Assessing Post-Bankruptcy Performance: An Analysis of Reorganized Firms' Cash Flows.
In this paper, we directly examine the post-bankruptcy performance of 89 firms by evaluating the total cash flows produced by the firm's assets for the five years after emerging from bankruptcy. In essence, we evaluate the rate of return received by investors who owned all the debt and equity claims on the firm as it emerged from bankruptcy. By examining cash flows rather than accounting data, and by allowing debt to be a risky claim, our approach produces an alternative view: firms neither underperform nor overperform following a Chapter 11 reorganization.
Debt and equity holders of a bankrupt firm have two choices: either let the firm reorganize and accept new claims on the reorganized firm, or vote to liquidate the firm. A logical benchmark against which to assess post-bankruptcy performance is, therefore, the return that could have been earned by liquidating the firm and placing the proceeds in an alternative investment.(2)
It is not easy to observe the value of a newly reorganized firm's assets. Assessing liquidation values is difficult because a complete asset sale may entail significant liquidation costs. This can occur if the company's assets are traded in thin secondary markets, either because the assets are highly specialized (Myers, 1977) or because the entire industry is distressed and therefore liquidity-constrained (Shleifer and Vishny, 1992).(3) What's more, the market value of a newly reorganized firm's assets is rarely observable because most debt claims are not publicly traded.
We thus use two different estimates of firm value. The first is the estimated going-concern (market) value of the firm's assets at the date of emergence from bankruptcy, and the second is the estimated liquidation value of the assets, both as presented in the firm's bankruptcy plan. Liquidation values in bankruptcy plans may be understated, and going-concern values may be overstated, since a firm has an incentive to try to convince creditors that the firm is worth more alive than dead. The going-concern value of assets may also overstate the true value of the firm, since it assumes that a complete asset sale would incur no liquidation costs whatsoever. We can therefore treat these figures as upper and lower bounds on the firm's true value, which generate lower and upper bounds on the true post-bankruptcy performance of the sample firms. Although our methods differ, our approach - to evaluate the total cash flows produced by the firm - is similar in spirit to Kaplan's (1989, 1994) analyses of the Federated bankruptcy, and to Andrade and Kaplan's (1998) analysis of 31 bankrupt LBOs.
If we assume that assets can be sold for their estimated going-concern value, a typical firm's return matches the return to a benchmark portfolio. If we assume that assets can be sold for the liquidation value estimated in the firm's bankruptcy plan, nearly 80% of the sample firms outperformed their benchmark. These figures provide lower and upper bounds for the true post-bankruptcy performance of the firms.
Most prior research has inferred that post-bankruptcy asset performance is poor because operating margins are low, debt levels are high, and second debt restructurings occur frequently. Our results, based on an analysis of the cash flows returned to the debt and equity holders of the sample firms, do not support this conclusion; rather, we conclude that firms neither under- nor overperform following bankruptcy. Robustness checks indicate that this finding is not sensitive to different methods of assessing firm value, or to various definitions of benchmark portfolios for the sample firms.
Cross-sectional results indicate that firms that avoid second restructurings and those that are acquired after emerging exhibit superior post-bankruptcy returns. We also find that post-bankruptcy performance is affected by the firms' investment-opportunity sets. High-growth-option firms generate superior returns when their net investment exceeds the industry median. In contrast, low-growth-option firms do not produce superior returns by investing more heavily in their industry.
The remainder of the paper is organized as follows. In Section I, we describe our data sources and valuation methods. In Section II, we present summary data for our performance measure and explore the sensitivity of our results to variations in our valuation methods. Section III presents a cross-sectional analysis of post-bankruptcy wealth creation, and Section IV concludes.
I. Data and Methods
In this section, we describe our sample selection procedure and demonstrate our valuation method with a case study of Resorts International.
A. Sample Selection
The study begins with a list of 201 firms that completed Chapter 11 reorganizations between 1983 and 1993. The list was compiled from the samples used by Altman and Nammacher (1985), Betker (1995), Eberhart, Moore, and Roenfeldt (1990), Franks and Torous (1989), Gilson, John, and Lang (1990), Lang and Stulz (1992), and LoPucki and Whitford (1990), as well as from the Bankruptcy DataSource. We obtained copies of the reorganization disclosure statement for 128 of these firms, either from the firms, their attorneys, or exhibits to 8K or 10K filings.(4)
To emerge from bankruptcy, a firm must show that its reorganization plan is in the best interests of all claimants. Each creditor class must receive at least as much value in reorganization as it would under the absolute priority rule in Chapter 7 liquidation (Bankruptcy Code, Section 1129(a)). Thus, firms must estimate what their assets would yield in a liquidation. We dropped 22 firms because they did not provide sufficient information to determine liquidation value.
Of the 106 remaining firms, 11 were excluded from the sample because they are privately held and were unwilling to provide post-reorganization financial statements. Data on six public firms could not be located, leaving 89 firms suitable for analysis.
B. Sample Characteristics
Table 1 contains summary data on the 89 firms in the sample. The table reports summary statistics for post-reorganization firm size and debt ratios, and the frequency distribution of the year firms in the sample emerged from Chapter 11.
The average sample firm has post-bankruptcy book value assets of $371 million, with a median of $167 million. The mean (median) liquidation value of assets is estimated at $278 million ($134 million). Similar to other samples of reorganized firms (e.g. Gilson, 1997), the sample firms carry a higher-than-average debt load following reorganization. The median ratio of long-term debt to total capitalization is 63.0%. The typical firm emerges from bankruptcy with a long-term debt to total capitalization ratio that is 31.9% higher than the industry median ratio.
The sample firms emerged from bankruptcy over the period 1983-1993; 62 of the 89 emerged from Chapter 11 between 1990 and 1993. The sample firms have 38 different two-digit SIC codes. Of the firms, 12 are in the oil and gas industry, eight are in the commercial machinery and computer industry, and five are in the communications industry. No other two-digit SIC code contains more than four firms. Twenty-one firms (or 24%) engaged in highly leveraged transactions prior to filing for Chapter 11 reorganization. Of the sample firms, 34 (or 38%) reorganized in a prepackaged bankruptcy filing, and 25 (28%) were acquired within five years of emerging from Chapter 11. Twenty-one (24%) completed second debt restructurings, either in the form of prepackaged bankruptcies, workouts, or traditional Chapter 11s, after emerging from bankruptcy for the first time. Three firms (3%) liquidated within five years after emerging from bankruptcy. The "failure rate" of reorganizations in this sample is thus slightly lower than the frequency of second reorganizations in LoPucki and Whitford (1993) (32%), Gilson (1997) (33%), and Hotchkiss (1995) (32%).
As in Hotchkiss (1995), we evaluate the operating performance of the sample firms for the first three years following emergence from bankruptcy, using the ratio of operating income to sales. We define operating income as earnings before interest, taxes, depreciation, and amortization (EBITDA). The year of emergence from bankruptcy (year 0) is defined as the fiscal year containing the date of plan confirmation. Years 1 through 3 are the first, second, and third post-reorganization fiscal years. Table 2 reports mean and median industry-adjusted figures that summarize the difference between the operating margin of an individual firm and the concurrent median operating margin of firms in the same two-digit SIC code.
During each of the three fiscal years following emergence from bankruptcy reorganization, at least 60% of the firms with available information reported operating profit margins that fell below the median level for the industry. Thus, our sample is similar to the firms examined by Hotchkiss (1995) and Hotchkiss and Mooradian (1997), in that the majority of firms exhibit poor accounting performance.
For example, during the first post-reorganization year, both measures of industry-adjusted operating margin were negative and different from zero at the 0.05 significance level. The median industry-adjusted ratio of EBITDA/Sales during Year 1 for the sample was -3.35%, with a mean value of -4.77%. Of the 84 sample firms with available information, 69% had EBITDA/Sales ratios that fell below the industry median. Operating performance for the second and third post-reorganization years is similar in magnitude and statistical significance.
C. Demonstration of Valuation Method: A Case Study of Resorts International
We illustrate our method of valuing the reorganized firms' cash flows by examining in detail the case of Resorts International. Data for this case are contained in Table 3. The firm emerged from bankruptcy on September 17, 1990. Data on cash flows produced by reorganized Resorts International were obtained from the firm's cash flow statements and from supplemental disclosures of cash flow (e.g., cash interest paid during the year) in the footnotes to the financial statements. Net cash flows to all claimholders are defined for each year (or part of a year) as:
Net cash flows from operations + Net cash flows from investment + Cash interest paid - Change in cash - Other cash flows from financing = Net cash flows paid to all claimholders (1)
"Other cash flows from financing" includes items such as investment banking and professionals' fees that are not returns to the debt or equity holders of the firm.(5) We add back cash interest paid since this figure [TABULAR DATA FOR TABLE 1 OMITTED] is deducted from "net cash flows from operations," yet it is a distribution to the firm's claim holders. We subtract the change in cash each year since additions to the firm's cash balance are not cash flows paid directly to claimholders. We also note whether the firm raised new capital or paid out cash to claimholders when it emerged from bankruptcy.
Table 2. Industry-Adjusted Operating Performance in the Three Years Following Emergence from Bankruptcy The table shows summary statistics for EBITDA/Sales minus the industry median ratio. Industry data are compiled at the two-digit SIC code level. Year 0 is the fiscal year containing the date of confirmation of the firm's reorganization plan, so Year 1 is the first full fiscal year after emerging from bankruptcy. EBITDA/Sales Year 1 Year 2 Year 3 Mean -0.0477 -0.0560 -0.0577 Median -0.0335 -0.0338 -0.0258 t for Mean -3.40(***) -3.39(***) -2.40(**) Number of Firms 84 79 67 Percent Negative 69.0% 64.6% 62.9% *** Significant at the 0.01 level. ** Significant at the 0.05 level.
1. Post-Bankruptcy Cash Flows
Panel A of Table 3 reports cash flow data for Resorts International. The firm raised $11 million in new equity capital when it emerged from bankruptcy, and we account for this as a negative cash flow to claimholders at that date.
Table 3 also indicates that Resorts experienced negative "other" financing cash flows in the first four periods after emerging from bankruptcy. These represent outlays for professionals' fees. In the year ending on December 31, 1994, there was a $60 million return to bondholders that appeared in Resorts' cash flow statement. Resorts filed for bankruptcy again in 1994, and $60 million represents the estimated fair market value of property distributed to bondholders in the firm's second bankruptcy case.
Finally, Panel A reports Resorts' industry-adjusted EBITDA/sales. Like most firms in the sample, Resorts' accounting profitability was sub-par compared to the typical firm in its industry.
2. Terminal Value of the Firm
Panel B of Table 3 reports our estimate of the "terminal value" of Resorts International at December 31, 1995. [TABULAR DATA FOR TABLE 3 OMITTED] We obtain the market value of common and preferred stock (if any) from CRSP or the Standard & Poor's Stock Guide. If these sources do not list a price, we use the average of the high anti low price for the last quarter of the firm's fiscal year as reported in its 10-K. If we are unable to find a stock price from any source, we assume the stock price is zero (this occurred in two cases).
Prices of public debt are obtained from the Standard & Poor's Bond Guide. Many firms also report the fair market value of their debt instruments as required by Statement of Accounting Standards 107, "Disclosures about Fair Value of Financial Instruments." If debt prices are available in both S&P and the firm's 10-K, we take the lower of these two reported values. If neither source reports debt values, we try to obtain the firm's debt rating and value the debt by discounting the promised cash flows at the yield on similarly-rated debt as reported in the S&P Bond Guide (this occurred in four cases). Finally, if we cannot obtain the firm's debt rating, we discount the promised cash flows at the yield on Merrill Lynch's high-yield debt index as reported daily in the Wall Street Journal (this occurred in two cases). The goal of this valuation protocol is to obtain a conservative estimate of the value of the firm's debt. Finally, we value nontraded debt (bank and trade debt) at its book value. We investigate the importance of this assumption in a later section.
We were able to obtain market prices for $212.5 million face value of Resorts' debt as well as for the firm's common stock. Panel B indicates that the market value of Resorts' assets as of December 31, 1995 was about $382 million.
3. How Did Resorts International Perform?
Panel C of Table 3 summarizes the results regarding whether the claimholders of Resorts were better off investing in the reorganized company or an alternative investment. This question is evaluated by examining whether Resorts International should have liquidated or continued operating following its first reorganization. We take the cash flows returned to claimholders each year and compound them out to December 31, 1995 by assuming that the cash flows were invested in the S&P 500 index. (We later report that our results are not sensitive to the choice of benchmark portfolio.) Additional cash contributed by claimholders is compounded at the rate of return on the S&P 500 to reflect the opportunity cost of the additional investment in the firm.(6) The total cash flow to the claimholders at December 31, 1995 is equal to the sum of the net cash flow to the claimholders from Panel A and the market value of the firm at that date from Panel B.
We compare the annualized return on the reorganized company with the annualized return on the Standard & Poor's 500 index. The realized return on the reorganized company relates wealth at the terminal date to the value of the firm at the date it emerged from bankruptcy. We employ two alternative assumptions with respect to firm value at the reorganization date. The first method assumes that firm value equals the estimated going-concern value of the company at the reorganization date, in effect assuming that there are no rents available to a buyer since the firm can liquidate at a zero discount to its going-concern value. In contrast, the second method assumes that the firm value equals the estimated liquidation value of the company. This approach recognizes that a complete asset sale may generate rents for the buyer, due to asset illiquidity or a lack of competing bidders (Schleifer and Vishny, 1992).
Going-concern values were estimated in one of two ways. Sixty-six firms reported the market value of the firm's assets and liabilities as estimated by investment bankers or valuation consultants, who use standard techniques such as discounted cash flows, multiples of cash flow, and comparable firm valuations. Firms produce these going-concern value estimates in order to facilitate comparison with their liquidation value, and/or because they adopted fresh-start or quasi-reorganization accounting procedures that require the restatement of balance sheet accounts to their fair-market value,v For nine firms, all of the claims on the reorganized firm were publicly traded, and we used the market value of these claims just after emergence from bankruptcy as an estimate of firm value. Results in the paper do not differ if we exclude these nine firms from the sample. Fourteen firms did not include estimates of going-concern value in their bankruptcy disclosure statements, and so we omit these firms in our first approach.
The valuation exercise indicates that by reorganizing, Resorts had created wealth of about $638 million at December 31, 1995. In contrast, the going-concern value of Resorts' assets at September 17, 1990 was estimated at $510,040. This value can be interpreted as the value that the assets would fetch in liquidation in the absence of agency and illiquidity costs.
Had Resorts been able to liquidate costlessly (i.e., for its full going-concern value), then the decision to continue meant forgoing $510,040 at that date. Since total wealth created by December 31, 1995 totaled only $637,885, Resorts' performance by this standard was poor.
We standardize these figures for cross-sectional comparisons by stating them as annualized rates of return. Forgoing $510,040 in September of 1990 in order to obtain $637,885 in December of 1995 represents a 4.35% annualized return: [(637,885/510,040).sup.(12/63)] = 1.0435, where 63 is the number of months between emergence and December 31, 1995. Similarly, the annualized return on the S&P 500 from 9/90 to 12/95 was 14.25%. We thus define the "excess return on assets relative to going-concern value" for Resorts International's decision to continue as 4.35% - 14.25% = -9.90%.
We also compute the return to continuation using Resorts' estimated liquidation value at September 17, 1990 as the denominator. The liquidation value of Resorts' assets at September 17, 1990 was estimated at $310 million. This figure includes a "fire-sale" discount that would likely occur in total-asset liquidation (Schleifer and Vishny, 1992). However, it is also likely that firms have an incentive to understate the liquidation value of their assets in order to convince creditors that the firm is worth keeping alive,(8) or that the liquidation value represents a worst-case scenario that is unlikely to be realized. We therefore interpret the firm's performance relative to its liquidation value as an upper bound on true post-bankruptcy performance.
Had Resorts in fact liquidated, and its claimholders placed $310 million in the S&P 500 at the reorganization date (September 17, 1990), that investment would have grown to about $624 million by December 31, 1995. Thus, the wealth created by continuation exceeded the wealth created by the S&P 500 Index despite the fact that Resorts International entered bankruptcy for a second time in 1994.
Forgoing $310,000 in September of 1990 in order to obtain $637,885 in December of 1995 represents a 14.73% annualized return: [(637,885/310,000).sup.(12/63)] = 1.1473. Similarly, the annualized return on the S&P 500 from 9/90 to 12/95 was 14.25%. We thus define the "excess return on assets relative to liquidation value" for Resorts International's decision to continue as 14.73% - 14.25% = 0.48%.
II. Results for the Full Sample
Next, we present the results of our valuation exercise for the entire sample of 89 firms.
A. Analysis of Cash Flows
We examine the statement of cash flows for each of the sample firms for up to five years following the year in which the plan of reorganization was confirmed. The terminal date of the valuation exercise is the end of the fifth full fiscal year after emerging from bankruptcy. We collect the net cash flows to all claimholders, as defined in Equation (1), during the year. The accumulated value of the reorganized firm is then computed at the terminal date. Total accumulated value is equal to the sum of the intermediate net cash flow to claimholders, reinvested in the S&P 500 index at the end of each year, plus the market value of the company's securities at the terminal date. If a firm is liquidated or acquired prior to the terminal date (recall that this happened to three and 25 firms, respectively), we assume that the liquidation or acquisition proceeds are invested in the benchmark portfolio until the terminal date.
The realized return relative to going-concern value for a firm is found by annualizing the ratio of the company's reorganized terminal value to its estimated going-concern value. Similarly, the realized return relative to liquidation value for a firm is found by annualizing the ratio of the company's reorganized terminal value to its estimated liquidation value. Excess returns are computed as the difference between the annualized return to continuation and the annualized return on the S&P 500 index over the same period.(9)
Table 4 reports the summary staff sties for the annual excess return to continuation for the sample of reorganized companies. The first column of Table 4 presents summary statistics for excess returns computed relative to going-concern value. Fourteen firms did not provide estimates of going-concern value, leaving 75 firms with usable data. Of the firms, 52% outperformed the Standard & Poor's index under this method. The mean (median) firm earned a return on assets that was about 4.6% (0.3%) greater than the return available from costlessly liquidating and placing the proceeds in the S&P 500. Neither the mean nor the median abnormal returns are significantly different from zero at the 0.05 level of significance. We therefore conclude that the companies in our sample matched the performance of the S&P 500 Index following reorganization - a sensible result in an efficient market.
For the sample of 89 firms with available information, the annualized return computed on liquidation value [TABULAR DATA FOR TABLE 4 OMITTED] exceeds the return on the market index by an average of 27.1%. The median excess return is 13.7%. Both the mean and median returns are significantly greater than zero at the 0.01 level. The excess return is positive for 66 sample firms, indicating that the realized return following reorganization exceeded the return available on the S&P 500 for nearly 75% of the sample.
As noted in the discussion of Resorts International's case, it is likely that firms underestimate the liquidation value of their assets. This may occur as managers of inefficient firms try to convince creditors that continuation is the preferred option. However, even if we assume that liquidation is costless (i.e., the firm can liquidate for its full going-concern value), we reach the conclusion that firms did no worse by operating rather than liquidating following reorganization.
Like Hotchkiss (1995), our analysis cannot address the fundamental capital budgeting issue that firms contemplating reorganization must answer, which is whether reorganization creates more wealth than liquidating at the onset of financial distress. Answering that question would require an accurate estimate of the firm's liquidation value at the onset of financial distress (an estimate that firms do not provide), as well as data on total cash flows produced during bankruptcy (which are often unavailable).
However, these two sets of results do provide lower and upper bounds on the post-bankruptcy performance of the sample firms. Since assuming costless liquidation is likely to overstate the proceeds that would have been available in liquidation, we can treat the going-concern-value-based estimates as lower bounds on the firms' true performance. If bankruptcy plans tend to underestimate liquidation proceeds, or if complete asset sales are infeasible because they would generate excessive rents for potential buyers, then the liquidation-value-based estimates provide upper bounds on post-bankruptcy performance.
B. Robustness Checks
In this section, we explore the robustness of our results to different methods of assessing the firm's terminal value and different benchmark portfolios. We also assess the unbiasedness of the going-concern and liquidation value estimates used in the study.
1. Assessing the Validity of the Liquidation Value and Going-Concern Value Estimates
As noted earlier, we have no way of directly assessing the validity of firms' liquidation value estimates. If firms understate their liquidation values, the liquidation-value-based excess returns will be overstated. Similarly, firms can "sell" the plan to creditors by overstating their going-concern value, thereby understating the excess returns computed under the assumption of costless liquidation.
We can, however, directly assess the validity of firms' going-concern value estimates. When a firm estimates the market value of its assets and liabilities for fresh-start accounting purposes, by implication it also forecasts its post-bankruptcy market value of equity (MVE). We can then compare each firm's actual post-bankruptcy MVE with the firm's estimate. Of the 75 firms with valid going-concern estimates, 60 had observable stock prices immediately after emerging from bankruptcy. We define a firm's estimation error as (estimated MVE - actual MVE)/actual MVE. The mean (median) estimation error was 0.139 (0.000). Neither the mean nor median error is significantly different from zero at conventional levels. We conclude that the going-concern value estimates are noisy but unbiased. Most importantly, the estimation errors are uncorrelated with both measures of excess returns to continuation. The correlation between the estimation errors and the going-concern-value-based (liquidation-value-based) excess returns is 0.12 (0.09). Neither figure is significant at the 0.10 level. Thus, firms that overestimate their market value do not systematically perform better or worse than firms that underestimate their market value.
We can also provide some indirect evidence on the validity of the liquidation value estimates. If management has an incentive to underestimate liquidation values when the firm's prospects are poor, then companies that perform poorly relative to their (unbiased) going-concern value estimates should exhibit strong performance relative to their (understated) liquidation values. We would therefore expect to find a negative correlation between the two excess return measures. In fact, the correlation between the two excess returns is 0.509 (significant at the 0.01 level). This result is not consistent with a deliberate understatement of liquidation values by firms with poor post-bankruptcy prospects.
If some firms underestimate liquidation values, we would also expect to find that the excess return assuming costless liquidation is negatively related to the gap between going-concern value and liquidation value. We define a firm's liquidation cost as (going-concern value - liquidation value)/going-concern value. Firms with artificially high liquidation cost estimates due to a deliberate underestimation of liquidation value should perform poorly relative to their unbiased going-concern value estimate. Instead, we find that the correlation between liquidation costs and the excess return assuming costless liquidation is -0.001 (not significant at conventional levels). These findings do not suggest a bias in the liquidation value estimates.
2. Computation of Terminal Values
The results in Table 4 are not sensitive to assumptions about the value of the firm's non-traded (bank and trade) debt. We conservatively estimated the value of non-traded debt to be the lesser of 90% of its book value or the value of the firm's traded debt. For example, if a firm's bonds traded at 75, we assume that bank and trade debt is worth 75 as well, while if bonds trade at 100, we assume that non-traded debt is worth 90. Under these assumptions, the mean excess return (relative to liquidation costs) is 0.2093, and 64 of the 89 firms (71.9%) have positive excess return measures. Cross-sectional inferences reported in Section III.B are similarly unaffected.
3. Risk-Adjusted Discount Rates
The excess returns in Table 4 are computed using cash flows to all of the firm's claimholders, not just the cash flow to equity. Since we are comparing the return on a firm's assets to the return on the S&P 500, we are implicitly assuming that each firm has an asset beta of 1.0. Thus, the comparison between the returns to the reorganized company and the Standard & Poor's 500 should be a conservative one - we are requiring the return on a firm's assets to exceed the return on levered equity.
To test the sensitivity of our results to this assumption, we follow Kaplan and Ruback (1995) and estimate the asset beta for each firm's industry in the year that the firm emerged from bankruptcy.(10) We estimate the value-weighted equity beta ([[Beta].sub.e]) for the industry (defined at the two-digit SIC code level) using two years of daily stock returns, returns on the S&P 500, and a Scholes-Williams (1977) correction. We unlever the industry equity beta using the value-weighted ratios of equity, preferred, and debt to total capital for firms in the industry. Industry ratios are calculated using COMPUSTAT data.
Equity betas are unlevered using the formula:
[[Beta].sub.u] = [[[Beta].sub.3] E + [[Beta].sub.p] P + [[Beta].sub.d] D (1 - [Tau])]/E + P + D(1 - [Tau])] (2)
where E is the market value of equity, P is the book value of preferred, and D is the book value of debt. As in Kaplan and Ruback (1995), we assume that the betas of preferred ([[Beta].sub.p]) and debt ([[Beta].sub.d]) equal 0.125, and set [Tau] equal to the combined federal and state marginal tax rate during the estimation period.
The mean (median) asset beta in the year of emergence from bankruptcy is 0.70 (0.67). The largest asset beta is 1.21; however, only four of the 89 sample firms (4.5%) were in an industry whose asset beta exceeded 1.0 in the year the firm emerged from bankruptcy. We conclude that assuming an asset beta of 1.0 for all firms biases against finding that the reorganized firm outperformed the S&P 500. Placing liquidation proceeds in an investment less risky than the S&P 500 would make liquidation at time zero a less attractive alternative than continuation, because the return to liquidation would be lower.
4. Alternative Benchmarks
We computed our performance measure using benchmarks other than the S&P 500. These included a) the CRSP value-weighted index of all NYSE, AMEX, and NASDAQ firms, b) value-weighted portfolios consisting of all firms in the same two-digit SIC code as the sample firm, and c) an equal weighted portfolio consisting of all firms in the same size-based portfolio as the sample firm. Size-based portfolios were constructed using the methods of Lyon, Barber, and Tsai (1999).(11) In each case, we assumed that intermediate cash flows produced by the firm were reinvested in the appropriate benchmark portfolio. We then compared the terminal value of the firm to the wealth that would have been created had the firm been sold at time 0 and the proceeds invested in the benchmark portfolio.
Inferences from Table 4 are not affected by the use of alternative benchmarks. The mean excess return (relative to going-concern value) ranges from 0.0121 to 0.0491, and these figures are not significantly different from zero at conventional levels. The mean excess return (relative to liquidation value) ranges from 0.2376 to 0.2750, significant at the 0.01 level.
5. Firms Without Post-Bankruptcy Cash Flow Data
Finally, the results in Table 4 may be affected by a selection bias that favors successful companies. As noted earlier, eleven firms were privately held after bankruptcy. While all eleven firms are still in existence at this writing, if private firms systematically create less wealth than public firms, this will bias performance measures upwards. In addition, there are six firms with missing post-bankruptcy cash flow data that liquidated after emerging from bankruptcy. It seems likely that the post-bankruptcy performance of these firms would be poor.
III. Cross-Sectional Evidence
The full sample analysis shows that the post-bankruptcy asset performance of our sample equaled or exceeded the performance of the Standard & Poor's 500, depending on the level of liquidation costs assumed. In this section we examine the sources of cross-sectional variation in our return measure. We have two goals in this analysis. First, to ensure that our results are not driven by the fact that we examine only firms that survived bankruptcy, we search for a relation with the factors thought to contribute to performance, such as industry sales growth, size, and investment opportunities. Second, we wish to examine (and control for) the relation between our measure of excess returns and the variables that have previously been used to characterize post-bankruptcy performance: accounting performance measures, post-reorganization debt levels, management changes, and subsequent financial restructurings. We do this by developing a regression model that controls for the possible sources of cross-sectional variation in our performance measure. We first discuss the variables used to explain variation in our performance measure and their expected relationship with post-bankruptcy returns.
EBITDA to Sales: we test whether a relationship exists between operating cash flow and the excess return to reorganization. As in Hotchkiss (1995), we assign a firm to the group "adjusted EBITDA<0" if industry-adjusted EBITDA/Sales was negative for either the first or second full fiscal year after the firm emerged from bankruptcy.
Firm Acquired: 25 of the 89 firms were acquired in the years following bankruptcy. Since in these cases we treat the purchase price of the firm as a cash flow to claimholders, and since acquired firms are often purchased at a premium, it may be that our performance measure is driven upward by the takeover premiums in these cases.
HLT: 21 firms conducted highly leveraged transactions (HLTs), including leveraged buy-outs (LBOs), in the years preceding formal bankruptcy. Jensen (1991) argues that an advantage of the LBO organizational form is that firms have an incentive to reorganize before too much going-concern value has been destroyed. If so, then these firms may be positioned to perform well after reorganizing their debt.
Second Restructuring: 21 sample firms restructured their debt a second time after emerging from bankruptcy. We expect that these firms will perform poorly compared to firms that did not restructure again.
Firm size: we control for firm size (log of total assets at emergence from bankruptcy) in case larger firms find it easier to survive in the post-bankruptcy environment.
Management Change: we examine performance as a function of management turnover. Hotchkiss (1995) finds that firms that retain their pre-bankruptcy CEO display poorer operating performance compared to firms that replace their CEO. She interprets this as evidence that the pro-debtor bias in the Bankruptcy Code allows pre-bankruptcy management to continue inefficient investment policies. In contrast, Hotchkiss and Mooradian (1997), using a later sample, find a positive relationship between performance and retention of the pre-distress CEO. We define a management change as occurring when the CEO who was in office two years prior to the bankruptcy filing is replaced before or at the date of emergence from bankruptcy.
Industry Sales Growth: the ability of a company to recover from financial distress may be determined in part by the economic strength of its industry. We test whether post-bankruptcy performance is related to industry sales growth during the post-reorganization period.
Post-Bankruptcy Debt Ratio: we examine performance as a function of the reorganized firm's long-term debt to total capitalization ratio (LTD/TC), using two measures: 1) LTD/TC or 2) industry-adjusted LTD/TC. We find that firms tend to emerge from bankruptcy with debt ratios above the industry median. If high debt impairs operating performance (Opler and Titman, 1994), then these firms may underperform firms with low debt. Alternatively, debt can control overinvestment by absorbing firm free cash flow and, in the event of default, providing creditors with a means of forcing liquidation should it be optimal (Jensen, 1989, and Harris and Raviv, 1990). Thus, firms with high debt may operate efficiently and exhibit superior performance.
Post-Bankruptcy Investment: firms with ample investment opportunities following bankruptcy should outperform firms with an unattractive investment-opportunity set. We therefore control for the investment-opportunity set in our cross-sectional analysis.
A. Summary Statistics for Explanatory Variables
Table 5 contains summary data for the explanatory variables described above. The industries represented by the sample firms exhibited strong sales growth that averaged 11.1% (median of 8.5%) per year during the post-reorganization period. Of the sample firms, 93% were affiliated with industries that experienced positive sales growth.
Long-term debt as a percentage of total capitalization was high among the sample firms when compared to the median debt/capitalization ratio of their respective industries. Long-term debt usage at reorganization exceeded the industry median by approximately 30%. Of the sample firms, 84% employed more financial leverage than the median firm in their industry.
The sample firms also tended to invest more than the median firms in their respective industries. Seventy percent of the companies in our sample were found to have a positive level of industry-adjusted net investment during the post-reorganization period. The median level of industry-adjusted net investment was 4.1%.
Sixty-four (72%) of the sample firms reported a level of EBITDA to sales that fell below the industry-median during at least one of the two years following reorganization. Twenty-one (24%) of the sample firms entered bankruptcy following a highly leveraged transaction (HLT), 25 firms (28%) were acquired following reorganization, and 21 firms (24%) were required to restructure their debt subsequent to the reorganization. Finally, 64 of the 89 companies (72%) experienced a management change at some time between the onset of financial distress and the emergence from bankruptcy.
B. Cross-Sectional Regression Evidence
Table 6 contains the results of cross-sectional regressions. The explanatory variables are described in the previous section. Recognizing the potential shortcomings of the liquidation value estimates, we present both sets of results in the discussion that follows, since debates over auctions versus reorganizations, as well as liquidation-cost-based theories of capital structure, are founded on the assumption that liquidation is not costless. The first two models use the excess return computed on the basis of going-concern value as the dependent variable. Models (3) and (4) employ returns computed on the basis of liquidation value. The cross-sectional regression results are similar for both specifications of excess return.
Firms that avoid second restructurings show significantly better performance compared to the rest of the sample. Table 6 also shows that post-bankruptcy performance based on cash flows is not related to industry-adjusted EBITDA, management changes, or the industry-adjusted debt level of the reorganized firm. We also find that acquired firms outperform firms that were not acquired.
In examining the effect of investment on performance, we must control for the investment-opportunity set facing the firms. Firms with poor investment opportunities should invest less than firms with good investment opportunities. We proxy the investment-opportunity set by the firm's liquidation costs, defined as the firm's (going-concern value of assets - liquidation value of assets)/going-concern value of assets. Firms with high liquidation costs are likely to have an abundance of firm-specific growth options, which by definition cannot be liquidated for their full going-concern value (Myers, 1977). Low-liquidation-cost firms, by contrast, are likely to have few growth options since their assets are worth nearly as much in liquidation as they are in their current use.(12) A portfolio of firms with high liquidation costs is likely to have a more attractive investment opportunity set compared to a portfolio of firms with low liquidation costs.
We therefore examine the effect of investment on performance separately for high- and low-growth-option firms. We look at the highest and lowest quartile of firms sorted on the basis of liquidation costs. Within [TABULAR DATA FOR TABLE 5 OMITTED] each quartile we define an indicator variable equal to 1 if the firm's net investment cash flows in the first year after bankruptcy (scaled by total assets at year 1) exceed the industry median, and zero otherwise. We also define a variable equal to 1 if the firm's net investment cash flows in the first two years after bankruptcy (scaled by total assets at year 1) exceed the industry median, and zero otherwise. This variable allows us to observe the firm's investment strategy over a longer time period than just the first post-bankruptcy fiscal year.
We find that in the high-growth-option quartile, performance is positively related to net investment - i.e., in high-growth-option industries, firms that expand relative to their industry outperform firms that shrink relative to their industry. In contrast, Models (2) and (4) in Table 6 indicate that low-growth-option firms that grow relative to their industry do not outperform firms that invest less than the median firm in their industry.
We examine the total cash flows produced by 89 firms that reorganized in Chapter 11 between 1983 and 1993. We estimate the annualized rate of return that was earned by investors who owned all of the debt and equity claims on the firm as it emerged from bankruptcy. We find that, despite having substandard accounting profitability, at least one-half of the sample firms generated a return that exceeded the return available on benchmark portfolios.
Cross-sectional results indicate that firms that avoid second restructurings or are acquired following reorganization are more successful under our performance measure. We also find that firms' investment behavior following bankruptcy affects their performance, but that this effect depends on the investment-opportunity set facing the firm. High-growth-option firms earn superior returns when their net investment exceeds the industry median. However, among low-growth-option firms, performance is unrelated to investment behavior.
Like Hotchkiss (1995) and Hotchkiss and Mooradian (1997), we examine a sample of firms that survived the Chapter 11 process and, as a group, produced weak operating margins in the post-bankruptcy environment. Operating margins do not tell the whole story, however, because they omit consideration of post-reorganization asset sales and other transactions that can cause EBITDA to differ from cash flow. We show that the total cash flows produced by the reorganized firms, and returned to both debt and equity holders, provided a return that was competitive with returns on benchmark portfolios.
The results of our study complement other research on financial distress. Schleifer and Vishny (1992) argue that resolving financial distress through an auction of the firm's assets rather than a Chapter 11 reorganization would provide rents to the buyers [TABULAR DATA FOR TABLE 6 OMITTED] when the secondary market for the firm's assets is illiquid. The excess returns computed relative to liquidation value demonstrate that those rents may be economically significant. Hotchkiss (1995) finds that managers of reorganized firms overstate projected post-reorganization performance. Our finding, that returns computed relative to going-concern value match the returns on benchmark portfolios, suggests that the market accurately discounts those projections.
Helpful comments were received from the Editors, Edward Altman, David Dubofsky, Jean Helwege, Edith Hotchkiss, Dan Keating, Robert Mann, Robert Mooradian, Tim Opler, Gabriel Ramirez, Walter Torous, three anonymous referees, and seminar participants at University of Missouri, University of Missouri-St. Louis, Ohio State University, and Saint Louis University. We thank the many attorneys and corporate officers who provided us with information on their firms' bankruptcy cases.
1 Gilson (1997) reports that after restructuring, about 80% of his sample firms have a debt ratio above the industry median. Hotchkiss (1995) and Hotchkiss and Mooradian (1997) find that a typical reorganized firm's operating profit margin falls short of its industry median following reorganization. These two authors, as well as LoPucki and Whitford (1993), find that one-quarter to one-third of reorganized firms subsequently restructure their debt. Betker (1997), Branch (1998), Eberhart and Weiss (1998), and Ravid and Sundgren (1998) have also recently examined the efficiency of the Chapter 11 bankruptcy process.
2 LoPucki and Whitford (1993) suggest this approach. Eisenberg and Tagashira (1994) examine 124 small-firm bankruptcies from Japan. They compute the present value of payments promised to creditors in the bankruptcy plan (not the post-bankruptcy cash flows actually realized) and compare it to the liquidation value of assets estimated by a court-appointed examiner. They find that, in the aggregate, promised payments to creditors exceed liquidation values and conclude that the Japanese bankruptcy system can create value compared to a system of forced liquidations.
3 Pulvino (1998) shows that distressed airlines incur significant liquidation costs, or "fire-sale" discounts, when selling used aircraft.
4 The disclosure statement for a bankruptcy plan contains a variety of information regarding the reorganized firm, including details of the firm's post-bankruptcy business plan and financial projections. It is mailed to all claimholders and is used to solicit votes on the bankruptcy plan. Under Section 1125(a) of the 1978 Bankruptcy Code, the disclosure statement must contain "adequate information...that would enable a hypothetical reasonable investor...to make an informed judgment about the plan."
5 Sources and uses of cash must balance, so that net cash flows from operations + net cash flows from investment + net cash flows from financing = change in cash . Thus, we can also write Equation (1) as: cash interest paid - net cash flows from financing - other cash flows from financing = net cash flows paid to all claimholders.
6 By assuming that intermediate cash flows are reinvested in the S&P 500 (i.e., at the hurdle rate) we avoid the "reinvestment rate assumption" problem inherent in the internal rate of return (IRR) computation. In capital budgeting, our valuation method is referred to as the "modified IRR" or "opportunity cost return" approach (Copeland and Weston, 1992, p. 321). Note that this comparison implicitly assumes that Resorts has an asset beta of 1.0. Since the average equity beta is 1.0, this quite likely overstates the risk of Resorts' assets. We are thus using a high hurdle rate for the sample firms by requiring the return on their assets to exceed the return to levered equity. We investigate the sensitivity of our results to this assumption later in the paper.
7 Later in the paper, we assess the accuracy of these fresh-start going-concern value estimates and find them to be noisy but unbiased estimates of the actual market value of the firm upon emergence from bankruptcy. Gilson, Hotchkiss, and Ruback (1999) also find that fresh-start accounting estimates of firm value provide unbiased estimates of the actual market value of the firm's post-bankruptcy assets.
8 Firms may also have an incentive to overstate the liquidation value of assets if they have secured debt in their capital structure. If collateral values are reported as high, there is a greater chance that secured creditors will be deemed oversecured, and thus prevented from foreclosing on the collateral during the reorganization proceedings. Under Section 362(d) of the Bankruptcy Code, secured creditors can petition to lift the automatic stay if their collateral is undersecured (value of collateral is less than face value of their debt).
9 We consider benchmarks other than the S&P 500 later in the paper.
10 It is not feasible to compute firm-level asset betas due to a lack of historical stock price and accounting data for the recently reorganized firms. Furthermore, due to changes in asset and capital structures that often accompany financial reorganizations, historical asset betas may be poor predictors of future asset betas. Kaplan and Ruback (1995) find that industry asset betas result in valuation estimates that are as close to actual market prices as valuation estimates produced using firm-level asset betas.
11 Lyon, Barber, and Tsai (1999) construct size decile portfolios each June 30, using cutoffs based on the universe of NYSE and AMEX firms. Size decile 1 (smallest firms) is further partitioned into quintiles, resulting in 14 size-based reference portfolios. Lyon et al. further recommend constructing book-to-market equity-matched portfolios, but the asset sales and financial restructuring that typically accompany a Chapter 11 case make measuring the book-to-market ratio for newly reorganized firms problematic.
12 Alderson and Betker (1995) find that liquidation costs can explain the post-bankruptcy capital structures chosen by a sample of reorganized firms. Firms with low liquidation costs choose capital structures with high debt ratios and covenants which severely restrict new investment and the use of free cash flow. This is consistent with the argument that low-liquidation-cost firms have few good investment opportunities - creditors restrict the ability of managers to make new investments if there are no good investments to make.
Alderson, M.J. and B.L. Betker, 1995, "Liquidation Costs and Capital Structure," Journal of Financial Economics (September), 45-69.
Altman, E.I., A.C. Eberhart, and R. Aggarwal, 1999, "The Equity Performance of Firms Emerging from Bankruptcy," Journal of Finance, forthcoming.
Altman, E.I. and S.A. Nammacher, 1985, "The Default Rate Experience on High-Yield Corporate Debt," Financial Analyst's Journal (July/August), 25-41.
Andrade, G. and S.N. Kaplan, 1998, "How Costly is Financial (Not Economic) Distress? Evidence from Highly Leveraged Transactions that Became Distressed," Journal of Finance (October), 1443-1494.
Betker, B.L., 1995, "Management's Incentives, Equity's Bargaining Power and Deviations from Absolute Priority in Chapter 11 Bankruptcies," Journal of Business (April), 161-183.
Betker, B.L., 1997, "The Administrative Costs of Debt Restructurings: Some Recent Evidence," Financial Management (Winter), 56-68.
Branch, B., 1998, "Streamlining the Bankruptcy Process," Financial Management (Summer) 57-69.
Copeland, T.E. and J.F. Weston, 1992, Managerial Finance, 9th ed., Fort Worth, TX, Dryden Press.
Eberhart, A.C., W.T. Moore, and R.L. Roenfeldt, 1990, "Security Pricing and Deviations from the Absolute Priority Rule in Bankruptcy Proceedings," Journal of Finance (December), 1457-1469.
Eberhart, A.C. and L.A. Weiss, 1998, "The Importance of Deviations from the Absolute Priority Rule in Chapter 11 Bankruptcy Proceedings," Financial Management (Winter), 106-110.
Eisenberg, T. and S. Tagashira, 1994, "Should We Abolish Chapter 117 The Evidence from Japan," Journal of Legal Studies (January), 111-157.
Franks, J.R. and W.N. Torous, 1989, "An Empirical Investigation of U.S. Firms in Reorganization," Journal of Finance (July), 747-769.
Gilson, S.C., 1997, "Transactions Costs and Capital Structure Choice: Evidence from Financially Distressed Firms," Journal of Finance (March), 161-196.
Gilson, S.C., E. Hotchkiss, and R. Ruback, 1999, "Valuation of Bankrupt Firms," Harvard University Working Paper.
Gilson, S.C., K. John, and L.H.P. Lang, 1990, "Troubled Debt Restructurings: An Empirical Study of Private Reorganization of Firms in Default," Journal of Financial Economics (October), 315-354.
Harris, M. and A. Raviv, 1990, "Capital Structure and the Informational Role of Debt," Journal of Finance (June), 321-349.
Hotchkiss, E., 1995, "The Post-Bankruptcy Performance of Firms Emerging from Chapter 11," Journal of Finance (March), 3-21.
Hotchkiss, E. and R. Mooradian, 1997, "Vulture Investors and the Market for Control of Distressed Firms," Journal of Financial Economics (March), 401-432.
Jensen, M.C., 1989, "Active Investors, LBOs, and the Privatization of Bankruptcy," Journal of Applied Corporate Finance (Spring), 235-244.
Jensen, M.C., 1991, "Corporate Control and the Politics of Finance," Journal of Applied Corporate Finance (Spring), 13-33.
Kaplan, S.N., 1989, "Campeau's Acquisition of Federated: Value Destroyed or Value Added," Journal of Financial Economics (December). 191-212.
Kaplan, S.N., 1994, "Campeau's Acquisition of Federated: Post-Bankruptcy Results," Journal of Financial Economics (February), 123-136.
Kaplan, S.N. and R.S. Ruback, 1995, "The Valuation of Cash Flow Forecasts: An Empirical Analysis," Journal of Finance (September), 1059-1094.
Lang, L.H.P. and R. Stulz, 1992, "Contagion and Competitive Intra-Industry Effects of Bankruptcy Announcements: An Empirical Analysis," Journal of Financial Economics (August), 45-60.
LoPucki, L.M. and W.C. Whitford, 1990, "Bargaining over Equity's Share in the Bankruptcy Reorganization of Large, Publicly Held Companies," University of Pennsylvania Law Review (November), 125-196.
LoPucki, L.M. and W.C. Whitford, 1993, "Patterns in the Bankruptcy Reorganization of Large, Publicly Held Companies," Cornell Law Review (May), 597-618.
Lyon, J.D., B.M. Barber, and C.L. Tsai, 1999, "Improved Methods for Tests of Long-Run Abnormal Stock Returns," Journal of Finance (February), 165-202.
Myers, S.C., 1977, "Determinants of Corporate Borrowing," Journal of Financial Economics (November), 147-175.
Opler, T.C. and S. Titman, 1994, "Financial Distress and Corporate Performance," Journal of Finance (July), 1015-1040.
Pulvino, T.C., 1998, "Do Asset Fire Sales Exist? An Empirical Investigation of Commercial Aircraft Transactions," Journal of Finance (June), 939-978.
Ravid, S.A. and S. Sundgren, 1998, "The Comparative Efficiency of Small-Firm Bankruptcies: A Study of the US and Finnish Bankruptcy Codes," Financial Management (Winter), 28-40.
Schleifer, A. and R.W. Vishny, 1992, "Liquidation Values and Debt Capacity: A Market Equilibrium Approach," Journal of Finance (September), 1343-1366.
Scholes, M. and J. Williams, 1977, "Estimating Betas from Nonsynchronous Data," Journal of Financial Economics (December), 309-328.
Michael J. Alderson and Brian L. Betker are Associate Professors of Finance at Saint Louis University.
|Printer friendly Cite/link Email Feedback|
|Author:||Alderson, Michael J.; Betker, Brian L.|
|Date:||Jun 22, 1999|
|Previous Article:||Does NYSE Listing Affect Firm Visibility?|
|Next Article:||Making Market Microstructure Matter.|