Jerold B. Warner,Bankruptcy Costs Some Evidence

Bankruptcy Costs: Some EvidenceAuthor(s): Jerold B. Warner

Source: The Journal of Finance, Vol. 32, No. 2, Papers and Proceedings of the Thirty-FifthAnnual Meeting of the American Finance Association, Atlantic City, New Jersey, September16-18, 1976 (May, 1977), pp. 337-347

Published by: Blackwell Publishing for the American Finance AssociationStable URL: http://www.jstor.org/stable/2326766Accessed: 19/02/2009 02:51

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at

http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unlessyou have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and youmay use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained athttp://www.jstor.org/action/showPublisher?publisherCode=black.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printedpage of such transmission.

JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with thescholarly community to preserve their work and the materials they rely upon, and to build a common research platform thatpromotes the discovery and use of these resources. For more information about JSTOR, please contact support@jstor.org.

Blackwell Publishing and American Finance Association are collaborating with JSTOR to digitize, preserveand extend access to The Journal of Finance.

http://www.jstor.org

THE JOURNAL OF FINANCE VOL. XXXII, NO. 2 MAY 1977 SESSION TOPIC: CORPORATE FINANCE-EMPIRICAL SESSION CHAIRPERSON: MARTIN J. GRUBER* TESTS BANKRUPTCY COSTS: SOME EVIDENCE JEROLD B. WARNER** INTRODUCTION AND SUMMARY ASSUMPTIONS ABOUT THE magnitude of bankruptcy costs will have a considerable bearing on the issue of how much debt it is optimal for the firm to have in its capital structure. For example, in the original Modigliani-Miller model (1958), which abstracted from both corporate taxes and the possibility of bankruptcy, no debt/equity ratio could be regarded as optimal. Given perfect markets and rational investor behavior, they showed that the value of the firm would be invariant to its capital structure. Stiglitz (1969) has shown that the invariance result holds even when there is a positive probability of bankruptcy, but only as long as there are no transactions costs associated with bankruptcy.' Relaxing the assumption that bankruptcy is costless and introducing a corporate tax in which interest payments can be deducted from net income reopens the possibility of optimal debt/equity ratios. Kraus and Litzenberger (1973) have developed a formal model dealing with this case, and on a more general level a central theme in textbook discussion of capital structure policy has become the presumed trade-off between tax savings and bankruptcy costs.2 This paper considers some issues surrounding the role of bankruptcy costs in models of capital structure. Evidence on the direct costs of corporate bankruptcy is presented for a number of railroad firms which were in bankruptcy proceedings under Section 77 of the Bankruptcy Act between 1933 and 1955. Elsewhere,3 I have examined the risk and return characteristics of defaulted debt claims of firms in the railroad industry; the railroad firms whose bankruptcy costs are discussed here are a subsample of the firms whose bond returns were the subject of those studies. The ratio of direct bankruptcy costs to the market value of the firm appears to fall as the value of the firm increases. As measured here, the cost of bankruptcy is on average about one percent of the market value of the firm prior to bankruptcy. That result seems in striking contrast to the figure of 20 percent reported in a study * New York University. ** University of Rochester. The author wishes to thank M. Gruber, R. Hamada, F. Jen, M. Jensen, E. H. Kim, E. Kitch, M. Scholes, J. Siegel, C. Smith, B. Stone, H. Stoll, and J. Williams for their comments on previous drafts. I am indebted to N. Gonedes and especially M. Miller for their encouragement and for their criticisms of previous versions of the paper. 1. The work of Arrow (19), Debreu (1959), and especially Hirshleifer (1965) makes it possible to establish the 'irrelevance' proposition within the context of a 'states of the world' model. For a recent treatment, see Fama and Miller (1972) and Milne (1975). 2. See, for example, Van Horne, (1974), p. 267. 3. Warner (1976, 1977). 337 338 The Journal of Finance by Baxter (1967), although it should be kept in mind that Baxter's data referred to personal bankruptcies and to dollar amounts of individuals' assets which are much smaller than those of the railroad sample. While studies by Stanley and Girth (1971) and Van Horne (1976) also report figures similar to Baxter's, they too deal with entities of much smaller dollar size than the railroad firms. Given the evidence on a \"scale\" effect, it is important to realize that the findings presented here are by no means inconsistent with previous studies. In interpreting the results, it must be emphasized that not all bankruptcy costs are measurable, direct costs. Some of the omitted indirect costs may be substantial. Furthermore, it is by no means clear that these findings can be generalized to other industries. The most that can safely be concluded at this point is that the direct costs of bankruptcy, such as legal fees, appear to be lower for large firms than the conventional wisdom suggests. THE COSTS OF BANKRUPTCY The costs of bankruptcy discussed in the literature are of two kinds, direct and indirect. Direct costs include lawyers' and accountants' fees, other professional fees, and the value of the managerial time spent in administering the bankruptcy. Indirect costs include lost sales, lost profits, and possibly the inability of the firm to obtain credit or to issue securities except under especially onerous terms. It is important to distinguish between these two classes of costs. For direct costs of bankruptcy to arise, it is sufficient that there be transactions costs associated with negotiating disputes between claimholders.4 But whether indirect costs arise depends upon the market setting. Suppose, for example, that the real operating characteristics of Ford and General Motors are identical, and that the cars they produce are perfect substitutes. If Ford is bankrupt but General Motors is not, the two firms will still face the same production/investment possibilities. If it is optimal (i.e. value maximizing) for General Motors to manufacture a particular product, it will also be optimal for Ford to do so, even if the latter is bankrupt. If it is optimal for General Motors to stay in the industry, it will also be optimal for Ford to do so. Bankruptcy would seem to be irrelevant under these conditions, and the bankrupt and nonbankrupt firm might each operate in the same way, engaging in identical activities to maximize the wealth of their claimholders. Circumstances can easily be imagined, however, under which the bankrupt firm cannot be treated as if it were \"equivalent\" but not bankrupt. Baxter (1967) and Jensen and Meckling (1976) point out, for example, that a firm's sales and profits 4. If there are costs associated with drawing up all-inclusive contracts, an optimal financing decision may involve issuing debt contracts which are actually ambiguous with respect to the rights of claimholders. Although additional costs are incurred conditional upon bankruptcy, it may be less costly to hire economic agents to negotiate claims in the bankruptcy than to hire them to design contracts to avoid bankruptcy in the first place. Gould (1973) discusses the conditions under which it will be optimal for the parties to a dispute to settle out of court. If there is agreement on the probable outcome of the bankruptcy proceeding and if side payments are permitted, then there would be no incentive for individual claimholders or the firm to take a dispute to court. When we do observe such cases in the courts, they are likely to be the result of disagreements as to the likely outcome. The cost of such disagreements is presumably taken into account when the firm makes it financing decision. Bankruptcy Costs: Some Evidence 339 may decline, and its market value fall, when potential buyers of the product perceive default to be likely. This might be the case because potential users of the product possess imperfect information and use their perception of the firm's financial condition to evaluate its operating characteristics. If so, the fact of bankruptcy is relevant in the sense that it conveys information about the longevity of the firm, and its ability to provide replacement parts or render other types of ongoing support services for the product it sells. Of perhaps more direct relevance are the indirect bankruptcy costs arising from the bankruptcy process itself. The bankruptcy trustee, as an agent of the court, has the authority to operate the firm. It is not clear that this agency relationship gives the trustee any incentive to run the firm efficiently and make decisions which are in fact value-maximizing. Unlike management, the trustee is responsible to the court and not directly to the firm's claimholders. He might not necessarily act in the claimholder's interests. To the extent that a trustee makes non-optimal decisions which would not have been made in his absence, the firm incurs an opportunity loss which can properly be regarded as a cost of bankruptcy. EVIDENCE ON DIRECT BANKRUPTCY COSTS FOR RAILROADS Because the indirect costs of bankruptcy are mainly \"lost opportunities,\" they are inevitably difficult, if not impossible to measure.' For the direct costs, however, some quantitative information is available, and the major concern in the remainder of this paper is an examination and evaluation of this evidence. One salient feature of the bankruptcy process is the extensive use of economic agents by each group of claimholders. Claimholders hire such agents in an effort to maximize the value of their respective claims when the court makes it decision on the terms of the reorganization. The agents include lawyers, accountants, various professional consultants, and \"expert\" witnesses. The bankruptcy trustee and his counsel are compensated by the firm, hence by the claimholders. The direct costs of a bankruptcy proceeding includes compensation which must be paid to the various parties just mentioned. By law, each party who performs services related to the bankruptcy must apply to and receive approval of the court before he can be paid. In a typical case, application will be made by the trustee, his counsel, and the counsel for each class of claimholder. Cases with as many as 100 or 200 separate parties applying for fees are not unknown. Fees applied for and awarded are a matter of public record. It would be exceedingly difficult to document these fees on a case-by-case basis by examining District Court records which are scattered throughout the country. Nor are the data on these costs generally collected in a manner which permits examination of the fees for a large number of companies. However, the Interstate 5. Another possible indirect cost of bankruptcy is the higher compensation that the managers of a highly-levered firm will receive because of the higher probability of unemployment they may face. Evidence from the railroad industry does not support the \"higher turnover\" argument. The study by Warner (1976) reports that in the five-year period following initiation of bankruptcy proceedings, the chief executive officers of the bankrupt firms he examined were replaced with a frequency of 8 percent per year. This compared to a figure of 9 percent per year for the non-bankrupt railroads of his control sample for a similar time period. 340 The Journal of Finance Commerce Commission has in the past collected bankruptcy cost data for a limited number of railroad firms. The data analyzed in this paper are the bankruptcy costs which the ICC reported for 11 bankrupt railroads. The cost data include payments to all parties for legal fees, professional services, trustees' fees, and filing fees. They do not include payments to the managers or employees to reflect the value of the time which they spent in administering the bankruptcy.6 Table 1 indicates the firms in the 11 railroad sample, the number of years each was in bankruptcy, and the bankruptcy costs for each firm as reported by the ICC. On average, the bankruptcies took about 13 years to settle. The most lengthy case, that of the Missouri Pacific Railroad, was in the courts from 1933 to 1955. The average direct cost of the bankruptcies was $1.88 million. TABLE 1 FIRMS IN THE SAMPLE Name Chicago and Northwestern Chicago, Indianapolis, and Louisville Chicago, Milwaukee, St. Paul and Pacific Chicago, Rock Island and Pacific Denver and Rio Grande Western Erie Railroad Minneapolis, St. Paul & Slt. Ste Marie Missouri Pacific New York, New Haven and Hartford St. Louis San Francisco Railway Western Pacific Railroad High Low Mean Median Number of Years in bankruptcy 1 1 14 11 16 13 4 7 23 13 15 10 23 4 12.5 13 Bankruptcy cost, millions of dollars 2.14 0.82 2. 2.00 1.37 2.22 0.95 2.54 2.15 2.34 1.24 2. 0.82 1.88 2.00 One problem in measuring bankruptcy costs is to relate these costs to the size of the firm. It is not appropriate simply to look at the bankruptcy cost as a fraction of the market value of the firm at the time of bankruptcy. What should be measured is the fraction of the market value of the firm which potential bankruptcy costs represented at the time the firm originally made its financing decision.7 This provides the relevant measure of the costs which the firm perceived at the time it for personal were but the data in his study 6. Note that Baxter also examined data on the legal costs, bankruptcy cases totalled about 20 bankruptcy cases. He found that the legal fees in large personal bankruptcy costs, Baxter assets. While he had no data on corporate percent of the individuals' percentage... but are smaller corporate cases the... costs may average a somewhat that \"for concluded far from insignificant.\" For purposes value. their present be discounted to reflect costs should 7. Strictly speaking, the future in to be zero. For a discussion of the problems the discount rate is implicitly assumed of this analysis, bankruptcy costs, see Warner (1974). with expected dealing Costs: Some Evidence Bankruptcy 341 decided on the tradeoff to make between bankruptcy costs and the tax advantage to debt. Given that firms may not be able to costlessly or instantaneously adjust their capital structures (due to regulatory constraints, floatation costs, and the like), it is necessary to look at the market value of the firm prior to bankruptcy, and before bankruptcy has become highly likely. Since it is not clear when this happens, it is appropriate to look at the market value of the firm at a number of points before bankruptcy. As long as bankruptcy is likely to be associated with a fall in the total value of the firm, failure to use such a procedure would result in measures of bankruptcy costs which are consistently biased upward when related to firm size. as the month in which the firm filed a bankruptcy petition. Define month \"O\" The total market value of each debt and equity issue of each firm has been calculated as of the last day of that month and for several other selected months in the 84 months preceding the bankruptcy filing. Table 2 shows the number of securities whose market values were calculated for each firm. All of the traded securities of each of the railroads were used. TABLE 2 SECURITIES USED IN COMPUTING MARKET VALUES OF FIRMS IN THE SAMPLE Name Chicago and Northwestern Chicago, Indianapolis & Louisville Chicago, Milwaukee, St. Paul and Pacific Chicago, Rock Island and Pacific Denver and Rio Grande Western Erie Railroad Minneapolis, St. Paul and Slt. Ste Marie Missouri Pacific New York, New Haven and Hartford St. Louis San Francisco Railway Western Pacific High Low Mean Median Total number of issues Debentures 14 9 15 12 5 10 8 15 12 11 4 15 4 10.5 11 12 7 13 9 4 7 6 13 10 9 2 13 2 8.4 9 Common & Preferred 2 2 2 3 1 3 2 2 2 2 2 Data on the outstanding amounts of each security were compiled from Moody's Transportation Manual. The source for price quotations was the Bank and Quotation Record. Transaction or bid prices were used in the calculations. As shown in Table 2, the calculations were made for an average of 8 securities per railroad. Table 3 uses the computed market values of the firms' debt and equity issues and shows the debt/equity ratio for each firm for various months prior to and including the month of bankruptcy. The debt/equity ratio for, say, month \"- 12\" is the ratio of the total market value of the firms' traded debt to the total value of its traded equity on the last trading day of the 12th month before the month in which a bankruptcy petition was filed. The data indicated that the mean debt/eq- uity ratio for firms in the sample rises sharply, starting at 3.8 in month \" - 84\" and eventually increasing to 24.3 in the month of bankruptcy. This is consistent with 342 The Journal of Finance TABLE 3 IN THE SAMPLE DEBT/EQUITY RATIOS OF FIRMS Name Chicago and Northwestern Chicago, Indianapolis, and Louisville Chicago, Milwaukee, St. Paul & Pacific Chicago, Rock Island and Pacific Denver and Rio Grande Western Erie Railroad Minneapolis, St. Paul and Slt. Ste Marie Missouri Pacific New York, New Haven and Hartford St. Louis San Francisco Railway Western Pacific High Low Mean Median 0 33.9 16.9 14.4 23.9 50.3 4.4 51.0 33.9 3.7 31.7 3.7 50.3 3.7 24.3 23.9 TABLE 4 -12 8.7 20.8 27.7 20.9 29.9 3.9 31.5 11.7 3.5 38.8 2.8 38.8 2.8 18.2 20.8 Month -36 21.2 2.4 15.0 1.7 43.8 5.2 50.8 1.3 2.4 3.3 3.6 50.8 1.3 13.7 3.6 -60 1.9 2.0 3.0 1.4 14.9 5.1 47.8 1.5 0.7 4.7 3.6 47.8 0.7 7.8 3.0 -84 1.3 1.9 2.6 2.0 9.6 1.9 15.3 1.6 0.9 3.9 1.0 15.3 0.9 3.8 2.0 MARKET MILLIONS OF DOLLARS VALUES OF FIRMS, Name 0 Month -60 - 12 -36 % change in value - 84 over observation period -78.3 -74.8 -.8 -67.3 -83.6 -73.5 -88.6 -61.0 -82.5 -85.2 -83.2 -.8 -61.0 -78.9 80.4 155.4 88.6 377.2 370.6 Chicago and Northwestern 12.5 8.7 43.5 50.9 49.7 Chicago, Indianapolis and Louisville Chicago, Milwaukee, St. Paul & Pacific 75.2 140.5 78.5 357.7 737.5 114.7 103.0 450.0 4.0 350.7 Chicago, Rock Island and Pacific Denver and Rio Grande Western 15.4 24.7 26.9 86.0 93.7 57.2 205.9 171.4 90.3 216.3 Erie Railroad 10.4 32.5 25.9 29.3 91.2 Minneapolis, St. Paul & SSM. 76.9 148.1 392.3 293.4 197.6 Missouri Pacific New York, New Haven and Hartford 54.3 97.1 140.8 202.0 309.6 St. Louis San Francisco Railway 39.2 23.9 277.4 378.5 2.2 14. 19.1 21.0 54.4 83.2 Western Pacific High Low Mean 114.7 205.9 450.0 378.5 737.5 10.4 8.7 21.0 29.3 49.7 50.0 87.1 156.0 216.7 251.3 expectations for firms declaring bankruptcy, namely a greater percentage fall in the market value of the firms' equity than of their debt. Table 4 shows the total market value of each firm, assumed to be represented by the total market value of its traded securities. 84 months prior to bankruptcy, the average firm in the sample has a market value of $250 million. In the month of bankruptcy, the average firm has a market value of about $50 million. The value of every firm declines over the observation period. The mean percentage change in the market value of the firm between month \"- 84\" and \"0\" is - 78.9. Bankruptcy Costs: Some Evidence 343 Table 5 displays the ICC reported bankruptcy costs as a percentage of the total value of the firm, Using the market market value of the firm in the 84th month 1 percent prior to bankruptcy, bankruptcy costs are on average of the value of the firm. If instead the market value of the firm 36 months prior to bankruptcy is used, 2.5 percent bankruptcy costs average of the total market value of the firm. Of the total change in the market value of the firm between months \"-84\" and \"0\1.3 bankruptcy costs represent percent. It should be noted that the percentages given in Table 5 are subject to several possible biases. One problem which has been ignored is that of non-traded are privately placed, and their market value is securities. Many short-term liabilities of these that the value of such securities is positive, exclusion unknown. Assuming from the market value computations biases the percentage securities figures up- ward. On the other hand, as discussed certain earlier, bankruptcy costs have not been taken into account. The fact that some components of the costs are not included tends to bias the results downward. TABLE 5 I.C.C. REPORTED BANKRUPTCY COST AS A PERCENTAGE OF MARKET VALUE Name Chicago and Northwestern Chicago, Indianapolis & Louisville Chicago, Milwaukee, St. Paul & Pacific Chicago, Rock Island and Pacific Denver and Rio Grande Western Erie Railroad Minneapolis, St. Paul & SSM. Missouri Pacific New York, New Haven and Hartford St. Louis San Francisco Railway Western Pacific High Low Mean 0 2.7 6.6 3.8 1.7 8.9 3.9 9.1 3.3 3.9 6.0 8.8 -12 1.4 9.4 2.1 1.9 5.5 1.1 2.9 1.7 2.2 9.8 6.5 Month -36 -60 2.4 1.9 3.7 0.4 5.1 1.3 3.7 0.6 1.5 0.8 5.9 5.9 0.4 2.5 0.6 1.6 0.8 0.4 1.6 2.5 3.2 0.8 1.1 0.6 2.3 3.2 0.4 1.4 -84 0.6 1.6 0.4 0.6 1.5 1.0 1.0 1.3 0.6 0.9 1.5 1.6 0.6 1.0 Cost as % of change in value 0.7 2.2 0.4 0.8 1.7 1.4 1.2 2.1 0.8 1.0 1.8 2.1 0.4 1.3 9.1 9.8 1.7 1.1 5.3 4.0 COSTS IN RELATION TO THE MARKET VALUE OF THE BANKRUPTCY FiRM cost of bankruptcy, and the corresponding Figure 1 is a plot of the ICC reported in absolute month \"0\" value of each firm. Bankruptcy costs tended to be higher, for the high market-value railroads than for the low market-value railroads. terms, market For example, the two railroads in the sample with the smallest values, the and Minneapolis, St. Paul and Sault Ste Marie and the Chicago, Indianapolis costs of values of about $10 million and bankruptcy each had market Louisville, and Northwestern and the the Chicago under The two largest $1 million. railroads, values of about $100 million, Chicago, Rock Island and Pacific, each had market and bankruptcy costs of about $2 million. 344 The Journal of Finance 3. X 2. x > o X cn 0~ X>4o 30 60 90 120 OF DOLLARS IN MONTH MIJ.LIONS MAR}ET VALUE \"0O\FIGURE 1. Bankruptcy costs and market value of firms o ,lo% 8_ 2 30 MARKE FIGURE 60 T VALUE, MILLIONS OF DOLLARS 90 120 2. Percentage bankruptcy costs and market value of firms Costs: Some Evidence Bankruptcy 345 While the higher market-value railroads generally did incur higher bankruptcy costs, the costs do not appear to be directly proportional to market value. Figure 2 shows the cost of bankruptcy, expressed as a percentage of the month \"0\" value, for each of the 11 firms. The percentage appears to decline for railroads with relatively high market values. The two smallest railroads, in terms of market values, had bankruptcy costs which were 9.1 and 6.6 percent of their values, while the two largest railroads had bankruptcy costs which represented 2.7 and 1.7 percent of their respective values.8 This evidence suggests that there are substantial fixed costs associated with the railroad bankruptcy process, and hence economies of scale with respect to bank- ruptcy costs. Moreover, when the relationship between the value of the firm and the length of time spent in bankruptcy proceedings was investigated, it was found that the two were uncorrelated. Thus it does not appear that high-market value firms incurred bankruptcy costs simply because they tended to be in bankruptcy proceedings for a longer period of time than low market value firms. THE EXPECTED COSTS OF BANKRUPTCY From the standpoint of a firm choosing its capital structure, it is the expected cost of bankruptcy that is the relevant measure of bankruptcy costs. These expected costs of bankruptcy cannot be inferred directly from the data presented, since the ex ante probability of going bankrupt is not known. However, what these results would imply about expected bankruptcy costs can be shown under different assumptions about the likelihood of bankruptcy. Suppose, for example, that a given railroad picks a level of debt such that bankruptcy would occur on average once every 20 years (i.e. the probability of going bankrupt is 5 percent in any given year). Assume that when bankruptcy occurs, the firm would pay a lump sum penalty equal to 3 percent of its now current market value. If anything, these numbers tend to overstate the frequency and apparent direct cost of bankruptcy. Given this background, the firm's expected cost of bankruptcy is equal to fifteen one-hundreths of one percent of its now current market value. If the cost of bankruptcy is doubled to 6 percent and the probability of bankruptcy increased to 10 percent, the expected costs would still be only six-tenths of one percent of the current value of the firm. These numbers are small indeed. This is not to say that they are small enough to be neglected completely in discussions of capital structure policy. But it would not seem unreasonable to conclude that for firms of the size under consideration, the expected direct costs of bankruptcy are unambiguously lower than the tax savings on debt to be expected at present tax rates in standard valuation models.9 8. The squared value of the correlation coefficient between percentage bankruptcy costs and market value is .88. At the .01 level of significance, the hypothesis that the two are uncorrelated must be rejected. 9. See Miller and Modigliani (1963). The analysis here assumes that the relative price of going bankrupt has not changed since the sample bankruptcies took place. This assumption may not be realistic if the price of legal services has increased in relative terms, or if the complexity of the legal system and of the bankruptcy process has increased. The analysis here also ignores the question of 346 The Journal of Finance APPLICABILITY TO OTHER INDUSTRIES The extent to which the results are applicable to firms other than those in the railroad industry is uncertain, since the magnitude of the costs and of the \"scale effect\" may be industry-specific. One reason that the figures could differ for nonrailroad firms is that the Interstate Commerce Commission plays an active role only in railroad bankruptcies. Some of the railroads' bankruptcy costs may in effect be subsidized. The subsidy exists to the extent that the ICC performs functions which in its absence would have been performed by the railroads or their claimholders. The resources expended by the ICC are not reflected in the cost figures. From one standpoint, this is unimportant. As long as the bankruptcy costs which the firm itself must bear are known, the fact that the government absorbs certain other costs does not influence the firm's financing decision. The only bankruptcy cost which the firm perceives is the cost which is incident upon the firm and its claimholders. However, if these data are to be used to infer that bankruptcy costs in other industries are similar, and exhibit a similar scale effect, additional assumptions must be made. It must be assumed that the size of the subsidy relative to total bankruptcy costs does not differ across industries, and that total bank- ruptcy costs are the same for railroad and nonrailroad firms, assuming equivalent size. There is no obvious factual basis for making such assumptions. While non- railroad firms receive no subsidy from the ICC, they do receive a similar type of subsidy from the Securities and Exchange Commission. It is not clear whether this subsidy differs from the one which the ICC offers, or whether railroad bankrupt- cies are any more (or less) expensive than bankruptcies for similar-sized non- railroad firms. Moreover, the differences between railroads and other firms may well be even more important in the case of indirect costs. Restrictions on entry by competitors, owing to ICC regulation, could serve to keep opportunity losses lower for bankrupt railroads than might otherwise be the case. For these reasons, the findings must be regarded as merely suggestive of bankruptcy costs in other industries. REFERENCES Kenneth Arrow. \"The Role of Securities in the Optimal Allocation of Risk-Bearing,\" Review of Economic Studies, 31 (April, 19), 91-96. Bank and Quotation Record. 1926-1955, William B. Dana Company, New York. Nevins D. Baxter. \"Leverage, Risk of Ruin, and the Cost of Capital,\" Journal of Finance, 22 (September, 1967): 395-404. Gerard Debreu. The Theory of Value, New York: John Wiley & Sons, Inc., 1959. of Finance, New York: Holt, Rinehart and Winston, 1972. Eugene Fama and Merton Miller. The Theory of Legal Studies, 2 (June, 1973), 279-300. of Legal Conflicts,\" Journal John Gould. \"The Economics Jack Hirshleifer. \"Investment Decisions Under Uncertainty: Choice Theoretic Approaches,\" Quarterly Journal of Economics, 79 (November, 1965), 509-536. Michael C. Jensen and William H. Meckling. \"Theory of the Firm: Managerial Behavior, Agency Costs, Journal of Financial Economics, 3 (October, 1976). and Capital Structure,\" whether or not at the margin (and not just on average) the direct bankruptcy costs offset the tax advantage. Bankruptcy Costs: Some Evidence 347 Alan Kraus and Robert Litzenberger. \"A State Preference Model of Optimal Financial Leverage,\" Journal of Finance, 28 (September, 1973), 911-922. Frank Milne. \"Choice Over Asset Economies: Default Risk and Corporate Leverage,\" Journal of Financial Economics, 2 (June, 1975), 165-185. Merton Miller and Franco Modigliani. \"The Cost of Capital, Corporation Finance, and the Theory of Investment,\" American Economic Review, 48 (June, 1958), 261-297. and . \"Corporate Income Taxes and the Cost of Capital: A Correction,\" American Economic Review, 53 (June, 1963), 433-443. Moody's Railroad Manual. 1925-1970, Moody's Investors Service, New York. James Scott. \"A Theory of Optimal Capital Structure,\" Bell Journal of Economics and Management Science, 7 (Spring, 1976), 33-54. David T. Stanley and Marjorie Girth. Bankruptcy: Problem, Process, Reform, Washington, D.C.: The Brookings Institution, 1971. Joseph Stiglitz. \"A Reexamination of the Modigliani-Miller Theorem,\" American Economic Review, 59 (September, 1969), 784-793. United States Congress, House. Hearings, Committee on Interstate Commerce, H.R. 2298, 80th Congress, 1947. United States Government. Bankruptcy Laws of the United States, Washington, D.C.: U.S. Government Printing Office, 1972. James Van Horne. Fundamentals of Financial Management, Englewood Cliffs, New Jersey: Prentice- Hall, 1974. . \"Corporate Bankruptcy and Liquidity Costs,\" Research Paper No. 205, Stanford Graduate School of Business, 1976. Jerold B. Warner. \"Corporate Bankruptcy,\" unpublished manuscript, University of Chicago, 1974. . \"Bankruptcy Absolute Priority, and the Pricing of Risky Debt Claims,\" Journal of Financial Economics, forthcoming. . \"Bankruptcy Costs, Absolute Priority, and the Pricing of Risky Debt Claims,\" unpublished Ph.D. thesis, University of Chicago, 1976.