The Odyssey of Financial Restructuring Under the US Bankruptcy Code - DecisionBoundaries

# The Odyssey of Financial Restructuring Under the US Bankruptcy Code

## The Odyssey of Financial Restructuring Under the US Bankruptcy Code

###### December 15, 2019

Abstract

I present a model of a financially distressed firm with outstanding bank and bond debt. I rely on the objective criterion of investment efficiency to lead the reader through the journey and choices facing the distressed firm and its creditor through the many choices facing each constituency along the long and tortuous way – seen from the perspective of US corporate and bankruptcy law – culminating in the firm’s financial restructuring either in- or out-of-court) or its liquidation.

During the late 1980s there was a dramatic increase in the leverage of U.S. corporations, raising concerns about the corporate sectors’ financial stability. Indeed, by June 1990, 156 ($24%$) of the 662 companies that issued high-yield bonds between 1977 and 1988 had either defaulted, gone bankrupt, or restructured their public debt. The face value of those distressed bonds amounted to nearly $21 billion. The central question raised by those distressed firms is easy to put but hard to answer: What is the effect of financial distress on a firm’s operating performance? There are two competing views. The first, an application of the Coase Theorem, holds that there are no real effects of financial distress. Critical to this view is the distinction between financial and economic distress. Admittedly, most firms in financial trouble also suffer from poor operating performance. No amount of financial maneuvering can save these economically distressed firms. If, however, a firm’s capital structure prevents from pursuing its value-maximizing operating strategy, creditors will restructure their claims to maximize firm value. We should expect financially distressed firms to do poorly on average, but no worse than if they had no leverage. The second view — implicit in the leading theory of capital structure — is that financial distress hampers operating performance. In this view, the Coase Theorem fails; financial renegotiation is inefficient and operating distortions are introduced. Distinguishing between these two views is important for understanding a variety of issues: capital structure decisions; the costs of tax policies which affect the level of corporate debt; the impact of wide-scale financial distress during a recession; and the role and effects of specific provisions of bankruptcy law. Unfortunately, it is difficult to distinguish empirically between financial and economic distress. Is a financially distressed firm liquidated because renegotiation is inefficient or because the firm is not economically viable? Is a firm’s poor operating performance the result of underlying business problems or an inappropriate capital structure? Unfortunately, the empirical attempts to distinguish between financial and economic distress are limited to specific environments in which it is relatively easy to make such distinction. The theoretical distinction between financial and economic distress emerges in the important work of Bulow and Shoven (1978) and the follow-up work of White (1980, 1983). These models demonstrate how conflicts among creditors can lead to inefficiencies when a firm is in financial distress. The impediment to efficient renegotiation in these models is the assumption that the firm cannot renegotiate with bondholders, although they can renegotiate efficiently with a bank. On the one hand, because public bondholders claim part of the cash flows from new investment, distressed firms can have difficulty issuing equity or debt for new investment. This, they may pass up positive net present value investments. This article pursues two research goals. The first is to show that these investment inefficiencies are still a problem even when firms can renegotiate with bondholders. We analyze the implicit renegotiation that takes place when firms offer a basket of new securities and cash in exchange for the original public bonds. Public debt restructurings almost always take the form of exchange offers because the Trust Indenture Act of 1939 requires unanimous bondholder consent before a firm can alter the principal, interest, or maturity of its public bonds. Exchange offers effectively alter these features but, since nontendering public bondholders maintain their original claim for payments on the firm, the Trust Indenture Act is not violated. Despite the frequency with which exchange offers have been made — 73 of 156 distressed junk bond issuers have successfully completed exchanges between 1977 and 1990 — there is at least one substantial obstacle to successfully completing an exchange. Those bondholders who do not tender can see the value of their bonds rise if the exchange offer is successful, since tendering bondholders forgive part of the debt and reduce the default risk of the issuer. Although public bondholders as a group would be better off if the exchange offer goes through, those with small stakes have an incentive to hold out. Thus, it can be very difficult to complete an exchange. This free-rider problem can be, and often is, mitigated by offering a more senior security in exchange for the public bonds, one with shorter maturity, or, when it is available, cash. Moreover, in these types of exchanges bondholders may be willing to tender at below-market prices because they fear that holding out will make them effectively junior to the new securities. But, the important point is that even though these types of offers enable firms to restructure their public debt profitably, they do not, in general, result in efficient investment. The problem is that in deciding whether to tender, public bondholders take the firm’s investment policy as a given. Thus, individual bondholders — each with small stakes — fail to take into account their effect on the firm’s investment decision, despite the fact that their decisions, taken as a whole, affect investment behavior. The second principal goal of this article is to analyze the effects of bankruptcy law on investment. I conclude that the key features of the law — the automatic stay, the voting rules for plan approval, and the power of shareholders to retain value for themselves — all act to increase investment both in and out of bankruptcy. Whether this increases efficiency depends on whether the firm would otherwise have underinvested or overinvested as a result of financial distress. I characterize the aspects of the firm’s debt structure — the seniority of bank debt relative to public bonds, the maturity structure, and the existance of covenants restricting debt issues — that lead to underinvestment or overinvestment. I am then able to identify the situations in which Chapter 11 reorganization increases or decreases investment efficiency. This article is organized as follows. Section 1 presents my benchmark model of workouts when bonds restructurings are not possible and bankrupt firms are liquidated, not reorganized. I build on the Bulow and Shoven model to analyze the effects of priority and maturity on investment after the onset of financial distress. Section 2 introduces the possibility of public bond restructurings through exchange offers and compares the results of this model to those of Section 1’s benchmark model. I show that if there is no restriction on senior debt issues, exchange offers do not affect the costs of financial distress but do place more of the burden of distress on bondholders. If covenants restricting the issue of senior debt exist, however, exchange offers can be used to eliminate them and, thereby, increase investment. In this case, exchange offers may reduce the debt burden so much as to lead to overinvestment and exacerbate inefficiencies. We show that it is sometimes efficient to eliminate seniority covenants, but investment efficiency is greater if a firm can only remove them with a vote that is separate from an exchange offer. Section 3 introduces the possibility of reorganization rather than liquidation upon default. I review some of the key legal issues of Chapter 11 and analyze their effects on investment. Section 4 concludes. 1 A Simple Model of Workouts and Investment In this section, I consider a simple model of a financially distressed firm with both private and public debt. I think of the private debt as bank debt and the public debt as unsecured bonds. I model the concept that it is easier to renegotiate with a bank (or a small syndicate of banks) than with numerous bondholders by assuming at first that the firm cannot renegotiate with bondholders. I relax this assumption in Section 2, where I present a model of exchange offers. The debt’s maturity structure affects the firm’s ability to work itself out of distress and is, therefore, an important issue. I assume that all of the bank debt, with face value $B$, is short-term, maturing at date 1. By contrast, fraction $q$ of the face value of the bond debt, $D$, is due at date 1, and fraction $1-q$ is due at a later date 2. This timing reflects the fact that the bank debt generally has a shorter maturity than the bond debt. The firm has two assets: cash and/or liquid assets of $Y$, and an investment project which require an investment of $I$ at date 1 and returns a stochastic cash flow of $X$ at date 2 distributed over the support $[0,\infty)$. I denote the cumulative distribution of $X$ as $F(X)$, the density as $f(X)$, and the mean as $\bar X$. For simplicity, I assume that the firm has no fixed assets such as plant and equipment. All parties are risk-neutral, and the riskless interest rate is zero. Finally, I assume that the firm is in financial distress at date 1; its assets in place are worth less than the face value of its obligations: $Y. Thus, if the firm is liquidated, and the absolute priority rule is followed, shareholders receive nothing, and bondholders and the bank share $Y$ between them. Assuming equal priority of bank and bond debt in liquidation, the bank gets $[B/(B+D)]Y$, which I denote $L_B$, and the bondholders get ${[D/(B+D)]Y}$, which I denote $L_D$ If the firm is liquidated, the bonds maturing at date 2 are accelerated to date 1, consistent with the bankruptcy code. In this section, I assume that bankruptcy is equivalent to liquidation and that Chapter 11 reorganization is ruled out. In Section 3, I analyze how the bankruptcy law’s reorganization mechanism affects investment incentives in this model. The central question is whether the financially distressed firm invests in the project at date 1. If $Y>I+B+qD$, the firm has enough cash to invest in the project and pay off the bank and bond debts maturing at date 1. In this case, the firm invests regardless of whether the project has positive or negative net present value: if the firm does not invest, the equity gets nothing; if the firm does invest, there is some chance that the equity’s payoff would be positive. I assume that $Y so that the firm needs an additional $I+B+qD-Y$ to meet its date-1 obligations and invest in the project. The firm has several options in meeting the cash shortfall. It can try to raise new funds by issuing debt or equity, or it can try to restructure its existing bank or bond debt. I focus on the debt restructurings. I show later that the firm prefers to restructure than to issue new debt or equity. 1.1 Bank Debt Restructurings I consider bank debt restructurings first because they are substantially easier to organize than public bond restructurings. public bondholders from changing the principal, interest, or maturity of public bonds without the public bondholders’ unanimous consent. Even without the Trust Indenture Act, free rider problems can impede successful renegotiation. For example, if some public bondholders forgive part of their debt, the value of the remaining debt rises. If each public bondholder is small, and thus has no effect on the outcome of the negotiations, then each will refuse to restructure its portion of the debt. I discuss these issues in detail in Section 2. In a bank restructuring the firm effectively rolls over its initial loan of $B$ and borrows and additional $1+qD-Y$ for the investment and to pay off the bond debt due on date 1. The analysis is simplified if we assume that the interest on this loan has lower priority than all outstanding debt while the principal has equal priority. This assumption is not realistic since bankruptcy law does not distinguish between principal and matured interest. But any other assumption complicates the analysis because the fraction of the firm that the bondholders get depends on the interest rate on the new loan. On the other hand, if we assume that the new interest has lower priority, the combined return to the bank and the firm is independent of the interest rate. This permits us to complete the analysis without determining the interest rate on the bank debt. The issue this raises for the ability to renegotiate with the bank is interesting, but it is an unnecessary complication for the basic analysis. If the firm invests, and $X, the bank receives: $\frac{I+B+qD-Y}{I+B+D-Y}X$ If $X>I+B+D-Y$, the shareholders and the bank together get to split $X_(1-q)D$. The bank agrees to finance the firm provided that: where $Z\equiv I+B+D-Y$ The right-hand side of inequality (1) is what the bank receives in liquidation. There are two important assumptions implicit in this formulation. First, the firm liquidates and cannot invest in bankruptcy. In Section 3 I introduce the possibility of an investment in Chapter 11. Second, I assume that if the bank does not lend money, the firm goes bankrupt: the firm cannot raise the necessary cash from other sources. We will see below that, although it may be possible to raise outside funds, the bank has a greater incentive to provide funds than any outsider. Since I wish to derive the conditions under which the investment occurs, not how the gains from the investment are split, my analysis is unaffected by this assumption. Inequality (1) is equivalent to: The first three terms of the right-hand side add up to the market value of the bond debt conditional on bank lending and investment. Therefore, we can write (2) as: where $V_D$ is the market value of the bonds in this case. Inequality (3) captures a simple but important idea. $V_D$ is the value of the bonds conditional on investment, while $L_D$ is their value if no investment occurs. So the difference between the two measures the value transfer from the bank and shareholders to the bondholders if the firm invests. If the NPV of the project, $\bar{X}-I$, is greater than this transfer, then the firm restructures its bank debt and invests. Interestingly, This transfer can be positive or negative. If it is positive, the firm will tend to forego positive NPV projects, those with NPV between zero and $V_D-L_D$; the debt obligations act as a tax on the project, discouraging investment. If it is negative, the firm may adopt negative NPV projects, those with NPV between $V_D-L_D$ and zero; creditors effectively subsidize the project, encouraging investment. So, inefficiencies can involve either underinvestment or overinvestment. This wedge is introduced because the value of the bonds conditional on investment can be greater or less than its liquidation value. If, for example, $Y$ is close to zero, the bonds are worth almost nothing in liquidation, so bondholders benefit from the investment. In this case, the existence of the bond debt discourages investment. By contrast, if $Y$ is close to $B+D$, bondholders would get repaid nearly in full if the firm is liquidated. But, if it is not liquidated, bondholders own a risky claim , the value of which could well be below $D$. In this case, the existence of bond debt encourages investment, though it may be inefficient. This discussion suggests that there are two effects at work. On the one hand, the debt obligations tend to make investment look unattractive because existing creditors can siphon off cash from the project. This is Myers’ (1977) well-known argument; the existence of a “debt overhang” discourages investment. On the other hand, debt obligations can lead the firm to take excessive risks: the equity receives nothing if the firm is liquidated but has some value if the firm invests, even if it is in a negative NPV investment, a point made clear by Jensen and Meckling (1976). The maturity structure of the debt has important effects on the efficiency of investment. As the maturity of the bonds becomes shorter (i.e., $q$ increases), its value increases because the date-1 portion is safe and the date-2 portion is risky: $dV_D/dq>0$. This increases the transfer of value to bondholders and reduces the firm’s incentive to invest. At the limit, as all the public debt becomes due at date 1, the transfer approaches $D-L_D>0$. In this case, the firm may pass up positive NPV investments but will never choose negative NPV investments. The efficiency of shortening the maturity of the bond debt is ambiguous. The increase in $q$ may force the firm to pass up positive NPV projects, but it also deters investment in negative NPV projects. An increase in bank debt, holding total indebtedness, $B+D$, constant has an unambiguously positive effect on efficiency. The increase in $B$ decreases the right-hand side of inequality (3) if it is positive and increases it if it is negative. So, the shift toward bank debt away from bond debt can either induce the firm to take positive NPV projects it would not have taken or turn down negative NPV projects it would have taken. Clearly, if all debt were held by the bank, investment would always be efficient; bank renegotiation is assumed to be costless so the conditions of the Coase Theorem are satisfied. 1.2 New capital Infusions Instead of restructuring its bank debt, the firm could try to raise new money from another bank or by issuing equity. Neither of these alternatives is as attractive as a restructuring. Like in a restructuring, the new bank lends $I+B+qD-Y$ and receives the same date-2 payoffs. But, unlike in a restructuring, some of the new money goes to pay off the existing bank debt of $B$ at face value. We can show that the firm will be able to raise new debt financing provided that: or, in other words, if the NPV of the investment exceeds the sum of the transfer to bondholders, $V_D-L_D$ and the transfer to the bank, $B-L_B$. The condition differs from that for a bank restructuring because, in a restructuring, the bank accounts for the fact that the debt is worth only $L_B in a liquidation. If the firm obtains new bank financing, the original bank receives a transfer of $B-L_B>0$. This subsidy means that the set of investment projects that can be financed without outside debt is a strict subset of those that can be financed with a bank restructuring. Investment is even less attractive if the firm issues equity rather than debt. The bank continues to receive a subsidy of $B-L_B$, but the transfer to bondholders increases. The bond debt conditional on investment is worth more because the date-2 portion of the debt is paid off before the equity is paid anything. By contrast, when the firm issues debt, the bondholders and the new bank are on equal footing on date 2. So, the condition for investment takes the same form as inequality (4), except that $V_D$ is greater when the firm issues new equity. This analysis implies that the firm never issues equity since an equity issue transfers value to bondholders which would not be transferred by a debt issue. The prediction is less clear about the choice between debt issues and a bank loan restructuring. Clearly, when inequality (3) is satisfied but inequality (4) is not, the firm will restructure its bank debt. But if both inequalities are satisfied the model has no prediction. The bank knows that if there is no restructuring the firm will issue new debt and the bank will receive $B$. So, in a restructuring, the bank will settle for nothing less than $B$. As a result, equity holders are indifferent between a debt issue and a bank debt restructuring because they must transfer $B$ to the bank in both situations. 1.3 The Effects of Priority So far, we have assumed that all debt has equal priority in bankruptcy. However, firms can explicitly contract for certain debts to be paid ahead of others in bankruptcy. There are two way in which priority can affect the ability of distressed firms to raise capital in our model. First, the seniority of the existing bank debt affects what the bank would get in bankruptcy liquidation if it did not lend new money, thereby determining the value of the bank’s next best alternative. The more junior the existing bank debt, the worse off the bank is in liquidation, so the more willing to lend. Second, the seniority of the new bank debt affects what the bank can get if it lends new money. In general, the more senior the new bank debt, the better off the bank is at any chosen interest rate. Thus, if they could, the firm and the bank would would like to issue debt that is senior to the existing bonds. Of course, there are often constraints on their ability to do so; the bonds may contain covenants restricting the issuance of any debt senior to the bonds. These restrictive covenants may prohibit such issuance altogether, may limit its amount, or may allow it subject to certain financial ratios being achieved. To see this more formally, suppose there is no restrictive covenant prohibiting senior debt issuance. Then the interest rate on the new bank debt can be set at such a high level that the firm always defaults on date 2 and the senior debt gets all the date-2 cash flow $X$. This means that the value of the bonds conditional on new senior lending is just $qD$ and bondholders receive their date-1 payment. The value of the bonds if the firm is liquidated is $L_D$, assuming, as before, that the existing bank debt and bonds have equal priority. Based on the previous section we know that the project’s NPV must exceed the net subsidy from investment to bondholders. So the bank will be willing to lend, provided that: The right-hand side of inequality (5) is strictly less than the right-hand side of inequality (3), since $qD; the firm is more inclined to invest when there is no covenant restricting senior debt issuance. It can have positive or negative efficiency effects by reducing the underinvestment problem or exacerbating the overinvestment problem. This analysis can tell us something about the interaction between maturity structure and seniority covenants. If the bond debt has a relatively short maturity ($q$ near $1$), the firm is likely to underinvest. In this case, a seniority covenant tends to worsen the problem, making it more difficult for the firm to raise capital. If the firm leaves out the covenant, we would expect to see the bank lend new money that is senior to the old bonds. The ability to issue such debt can counteract the efficiency created by the short maturity of the bond debt. In contrast, if the debt has a relatively long maturity, the firm is more prone to overinvest. In this case, a seniority covenant makes it more difficult to raise capital and could eliminate the tendency toward overinvestment. Thus, if the capital structure is chosen to minimize the costs of financial distress, we would expect long-term bond debt to contain seniority covenants in the indentures and short-term bond debt to omit such covenants. This framework can also tell us something about the interaction between the interaction between bond debt maturity and the priority of the existing debt. Suppose that there is no seniority covenant. Then, if the original bond debt is pari passu with the bank debt, the investment condition is given by inequality (5). But if the initial bank debt is senior to the bond debt, the condition becomes: because the value of the junior bond debt in liquidation is now $max(Y-B,0)$. Since this is less than $L_D \equiv [D/(B+D)]Y$, the value of the bond debt if it is pari passu with the old bank debt, the firm is now less prone to invest; the bank does better in liquidation, so financing new investment is less attractive. The shorter the maturity of the bond debt, the more likely the firm is to underinvest. Thus, the model suggests that when the bond debt is relatively short term, existing senior bank debt is likely to worsen the underinvestment problem. But when the bond debt is long term, the seniority of the bank debt can be a useful way of curbing the overinvestment problem. If the costs of financial distress drive capital structure choices, my model predicts that the bank debt will be senior if the bond debt is long term and junior if it is short term. Although the model predicts that the bank debt will be junior if the bond debt is short term, it is difficult to make short-term bank debt junior in practice. To see this, suppose that if the firm does not invest and is not liquidated at date 1, it nevertheless has positive, stochastic cash flows at date 2. Thus, unlike the model above, if the firm pays off its debts at date 1, the value of the equity is positive even of the firm does not invest. The firm has three alternatives: invest, continue without investing, or be liquidated. Now suppose that $Y \geq qD+B$ so that it is feasible for the firm to meet its date-1 obligations and continue in operation without investing. The value of the bank debt is $B$, which is what it is worth in liquidation if the bank debt is senior. The bank refuses to provide new funds for investment, but demands payment of $B$ in period 1. This is more than $max(Y-D,0)$, the bank’s payoff if the firm is liquidated and the bank debt is junior to the bond debt. Thus, even though the bank debt is contractually junior to the bond debt, the bank acts as its senior. This makes the bank reluctant to lend new money, a more efficient outcome. So, in this model, if $q$ is small enough so that $Y>qD+B$, the bank acts as a senior lender. But, if $q$ is very close to one, it is possible to induce the bank to subordinate itself to the bond debt. 2 Distressed Exchange Offers for Bond Debt So far I have assumed that it is impossible to renegotiate with bondholders. This assumption is not too far off the mark; the Trust Indenture Act’s prohibitions on changes in the timing or amount of bond debt payments forces bond restructurings to take the form of exchange offers. In this section, we analyze the extent to which this limited form of renegotiation affects the inefficiencies discussed in the previous section. The key assumption of the model is that each bondholder’s stake is small enough that they ignore the effect of their tender decision on both the firm’s investment decision and the value of the firm’s securities. This assumption is unrealistically strong for firms with a large portion of their bond debt held by just a few institutional investors, an admittedly common situation. I make this assumption to highlight the problems that arise when creditors cannot fully coordinate their actions. I believe that similar effects would be present in a model in which bondholders do have substantial stakes. I proceed in two stages. First, I analyze the profitability of exchanges assuming that the firm has ample cash to finance the investment even without a debt restructuring. I will show that an exchange is profitable only if the debt is exchanged for cash or for debt that has higher priority than the original debt. Although this analysis has no efficiency implications — the firm invests even without an exchange — it is helpful in answering the second, more interesting, question: when can an exchange reduce cash obligations and enable the firm to invest? I will show that the bank is generally better off if the firm can exchange its bonds, that investment incentives are unaffected by the ability to exchange debt in most circumstances, and that the ability to exchange is not equivalent to efficient renegotiation of the bond debt. 2.1 Exchanges Assuming No Cash Shortage In this subsection I assume that, while the firm is in financial distress, it does not need an exchange or a bank concession in order to invest and meet its date-1 obligations: $Y>I+B+qD$. I first consider an exchange for debt due at date 2 with a face value of $p$ for each$ in face value of the existing bond debt. Let $X_b$ be the breakeven value of $X$, so the firm defaults at date 2 for all $X < X_b$. Shareholders receive nothing if $X and receive $X-X_b$ otherwise. Thus, an exchange is profitable if and only if it lowers $X_b$.

Let $\beta$ denote the fraction of bond debt the firm exchanges. Without an exchange, $X_b=I+D+B-Y$. By contrast, if the firm exchanges, it owes the nontendering bondholders $(1-\beta)D$ and the tendering bondholders $\beta pD$, so $X_b=I+(1-\beta)D+\beta pD+B-Y$. Here, $X_b$ is decreasing in $\beta$ if and only if $p<1$, i.e., the firm can exchange a $of old debt for less than a$ of new debt. So, if $p<1$, an exchange is profitable and. if $p>1$, an exchange is unprofitable.

Proposition 1

It is unprofitable to offer an exchange for new debt with equal priority to the old bond debt

Proof.
For a given $p$ and $\beta$, the value of the firm in default is: $X+Y-I_(1-\beta)qD$. Total outstanding claims at date 2 are $(1-\beta)(1-q)D+\beta pD$, of which tendering bondholders collectively receive a fraction $\beta pD/[(1-\beta)(1-q)D+\beta pD]$. Thus, the value of each of the $\beta D$ tendered bonds is:

Each nontendering bondholder receives a certain payment, $q$ at date 1 and a risky claim at date 2 comprised of their share of the insolvent firm if $X and full payment of $(1-q)$ if $X>X_b$:

Equating (7) and (8) determines, for any given $\beta$, the $p$ at which bondholders are just indifferent between tendering and not tendering. This equation can be rewritten as:

At $p=1$, the left-hand side is:

The integrand is 0 at $X=X_b$ and positive for $X, so (A4) is positive at $p=1$. Since the left-hand side of (9) is decreasing in $p$, the $p$ that solves (9) is greater than one.

The exchange is unprofitable because of a classic holdout problem. Will the holdout have an incentive to tender, assuming that all the other bondholders tender? If so, then it is an equilibrium for all bondholders to exchange.

The answer depends on the payoffs of the two bonds when the firm is in default at date 2. If the firm does not default, the bondholder is just as happy with the new bond as with the old. But if the firm does default at date 2, the payoffs are quite different. Those who tender receive their pro-rata share of the firm at date 2, $(X+Y-I-B)/D$, but the holdout receives $q$ at date 1 and receives a pro-rata share of the firm at date 2, $(1-q)(X+Y-I-B)/D$. since $(X+Y-I-B)/D<1$, the bondholder is better off holding out.

The holdout is better off because the earlier payment on the old bond is effectively senior to the new bond. Tendering bondholders share ratably in a risky date-2 claim. But, holding out, the bondholder receives a safe date-1 payment while still sharing pro-rata in the date-2 portion of payoffs.

This logic rests crucially on the assumption that the bondholders do not act collectively. Suppose they could. Then the question becomes: are we all better off if we all tender than if we all hold out? This is quite different from the individual in question: am I better off if I tender than if I hold out assuming everyone else tenders? In the collective case if everyone tenders then the payoff is again $(X+Y-I-B)/D$ when the firm defaults. But, if no one tenders then the payoff is $q$ at date 1 and $(X+Y-I_B-qD)/D$ at date 2. This is equal to the payoff from tendering, so bondholders as a group are indifferent between the two options when $p=1$.

The holdout problem is even more pronounced if the firm offers to exchange junior debt or equity for the old bonds. There are now two reasons why bondholders would want to hold out. As before, holdouts are senior in that some of their claim is paid at date 1, before the uncertainty is realized and tendering bondholders are paid. In addition, holdouts also have seniority at date 2 since the new security is junior debt or equity. If all bondholders tender, a holdout’s claim would be riskless since the holdout gets $q$ at date 1, and the $1-q$ that is owed at date 2 is senior to the claims of all tendering bondholders, making it riskless as well. Thus, a corollary of Proposition 1 is that exchange offers for junior debt or equity are also unprofitable.

Quite the opposite result holds true if the firm can offer a more senior bond in exchange for the old bonds. These types of exchanges are quite common.

Proposition 2

It is profitable to offer an exchange for new debt which is senior to the old bonds

Proof.

The value of the old bonds, given $\beta$ and $p$, when the firm exchanges for senior debt is given by:

where $X_1\equiv I+B+(1-\beta)qD=\beta pD-Y$ is the cutoff value $X$ above which the new bonds are paid in full and $X_b\equiv I+B+(1-\beta)D+\beta pD-Y$ is the cutoff value of $X$ above which the old bonds are paid in full and the firm is solvent: $X_1\leq X_b$, with equality if and only if $\beta=1$. Since the new debt is senior to the old debt, for $X$ between $X_1$ and $X_b$, holdouts share $X+Y-I-B-(1-\beta)qD-\beta pD$, the cash left after date-1 payments and date-2 payments of $\beta pD$ to the new senior bonds.

Tendering bondholders do not receive $q$ at date 1 but do receive a senior claim at date 2. The value of their debt is:

We now show that the value to the firm of an exchange is increasing in $\beta$. Equating (11) and (12) and combining them gives:

Since $X_b=I+B-Y+(1-\beta)D+p\beta D$,

Differentiating (13),

Multiplying by $\beta$ and substituting from (13),

and

The denominator is clearly positive. The numerator is equal to:

At $X_1$ the integrand is $(1-\beta)(q-1)D$, which is negative. Since the integrand is increasing in $X$, the numerator is negative, and $X_b$ is decreasing in $\beta$. Thus, once we determine whether an exchange offer is profitable by checking to see whether $p$ is greater or less than one at $\beta =1$.

If we set $\beta=1$, $X_1=X_b$, then we can rewrite (13) as:

Since $Y>I+B+qD$ by assumption, the first term in (20) is positive.
Thus, to satisfy (A13), $p$ must be less than one.

There are two competing effects at work. Again, the difference in the payoffs from tendering and holding out depends on the payoffs of the old and new bonds when the firm is in default at date 2. As before, consider the decision facing the holder of $1 of bonds, assuming that all others tender when $p=1$. On the one hand, the holdout’s date-2 claim is worthless when the firm defaults. Since the new debt is senior, each new bondholder is paid $(X+Y-I-B)/D$ and there are insufficient funds to pay the old junior bondholder. On the other hand, the portion, $q$ of the holdout’s claim is paid at date 1, making it effectively senior to the new bonds. On the whole, given my assumptions that $X>0$ and $Y>I+B+qD$, the increased seniority at date 2 is worth more than the earlier maturity of the $q$ portion of the claim. Instead of a holdout problem there is a hold-in problem; bondholders would tender for $p<1$ despite the fact that they’re made worse off as a group. The hold-in problem is more severe when the bond debt is relatively long-term. Very short maturity debt is paid off almost in full at date 1. So only a small portion of the debt can be leapfrogged in the capital structure. The short maturity of the bonds effectively gives them a degree of seniority that cannot be negated by a senior debt issue. Indeed, one can show that, as the bonds become shorter-term, $p$ increases and exchanges become less attractive to the firm. I have shown that the firm prefers exchanges for senior debt to exchanges for pari passu or junior debt. But in many cases there are seniority covenants in the bonds prohibiting senior debt issues. Yet firms with such covenants do issue more senior debt in exchanges.How is this possible? Bond indentures typically specify that covenants can be changed or eliminated by either a simple majority or a super majority vote of the face value of the bonds. The exact opposite is true if the firm can offer a more senior bond in exchange for the old debt. Empirical evidence supports the proposition that these latter types of exchanges are quite common. In Section 2.2, below, I discuss the efficiency consequences of tying the covenant waiver to the exchange offer via an exit consent. There are at least two other ways firms commonly structure an exchange. One is to offer cash instead of a security such as debt or equity. Another is to offer debt with a shorter maturity than the existing debt. It turns out that in my two-period model these alternatives are equivalent. Debt due at date 1 is paid off with certainty, so exchanges for short-term debt are equivalent to cash repurchases. Proposition 3 It is profitable to offer an exchange for cash Proof. Suppose the firm offers to exchange each$ of old debt for $V$ \$s of cash or new short-term debt. Bondholders will be indefferent between tendering and not for any $V$ and $\beta$, provided that:

where $X_b \equiv I+B+(1-\beta)qD+V\beta D-Y$. We can rewrite (21) as:

Totally differentiating (22) yields:

where the second equality follows from substituting (23) into (24). From (22), $V<1$, so (24) implies that $V^{'}(\beta)<0$. The cost to the firm of the exchange is $\beta DV(\beta)$. Since bondholders are indifferent between tendering and not, the expected payments to the nontendering bondholders must be $(1-\beta)DV(\beta)$. Adding, the expected payments to all bondholders is $V(\beta)D$. So, the firm maximizes profits by choosing $\beta$ to minimize $V(\beta)$. Since $V^{'}(\beta)<0$, an exchange for cash is profitable.

Exchange offers for cash are profitable for similar reasons that senior debt exchanges are profitable. As more bondholders tender, more cash is paid out at date 1, reducing the value of the old debt at date 2. Tendering bondholders are paid cash for the $1-q$ portion of their claim at date 1. Since this is paid before a holdout receives payment on the $1-q$ portion of their claim, the tendering bondholders are effectively senior to the nontendering ones. As a result, the date-2 portion of the old debt claim is less valuable. Faced with this hold-in problem, old bondholders are willing to tender at a low price.

Recall that throughout the analysis we have assumed that the firm does not have a cash shortage. If the firm does not have sufficient cash, it will use all of its cash in excess of $B+I$ to buy back debt. It is important to note that the firm would not find it profitable to issue outside equity or debt (with equal or junior priority to the old debt) in order to buy back bonds. The outside capital would not be senior to the untendered debt, so the required return on the outside capital would more than negate the savings on the exchange offer.

In this model, the ability to exchange for cash does not lead to any added inefficiencies since the firm will always invest in the single project. However, in a model in which there are several projects or the level of investment is a choice variable, significant inefficiencies can result. The firm may choose to use cash which could be invested in positive-NPV to buy back bonds if the cash flow relief to creditors exceeds the NPV of the project. But this inefficiency is limited in scope; financially distressed firms tend not to have a great deal of excess cash available for this type of activity.

As I discussed above, there is no difference between an exchange for cash and one for short-maturity debt of any priority in this model. In a model with more than two periods, there may be a difference because the firm may not have enough cash to exchange all the debt for cash immediately but may be able to achieve a similar effect with an exchange for shorter maturity debt. My analysis suggests that an exchange for shorter maturity debt is profitable when the firm can make the bondholders who tender effectively senior to those who don’t. This is possible if the realization of the risky project occurs after the new debt matures, but there are some relatively certain cash flows before the new debt matures. This allows the new debt to have low default risk and be paid off before the old debt matures.

2.2 Exchanges When There Is a Cash Shortage

The above analysis assumes that the firm does not need to restructure its debt in order to invest at date 1. Exchanges have no effect on efficiency; they just redistribute value from bondholders to shareholders. We now suppose that the firm needs a concession from either the bank or the bondholders to invest at date 1. We start by assuming that $I+B; the firm needs some concession to invest but has enough cash to pay off the bank and invest. I explicitly model bank renegotiation and bond exchanges.The firm approaches the bank seeking a concession. It makes a take-it-or-leave-it offer to postpone some or all of $B$ until date 2, perhaps along with some debt forgiveness. The firm has the option of offering an exchange to bondholders. This timing captures the idea that a firm is unable to commit to the bank not to pursue a profitable exchange offer. Suppose the bank refuses to give the firm a concession. At this point, the firm can propose to exchange the bonds for more senior ones. (As we saw above, this is preferred to offering new bonds that are pari passu with the old ones.) I assume for the moment that there is no seniority covenant. Because the new bonds are senior to the old, the firm can set $p$, the face value of the new bonds, so that it is paid all of the date-2 cash flows. Thus, the maximum value of a unit of the new bond is $(\bar{X}-Y-I-B)/D$, provided that the firm buys back all the bonds. $X+Y-I-B)/D>q$ or, equivalently, if:

an exchange offer for senior debt is feasible. In this case, the firm will want to buy back its bonds because the alternative is liquidation in which case shareholders get nothing.

Now consider the first stage of the model in which the firm approaches the bank to receive a concession. The bank knows that if it turns down the firm’s offer, the firm will be able to exchange its bonds provided that inequality (7) is satisfied. In this case, the bank receives $B$. So, the bank will turn down any offer which has an expected value less than $B$.

It is possible that the firm might prefer to renegotiate with the bank to receive some date-1 debt relief rather than restructure its bonds. As long as it can defer enough of its bank debt to pay off the date-1 portion of the bonds, this strategy is feasible. So, suppose that the bank extends the maturity of its loan but requires the firm to pay $B'$ at date 2. Assume for the moment that there is no bond covenant prohibiting the issuance of senior debt; $B'$ can be senior to the date-2 payments on the bonds. In addition, if $Y the bank has to provide new money in the amount of $I+qD-Y$. If $Y>I+qD$, the remaining cash of $Y-I-qD$ is available to pay off the bank at date 1. Since the new bank debt is senior, the minimum $B'$ that the bank would accept satisfies:

Proposition 4

If $I+Band there are co contractual restrictions on issuing senior debt, the firm prefers a bond exchange to a bank restructuring.

Proof.

From (A13), the exchange offer terms for senior debt are determined by:

where $X_b=X+Y-I-B-pD$.

In an exchange, the shareholders receive $X-X_b$ if it is positive and zero otherwise. In a bank restructuring, shareholders receive $X-B^{'}-(1-q)D$ if it is positive and zero otherwise. So an exchange is more profitable provided that $X_b.

To show that this is indeed the case, we assume , to the contrary, that $X_b\geq B^{'}+(1-q)D$. Thus, let $B'\equiv X_b-(1-q)D-\epsilon,\;\epsilon\geq 0$. Then, equation (8) can be rewritten as:

Using the definition of $X_b$ and rearranging, (28) becomes:

Now compare (29) and (27). The only difference is the limits of integration. Note that the left-hand side of (27) is increasing in $X_b$ and, since $X_b-(1-q)D-\epsilon, the left-hand side of (29) is less than that of (27). Thus, if (27) is satisfied with equality, (29) must be violated. Therefore, $B^{'}$ must be greater than $X_b+(1-q)D$

In both an exchange offer and a bank restructuring, the bank ends up with a claim worth $B$. However, the exchange is less costly because the firm can take advantage of the hold-in problem; by exchanging for senior debt and leaving holdouts with a junior security, the firm induces bondholders to tender for a claim that the bank would not accept.

Now suppose instead that $\bar{X}-I, so expression (7) is violated. In this case, an exchange offer is not feasible without a bank concession. Thus, if the bank turns down the firm’s take-it-or-leave-it offer, the firm is liquidated and the bank gets $L_B$. This means that the firm can offer the bank a claim worth $L_B$ and the bank will accept the offer. Also note that when $Y the bank would also accept an offer of $L_B$ because, without such a writedown, the firm would be unable to invest at date 1.

Given an offer worth $L_B$ and the bank’s acceptance, the firm may be able to exchange its bonds. In the exchange, the maximum value of each new senior bond is $(\bar{X}+Y-I-L_B)/D$, while each untendered bond is worth $q$ because there will be no funds available at date 2 to pay off the untendered junior bonds. Thus, the firm can complete an exchange provided that:

Not that if the echange is successful, the firm will be able to make its date-1 bank payment of $L_B$ and invest $I$ since I have assumed that $Y>I+B>I+L_B$. If (9) is violated, however, the firm does not offer to exchange and is therefore liquidated at date 1.

There will tend to be underinvestment if the current portion of the bond debt, $qD$, exceeds the liquidation value, $L_D$, and overinvestment if the current portion is less than the liquidation value. The minimum transfer to the bondholders from investment is the least that they can be given with investment, $qD$, minus what they get in liquidation, $L_D$. If the transfer is positive, there is underinvestment, and if the net subsidy is negative, there is overinvestment.

The condition for investment is exactly the same as in the model in Section 1 in which exchange offers were ruled out, but it is possible for the firm to issue senior bank debt. In both cases, investment occurs if the NPV of the project exceeds $qD-L_d$. Although investment behavior is no different, the parties who pay for the investment are different. If (9) is a strict inequality, the bondholders are worse off with an exchange than with a bank debt restructuring. In the bank restructuring they keep their old securities, while in the exchange the hold-in problem leads bondholders to accept a lower value security. Since the bank gets a claim worth $L_B$ in both cases, the equity is the beneficiary of the exchange offer.

Thus, exchange offers can be profitable to the firm if it is able to exchange the bonds for more senior securities or has excess cash it can use to buy back the bonds for cash. But note that the ability to exchange does nothing to improve the efficiency of investment decisions of financially distressed firms if there is no seniority covenant in the bonds; it just affects who bears the costs of financial distress. The reason is that bondholders take the success of the exchange as a given in making their tender decision. Therefore, they do not consider how a change in operating policy made possible by the exchange affects their claim.

We summarize these results in the following proposition:

Proposition 5
If the firm has insufficient cash to invest, there are three possible outcomes. If the NPV of the investment, $\bar{X}-I$, is sufficiently large, the bank is paid in full, the bondholders accept an exchange, and the firm invests. For intermediate NPVs, the bank debt is forgiven to $L_B$, the bondholders accept an exchange, and the firm invests. If the NPV is sufficiently small, the firm is liquidated and does not invest. The possibility of a bond exchange does not alter investment when there are no covenants prohibiting senior debt issues.

The analysis assumes that there is no covenant in the bonds prohibiting the issuance of senior debt. As discussed in Section 2.1, however, firms can get around this covenant through an exit consent in which bondholders simultaneously tender their bonds for more senior ones and, as condition of the exchange, vote to remove the seniority covenant from the original bonds. The condition for investment continues to be given by inequality (9).

Thus, exchange offers combined with exit consents can be used to strip seniority covenants that would otherwise prevent a bond exchange and constrain investment; in this case, exchange offers have real investment effects. But the firm can go too far; exit consents and exchange offers can reduce the value of the bonds so much that the firm actually overinvests. Coffee and Klein (1990) have argued that the “coercive” character of exit consents leads to inefficiencies and have called for their ban. The ban they propose would be such that tendering bondholders could still vote to remove covenants but such vote would not be a condition for tendering in an exchange.

Such a ban on exit consents is efficient in my model. To see this, suppose that there is a seniority covenant on the bonds. The interesting caseis where the firm cannot raise new bank financing that is pari passu with the existing debt: $\bar{X}-I and $V_{DLD}>0$, so that the firm potentially underinvests. If the firm could renegotiate directly with bondholders they would be willing to reduce the value of their debt conditional on investment to $L_D$ through a reduction of principal or interest. Of course, the Trust Indenture Act does not permit bondholders to reduce $V_D$ by voting to waive the seniority covenant. A the same time, the bank lends new money senior to the bonds, and the interest rate is chosen so that the value of the bonds, $V_D^{'}$ is anywhere from a minimum of $qD$ to a maximum of $V_D$. (Note that $V_D^{'}$ cannot be below $qD$ because if the firm invests the payment of $qD$ is required.)

Bondholders will accept a covenant waiver only if they know that they will receive at least $L_D$ as a result of the exchange. If $qD>L_D$, the value of the bonds cannot be reduced all the way to $L_D$, and the offer will be $qD$. So, $V_D^{'}=max\{qD,L_D\}$. Thus, the condition for investment with a covenant waiver is:

Contrast this condition to inequality (9) which determines investment when exit consents are possible. The two conditions are the same when $qD>L_D$. In both cases underinvestment may result because there are limits on how much debt reduction is feasible via exit consents or covenant waivers. But when $qD, exit consents allow some negative NPV projects to be taken while covenant waivers do not. The firm can reduce the value of its bonds below their liquidation value when exit consents are possible but cannot do this when bondholders vote separately on the covenant waiver. Thus, in some situations, exit consents go too far in lowering the debt burden. We have only focused on the case where $\bar{X}-I and $V_D-L_D>0$, but in the other cases covenant waivers also lead to weakly more efficient investment outcomes than exit consents.

The conclusion is that exchange offers only alter investment behavior when there is a covenant in the bonds prohibiting senior debt issues. In these cases, firms can use exit consents to remove covenants, issue senior debt, and increase investment. By contrast, if the firm is prohibited from using exit consents and instead must ask a separate vote to waive a seniority covenant, investment decisions are improved.

The results of this section indicate that the firm would never propose an exchange for more junior securities. This is difficult to reconcile with empirical observations. There are two promising explanations. First, the firm has private information, it may signal its information by the type of security offered in an exchange. As Myers and Majluf (1984) show, equity issues can signal the value of the firm is low. The firm may then offer an exchange for equity so that bondholders lower the value of the claim they require in exchange. This may offset the losses the firm incurs from the holdout problem created by an exchange for a more junior security.

A second reason why firms may offer junior securities is that bondholders are not really atomistic. In this case, the firm may be able to convince a sufficient number of large bondholders that their acceptance of equity is necessary for a successful restructuring. Equity may be preferred because it reduces the cash drainage from the firm.

Finally, we note two historical developments that have made exchanges less attractive. In the LTV bankruptcy, Judge Lifland disallowed a portion of the bondholder claims who had participated in a previously completed exchange. He ruled that the admissible claim was the market value of the bonds at the time of the exchange, not their face value. Thus, there may be some reluctance to exchange for fear that the firm could subsequently file for Chapter 11. In addition, the tax treatment of exchanges was changed as part of the Revenue Reconciliation Act of 1990, requiring the firm to recognize cancellation of debt income based on the market value of the new securities, not their face value. Firms may be able to avoid this tax liability in Chapter 11.

3 Reorganization Law and Investment

In the U.S., financially distressed companies often seek court protection under the provisions of Chapter 11 of the Bankruptcy Code. These provisions in the Code are intended to promote reorganizations of economically viable firms as going-concerns and thereby avoid inefficient liquidation of distressed firms. When a firm files for bankruptcy, all its debts become due, but an automatic stay is invoked stopping all principal and interest payments, and secured creditors lose the right to take possession of their collateral.

In Chapter 11, control of a firm, known as the debtor in possession, typically remains with the current management and board of directors. This contrasts with Chapter 7 bankruptcy proceedings in which a trustee takes control and manages the company while organizing a piecemeal liquidation or sale of the firm as a going concern. Creditors are paid in accordance with the absolute priority rule, so equity gets nothing unless all creditors are paid in full. In Chapter 11, management is permitted to continue operating the business, but all significant decisions are subject to court review and legal motions by creditors to disallow the proposed policy. In reviewing the debtor’s policies, the court’s objective is to approve policies which increase the value of the estate. The court has the charge of promoting “equitable” resolutions. This gives the court significant latitude in overseeing the debtor’s operations. In addition, the fiduciary responsibility of management is to maximize the value of the estate, not the value of equity.

Operations proceed with court oversight until a plan of reorganization is approved through a voting procedure of creditors or the firm is liquidated (piecemeal or as a going concern), either in Chapter 11 or after conversion to Chapter 7. A plan of reorganization specifies a new capital structure for the firm, delineating how creditors are paid in terms of cash or securities of the reorganized firm.

In this section, I focus on the three aspects of Chapter 11 that I believe are fundamental to understanding its effect on operating and investment decisions: the automatic stay, the voting rules that determine whether a plan of reorganization is approved, and the maintenance of equity value despite the fact that creditors are not paid in full. In general, Chapter 11 has ambiguous effects on efficiency, but the analysis characterizes the situations in which efficiency is enhanced or diminished.

3.1 The Automatic Stay

The automatic stay increases the firm’s incentive to invest. To see this, suppose that the firm files for Chapter 11 and that the automatic stay is the only feature of Chapter 11. Bondholders’ claims are deferred until date 2, at which time they are either paid in full or share the firm’s assets with the bank if the firm is unable to make its debt payment.

Effectively, the automatic stay extends the maturity of the bonds from $q>0$ to $q=0$. As we are aware, the firm has a greater incentive to invest when the debt has longer maturity. There are two separate effects. First, the firm may now have the cash needed for investment, so it may not have to borrow funds at date 1: $Y$ may be less than $I+B+qD$ but greater than $I+B$. And even if the firm must borrow $(Y, investment is more attractive because the automatic stay forces bondholders to bear more risk.

The firm may be more willing to invest, but it is not necessarily efficient for it to do so. Bondholders may be forced to bear too much risk, leading the firm to overinvest. The oversight of the court and the ability of bondholders to object to the firm’s investment plans may prevent large abuses of this type.

This analysis assumes that the new money comes from the bank and is pasri passu with the outstanding bonds. But the debtor will generally seek the court to approve financing senior to all existing debt. Such financing — known as debtor in possession (DP) financing — is considered an administrative cost which is paid ahead of all other creditors. The court can even make post-petition debt senior to other administrative costs. In addition, the court can approve a cash collateral agreement, allowing the debtor to use liquid assets to finance its operations even if these assets are pledged as collateral to a creditor. Thus, the court can effectively strip seniority covenants and security from existing debt. This leads to even greater investment incentives, although the junior creditors who are potentially harmed by the new senior investment can oppose the new financing in court.

The automatic stay also affects the ban’s incentives to lend outside of bankruptcy. Since the subsidy to the bondholders is reduced by the automatic stay, the bak and the firm have an incentive to restructure inside bankruptcy rather than outside. If the deadweight losses associated with bankruptcy are less than the reduction in the net subsidy to bondholders, firms will file for bankruptcy even though they could have successfully restructured outside of bankruptcy. In this case, the Chapter 11 option can reduce efficiency. Investment is unchanged by the filing, but the firm is willing to incur a deadweight cost to extract value from bondholders.

3.2 Chapter 11 Voting

Investment inefficiencies arise in my model because of the inability to negotiate directly with bondholders. The underlying problem is that, unlike the bank, bondholders do not take into account their effect on the firm’s investment policy.

Chapter 11 voting procedures can get around this problem. Plans of reorganization must be approved by all classes of creditors and the court. Classes are determined by grouping creditors with essentially equivalent claims. So, for example, secured and unsecured creditors are always assigned to different classes. A class approves a plan if two-thirds of the allowed monetary interests and a majority in number within the class accepts the plan. A dissenting member of a class can object to a plan if they get a claim worth less than their claim in liquidation.

To see how the voting procedure affects restructuring and investment, suppose that the firm files for Chapter 11 and immediately proposes a plan of reorganization that gives bondholders a claim on the reorganized company which, conditional on investment, is worth $L_D+\epsilon$, a little more than the bondholders’ recovery under liquidation. Furthermore, suppose that this is a take-it-or-leave-it offer and that if the plan is rejected the firm is liquidated. In deciding how to vote, a bondholder compares their return if the plan is successful with their recovery if it is not. If the plan is successful, all bondholders share $L_D+\epsilon$. If the plan is unsuccessful, all bondholders share $L_D$ in liquidation. Thus, they all vote for the plan. The debtor can offer the bondholders a claim worth just above its liquidation value, so there is no subsidy to or from bondholders. The result is efficient investment.

Why does the voting mechanism work while an exchange offer does not? The answer is that the voting procedure does not allow bondholders to be treated differently depending on their vote, whereas tendering and nontendering bondholders are treated differently. In an exchange offer , a bondholder compares the value of the new claim with the value of the old claim conditional on success of the exchange offer because it is possible for the bondholder to retain their old claim even if the tender offer is successful. But if the conditions for acceptance under the voting procedure are met, those who do not vote for the plan are compelled to accept the offer.

The voting procedure is unlikely to work as smoothly as we have modeled it. In practice, the debtor does not have all the bargaining power. The threat to liquidate the firm if the plan is rejected may not be credible; the debtor may choose to continue operating the firm under Chapter 11. Asymmetric information may lead to inefficiencies through strategic behavior and delay.Nevertheless, an important feature of voting is its capacity to overcome the holdout and hold-in problems.

This analysis raises a natural question: if Chapter 11 voting procedures enhance efficiency, why can the firm not include in its debt covenants a provision that mimics Chapter 11 voting procedures for exchange offers by the firm. The answer is that, as discussed above, the Trust Indenture Act of 1939 prohibits it.

This voting rule can help the firm obtain concessions from bondholders. Even if the bank is willing to lend outside of Chapter 11, the firm may be better off filing for bankruptcy and taking advantage of the voting procedure to obtain a transfer from bondholders. This is likely to be the best strategy when these concessions are large. Thus if the bonds are relatively short term, senior, or protected by seniority covenants, the bonds are generally more valuable outside of Chapter 11 than inside. In these cases, we would expect firms to file.

3.3 Maintenance of Equity Value

One of the most salient features of Chapter 11 reorganizations is that shareholders typically retain a stake in the firm, even though creditors are not paid in full.

The debtor’s bargaining power in Chapter 11 is derived from a number of procedural rules on the formation and acceptance of a plan or reorganization. The debtor in possession has the exclusive right to propose a plan for the first 120 days after filing the bankruptcy petition. This exclusivity period can be, and often is, extended by the judge for up to 18 months. Only after exclusivity is lifted can creditors propose a plan.

The debtor’s threat to delay a plan is often credible; the debtor wishes to protect bankruptcy proceedings on the chance that the debtor will turn solvent and that shareholders will receive a larger payoff in liquidation or reorganization. These debtor bargaining powers help explain why shareholders typically retain a stake in the reorganized firm even though creditors are nit paid in full.

This threat is damaging to creditors because they usually want the proceedings to end as soon as possible in order to receive principal and interest payments on their claims. In addition, all creditors also face the risk that the estate’s value will decline dramatically during bankruptcy. Secured creditors also face the risk that their security will depreciate during Chapter 11.

Clearly, the decision to accept or reject a plan depends on what happens if the plan is approved, i.e., on the threat points in this game.One threat point of a plan’s sponsor is that the plan will be approved by the court even in the presence of a dissenting class of creditors. The procedure is referred to as cramdown. Section 1129 of the Bankruptcy Code provides for cramdown if a class receives a claim with a value equivalent to full payment or if every class junior to the dissenting class receives nothing.

Creditors also have threats. They can propose a plan of their own which can be crammed down on the equity holders.Perhaps, even more important, secured creditors can move to lift the automatic stay. Creditors can oppose management’s operating and investment decisions. They can refuse to lend new money, and they can oppose asset sales.

The fact that equity retains value in many reorganizations even if creditors are not paid in full can have important implications for behavior outside of bankruptcy. In my model, the firm has only two alternatives: to obtain new funds and invest, or to go bankrupt and liquidate. In practice, however, there’s a third option: to file for Chapter 11 protection, invoke the automatic stay, and maintain control, continuing in operation without new funds for investment. This threat is often both harmful to creditors and perfectly credible: equity value is almost certain to be wiped out in liquidation, while in Chapter 11 equity value is positive if there is any possibility of solvency. Faced with this threat, the creditors’ best alternative may be to extend further funds for investment. Thus, reorganization law provides a distressed firm with a credible threat that increases the creditors’ incentives to provide new funds. In essence, the law affects the bargaining process outside of bankruptcy by changing not just how the surplus is split but the efficiency of outcomes as well.

To develop this idea in more detail, I consider the following simple extension of my model in Section 1. Suppose that if the firm continues in operation without investing, it receives a stochastic date-2 payoff of $X_c$ (with mean $\bar X_c$) in addition to the date-1 liquidation value of $Y$. In order to focus on continuation as a threat rather than a value-maximizing strategy, I assume that continuation is inefficient; total value is higher if the firm liquidates than if it continues without investment, $\bar X_c<0$. the value of the bonds if the firm follows the continuation strategy is $V_D^c$. The bank and the firm together get $\bar X_c+Y{VD}^c$ if the firm continues without investing. If the firm invests, their combined payoff, as before, is $\bar{X}-I+Y-V_D$. Finally, if the firm is liquidated, their combined payoff is $L_B$, with the equity getting nothing. Suppose that among these three alternatives liquidation is the most attractive to the bank and equity combined, so that:

$L_B>max(\bar{X}-I+Y-V_D,\bar X_c+Y-V_D^c)$

the bank will lend money for investment. If not, the firm will file for Chapter 11.

The option to file for Chapter 11 protection can increase efficiency. If (12) is satisfied, the firm will more prone to invest. This is efficient if $V_D-L_B>0$, the case in which the firms underinvests without Chapter 11. If, however, $V_D-L_D<0$, the firm would otherwise overinvest and Chapter 11 merely exacerbates the inefficiency. By contrast, if (12) is violated, Chapter 11 always reduces efficiency since the firm continues rather than liquidates, and $\bar X_c<0.

The overall efficiency effects of this aspect of Chapter 11 are ambiguous, but we can identify the situations in which it is likely to be helpful or harmful. First, when the bonds are short term, te bank debt is senior, and the bonds are protected by seniority covenants, underinvestment is likely to be a problem, and Chapter 11 can be helpful. Second, when investment is risky relative to continuation, investment tends to be more attractive to the bank and the equity because the bonds are worth less. In this case, the likely effect of Chapter 11 is to promote investment rather than to give the easy way of avoiding efficient liquidation.

Another out-of-bankruptcy effect of the maintenance of equity value in Chapter 11is to reduce the incentives to take risk. In a Chapter 7 liquidation, shareholders generally receive nothing, making Chapter 7 very unattractive to shareholders and management. So, as the firm’s financial position gradually deteriorates, management has a strong incentive to take risk-increasing investments and to pay out as much firm value as possible to themselves. This incentive is obviously diminished the higher the return to the equity and management in Chapter 11. Of course, if bondholders are aware of the law, they must be promised a higher interest rate to compensate them for their lower return when the firm is in distress. If the investment decisions of a financially distressed firm are more efficient, there will be more than enough increased value to pay the higher interest rates and yet increase equity value.

4 Conclusion

We’re at the end of the Odyssey.

This article outlined some of the characteristics of corporate financial structure that can make financial distress more or less costly. I focused on coordination problems among bondholders as the main source of inefficiency. This problem can lead to underinvestment when bank debt is senior, when bond debt is short-term, or when it is protected by seniority covenants. Overinvestment tends to be a problem when bank debt is junior, bonds are long-term, and when a firm can strip seniority covenants with exit consents.

Exchange offers can be used to restructure bond debt, but they do not, in general lead to efficient investment. So, financial distress may result in inefficient operating policy even though banks are perfectly informed and bond exchanges are possible. If bonds lack seniority covenants, exchange offers do not change the firm’s investment behavior but simply force bondholders to bear more of the burden of financial distress. If bonds do have seniority covenants, however, investment can be increased through an exchange offer that strips public debt of its covenant and enables a firm to issue senior debt to finance investment. However, such exchange offers can go too far, resulting in overinvestment in some cases. Efficiency is increased if exit consents are not allowed and, instead, bondholders vote voluntarily to eliminate seniority covenants.

The Trust Indenture Act gives rise to investment inefficiencies because it forces firms to conduct exchange offers rather than bargain directly with bondholders. In my model, all investment inefficiencies would be eliminated if the Trust Indenture Act was repealed. Of course, this result follows from my assumption of complete information, in which case bargaining is efficient in the absence of transaction costs. In a more realistic model with asymmetric information and other transaction costs, investment efficiencies are likely to result.

There are a number of practical implications to my model. First, the model predicts that, conditional on an out-of-court workout, distressed firms with senior bank debt, short-term bonds, and effective seniority covenants, will invest less. Second, the model predicts that exchanges are more likely when the bonds are relatively long-term. And, when possible, exchanges should shorten the maturity of existing bond debt, strip existing covenants, and offer more senior securities,

My model is also a useful starting point to think about the tradeoffs firms face in deciding whether to file for Chapter 11 rather than seek an out-of-court workout. I have outlined how the capital structure affects the payoffs from an out-of-court restructuring.

I conclude by noting that, while I have analyzed the effects of Chapter 11 on distressed firms, I have sidestepped an important point made by legal scholars. Roe (1983), Baird (1986), and Jackson (1986) have all argued that the manipulation that is possible in Chapter 11 can be avoided by eliminating Chapter 11 altogether and relying on Chapter 7. To be clear, the above scholars’ argument does not argue that every firm should be liquidated but, whenever possible, sold by a trustee and recapitalized. The reason I did not engage with the legal scholars’ argument is that I disagree with it and that, given that their construct is a nuanced and sophisticated one, it would have taken me an inordinate amount of ink to disprove a point with only negligible importance for the above conclusions.