Credit Rating and the Impact on Capital Structure

Inhaltsverzeichnis

Index of Abbreviations

Index of Appendix

Index of Symbols

1. Introduction

2. Credit Ratings in the Context of Traditional Capital Structure Theories
2.1 The Tradeoff and Pecking-Order Theory
2.2 Credit Ratings
2.3 The Credit Rating-Capital Structure-Hypothesis

3. Testing the Impact of Credit Ratings on Capital Structur Decisions
3.1 Specification of the Test
3.2 Data and Summary Statistics
3.3 Empirical Tests of the CR-CS

4. Incorporating the CR-CS in the existing Capital Structure Theory
4.1 Credit Rating Effects vs. Tradeoff and Percking-Order Theory
4.2 Nesting Credit Rating Effects into Tradeoff and Pecking-Order Tests

5. Complementing the Credit Rating-Capital Structure-Hypothesis
5.1 Capital Structure Decisions after a Rating Change
5.2 Testing the ex post Behavior

6. Conclusion

Appendix

References

Index of Abbreviations

illustration not visible in this excerpt

Index of Appendix

Annex 1: Rating Categories

Annex 2: Table 3, Sample Summary Statistics - Rating and Leverage

Annex 3: Figure 3, Capital market activity by rating, 1986-2001

Annex 4: Figure 3, Debt and Equity offerings by Ratings, 1986-2001

Annex 5: Table 4, Credit Rating Impact on Capital Structure Decisions - Plus or Minus Tests

Annex 6: Table 5, Credit Rating Impact on Capital Structure Decisions - Credit Score Tests

Annex 7: Table 5, Credit Rating Impact on Capital Structure Decisions - Investment-Grade to Speculative-Grade

Index of Symbols

illustration not visible in this excerpt

1. Introduction

The question about capital structure is one of the most important issues which the management of a company faces in implementing their daily business. Therefore, the question of which factors affect capital structure decisions attracts high attention in the past and recent literature on capital structure. There are many papers providing valuable insights into capital structure choices, starting with the paper of Modigliani and Miller (1958)¹. The MM-Theorem is generally considered a purely theoretical result since it ignores important factors in the capital structure decision like bank- ruptcy costs, taxes, agency costs and information asymmetry. Based on this paper many other theories which consider factors neglected by Modigliani and Miller have been evolved. Two major theories are the Tradeoff- and the Pecking-Order-Theory. The former loosens assumptions stated in the MM-Theorem by including bankruptcy costs and taxes while the latter introduces information asymmetry into the capital structure discussion. Chapter 2.1 will give a brief overview of these theories. For complexity reasons these models cannot capture all relevant factors affecting the capital structure policy of a company. However, all these theories disregard one cru- cial factor which plays an important role on capital markets all over the world. The significance of Credit Ratings is gradually increasing, and it is doing so in many re- spects. This paper focuses on the Credit Rating-Capital Structure-Hypotheses (CRCS) developed by Darren J. Kisgen as a modern approach to the capital structure discussion. The hypothesis argues that credit ratings have an impact on capital struc- ture decisions due to discrete costs (benefits) associated with a rating change. Firstly, reasons why credit ratings are material for capital structure decisions will be out- lined. Then, situations in which credit rating effects play a role will be examined. For this issue it is very important to show how it can be measured whether a firm is con- cerned about a rating change or not. Afterwards the CR-CS will be empirically tested. The traditional theories don’t explain the results obtained in these tests. Therefore credit rating effects will be combined with factors discussed in the Tra- deoff- and Pecking-Order-Theory. In subsequent empirical tests credit rating factors will be integrated into previous capital structure test to show that the results of the CR-CS tests remain statistically significant. The paper ends with an analysis on whether capital structure decisions after a rating change are driven by credit rating effects.

2. Credit Ratings in the Context of Existing Capital Structure Theories

2.1 The Tradeoff and Pecking-Order Theory

Since the assumptions made in the MM-Theorem do not hold true in practice, it is necessary to examine the factors which explain why capital structure is relevant for firms in an imperfect market. As opposed to the MM-Theorem, the Tradeoff-Theory includes the factors bankruptcy costs and taxes². Furthermore the theory refers to the idea that a value-maximizing firm will balance the value of interest tax shields against various costs of bankruptcy³. Hence, the tradeoff between the two forces de- scribed above determines an optimal level of leverage for each company.

illustration not visible in this excerpt

Figure1. Tradeoff Theory (own depiction in dependence on Myers (1984), p.577.)

In Figure1it is important to understand that, in this theory, firms are financed partly with debt and partly with equity; the leverage level is determined accordingly. Figure

1 implies that raising the leverage level is associated with both higher tax benefits of debt and higher costs of bankruptcy and financial distress. However, for higher leve- rage levels the bankruptcy costs increases above average compared to the interest tax shield benefits. Due to this fact a value-maximizing firm will target the optimal leve- rage level at L*. The Pecking-Order-Model, which was developed by Stewart C. Myers and Nicolas Majluf (1984), chooses a different approach including asymme- tric information. The theory says that companies prioritize their ways of financing from internal funds to equity due to asymmetric information⁴. This problem occurs because managers have an incentive to use private information to issue stocks when they are over-priced. Investors will take this asymmetric information problem into account and discount the firm’s new and existing shares. This behavior leads to asymmetric information costs because managers have to issue these stocks under- priced. Anticipating this behavior, managers may neglect profitable projects if they must be financed with under-priced stocks. To minimize asymmetric information costs and other financing costs, managers finance projects first with internal funds, which maintain no asymmetric information problem, then with low risk debt for which this problem is negligible and finally, under compulsion, with equity⁵.

2.2 Credit Ratings

After presenting the two capital structures theories, credit rating is now introduced as an important factor which is missing in the traditional literature about capital struc- ture. Credit Ratings are issued by credit rating agencies (CRA) like Standard& Poor’s, Moody’s and Fitch, just to name the most important ones. They assign rat- ings for several issuers (e.g. firms, nations and local governments) of specific types of debt. In this paper we focus solely on companies. Credit ratings try to capture the creditworthiness of corporations and provide an ordinal ranking of default risk across firms (see Annex 1). Such rankings are based on a complex rating process due to the fact that credit ratings provide information about the quality of a firm beyond public- ly available information⁶. To set up the respective ratings, the agencies accumulate and evaluate qualitative information, such as the market share of the company and the competiveness within its industry as well as quantitative factors. Moreover, the agencies have access to insider information; ratings therefore include estimates of the company’s development potential as well as their ways of financing. Consolidated credit ratings provide a lot of information on various aspects.⁷

2.3 The Credit Rating-Capital Structure-Hypothesis

Kisgen’s CR-CS links chapter 2.1 and chapter 2.2, claiming that credit ratings are a material consideration in managers’ capital structure decisions due to the discrete costs (benefits) associated with different rating levels. In addition, the hypothesis argues that concerns about the impact of credit rating changes directly affect capital structure decision making, with firms close a ratings change issuing less net debt relative to net equity than firms not close a ratings change⁸. This statement has to be proven by both economic facts and empirical tests. Hence, before conducting empiri- cal tests, reasons are outlined supporting the hypothesis that there is a relation be- tween credit ratings and capital structure. A simple but quite intuitive fact is that a utility-maximizing manager will consider credit ratings since higher ratings are asso- ciated with higher reputation. It is therefore likely that credit ratings affect capital structure policy. Another point is that nearly all financial regulators including public authorities that oversee banks, insurance company’s etc., rely on the ratings issued by rating agencies⁹. Thus, several regulations related to financial institutions and other intermediaries result from credit ratings. For instance, low credit rating levels prevent some investors (e.g. banks) from investing in a firm’s bond and cause stricter capital requirements for investor groups such as insurance companies or broker-dealers in- vesting in particular bonds¹⁰. Here it is important to mention that regulations general- ly do not border firms within a rating category (e.g. AA+, AA and AA-)¹¹they are rater focused on the distinction between broader rating categories. Since many regu- lations are tied to the downgrade from investment- to speculative-grade, the impact of this change should be decisive in the decision making process¹². Another reason for the significance of credit ratings is the information they provide for investors. As mentioned above credit ratings contain quantitative, qualitative and publicly unavail- able information. Credit ratings could therefore, reflect a corporation’s credit quali- ty. If the market regards ratings as a signal of a company’s quality, companies will be pooled with other companies in the same rating category. This implies that, carried to extremes, a downgraded BB firm (now a particularly good BB-firm) would have the same credit spread as a bad BB- firm since the market groups together all firms with the same rating. Thus, the rating change for the BB firm would result in discrete changes in its cost of capital. Likely, any rating category contains specific informa- tion, so as opposed to regulations, a potential rating change of any kind (including BB to BB-) should be significant for capital structure decisions¹³. Costs directly im- posed on firms and associated with different rating levels are the next point to be examined. Rating changes can affect a company’s operations in different ways. For instance, if a firm wants to enter into a long-term supply relationship the counterpar- ty may demand certain credit ratings to accept the contract. Furthermore, low ratings may have a negative effect on the employee or customer relationship¹⁴. In this con- text the most important point is, however, that a firm’s rating affects its access to other financial markets, and therefore may complicate the funding of the firm. More- over, disclosure requirements for bonds (e.g. speculative-grade bonds have the most stringent requirements) and bond covenants are influenced as well. Bond covenants can include rating triggers whereby a rating change can lead to a forced repurchase of the bonds or a change in coupon rates. Here is needs to mentioned that, just regu- latory effects, rating triggers are most prominent around the investment- to specula- tive-grade because no firm would accept bond covenants which are sensitive for small rating changes (e.g. from BB to BB-). The above presented facts strongly sup- port the CR-CS and provide evidence that credit rating concerns have an impact on capital structure decisions. In the next chapter the purely theoretical argumentation will be complemented by empirical tests of the CR-CS.

3. Testing the Impact of Credit Ratings on Capital Structure Decisions

3.1 Specification of the Test

The empirical framework and certain tests conducted by Kisgen (2006) will be ex- amined below. The findings of chapter 2.3 imply that capital structure decisions are not directly influenced by a specific rating level; they are rather affected by a firms concern about a potential rating change since these changes are linked to discrete costs (benefits). For the following tests this distinction is crucial. In addition the re- sults of chapter 2.3 indicate that there are three kinds of rating changes associated with concerns about discrete costs (benefits). Due to various issues mentioned above (e.g. regulatory effects, pooling effects or rating triggers), concerns about a rating change from investment- to speculative-grade, from one broad rating category to another (e.g. from A to BBB) or a change of any kind (e.g. from BBB+ to BBB) are material for capital structure decisions. On this basis it is essential to find out which facts indicate whether a firm is close to a rating change or not. This point is highly important as firms far away from a down- or upgrade will not be concerned about a rating change (e.g. an AA firm won’t be concerned about a change from investment- to speculative-grade as opposed to a BB+ company). The three kinds of rating changes and indicators of whether a firm is close to a rating change or not will be examined in the following. Figure 2 shows the three kinds of rating categories for which concerns about a rating change exist. Broad Ratings include the plus, middle and minus specifications for a certain rating (e.g. A+, A and A-).

illustration not visible in this excerpt

Figure 2. Three kinds of Rating Changes

In this case, firms are categorized as near a Broad Rating change if their rating is denoted with either a “+” or a “-“. Furthermore, Kisgen determined BB+, BB, BBB and BBB- firms as near to an IG/SG change, since the significance of this change implies that not only BB+ and BBB- firms will be concerned about this. To find out whether a firm is near a Micro Rating change, Kisgen calculated so-called “Credit Scores” for each firm to rank them into a high third, middle third and low third with- in a certain specification of a Broad Rating (e.g. BB+-). These “Credit Scores” try to capture the credit quality of each company. In order to obtain the equation to calcu- late such “Scores”, Kisgen regressed credit ratings on factors that are thought to pre- dict ratings. Eliminating redundant and non predictive factors, he finally obtained an equation with three explanatory variables that have an adjusted R² of 0,631 (i.e. ap- proximately 63,1% of the variability of the credit rating can be explained by these variables). Kisgen identified firms to be near a Micro Rating change if they are ranked in the high or low third within a Micro Rating due to their “Credit Score”.

Table 1 summarizes the results:

illustration not visible in this excerpt

Table 1. Indicators for Firms near a Rating Change

Regarding any potential rating change is an advantage of Micro Ratings whereas the measurement of the “Credit Score” adds noise to the empirical tests. A disadvantage of Broad Ratings is that they might not be specific enough in certain cases. For instance, a strong BB- firm may not be close to a downgrade; likewise a weak BB+ firm may not be near an upgrade. The test might therefore underestimate the true effect. To carry out a number of regressions in the empirical tests in chapter 3.3, dummy variables will be introduced which distinguish between firms near a rating change and others that are not.

Table 2 illustrates these variables:

illustration not visible in this excerpt

Table 2. Dummy Variables

The CR-CS states that credit ratings have an impact on capital structure decisions. Moreover the hypothesis specifies the impact on capital structures in the way that firms near a rating change will issue less net debt relative to net equity in order to either avoid a downgrade or increase the probability of an upgrade. To specify this hypothesis: “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”¹⁵. Therefore the dummy variables used in the regressions reflect the credit rating situa- tion at the beginning of each period. Then the firm’s capital decisions measures is calculated for the subsequent 12 months. To quantify the capital structure decisions, Kisgen introduces the NetDIss variable. The variable measures the amount of net debt relative to net equity issued for a firm i at time t. Book values are used because

illustration not visible in this excerpt

credit rating agencies emphasize these measures and they directly reflect managerial decision making.

[...]

¹ Myers (1984), p. 575.

²Often agency costs are included as well.

³Myers (1984), p. 577.

⁴ Spremann, Gantenbein (2005), p. 109f.

⁵Fama French (2002), p. 3f.

⁶Kisgen (2006), p. 1038.

⁷Hovakimian, Kayhan Titman (2009), p. 4.

⁸ Kisgen (2006), p.1037.

⁹Cantor and Packer (1994), p. 5.

¹⁰Kisgen (2006), p. 1036.

¹¹See Annex 1.

¹²Kisgen (2006) p. 1037f.

¹³ Kisgen (2006) p.1039.

¹⁴ Kisgen (2006) p. 1039f.

¹⁵ Kisgen (2006), p.1045.