changes in costs as firms increase their probability of bankruptcy. they will be pooled into the group of all firms in that worse credit class.firms. Millon and Thakor (1985) propose a model for the existence of “information gathering agencies” such as credit rating agencies based on information asymmetries.party relationship arguments are in some ways similar to arguments made in the financial distress literature1. Rating agencies may receive significant sensitive information from firms that is not public. credit rating effects imply discrete costs associated with a change in rating. Likewise. its credit spreads would not be lower than credit spreads of other BB.for example. Pooling Effects Credit ratings may provide information on the quality of a firm beyond what is discernable from publicly available information. firms that are near an upgrade will have an incentive to obtain that upgrade to be pooled with firms in the higher ratings category. CR-CS applies to financially strong firms as well. Firms would then be pooled with other firms in the same rating category. Additionally. where in the extreme all firms within the same ratings group would be assessed similar default probabilities and associated yield spreads for their bonds. Thus. They argue that credit rating agencies are formed to act as “screening agents” certifying the values of firms that approach them. in particular with regard to competitors. however. Otherwise. if they are given the lower rating (even though they are only a marginally worse credit). 6 . Firms that are near a downgrade in rating will then have an incentive to maintain the higher rating.

Section V includes tests that nest credit rating factors into the empirical tests of these theories. managers may target a higher credit rating than might be optimal for overall firm value. Enron faced $3. For example. the survey showed that at least 20% of the companies surveyed have exposure to some sort of contingent liability.S. whereby a downgrade would be compounded by provisions such as ratings triggers or covenants that could create a liquidity crisis. positive news (e. Further. it may be a disadvantage to come from a junk bond rated firm. For example. If credit ratings affect a manager’s reputation.g. forcing certain actions to be taken by the firm given a downgrade that may be costly). bond covenants may be directly tied to the credit rating of the firm.. F. I set forth empirical implications of these theories and consider generally how credit rating concerns might be integrated into these models. Ratings Triggers Firms may be concerned about credit ratings since triggers may exist for changes in ratings (for example.9 billion in accelerated debt payments as a result of a credit rating downgrade. if a manager wishes to change jobs. Negative credit rating developments may also have negative consequences for a financial manager with regard to his job security or compensation. Likewise. Standard and Poor’s (2002) recently surveyed approximately 1. and European investment-grade issues and found that 23 companies show serious vulnerability to rating triggers or other contingent calls on liquidity. or it might be an advantage to have worked at an AAA-rated company. III.E. an upgrade to AAA) may be considered positively in compensation decisions. Manager’s Utility Management’s own maximization of utility may create credit ratings concerns.000 U. Credit Ratings in the Context of Existing Capital Structure Theories This section examines how CR-CS complements current capital structure theories. specifically the tradeoff and pecking order theories. 7 .

a BB+ rating).g. managers will balance that cost against the costs and benefits implied by the tradeoff theory. for example). In certain cases. one could consider several disjoint jumps in value corresponding to several different credit ratings. An implication of the tradeoff theory is that a firm will tend to move back toward its optimal leverage to the extent that it departs from its optimum (see Fama and French (2002). C* is the new optimum considering credit rating effects as well. CR-CS is not restricted to this. shown as the point T*. If this cost is material. Tradeoff Theory The tradeoff theory argues that a value-maximizing firm will balance the value of interest tax shields and other benefits of debt against the costs of bankruptcy and other costs of debt to determine an interior optimal leverage for the firm. a firm may face the situation depicted in Figure 1a. 2 8 . To illustrate this. Further assume that the optimal leverage as implied by the tradeoff theory is a leverage that will cause the firm to have a high rating within junk bond ratings (e.A.2 If there is no discrete cost related to credit ratings. assume that the main discrete cost created by credit ratings is in the change from investment grade to junk bond status. the tradeoff theory factors may outweigh the credit rating considerations.. This situation is depicted in Figure 1b. The benefits from that rating might be large enough to give up whatever benefits are lost relating to tradeoff theory effects. Now consider a firm that faces a discrete cost (benefit) at the change from investment grade to junk bond status. that firm will be interested in For simplicity this discussion considers only one rating where a discrete cost occurs. CR-CS states that a credit rating change has a discrete cost/benefit. as is shown later. This graph depicts firm value as a function of leverage and illustrates a tradeoff between the benefits and negatives of higher leverage. A firm in this position might choose a smaller leverage than implied by the tradeoff theory to obtain an investment grade rating. the costs associated with a change in credit rating may then result in different capital structure behavior than implied by the tradeoff theory alone. In other cases. A firm in this situation will choose the leverage that maximizes firm value. T* corresponds to the optimal capital structure considering only tradeoff effects. Note also that a firm at C* is going to be cautious about its capital structure policy – in particular.

Figures 1c and 1d depict cases where CR-CS effects are outweighed by tradeoff theory effects. the firm may chose to stay at that rating because. Likewise.g. Similar graphs can be depicted where firms choose a different optimum as a result of any of the potential credit rating jumps. The pecking order model implies debt will 9 . it is possible that credit rating effects will be relevant for the firm. a CCC. The graph shows one example where credit rating effects create an optimum that is different from tradeoff predictions alone. Also consider a firm that is slightly to the right of C*. In this case. Figure 1e shows a more complete depiction of the tradeoff theory combined with credit rating effects by showing several jumps.. and only when internal funds have been extinguished and a firm has reached its debt capacity will a firm issue equity. B. since a large offering could create a downgrade for them. regardless of a firm’s optimal leverage as predicted by the tradeoff theory alone. however it is also consistent with CR-CS that these firms will still be concerned about the potential effects of a large debt offering. Here. a firm in that position is not affected by that potential credit rating cost. Firms will prefer to fund projects first with internal funds and then with debt.. Figure 1c depicts a firm whose optimal leverage as implied by the tradeoff theory implies a high rating for the firm (e. That firm will be very interested in getting an upgrade and will therefore be more likely to issue equity relative to debt. Pecking Order Theory The pecking order theory argues that firms will generally prefer not to issue equity due to asymmetric information costs (Myers (1984)). Note that firms that are somewhat farther away from a downgrade will have less concern for a small offering of debt. firms that are relatively far from an upgrade may consider a large equity offering to get an upgrade. although there are benefits to be obtained by achieving an investment grade rating for the firm.rating). however they would be less likely to issue smaller equity offerings versus firms that are very close to an upgrade. Figure 1d depicts a firm with an optimal leverage as implied by the tradeoff theory that implies a credit rating that is low-rated junk (e. If the only credit rating that creates a discrete cost is the change to junk bond status.g.avoiding debt issues relative to equity issues to avoid a downgrade. the benefits that the firm gives up as implied by the tradeoff theory may be more significant. an A rating).

Thus the dependent variable in the regressions that follow will be a measure of the amount of debt and/or equity raised or a binary decision variable indicating a choice between debt and equity. CR-CS has less direct implications regarding the absolute levels of leverage or debt for a company. As such. in some cases firms that are near an upgrade may choose to issue equity instead of debt in order to obtain the benefits of a higher rating and firms that are near a downgrade may avoid issuing debt to prevent the extra costs that result with a downgrade. 10 . The pecking order predicts a strong shortterm response of leverage to short term variations in earnings and investment. The book value measures are also under the direct control of managers and therefore directly reflect managerial decision-making. the tests examine if firms close to a ratings change issue less debt relative to equity (or simply less debt) to avoid a downgrade or increase the chances of an upgrade. the dependent variables in the tests reflect changes in debt and/or equity or a discrete decision to issue equity or issue debt.near to an upgrade as well as a downgrade. This conflict will exist most strongly for firms that are near a change in rating . in contrast to any concern for reverting to a target level. CR-CS directly predicts capital structure decisions over a subsequent period based on the credit rating situation a firm faces at a particular point in time. Since I am concerned with credit ratings. This is distinct from some tests of capital structure that may imply certain levels of leverage for a company based on company factors or other measures. Therefore contrary to the implications of the pecking order. IV. book values of equity and debt are used. Empirical Tests of CR-CS The main empirical tests examine if firms that are close to being upgraded or downgraded to a certain credit rating undertake different capital structure decisions compared to other firms. CR-CS implies that for some incremental change in leverage. a discrete cost/benefit will be incurred due to a credit rating change. whereas. a firm will face a tradeoff between the costs of issuing equity versus the discrete cost associated with a potential change in credit rating. as these are the variables credit rating agencies emphasize.increase for firms when investment exceeds internally generated funds and debt will fall when investment is lower than internally generated funds. In particular. Assuming that for some level of leverage both CR-CS and pecking order effects are material.

firms in the middle of a Broad Rating might reasonably be concerned with upgrades or downgrades to a plus or minus category. managers must care more about a change in Broad Rating than a change in a rating of any kind. For example.refers only to BBB-..and compute scores for each of them and separate those firms into high. For the first measure.. I define “Micro Ratings” as specific ratings that include a minus or plus modification if given. a strong BB. which would reduce the precision of the tests. I then separate firms within each Micro Rating into a high third. middle and low thirds. I distinguish firms as near a ratings change if their rating is designated with either a “+” or a “-“ within a Broad Rating and not near a ratings change if they do not have a plus or minus modifier within the Broad Rating (they are in the middle of the Broad Rating). BB.g. I compute a “Credit Score” for each firm that assigns a credit quality value to each firm based on firm data used by rating agencies. Using these Credit Scores. A disadvantage of this is that the distinctions might be too broad. a Broad Rating of BBB refers to firms with ratings of BBB+.Being close to a ratings change is measured in two ways. For example. For the second measure. middle third and low third ranked by their respective Credit Scores. BBB and BBB-. Tests using this measure are designated “Plus or Minus” tests (or “POM tests”). etc. I define “Broad Ratings” as ratings levels including the minus. interest coverage.firm may not be near a downgrade within the BB Broad Rating and likewise a weak BB+ firm may not be near an upgrade. and the Micro Rating of for example BBB. Thus the Micro Rating of BBB refers only to BBB firms. A second disadvantage of this measure is that for the tests using this measure to be most effective.and BB+ firms are defined to be near a ratings change and firms that are BB are not. such as debt/equity ratios. The second measure takes firms within each Micro Rating and ranks them within that rating based on the factors that tend to indicate credit quality. since the ratings themselves are used to distinguish firms. The advantage of this measure is that this measure should accurately reflect being near a change in rating. I take all of the firms that are BBB. middle and plus specification for a particular rating. For example. e. with weightings determined by regressing ratings on these factors. For example. For this. This will increase the noise in these tests. as opposed to BBB+ or BBB-. Firms that are in the high or low third of a Micro Rating are then 11 . within the Broad Rating of BB.

generally defined to be a net offering during the year that exceeds 10% of total assets. I also exclude financial companies and utilities (SIC codes 40004999 and 6000-6999). and the firm’s capital structure decision measures are computed for the subsequent 12 months. Consistent with prior literature. Lastly. and 216. The credit rating used is Standard & Poor’s Long-Term Domestic Issuer Credit Rating (Compustat data item no. 108. capital structure transactions also require time to execute. there may be several years where managers are not undertaking offerings of any kind for a firm.3 For the majority of the empirical tests. the credit rating situation for a firm could change during the middle of a year. if the Credit Scores are measured correctly. which should increase the precision of these tests. 6. The exclusion of large offerings is made for several reasons. 115. 12 . An advantage of this measure is that. 34. Furthermore. The measure will also account for all potential ratings changes thereby rectifying potential problems with the previous measure. For example. 280). so there may be a significant lag from a decision being made to the time it appears in the data. I exclude large offerings from the sample. 111. I also exclude firm years where the firm has missing data in the fields required regularly in the calculations of the tests in the paper. This rating represents the issuer’s senior debt rating. making the credit rating measure at the beginning of the year inaccurate. the test group of firms should be very close to a ratings change. and is usually the highest rating the company has on its long-term debt.considered to be near a change in rating whereas the firms in the middle third are not. 9. Credit rating dummy variables are created for firms at the end of each calendar year. 114. 13. and this may reduce the effectiveness of the test. A complicating factor in these tests is that changes in capital structure have material transactions costs and as a result capital structure changes are lumpy and sporadic. The decision to conduct a large offering is likely to 3 These are Compustat data item no’s. The sample period is 1986-2001 (1985 is the first year that this rating is available in Compustat). Tests using this measure are designated “Credit Score” tests. The disadvantage of this measure is that the measurement of the Credit Scores will be noisy in themselves. The beginning sample of firms for the tests is all firms with a credit rating in Compustat at the beginning of a particular year.

this restriction excludes approximately 14% of the firm years. I conduct robustness checks of these restrictions. When both large debt and equity offerings are excluded. reorganizations or changes in management. and if only large debt offerings are excluded. including large offerings could therefore confound the results. i and t subscripts are suppressed for the credit rating dummy variables): Dit = book long-term debt plus book short-term debt for firm i at time t (Compustat data item no. Since equity offerings involve larger transactions costs. a change in debt greater than 10% of total assets is likely to cause a change in Broad Rating for all firms. this restriction excludes approximately 16% of firm years. Since the dependent variable is a continuous measure of capital issuance. 13 . some of the more commonly used notation is defined as follows (for notational convenience. 111 minus data item no. CR-CS implies that firms not near a change in rating will conduct fewer large debt offerings conditional on conducting a debt offering of any kind than plus or minus firms (see the Appendix). and this could actually result in firms near a downgrade. 114 plus data item no. regardless of where they are in the Broad Rating. For the empirical tests detailed in this section and the following section. Lastly. 301). Also. such as reorganizations due to financial distress (see Asquith. As a result. for example. This is distinct from changes in debt that can occur more fluidly. Thus including large equity offerings may not have some of the undesirable effects of including large debt offerings. conducting more large debt offerings than firms in the middle (even though they conducted fewer offerings in total). Gertner & Scharfstein (1994)).have the same Broad Rating consequences for all firms – that is. I consider restrictions on both large debt and equity offerings as well as restrictions on large debt offerings only. large debt issues as opposed to large equity offerings more typically accompany some of the events described above. they happen less frequently and are on average larger. 34). there should be no distinction among the different credit rating groups for these types of offerings so including them would make the results noisier. Also. large offerings might be associated with acquisitions. 9 plus data item no. ∆Dit = long-term debt issuance minus long-term debt reduction plus changes in current debt for firm i from time t to t+1 (Compustat data item no. and it is less likely that credit rating changes will be significant in these contexts.

Note that this variable reflects only changes in capitalization resulting from capital market transactions. 111 minus data item no. or for the previous year. Table I shows statistics for debt to total capitalization ratios by credit rating See Rajan and Zingales (1995) for a discussion of these control variables. CRHigh = dummy variable for firms that are in the top third of their Micro Rating with regard to their Credit Score at the beginning of the period. CRMinus = dummy variable for firms that have a minus credit rating at the beginning of the period. Kit = set of control variables. This excludes changes in equity resulting from earnings for the year. as described above. as described above. NetDIssit5 = (∆Dit . including Debt/Total Capitalization: Dit/(Dit+Eit) and EBITDA/Total Assets: EBITDAit/Ait (EBITDA is Compustat data item no. where appropriate).∆Eit)/ Ait A. as described above. CRLow = dummy variable for firms that are in the bottom third of their Micro Rating with regard to their Credit Score at the beginning of the period. ∆Eit = sale of common and preferred stock minus purchases of common and preferred stock for firm i from time t to t+1 (Compustat data item no. The factors are lagged values (calculated at the end of the previous year. Eit = book value of shareholders’ equity for firm i at time t (Compustat data item no. Summary Statistics Summary statistics for the sample are shown in Tables I and II. Ait = beginning of year total assets for firm i at time t (Compustat data item no. 216). 5 4 14 . CRPOM = CRPlus + CRMinus = dummy variable for firms that have a minus or plus credit rating at the beginning of the period. CRHOL = CRHigh + CRLow = dummy variable for firms that are in the top or bottom third with regard to their Credit Score at the beginning of the period. as I am interested in capital structure decision making. 13)4. 114). 108 minus data item no. 6). not changes in leverage that are a result of firm performance. CRPlus = dummy variable (set equal to 1) for firms that have a plus credit rating at the beginning of the period. The sample contains 6.∆LTDit = long-term debt issuance minus long-term debt reduction for firm i from time t to t+1 (Compustat data item no. as described above. 115).906 firm years. as described above.

This demonstrates that that the empirical results in this paper will encompass credit ratings as a whole. approximately 90% of firms will issue debt only or equity only versus issuing both during the year. 10 of the 17 ratings categories have between 250 and 600 firm years. The variances within each rating for the debt to total capitalization ratios are generally high. It appears that the levels for debt to total capitalization for AAA firms are inconsistent when compared to AA firms. 15 . Table II shows that nearly 40% of the sample raised debt only for the firm year compared to less than 10% issuing equity only. the potential for differences within each rating is significant. Although the range is from 65 AA+ firm years to 899 B+ firm years. with approximately 30% reducing debt levels only compared to 17% reducing equity levels only. Table II also shows that firms are more likely to use one form of financing (debt or equity) during the year as opposed to using both debt and equity. leaving again approximately half of the sample not reducing capital. Within offerings. For example. Size is a significant determinate for credit ratings as well as leverage. indicating that although a relationship exists between debt to total capitalization and ratings. not how much of both it should issue during the year. the top 4 credit ratings have median debt to total capital ratios ranging from 19% to 33% whereas the bottom 4 credit ratings have median debt to total capital ratios ranging from 74% to 86%. A firm’s decision for a particular year appears to be more a decision whether to issue debt or equity. conditional on an offering taking place. they are not doing this on an annual basis using both debt and equity offerings. Once again both a debt and equity reduction is rare (7%). To the extent firms are following the tradeoff theory and targeting a specific debt to capitalization level.1 billion compared to AA firms with median total assets of $5. Table I indicates the sample seems relatively well distributed by rating. The debt to total capitalization ratios have the expected relationships to ratings – for example. Table II shows the capital raising and reducing activity within the sample. and it also indicates the number of firm years by rating.within the sample. A small number issued both equity and debt (5%).6 billion). with an offering defined as a net amount greater than 1% of total assets. leaving nearly half of the firm years with no offerings. The propensities are similar for capital reductions. however AAA firms are on average significantly larger than AA firms (AAA firms have median total assets of $17. not just specific ratings categories.

firms that are near a change in credit rating are less likely to issue debt relative to equity than firms in the middle.As debt is the more employed instrument for changing capital levels by firms.1) (4.3)) ranging from –3. with t-statistics on the POM dummy variable (equation (4. I conduct tests using the Plus or Minus measure of being close to a change in rating. Throughout these tests. as implied by CRCS. The minus rating dummy variable might be biased toward being more significant 16 . B. Plus or Minus Tests In this section. CR-CS implies that firms with a minus or plus rating will issue less debt relative to equity than firms that are in the middle.2) (4. Equation (4. predicts that βi < 0. Results of these regressions are shown in Table III. The following three regressions are run to test this hypothesis: (4. Note that the null would be accepted for any positive values of the coefficient.2.1.3.1) and (4. The null hypothesis is βi ≥ 0.3) NetDIssit = α + β 0 CR POM + φK it + ε it NetDIssit = α + β 1CR Plus + β 2 CRMinus + φK it + ε it NetDIssit = α + β 3CR POM + ε it These equations test whether a firm’s net issuance of debt versus equity for a particular year is affected by how near that firm is to an upgrade or downgrade in their credit rating at the end of the previous period. The implication that firms that are near a significant ratings change will have more conservative debt financing policies versus firms in the middle. I will also examine the debt decision specifically in the empirical tests. i = 0.94. with dummy variables constructed for firms with a minus or plus credit rating within their Broad Rating (or both). The sign for the coefficient is as predicted as well.05 to –5. the null that firms are indifferent to being near a credit rating change for capital structure decisions is rejected at the 1% level. These results provide strong rejection of the null in favor of CR-CS.2) examines if firms are more sensitive to being near an upgrade versus a downgrade. so technically this is a one-sided statistical test and t-statistics will be interpreted as such.

This test is also of interest generally as a benchmark against future tests in this paper. The sample contains only 144 firm years for the first year.5% more equity net of debt as a percentage of total assets) than firms in the middle. These results alleviate concern that the results of these tests are driven by the minus rating proxying for financial distress. however the results generally seem to support CR-CS. consistent with CR-CS. This indicates that the results are not only statistically significant. compared to an average of 388 firm years for 17 .3) is a regression without control variables.3) shown in column 3 of Panel B indicates that firms with a plus or minus rating annually issue approximately 1.5% less debt net of equity as a percentage of total assets (or 1. The value for the coefficient β3 of equation (4. the coefficient on the plus dummy is larger and even more significant than the minus dummy when only large debt offerings are excluded. For 14 out of 16 years. Applying this approach to Equation (4. and this may confound the results. it is difficult to provide strong theoretical support that the control variables are necessary. Panel B. Equation (4.1) yields a t-statistic of –3.22 to –4. Interestingly. the coefficient on CRPOM is the correct sign.74. however the control variables themselves proxy for financial condition. they will be related to whether a firm has a plus or a minus rating. using averages of the year-by-year cross-section regression coefficients to calculate overall coefficients. Since the sample sizes of each of the yearly tests are smaller than the overall sample. the coefficients on both the plus and minus dummy variables are statistically significantly negative with t-statistics ranging from –2. more robust standard errors can be generated using a procedure similar to that in Fama and MacBeth (1973). In both Panel A and B of Table III.33. lower than the tstatistic of –5. smaller t-statistics could be expected. Table IV shows results of equation (4.versus the plus rating dummy variable to the extent that the minus rating is a proxy for financial distress. and 8 out of 16 are significant at 10%. Fama and French (2002) suggest that in pooled time-series cross-section regressions such as these. and by using the time-series standard errors of the slopes for inferences. as they don’t control for effects that should necessarily bias the results.1) on a year-by-year basis to check against statistical complications of the pooled time-series cross-sectional results. Since the control variables are related to the financial condition of the firm. Additionally. but still significant at the 1% level. 1986. but economically significant as well. for 5 out of 16 years the coefficient is statistically significant at the 5% level.02 shown in Table III.

as the coefficient is insignificant in both cases. but this is consistent with the previous discussion of the implications of CR-CS with respect to large and small offerings. Excluding 1986 in this approach. In all cases the two columns represent large offering restrictions being placed on both debt and equity or debt only. The results using this approach show slightly greater statistical significance for the dummy variables. so this test may capture effects not captured in my lagged rating structure. The results are not however robust to the inclusion of the entire sample for the POM tests. I also examine large versus small firms by examining separate regressions for firms with total assets greater than $2 billion and less than $2 billion ($2 billion is chosen to split the sample roughly in half). however in the specific Fama and French (2002) tests of Section V. Robustness Checks Table V shows t-statistics from several regressions that modify certain assumptions from the previous regressions to gauge the robustness of the results. dramatically for the debt and equity cutoff sample (from a t-statistic of –3.15. I obtain a t-statistic of –5. capital structure decisions are lumpy and credit ratings can change in the middle of a year. Statistically this test is less appealing however due to endogeneity issues. The last row shows results from a regression using the firm’s contemporaneous rating as opposed to the lagged level. I report standard OLS coefficients and standard errors. and the fourth is of the correct sign. The results show that the findings are robust to moderate changes on the 10% assumptions both above and below that cutoff. I use the Fama and MacBeth method. Throughout this section. and they proceed with standard OLS regressions.05 to –4. The results in fact improve when the cutoff is changed to 5%.the other 15 years of data. 3 out of the 4 coefficients are statistically significant. Furthermore. The results are robust to this modification. Firms may consider expected ratings more than past ratings in their decision-making. 18 . a result consistent with the initial results. The tests in the first four rows modify the exclusion of large offerings from the sample. Frank and Goyal (2002) also find little difference when employing this alternate approach. The first results include all SIC codes rather than excluding SIC codes 4000-4999 and 6000-6999.94). The remaining tests revert back to the 10% cutoff assumption and modify other assumptions. as mentioned previously.

82). ∆Dit/Ait. The credit rating dummy variables remain statistically significantly negative in each of these cases as well (results not reported for brevity). and firms with a rating of BBB. Investment Grade versus Junk Bond Ratings Change Several of the explanations for why credit ratings are significant outlined in Section II imply that credit rating concerns should be most prominent around the change from investment grade to junk status (i. The coefficient on CRIG/Junk is negative for both measures in equation (4. from BBB. with binary decision variables reflecting both a decision to issue equity only versus debt only and a decision to issue debt only versus not issuing debt. firms with BBB and BB ratings might be concerned with that change in rating as well as firms with a BBB. in place of NetDIssit in equations (4. BBB-. Both coefficients in Regression 2 of Panel B imply over 1% less debt relative to equity as a percentage of total assets annually for those firms.3). BB+ or BB (given the significance of this potential change in rating.02 to –5.5) NetDIssit = α + δCR IG / Junk + φK it + ε it NetDIssit = α + δCR IG / Junk + βCR POM + φK it + ε it Table VI shows results of these two equations with the 2 different definitions of CRIG/Junk.e.to BB+). For the first measure.93) and it also increases the statistical significance of the coefficient on CRPOM (from –5. with statistical significance at the 10% level in the first case and 1% level in the second case.4) (4. I define this in two ways: firms with a credit rating of BBB.4).and BB+ rating).Additional robustness tests conducted were logit regressions.1)-(4. 19 .or BB+. CRIG/Junk is not incrementally significant in equation (4. C. CRIG/Junk is incrementally significant (with a t-statistic of –3. whereas for the second measure. and tests with a dependent variable that reflects net debt offerings only. I denote this additional dummy variable as CRIG/Junk and I conduct the following 2 tests for each definition: (4. suggesting economic significance as well for both dummy variables.. This section therefore examines this change specifically by introducing an additional credit rating dummy variable that indicates firms are near that change in rating.5).

with an average of . EBITDA/Interest Expense and EBIT/Interest Expense. 6 20 .7) Credit Score = 1. Debt/Total Capitalization. The sample used for calculating this Credit Score equation is the same as in the previous section.6 In initial regressions I found that some of these variables were redundant or had counterintuitive signs for coefficients. my goal here is simply to obtain a sufficiently predictive measure within my sample of firms that is also theoretically consistent.703. To derive the equation to calculate scores. Kennedy and Suan (2001). Previous studies also often look at ratings of new issues – the interest in this paper Previous studies often include a dummy variable for subordination status as well. I evaluate the concern for a potential change in Micro Rating.up to a value of 18 for a rating of AA+. EBITDA scaled by assets. however I also exclude any firm whose debt/total capitalization is greater than 1 or less than zero. net income scaled by assets. Ederington (1985) and Kaplan and Urwitz (1979)). and they find that previous studies have ranged in R2 from . EBITDA/Interest Expense and Debt/Total Capitalization had an adjusted R2 of .631. Debt/Total Capitalization squared.D. The dependent variable is equal to 1 for a rating of CCC. To create credit rating dummy variables. this distinction does not apply. approximately the same as a regression with all explanatory variables. Credit Score Tests In this section. Kamstra. I compute a Credit Score for each firm and rank firms within each Micro Rating. I consider as explanatory variables those from previous studies that have predicted credit ratings and also variables that the ratings agencies themselves argue are significant factors. Kennedy and Suan (2001) survey previous credit rating prediction research. I use the coefficients from this parsimonious regression to calculate the score as follows7: (4. These values are outliers that obscure the calculation of the Credit Score equation if included. Kamstra. The results I achieve appear consistent with the success found in previous literature. These include Log of Total Assets.6702 EBITDA/A – 6. I run a regression of credit ratings on factors that are thought to predict ratings. and the coefficients on each of the variables were the correct sign. A regression including only Log of Total Assets. but since I am looking at senior ratings for existing bonds in my sample.565 to . 7 Although an extensive literature exists on predicting credit ratings using various involved techniques (see for example.618.4501 Log (A) + 11.0462 Debt/Total Cap.

8) (4. I also create dummy variables for the individual high and low categories. since the measure for the Credit Score is measured with error.however is predicting ratings for existing bonds of firms. as that will not affect the ranking approach.are given a value of 1 for the dummy variable CRHOL and firms with credit scores that place them in the middle third are assigned a zero for that dummy variable. but this would require the (more unrealistic) assumption that the quality distribution of firms within each Micro Rating is constant over time. and I then rank firms within each Micro Rating into high thirds. but it also allows for potential additional inferences regarding the way in which managers care about credit ratings. I have a potential errors-in-variables complication. which requires the further assumption that the credit rating agencies maintain the same requirements for a particular rating throughout the sample period. so I use all firm ratings across the sample. For example. so the effects of the errors-in-variables complication should be reduced given this approach. Creating dummy variables using Credit Scores allows us to examine effects within a specific rating category. As before. This not only provides an additional test generally of CR-CS. depending on the rating level). initially I also exclude debt offerings greater than 10% of total assets for the year. as opposed to the previous section where effects within Broad Ratings were examined. I then run the following regressions:9 (4.8 Dummy variables are then constructed from these thirds for inclusion in the test equations. I calculate Credit Scores for each firm year with this equation. although some additional observations are lost as in some cases there are missing values for the terms required to calculate the Credit Score. The alternative would be to rank on a year-by-year basis. That is. within the Micro Rating A-. they are used to group firms into high and low thirds and create dummy variables based on these groupings.9) NetDIssit = α + β 0 CR HOL + φK it + ε it NetDIssit = α + β 1CR High + β 2 CR Low + φK it + ε it (4. managers may be concerned more with Broad Ratings changes or they may be more concerned with Micro Ratings changes (or both. middle thirds and low thirds.10) NetDIssit = α + β 3CR HOL + ε it The sample for these tests is as before. The intercept is dropped. 9 Note in using this approach. The ranking uses the entire sample for each Micro Rating. 8 21 . firms with a Credit Score that places them in the high or low third within A. The Credit Scores are not used directly however.

That is. credit ratings appear both statistically and economically significant. on average. I also conduct the test with the individual high and low dummy variables as well in equation (4. a correlation that implies that coefficient would be 22 .10). the individual variables used to calculate a score are highly related to the financial condition of the firm. since the control variables in the regression allow for a linear relationship with the particular variable and the dependent variable. Therefore once again CR-CS predicts that the coefficient β0 in equation (4. the Credit Score is highly correlated with the financial condition of the firm. less debt relative to equity. Including both the high third and low third within the dummy should negate this effect by offsetting one against the other.49.9). and the variables in the score calculation enter the regression only indirectly in the determination of the dummy variable (which by construction is not a linear relationship). with t-statistics of –3. This will allow for inferences regarding whether managers care more about upgrades or downgrades. however the coefficient on CRHigh is not statistically significant whereas the CRLow dummy variable is significantly negative at the 1% level. One potential problem with this however is that the control variables are the same or similar to the variables used in the score calculation.8)) as in the previous sections.The combined high and low dummy variable should mitigate the potential commingling effect of the financial condition of the firm. this should not be a problem. Another approach to mitigate the effects of the financial condition of the firm is to include control variables (equation (4. As mentioned above. The second regression of Panel A shows results with individual high and low dummy variables. This correlation would therefore likely produce a negative coefficient on the credit dummy for the lower third and a positive dummy for the coefficient on the high third. Table VII shows results for the regressions (4. The coefficients on both ratings are negative as predicted by CR-CS. as those firms closer to a change in rating will be more conservative (issue less debt relative to equity) with respect to their financing choices. to see how much the results are driven by one or the other. independent of the credit rating effects I am trying to identify.05 and –3. The size of the coefficient is similar to the previous section. On the other hand. Once again. The first and third regressions of Panel A show that the coefficient on CRHOL is statistically significantly negative at the 1% confidence level. where the third regression indicates that firms that are in the high or low third within a particular credit rating issue nearly 1% less net debt minus net equity than firms in the middle.8)-(4. and firms with worse financial condition issue.10) will be less than zero.

Further. For the Credit Score tests however. The results are similar with this modification. Incorporating CR-CS into Traditional Capital Structure Tests A. with 5 out of 6 coefficients significant at the 10% level. in 5 out of the 6 ratings categories. the deficit will be made up through 23 . Considering both tests’ results. The Credit Score test in particular seems consistent across ratings. I also examine a cutoff of 5%. For both sets of tests however. To consider this. Results by Rating. with results shown in Panel B of Table VII. the coefficients on the combined dummy variables are of the correct sign. the AA and B Broad Rating categories appear most significant. the coefficients on the Credit Score factors are also statistically significantly negative when including the entire sample. Shyam-Sunder and Myers (1999) (SSM) tests Pecking Order Test The SSM test for pecking order asserts that to the extent that a firm has a deficit in funds (DEF) beyond what can be met by internally generated funds. V. and firms in the middle third would not face an upgrade or downgrade if they undertook offerings of that size. with t-statistics of –2. as the sample sizes are smaller in each test. Since I control for two factors that might proxy for financial condition however. This result indicates that credit rating effects persist throughout the ratings spectrum. The POM tests excluded large offerings.10) (not reported). The power of these tests is reduced.32 on coefficients of CRHOL in both equation (4. The size of offerings included in the sample would have to be small enough such that firms in the high or low third would be concerned with an upgrade or downgrade when considering offerings of that size. this distinction would arguably need to be applied on a smaller scale. as firms conducting large offerings would face similar credit rating effects regardless of where they were within a Broad Rating.8) and (4. E.positive. Credit Score and POM Tests Table VIII shows results of the Credit Score test and POM test by Broad Rating. these results may indicate that firms are somewhat more concerned with downgrades than upgrades.

They use this sample because they want firms to have continuous data available during their sample period in order to implement other tests. the additional term in equation (5. the β coefficients should be equal to zero. Frank and Goyal (2002) and others have criticized this test. I therefore proceed with the SSM tests. and whether credit rating effects persist in the context of pecking order effects. Since I am only interested in 24 . but I also consider some modifications suggested by these other papers. after interest and taxes. this restriction would not be necessary. I run these tests using the sample of firms from the previous section. They note however that if they were only testing the pecking order theory (that is. DEF is defined as capital expenditures. Of interest is whether the model is enhanced with the introduction of credit rating considerations. CR-CS implies β coefficients less than zero. To determine if credit rating effects persist in the context of this pecking order test.3) As before.1) The pecking order theory implies that “b” is close to 1 and “a” is equal to zero.2) is a dummy variable for if a firm has a plus or minus within its Broad Rating and the second equation contains the High or Low Credit Score dummy variable. If the pecking order model is correct and credit ratings do not matter. the test that they undertake is: DEFit ∆LTDit = a+b + εit Ait Ait (5. however my intent is not to resolve the debate about the model. SSM however use a specific sample of only 157 firms. I consider the following modified tests: DEFit ∆LTDit = a+b + β 0 CRPOM + εit Ait Ait DEFit ∆LTDit = a+b + β 1CRHOL + εit Ait Ait (5.2) (5. the net increase in working capital and the current portion of long-term debt (at the start of the period) less operating cash flows. additionally dropping firm years where a variable that is required for calculating DEF is missing. Chirinko and Singha (2000). conducting the test I am interested in here).an issuance of debt instead of equity. dividend payments. Using this variable.

since this would allow firms with large deficits and small debt offerings to be included yet it would exclude firms with large deficits and large debt offerings. The sample is also different from Frank and Goyal. Regression 2 of Panel A of Table IX shows results including the POM credit rating dummy variable only. I would also like to exclude large offerings. including a broad sample of firms.87. however the sample is much broader. Frank and Goyal (2002) attempt to match the sample from SSM. The dependent variable is the change in debt. This test yields statistical significance for the 10 Note also that SSM use only long-term debt. It does appear however that the coefficient in Table IX matches fairly well within the findings of these previous tests. so the restriction is placed on large debt offerings only. in this case it would hurt the pecking order results to restrict the sample only to firms that do not undertake large offerings specifically. 25 . the credit rating variables are still statistically significantly negative with a t-statistic of -3. To maintain consistency with the tests of the previous section. However. Although the dependent variable has changed from the previous section to long-term debt only and the sample is slightly changed given missing data required for calculating the DEF term and the 10% DEF exclusion. and the coefficient on the deficit is equal to 0. limiting the size of the deficit to be less than 10% total assets. Further. equation (5.examining pecking order here in the context of credit ratings. most significantly as the time frame is 1986-2001 and the sample is restricted to firms that have credit ratings. This value is smaller than the 0. They therefore consider several modifications to the sample that SSM use. so again the results should be different from theirs.1)) with my sample.75 coefficient found in SSM.54 and it is statistically significant.10 Regression 1 of Panel A of Table IX shows results replicating the test from SSM (equation (5. for example.2). Regression 3 of Panel A of Table IX shows results including both the POM credit rating dummy variable and the DEF term.28 when they use a broader population of U. versus long and short-term debt as was used in the tests of the previous section. Therefore I also place the restriction on the deficit itself. firms from the period 1971 to 1989. but are unable to match the exact sample using several different sets of assumptions.S. I can extend their sample to include a greater number of firms. as well as limiting the size of the change in long-term debt to be less than 10%. Frank and Goyal get a coefficient of 0.

as the credit rating coefficient remains statistically significant when nested in the SSM pecking order test. and the R2 from the regression is largely unchanged with the inclusion of the credit rating dummy variable.00. Note that in comparing the CR-CS effects to the traditional effects. Therefore even when pecking order considerations are considered. As the POM and HOL credit rating factors persist in this context. Panel B of Table IX shows results using the Credit Score based dummy variables. 26 . Once again. it is not surprising to see strong statistical relationships between that factor and debt issued. and they find that empirically the current portion of long-term debt does not belong in DEF. Frank and Goyal argue that the current portion of long-term debt should potentially be excluded as part of DEF. I also run the tests excluding the current portion of long-term debt in DEF. there again is strong support for CR-CS.1). as SSM assume that this must be paid in the next year. The definition for DEF in the SSM test includes the level of the current portion of longterm debt. I obtain results similar to Frank and Goyal in regression of equation (5. whereby the coefficient and t-statistic on DEF are both larger with this modification (not reported). The t-statistic for the coefficient β1 on CRHOL in equation (5. so this variable is not predicting capital raising activity. and the credit rating variables add only marginal explanatory power to the test. The coefficient on DEF is largely unchanged from Regression 1 to Regression 3. The credit rating factors are lagged values and thus they are predicting future capital raising activity. credit ratings remain relevant for managers.3) is –3. The SSM pecking order test uses contemporaneous data in constructing their independent variable DEF. The implications regarding CR-CS are similar however to the tests including short-term debt. They run tests including it and excluding it. CR-CS is supported using this measure for being near a change in rating. Therefore the pecking order test withstands credit rating effects. As DEF is constructed with contemporaneous data in such a way that it is equal to debt issued during the period plus equity issued during the period. there is an important distinction in the nature of these different tests.dummy variable and the predicted sign.

given the extent of this test in the literature. it is instructive to incorporate the credit rating factors to see the effects. SSM test the tradeoff theory by running the regression: ( LTD *it − LTDit ) ∆LTDit = a+b + εit Ait Ait (5. regardless of whether the firm is near a change in rating. One common approach to estimating the target is to take historical averages of the debt to total capital ratios (see Taggart (1977). for example). This test can be modified with the introduction of the credit rating factors as follows: ( LTD * − LTDit ) ∆LTDit = a+b + β 0 CRPOM + εit Ait Ait ( LTD * − LTDit ) ∆LTDit = a+b + β 1CRHOL + εit Ait Ait (5. Proceeding as before. The tradeoff theory states that b will be greater than zero but less than one (it is less than one due to transaction costs associated with changing capital levels). However. except to revert back to the target leverage (holding the company’s marginal tax rate and distress costs constant).Tradeoff Theory Test The tradeoff theory implies that there should be no change in leverage from year to year. the test of the tradeoff theory conducted in SSM can also be modified to incorporate potential effects from credit rating concerns. even if they are currently below their target levels. however.6) CR-CS implies that firms that are near an upgrade or downgrade may be less willing to increase their debt levels. SSM show that this test can result in non-rejection of the model even when the model is false. A difficulty with implementing this test is that the target debt level is not observed. and this is the predominate approach SSM use. The debt level is determined by taking an average debt ratio for each firm for the sample period and multiplying that average by the firm’s total capital at the beginning of each year to obtain the target debt level for that particular firm year.5) (5. firms that 27 .4) LTD* is the firm’s target long-term debt level.

since they will be less concerned about a change in rating.4).24 compared to 0. although they only marginally change the R2. Regression 3 of both Panels of Table X nests the credit rating factors into the tradeoff theory test. which limits their results. Firms that are above their target will reduce their debt no matter where they are with regard to credit ratings. Furthermore. 28 . in addition to excluding firm years if the firm has a debt offering greater than 10% of total assets. and in equation (5. Their restriction reduces their sample to 157.11 SSM include a firm only if there is continuous data in which to measure the relevant funds-flow and balance-sheet variables during their entire sample. with the modified sample. The t-statistic for the coefficient β0 on CRPOM in equation (5. but the coefficient from my test seems reasonable considering the SSM results. making this requirement may bias the sample to large firms with conservative debt ratios.are far away from an upgrade or downgrade will be in a better position to increase their debt levels if they are below their target. For these reasons.23-0.77. I instead include a firm if it has 3 years of data in which to calculate the relevant variables. b. however their debt offering for the year would have to be less that 10% given the restriction placed on debt offering size. from 1971 and 1989. 11 Without this additional restriction. to be fair to the tradeoff theory I also exclude firm years if the difference between the firm’s debt level and its target debt level is greater than 10%. which would make testing credit rating effects difficult considering the dispersions of firms across ratings. firms that were over 10% away from their target might be included. The target adjustment coefficient.33 in SSM.34. My results are for a significantly larger sample and I place restrictions on debt offering size. as they point out. however they may be even more inclined to reduce their debt if they are near a change in rating. The second regression of both Panels shows results from a regression with the credit rating dummy variables alone.5) is -3.6) the t-statistic for the coefficient on CRHOL is –2. Thus CR-CS and the tradeoff theory combined imply the same results for b and a negative value for β0. The credit rating dummy variables remain statistically significant when nested in this test. Table X shows results considering POM and HOL dummy variables. in my sample is 0. and the coefficient on the credit rating dummy is statistically significantly negative in both cases. equation (5. In excluding large offerings in this case. Regression 1 of both Panels shows results of the same test as conducted in SSM. the results could differ.

there is an important distinction between the factors. which is a more difficult requirement. As this measure of leverage is different from what credit rating agencies use for determining ratings. L. I focus on the leverage tests they undertake. On the other hand. not only long and/or short-term debt.8) ( Lt +1 − Lt ) / At = a + b[TLt +1 − Lt / At ] + cP + ε t +1 In this regression. B. this predisposes this test to perform well. as this includes all liabilities. defined as total assets minus book equity. they use partial adjustment models as follows (the i subscripts are suppressed for notational convenience): (5. TL is target leverage for a firm and P is a vector of changes in current and lagged assets (At+1-A t and A t-A t-1) and earnings (ERt+1-ER t and ER t-ER t-1. The credit rating factors in contrast are lagged values. To further examine this point. I modify equation (5. so they are predicting future capital structure activity. The SSM tradeoff theory test uses contemporaneous and future data to calculate the target leverage level.7) Lt +1 / At +1 − Lt / At = a + b[TLt +1 − Lt / At ] + cP + ε t +1 (5. since it is strongly related to factors credit rating agencies use. The primary variable FF use in their dependent variable calculation is leverage.8) as follows: 12 It is left for future research to examine the impact of credit ratings on dividends. it should still allow for identification of credit rating effects.12 For these tests. Note also again in comparing the CR-CS factors to the traditional factors. This is a different measure of debt than tests in the previous sections of this paper. 29 . using this measure will bias the test somewhat against finding effects of credit ratings. Fama and French (2002) (FF) Tests FF empirically examine the implications on both leverage and dividend decisions of the tradeoff theory and pecking order models.These results support CR-CS in the context of the tradeoff theory. For purposes of this paper. where ER is defined as after-tax before-interest earnings). As SSM themselves show.

TL. To estimate equations (5. I therefore also maintain a test with their exact specification. The first stage regression is: (5. calculated from a target payout regression approach set forth in the paper.(5. RD_Dummy is a dummy variable set equal to 1 if a firm reports no R&D. FF conduct a two-step cross-sectional regression. as firms should not be concerned with moving towards a target leverage level. RD is R&D expenditures.9) ∆Dt / At = a + b[TLt +1 − Lt / At ] + cP + ε t +1 This is similar to their test. The tradeoff theory on the other hand predicts that firms will change their leverage levels such that they move closer to their target levels of leverage.7) and (5. as the pecking order model implies that the relationship between leverage and 30 . FF divide firms into dividend paying and non-dividend paying in both stages of the approach. FF find however that there is high collinearity between TP and ET.7) however. FF conclude that there is little evidence that TP is related to leverage as it does not remain statistically significant when both TP and ET are included. and therefore the tradeoff theory implies c=0. and TP is the target dividend payout ratio for a firm. Each year they regress book leverage at time t+1 on variables assumed to determine target leverage. The pecking order model predicts a strong short-term response of leverage to the variables in P.10) Lt +1 / At +1 = β 0 + β 1Vt / At + β 2 ETt / At + β 3TPt +1 + β 4 LogAt + β 5 RD / At + β 6 RD _ Dummyt + β 7 Dp / At + ε t +1 V is equal to L plus the market value of equity. I do not include TP as a factor in the first stage regression. The pecking order also predicts that b=0. Dp is Depreciation. however it considers changes in liabilities occurring only as a result of debt capital market activity. and thus the tradeoff theory implies that b>0.8). whereas ET remains statistically significant. They then use the fitted values from that equation to proxy for each firm’s target leverage. For these reasons. I do not modify equation (5. and as a result including both factors reduces the significance of both. The tradeoff theory implies that the variables of P are not significant with regard to leverage decisions. ET is earnings before interest and taxes.

I proceed similarly. the effects of credit ratings can be tested by including dummy variables for firms with a plus or minus credit rating or dummy variables for firms with a high or low Credit Score within their Micro Rating. These results support CR-CS in the context of both pecking order and tradeoff theory factors. In the second stage regressions.12) ∆Dt / At = a + b[TLt +1 − Lt / At ] + cP + βCR* + ε t +1 The TL values are those predicted from the first stage regressions. 2 of which are significant at the 5% level. and Panel B corresponds to equation (5. while the implications of the tradeoff and pecking order models remain the same. FF employ a statistical technique similar to that in Fama and MacBeth (1973) (this approach will henceforth be referred to as “FM”). FF require that several data variables be available for three consecutive years for each firm year. The sample for these tests is different given the data requirements for these tests. Instead of using a pooled time-series cross sectional approach as typically seen in the literature. Table XI shows results of equations (5. With these credit rating dummy variables denoted as CR* (where * = POM or HOL). I examine the following two tests: (5. 31 .investment may differ in the two groups. FF also conduct tests using market and book values of equity in deriving their dependent variables. Results of my first stage regressions yield coefficients similar to those found in FF (not reported). but I analyze the book value regressions only as they are more relevant for credit rating considerations. as discussed previously. The credit rating implication in both cases is that β<0. In all 4 regressions. Panel A corresponds to equation (5.11) for both POM and Credit Score dummy variables. They argue this approach will provide more robust standard errors.12) for dividend paying firms with coefficients and standard errors calculated using FM. I also as before exclude firm years where the firm conducted a large debt offering.12) for POM and Credit Score dummy variables. defined to be greater than 10% of total assets.11) and (5. the credit rating dummy variable coefficients have the predicted sign.11) Lt +1 / At +1 − Lt / At = a + b[TLt +1 − Lt / At ] + cP + βCR* + ε t +1 (5.

I also obtain a negative coefficient on the contemporaneous change in investment variable. a result FF find inconsistent with all other leverage tests. On an annual basis. but not statistically significant.11) tests. as well as the change from investment grade to junk. The results apply both to the potential for an upgrade as well as a downgrade and to both large and small firms. and are of the same sign as in FF. Conclusion Credit ratings directly affect capital structure decisions by managers. Similarly. the lagged changes in assets and earnings are no longer significant. however the HOL dummy variable coefficients are positive. Consistent with FF. one of which at the 10% significance level.firms near a credit rating change issue less debt relative to equity than firms that are not near a ratings change.5% less (more) net debt minus net equity (net equity minus net debt) as a percentage of total assets than firms that do not have a plus or minus rating. Managers appear to be most concerned with ratings changes around the AA and B credit rating levels. In equation (5.both near a Broad Ratings change and a Micro Ratings change . the coefficients on the POM dummy variables with both dependent variables are negative. In regressions including dummy variables that account for a firm being close to a ratings change .In the equation (5. the coefficients on all six of the FF independent variables are statistically significant. whereas CR-CS appears somewhat more robust. This indicates that my tests are consistent with their approach and findings. firms with a plus or minus rating within their Broad Rating issue approximately 1. For non-dividend paying firms (not reported). firms with Credit Scores in the high or low third within their Micro Rating issue approximately 1. Thus credit rating effects persist in the context of these other theories. While inclusion of credit rating dummy variables did not 32 .12) using a different dependent variable. This suggests that some of their results may not be robust to alternate specifications of the dependent variable. They also are similar in magnitude to the coefficients in FF.0% less (more) net debt minus net equity (net equity minus net debt) as a percentage of total assets than firms in the middle of their Micro Rating with respect to their Credit Score. VI. The credit rating dummy variables remain statistically significant when they are nested in the empirical tests of the tradeoff and pecking order capital structure theories found in ShyamSunder and Myers (1999) and Fama and French (2002).

whereas the credit rating dummy variables are constructed without looking ahead.Z) .Z) and PU(F. The value to the firm as a result of the financing including transactions costs.Z).PD(F. but excluding credit rating effects. both to ensure correct inferences in those tests. Clearly firms of type A will choose to issue equity rather than debt in more cases than firms of type B.Z)γ. The probabilities of a credit rating upgrade or a downgrade given a new offering are PD(F. Denote the size of the offering as Z.Z)γ + PU(F. is a function of the size and type of offering: V(F. Case 1: Consider two firm types: firms of type A are near a downgrade. whereas for B firms the probability of an upgrade or downgrade is zero for debt or equity. Appendix Consider a 1-period financing decision for a firm where the firm has exhausted its internal cash and therefore must issue outside capital. Assuming risk neutrality.materially affect the implications of these previous tests about their theories. 33 .Z)-γ whereas for a B firm the value is VB(D.Z). and that if A firms issue debt. Thus. 13 Upgrade results follow similarly.13 Assume both firms are considering an offering of size Z. for example. and the cost/gain from a change in rating is given by γ. and the financing choice as F (equal to D or E). a debt offering has a value to an A firm of VA(D. the probability of a downgrade is equal to 1. firms will then choose to maximize the expected value of undertaking an offering: V(F. for example). this is not surprising considering the structure of these previous tests. and firms of type B are not near a change in rating. and also to obtain more comprehensive depictions of capital structure behavior. The previous tests include contemporaneous explanatory variables for predicting capital structure decisions. Future capital structure empirical tests may benefit from including variables that account for firms’ credit rating situations.Z). debt only or equity only (to finance a new project. Thus the ability of credit ratings to predict capital structure decisions within these tests is an arguably stronger result than the ability of the contemporaneous factors from these tests to maintain predictive ability.

A firms (near a downgrade) issue fewer small debt offerings than B firms (32.0% 36.0% for γ=. CR-CS implies A firms issue more large debt offerings than B firms in all cases as well (32.S) and zero.5.γ.L) and V(D.0% 30.S). a firm of type B will choose the maximum of V(D.5 γ +.L) and V(D. since for values larger than . Small Debt.1). Firms will not issue debt if the total value to the firm is less than zero for both sizes of offerings. However. and for the example of γ=0.5 to .γ and zero.1).L). Here the firms are considering two sizes of offering. thus a firm may choose not to issue debt at all.0% versus 64.14 General Case Firm Type A B Example. 34 .1 (γ + .25 32.1).0% 28. a positive credit rating dummy variable 14 The general percentages assume a value of γ between 0 and 0.5.1. implications consistent with Case 1. Large Do Nothing 3 γ γ2 − − 8 2 2 3 γ γ2 + − 8 2 2 3 γ γ2 − − 8 2 2 3 γ2 −γ + 8 2 For all relevant values of γ.S). from -. V(D.0% Debt. γ=0. Small Debt.Case 2: Consider firms of type A and B. Therefore.0% for γ=.5) 2 .1).0% for γ=. The table below depicts the percentages of firms that will choose the three alternatives for a general value of γ.0% versus 42.5. large and small. Large Do Nothing Debt. V(D.L). but zero for a small offering.0% versus 28. but now assume that debt is the only financing choice they are considering. Assume that A firms have a probability of a downgrade of 1 for either large or small debt offerings.0% 42. A firms would never undertake a project. Likewise. Assume V values are independently uniformly distributed.γ. for firms within each type and for each financing choice. given values of V(D.S).0% 32. Therefore if firm years with large offerings are included in tests of credit rating effects such as in equation (4. given values of V(D. and B firms issue more debt offerings of any kind than A firms (70. a firm of type A will choose the maximum of V(D. whereas B firms undertaking a large debt offering have a probability of 1 for a downgrade. somewhat non-intuitively.

but the probability of a downgrade is 1 for a large debt offering. 1994. Large (1 − γ )3 (1 − γ ) 4 − 3 12 Equity. For B firms. Anatomy of financial distress: an examination of junk-bond issuers. it can be shown that the coefficient would be positive if the large offering size divided by the small offering size exceeded the difference in the percentage of small offerings conducted by B versus A firms divided by the difference in the percentage of large offerings conducted by A versus B firms. Robert Gerner. 35 . REFERENCES Asquith. Under these assumptions. Small (1 − γ ) 3 (1 − γ ) 4 − 3 12 1 5γ 3 −γ 2 + 4 6 Debt. whereas the probability of an upgrade or downgrade given a large or small equity offering is zero. In this case. I allow V to vary uniformly between 0 and 1 implying an offering will be undertaken (making computations simpler). Cantor. For A firms. and David Scharfstein. In the case of γ=0. Richard and Frank Packer. Federal Reserve Bank of New York Quarterly Review 19.5): Type A B Debt .1. assume the probability of a change in rating is zero for a small offering of debt or equity. Large γ4 12 + 7(1 − γ ) 4 (1 − γ )3 − 12 3  8 − 6γ + γ 3  1  8 − 6γ + γ 3  1 +γ  +γ    4 12 12   4   3 4 1 5γ 1 (1 − γ ) 7γ 4 −γ 2 + +γ − − 4 6 3 12 12 Inspection of these equations reveals that this more general case has the same testing implications as Case 2.coefficient could appear even though firms near a ratings change issue debt less frequently. firms will choose different financings in the following percentages by type (these percentages once again require a value of γ between 0 and 0. the probability of a downgrade is 1 for either a large or small debt offering. Case 3: Now assume firms of type A and B are considering debt or equity offerings of large or small size. 1-26.15 These implications indicate that the most effective tests of CR-CS should exclude larger offerings. 15 Under Case 2 assumptions. The credit rating industry.5). Quarterly Journal of Economics 109. and the probability of an upgrade is 1 for a large equity offering. 1994. Paul. Small Equity. 625-658. this size ratio would be equal to 2.

569-585. The effect of bond rating agency announcement on bond and stock prices. Testing tradeoff and pecking order predictions about dividends and debt. Murray Z.. John R. Goyal. 139-178. Singha. and Kenneth French. The Financial Review 20. Differences of opinion and selection bias in the credit rating industry. Eugene. and J. Graham. Masulis. and Mitchell A. 237-261. Moral hazard and information sharing: a model of financial information gathering agencies. Journal of Banking and Finance 21. Hand. Frank. Thakor. 2002. 1987. 2001. 261-275. Chirinko. 187-243. Kamstra. Journal of Financial Economics 60. Leland. 1977. Journal of Financial Economics 8. Post. 2001. Merton. Journal of Financial Economics. and Anuju R. Classification models and bond ratings. 75-96. 2177-2207.S. The theory and practice of corporate finance: evidence from the field. and Jeremy C. Louis H. The effect of capital structure change on security prices. and Anjon V. Ederington. 2002. Spencer Martin. 1403-1422. Journal of Finance 56. 1395-1417.. Richard W. 1985. Collin-Dufresne. Journal of Finance 49. 733-752.. Crabbe. Richard and Frank Packer. The effect of a rating downgrade on outstanding commercial paper. and Campbell R. The Journal of Financial Research 10. forthcoming. 1998... 39-56. Robert S. Jess B. Goh. Testing the pecking order theory of capital structure. Goldstein. Journal of Financial Economics 58. Ederington. 2001. Louis H. Miller. Leftwich. Bond rating agencies and stock analysts: who knows what when?. and G. 231-261. Roberts. 211-226. Holthausen. and Teck-Kin Suan. Yawitz and Brian E. Combining bond rating forecasts using logit. John R. Statistical models of bond ratings: A methodological inquiry. 1980. Peter Kennedy. R. 36 . Ederington. Journal of Finance 32. and Vidhan K.. Mark. Testing static tradeoff against pecking order models of capital structure: a critical comment. 2000. Debt and taxes. 1985. Fama. Pierre. Marcia H. Harvey.Cantor. Review of Financial Studies 15. 1994. Robert S. M. 1979. Millon. 417-425. The Financial Review 37. 1-33. Kaplan. 1992. Louis H. Journal of Finance 47. 1997. Urwitz. The informational content of bond ratings. Robert W. and Richard W. Journal of Finance 40. The determinants of credit spread changes. Journal of Business 52. Journal of Financial and Quantitative Analysis 33.

619-715. Corporate financing and investment decisions when firms have information that investors do not have. 159-168.S. and Stewart C. Stewart C. 1977. 1958. Bond ratings.html Taggart. Testing static tradeoff against pecking order models of capital structure.. The paradox of credit ratings. 1984.. Franco. 37 . Majluf. 1973. 2002. Washington University Law Quarterly 77. 261-297. 2002. Partnoy.com/europe/francais/Fr_resource_center/Articles_Resource_Center/ Identifying-Ratings-Triggers-and-Other-Contingent-Calls-on-Liquidity-Part2_15-05-02.A. Stewart C. 14671484. and Merton H. 147-155. 575-592. 20 (working paper) Rajan. Myers. The cost of capital. “Identifying Ratings Triggers and Other Contingent Calls on Liquidity – Part 2”. Journal of Financial Economics 5. The capital structure puzzle. Frank. Determinants of corporate borrowing. What do we know about capital structure? Some evidence from international data. Myers. 219-244. Journal of Finance 39. bond yields and financial regulation: some findings. Partnoy. 187-221.standardandpoors. corporation finance. and Luigi Zingales. and N. Stewart C. 1999. and the theory of investment. Shyam-Sunder. Journal of Finance 32. 1984.. 1995. Lakshmi. Miller. R. 1999. American Economic Review 48. Law and Research Paper No. A model of corporate financing decisions. Journal of Law and Economics 16. Standard and Poor’s. Journal of Financial Economics 51. Raghurum G.Modigliani. Journal of Financial Economics 13. Myers. Frank. West. 1977. Journal of Finance 50. 1421-1460.. at http://www. Myers. The Siskel and Ebert of financial markets?: two thumbs down for the credit rating agencies. Richard R.

9% 62.3% 83.7% 14. The sample is Compustat firms from 1986 to 2001.4% 16. medians and standard deviations of debt/(debt+equity) by credit rating within the sample.Table I Sample Summary Statistics .2% 40.9% 22.8% N/M 38 .Ratings and Leverage Means.0% B 331 61.7% 47.0% 73.6% 59. and the number of firm years (out of the total sample of 6. Debt/(debt+equity) is book long-term and short-term debt divided by book long-term and shortterm debt plus book shareholders' equity.906 firms years) that had the indicated rating at the beginning of the firm year.6% 42.7% 54.0% BBB+ 510 32. 606 66.8% BBB 562 34.5% CCC+ or below 142 58.9% N/M 85. total assets. and equity).0% 77.0% BB- 42.2% 61.1% B+ 897 47. 171 AA+ 65 AA 254 AA250 A+ 454 A 754 25.9% A- 19. excluding firms with SIC codes 4000-4999 and 6000-6999.5% 49.6% 37.7% 18.5% 45. and excluding firms with missing values for regularly used variables in the empirical tests of the paper (these include credit ratings.3% 19.4% 18.3% 15.2% BB+ 327 38.1% 55. 488 41.3% 33.9% 10. debt.4% BB 495 Number of Firm Years Debt/(Debt+Equity) Mean Median Std Dev.2% 31.6% 90. AAA Number of Firm Years Debt/(Debt+Equity) Mean Median Std Dev.8% 16.2% 81.2% 37.3% 19.9% BBB478 38.9% 19.8% 63.2% 27.2% B122 62.2% 91.0% Number of Firm Years Debt/(Debt+Equity) Mean Median Std Dev.4% 94.4% 46.

0% 39 . and firms with missing values for regularly used variables in the empirical tests of the paper (these include credit ratings.046 1.274 6.0% Reductions N 2.0% 46. and excludes firms with SIC codes 4000-4999 and 6000-6999.6% 17.906 % 29.4% 100.Capital Activity Number of firm years in the sample with the indicated capital activity. The sample is Compustat data covering security issuance from 1986 to 2001. debt.202 6.0% 7. total assets.9% 8. and equity). Offerings N Debt Only Equity Only Debt and Equity Neither Total 2. A Debt or Equity Offering or Reduction is defined as a net amount raised or reduced equal to 1% of total assets or greater for the calendar year.Table II Sample Summary Statistics .683 581 368 3.4% 5.172 486 3.4% 100.3% 47.906 % 38.

0524 5788 0.0058 (-3.44) -0. Panel A: Excluding large debt and equity offerings 1 2 3 Intercept t-statistic CRPOM t-statistic CRPlus t-statistic CRMinus t-statistic D/(D+E) t-statistic EBITDA/A t-statistic -0. CRPOM is a credit rating dummy variable with a value of 1 if the firm has either a plus or minus credit rating and equal to zero otherwise. book debt divided by book shareholder's equity plus book debt.76) -0. and zero otherwise.0050 (-2.0451 5969 0.02) -0.98) -0. previous year's EBITDA divided by total assets. and EBITDA/A.97) 0.0173 (-3.30) -0.96) -0.41) -0.0455 5969 0.77) -0.76) -0. R2 N 0.17) -0. The sample covers security issuance from 1986 to 2001.22) -0.0159 (-4.0172 (-3.0153 (-5.94) Adj.81) 0.35) -0.1831 (9. t-statistics are calculated using White's consistent standard errors.1829 (9.0068 (-2.0023 (-1.0052 5969 40 . A large offering is defined as an offering greater than 10% of total assets in the year.0175 (-1. and excludes firms with SIC codes 4000-4999 and 6000-6999.0025 5788 0.97) -0.2077 (5. and firms with missing values for any of the variables.0097 (-3.0130 (-5.0229 (-2.05) -0. respectively.2069 (5.0175 (-1.97) 0.0113 (-1. The control variables include D/(D+E). CRPlus and CRMinus are credit rating dummy variables with a value of 1 if the firm has a plus or minus rating.0078 (-4.01) Panel B: Excluding large debt offerings only 1 2 3 -0.74) -0.0228 (-2.81) 0.Plus or Minus Tests Coefficients and t-statistics from pooled time-series cross-section regressions of net debt raised for the year minus net equity raised for the year divided by beginning of year total assets on credit rating dummy variables and on control variables measured at the beginning of each year.0113 (-1.Table III Credit Rating Impact on Capital Structure Decisions .0529 5788 0.74) 0.0032 (2.

The samples exclude firms with SIC codes 4000-4999 and 6000-6999.17 -1. book debt divided by book shareholder's equity plus book debt.02 -0.56 -2.11 1987 1988 1989 1990 1991 1992 1993 -0.59 -0. and EBITDA/A.10 -2.25 -2.16 -1.53 -1.27 -1.60 1.25 -0.11 -1. Regression 1 includes the CRPOM dummy variable and Regression 2 includes the CRPlus and CRMinus credit rating dummy variables. and zero otherwise.89 -2.53 -1.84 -0.26 -0.40 -1. The sample also excludes a firm year if the firm had a debt offering greater than 10% of total assets in the year.48 -0.58 -0.00 -1.23 -2.70 1995 1996 1997 1998 1999 2000 2001 -2. The control variables (not shown) are D/(D+E).33 -0.74 -1.40 -2.76 41 .63 1994 Regression 1: CRPOM Regression 2: CRPlus CRMinus -3.03 -1. EBITDA divided by total assets.86 -1. 1986 Regression 1: CRPOM Regression 2: CRPlus CRMinus 0.84 -0.66 -0.21 -1.23 -0.61 -2.80 0.18 -0.17 0.50 -0.91 1.53 -0. respectively.Table IV Credit Rating Impact on Capital Structure Decisions . and firms with missing values for any of the variables.33 -0.58 -2. CRPlus and CRMinus are credit rating dummy variables with a value of 1 if the firm has a plus or minus rating. credit rating dummy variables and control variables measured at the beginning of each year. CRPOM is a credit rating dummy variable with a value of 1 if the firm has either a plus or minus credit rating and equal to zero otherwise.42 -1.51 -1.26 0.96 -2. t-statistics are calculated using White’s consistent standard errors.61 1.POM t-statistics by Year t-statistics from cross-sectional regressions by year of net debt raised for the year minus net equity raised for the year divided by beginning of year total assets on a constant.

788 -3.09 5.837 0. Offering restriction applied to D&E Large offering defined as >10% N Large offering defined as >20% N Large offering defined as >5% N All offering Sizes N Large Offering defined as >10% and: Including all SIC codes N Assets greater than $2 billion N Assets less than $2 billion N Credit Rating measure 1-year forward N -2.65 6.080 -5. CRPOM is a credit rating dummy variable with a value of 1 if the firm has either a plus or minus credit rating and equal to zero otherwise.906 Offering restriction applied to D only -5. EBITDA divided by total assets. The two columns correspond to restrictions of the sample whereby the sample excludes a firm year if the firm has a large debt offering or a large debt or equity offering defined as indicated.Table V Credit Rating Impact on Capital Structure Decisions – Robustness Tests t-statistics for the coefficient on CRPOM from pooled time-series cross-section regressions of net debt raised for the year minus net equity raised for the year divided by beginning of year total assets on CRPOM and on control variables measured at the beginning of each year.925 -3. and EBITDA/A.52 2.36 5. The control variables include D/(D+E).54 6.03 2.65 6.36 2.092 -2.864 -0.788 -2.969 -3.906 -4.423 -4. t-statistics are calculated using White's consistent standard errors. book debt divided by book shareholder's equity plus book debt.38 3.344 -2.05 5.94 4.02 5.82 10.514 -5. and firms with missing values for any of the variables.890 -3.165 0. The sample covers security issuance from 1986 to 2001.969 42 . and excludes firms with SIC codes 4000-4999 and 6000-6999 (except where indicated).76 10.04 6.56 5.

0300 (-4.41) Panel B: BBB.27) -0. R2 N 0. book debt divided by book shareholder's equity plus book debt.0114 (-3.or BBB+ in Panel A or BBB. The sample covers security issuance from 1986 to 2001.Table VI Credit Rating Impact on Capital Structure Decisions – Investment Grade to Junk Coefficients and t-statistics from pooled time-series cross-section regressions of net debt raised for the year minus net equity raised for the year divided by beginning of year total assets on credit rating dummy variables and on control variables measured at the beginning of each year.95) 0.0066 (-1.48) -0. and excludes firms with SIC codes 4000-4999 and 6000-6999.96) 0.43) -0.96) -0.0175 (-1.62) -0.042 5969 0.0285 (-3.2068 (5.0153 (-5. BB+. BB+ and BB in Panel B.23) -0.68) -0. BB 1 2 -0. and firms with missing values for any of the variables.99) 0.0180 (-1. CRPOM is a credit rating dummy variable with a value of 1 if the firm has either a plus or minus credit rating and equal to zero otherwise. and EBITDA/A.29) Adj.045 5969 0.82) -0.75) -0. Panel A: BBB-.0010 (-0. BBB-.0076 (-2. EBITDA divided by total assets.2074 (5.0228 (-2. BB+ 1 2 Intercept t-statistic CRIG/Junk t-statistic CRPOM t-statistic D/(D+E) t-statistic EBITDA/A t-statistic -0.66) -0.047 5969 43 .0181 (-2.2038 (5. The control variables include D/(D+E).042 5969 0.2088 (5.93) -0. CRIG/Junk is a credit rating dummy variable with a value of 1 if the firm has a rating of BBB.01) -0.0180 (-1. BBB-. The sample also excludes a firm year if the firm had a debt offering greater than 10% of total assets in the year. t-statistics are calculated using White's consistent standard errors.99) 0.0173 (-1.0128 (-4.

84) 0.30) 0.0095 (-3. and EBITDA/A.05) -0.0268 (-3. CRHOL is a credit rating dummy variable with a value of 1 if the firm's Credit Score is in the high or low third of its Micro Rating.0079 (-2.0168 (-4.0095 (-3.0503 5137 0.0548 5938 0.0334 (-4.0144 (-4.1639 (3. book debt divided by book shareholder's equity plus book debt.84) -0.0017 5938 0.0021 (-0. R2 N 0.02) 0. and firms with missing values for any of the variables. The sample covers security issuance from 1986 to 2001. t-statistics are calculated using White's consistent standard errors.45) -0. and excludes firms with SIC codes 4000-4999 and 6000-6999.08) -0.0544 5137 0.1715 (4.0017 5137 44 .0081 (-3.32) -0.05) -0.0333 (-4. Panel A: Excl. debt offerings > 10% 1 2 3 Intercept t-statistic CRHOL t-statistic CRHigh t-statistic CRLow t-statistic D/(D+E) t-statistic EBITDA/A t-statistic -0.13) -0.11) -0.72) -0.0528 5938 0.Table VII Credit Rating Impact on Capital Structure Decisions .02) 0.0207 (-9.Credit Score Tests Coefficients and t-statistics from pooled time-series cross-section regressions of net debt raised for the year minus net equity raised for the year divided by beginning of year total assets on credit rating dummy variables and on various control variables measured at the beginning of each year.0169 (-2.32) Adj. The control variables include D/(D+E).73) -0.0260 (-3. EBITDA divided by total assets.49) Panel B: Excl.0065 (-3.38) -0.96) -0. CRHigh and CRLow are credit rating dummy variables with a value of 1 if the firm's Credit Score is in the high or low third.0299 (-3.1956 (4. debt offerings > 5% 1 2 3 -0. The sample excludes a firm year if the firm had a debt offering greater than the indicated percentage of total assets in the year. within its Micro Rating and zero otherwise.0302 (-4.45) -0.0003 (0. respectively.50) -0.01) -0.0163 (-2.2012 (5.74) 0.

respectively. The control variables (not shown) are D/(D+E).04 -3.Table VIII Credit Rating Impact by Broad Rating t-statistics from pooled time-series cross-section regressions by Broad Rating of net debt raised for the year minus net equity raised for the year divided by beginning of year total assets on a constant.16 -3.10 -0. CRHigh and CRLow are credit rating dummy variables with a value of 1 if the firm's Credit Score is in the high or low third.87 -1.72 -3.30 -0.49 -0.32 0.16 -2.09 0. The sample also excludes a firm year if the firm had a debt offering greater than 10% of total assets in the year.46 -0. Regressions 1 and 2 distinguish different credit rating dummy variables included the regressions.51 0.26 -2.36 Panel B: Plus or Minus Tests Regression 1: CRPOM Regression 2: CRPlus CRMinus 0.27 -2. The sample covers security issuance from 1986 to 2001. and excludes firms with SIC codes 4000-4999 and 6000-6999. AA Panel A: Credit Score Tests Regression 1: CRHOL Regression 2: CRHigh CRLow -1. within its Micro Rating. credit rating dummy variables and control variables measured at the beginning of each year.07 -1.26 -1.48 -1. and EBITDA/A.04 -1.12 A BBB BB B CCC -1.68 -0.27 0.10 -2.44 -1.24 45 .48 1.13 -1.35 -1. CRHOL is a credit rating dummy variable with a value of 1 if the firm's Credit Score is in the high or low third of its Micro Rating.87 0. EBITDA divided by total assets. respectively. CRPOM is a credit rating dummy variable with a value of 1 if the firm has either a plus or minus credit rating. t-statistics are calculated using White's consistent standard errors.23 0.44 -2.88 -3.84 2.43 1.35 -1. book debt divided by book shareholder's equity plus book debt.58 2. CRPlus and CRMinus are credit rating dummy variables with a value of 1 if the firm has a plus or minus rating. and firms with missing values for any of the variables.

3150 4767 0.61) Panel A: POM tests 2 -0.34) 0.0166 (-12.0198 (-26. CRHOL is a credit rating dummy variable with a value of 1 if the firm's Credit Score is in the high or low third of its Micro Rating.87) -0.3130 4767 0.0047 (-3.0166 (-14.0024 4613 0.0195 (-26.3165 4613 0.00) Adj.0067 (-3. The sample covers security issuance from 1986 to 2001. as capital expenditures plus dividend payments plus net increase in working capital and the current portion of long-term debt minus operating cash flows after interest and taxes.5453 (46.0111 (-7.0048 (-3.23) Panel B: HOL tests 2 -0.5444 (46.15) -0.81) 0.48) -0.0029 4767 0.3177 4613 46 .12) 3 -0. and excludes firms with SIC codes 4000-4999 and 6000-6999.0109 (-7.5384 (46. and firms with missing values for any of the variables.52) 1 -0. CRPOM is a credit rating dummy variable with a value of 1 if the firm has either a plus or minus credit rating and equal to zero otherwise.17) -0.55) 0. DEF is defined as in Shyam-Sunder and Myers (1999).5396 (46.0070 (-3. 1 Intercept t-statistic DEF t-statistic CRPOM t-statistic CRHOL t-statistic -0. R2 N 0.Table IX SSM Test of Pecking Order with Credit Rating Factors Coefficients and t-statistics from pooled time-series cross-section regressions of net long-term debt raised for the year divided by beginning of year total assets on credit rating dummy variables and on DEF.17) 0.73) 3 -0. The sample also excludes a firm year if the firm had a debt offering greater than 10% of total assets in the year or if DEF is greater than 10% of total assets for the year.

58) Panel B: HOL tests 2 -0.62) 0.67) 0.73) 3 -0. R2 N 0.0016 (-1. CRPOM is a credit rating dummy variable with a value of 1 if the firm has either a plus or minus credit rating and equal to zero otherwise.0013 (-1. and firms with missing values for any of the variables.0057 (-3.0069 (-4.2357 (27.LTD) is greater than 10% of total assets for the year.Table X SSM Test of Tradeoff Theory with Credit Rating Factors Coefficients and t-statistics from pooled time-series cross-section regressions of net long-term debt raised for the year divided by beginning of year total assets on credit rating dummy variables and on (LTD* .43) 1 -0.34) -0.16) 0.0052 (-3.10) 0.1154 5820 0.LTD) t-statistic CRPOM t-statistic CRHOL t-statistic -0. as target debt level minus current debt level divided by total assets.0060 (-4.0077 (-4. and the sample includes firms with at least 3 years of data.1169 5820 0.34) -0.57) Panel A: POM tests 2 -0. The sample also excludes a firm year if the firm had a debt offering greater than 10% of total assets in the year or if (LTD* .35) -0. (LTD* -LTD) is defined as in Shyam-Sunder and Myers (1999).2373 (27. CRHOL is a credit rating dummy variable with a value of 1 if the firm's Credit Score is in the high or low third of its Micro Rating.0046 (-2.2243 (27.2255 (27.1209 5581 47 .0029 5820 0.1198 5581 0. and excludes firms with SIC codes 4000-4999 and 6000-6999.0044 (-5.94) 3 -0.0032 5581 0. The sample covers security issuance from 1986 to 2001. Target debt levels are calculated using historical averages. 1 Intercept t-statistic (LTD* .23) -0.77) Adj.LTD).0046 (-5.

34) 0. and excludes firms with SIC codes 4000-4999 and 60006999.71) -0. CRPOM is a credit rating dummy variable with a value of 1 if the firm has either a plus or minus credit rating and equal to zero otherwise.85) -0.0647 (-4. Panel A: Lt+1/At+1 .69) -0.0568 (2.34) -0. Coefficients are means of 16 cross sectional regressions from 1986-2001.1423 (6.54) -0.0012 (-0.1407 (5.0725 (3.57) 4257 0.44) -0.2094 (-6. Other dependent variables include changes in current and lagged Assets and changes in current and lagged Earnings (where ER denotes after tax earnings).12) -0.53) -0.82) -0.5211 (-9.0640 (-5.97) -0.0193 (-0.0032 (-0.0747 (3.82) 0.Table XI FF Test of Tradeoff and Pecking Order Theories with Credit Ratings .96) -0.0561 (2.65) -0.1671 (5.0408 (-4.2023 (-5.0018 (-0.5.0026 (-0. and net debt issued divided by total assets in Panel B.Lt/At as Dependent Variable 1 2 Panel B: Net Debt Issued/At as Dependent Variable 1 2 CRPOM t-statistic CRHOL t-statistic TLt+1 t-statistic Lt/At t-statistic (At+1-At)/At+1 t-statistic (At-At-1)/At+1 t-statistic (ERt+1-ERt)/At+1 t-statistic (ERt-ERt-1)/At+1 t-statistic N -0. firms with missing values for the regression variables.0026 (-1. CRHOL is a credit rating dummy variable with a value of 1 if the firm's Credit Score is in the high or low third of its Micro Rating.78) -0.95) 4639 0.0402 (-4.0166 (-0.0039 (-2.0667 (-5.2600 (-5.5302 (-9.98) 0.2052 (-5.82) -0.75) 4415 -0. Dependent variables are change in leverage in Panel A (where L=total assets minus book equity and A=total assets).10) -0.0684 (-5.29) 0. and firm years if the firm had a debt offering greater than 10% of total assets in the year. and t-statistics are time series standard deviations of the coefficients divided by 16.56) 0.46) 4056 48 .91) -0.24) -0.1404 (6.36) 0. The sample includes firms with dividends at time t-1.Dividend Paying Firms Fama and MacBeth time-series cross-section regressions coefficients and t-statistics.39) -0.

9 1 0 0.8 0.9 1 Debt/Total Capital Debt/Total Capital e) Credit Rating Effects with Tradeoff Theory .4 0.8 0.6 0.3 0.5 0.5 0.2 0.1 0.2 0.8 0.7 0.9 1 Debt/Total Capital Debt/Total Capital c) Credit Rating Jump d) Credit Rating Jump T* Firm Value Firm Value T* 0 0.6 0.3 0.4 0.6 0.2 0.1 0.7 0.3 0.3 0.4 0.1 0.7 0.1 0.5 0.Figure 1 – Tradeoff Theory with Credit Rating Effects a) No Credit Rating Effects b) Credit Rating Jump C* Firm Value T* Firm Value T* 0 0.3 0.7 0.8 0.Multiple Ratings C* T* Firm Value 0 0.9 1 0 0.8 0.6 0.4 0.7 0.2 0.2 0.1 0.9 1 Debt/Total Capital 49 .5 0.5 0.4 0.6 0.