Impact of Receivership Costs on the Optimal Capital Structure for Small Businesses

Transcription

1 Impact of Receivership Costs on the Optimal Capital Structure for Small Businesses By Ed Vos and Philippa Webber Small Enterprise Research, Vol 8 No 2, 2000, pp University of Waikato Department of Finance Private Bag 3105 Hamilton, New Zealand Phone: ext 8110 Fa: evos@waikato.ac.nz

2 Impact of Receivership Costs on the Optimal Capital Structure for Small Businesses Abstract: If the costs of bankruptcy are significant, then they will have an influence on a firm when it is determining its optimal capital structure. If anticipated bankruptcy costs are high, lower levels of debt can be anticipated. By studying all firms who were put into receivership in the Hamilton region of New Zealand between 1990 and 1996, this study shows that the direct costs of receivership are approximately 23.5% of receivership proceeds, but only 3.64% of the book assets of the healthy company. Results of other bankruptcy studies are reviewed and help confirm this study s findings of a scale effect of direct bankruptcy costs as a percentage of the value of the firm at bankruptcy, but reject a scale effect between direct bankruptcy costs and the healthy value of the firm. Thus, costs of bankruptcy are not seen to be a significant consideration in capital structure determination for a healthy small firm. Introduction Nearly forty years ago Modigliani and Miller (1958) showed that in a perfect world the value of the firm is not a function of the capital structure (relative proportions of debt and equity used to finance the assets) of the firm. In their correction paper (1963) they show that when interest payments on debt are tax deductible (as they are in the real world) and debt is default free, the value of the firm is maximised when the level of debt financing is maximised. Since then researchers have examined why firms do not exhibit these optimal debt levels. Agency costs (Jensen, 1976), financial slack (Myers and Majluf, 1984), information asymmetry, personal taxes (Miller, 1966) and bankruptcy costs (Robichek and Myers, 1966) are all theories which can be used to explain the difference between tax maximising levels of debt and actual levels of debt. Researchers 1

3 have also shown that small businesses have even smaller debt levels than that of their larger counterparts. The purpose of this paper is to see if bankruptcy costs have an impact on the optimal capital structure of small businesses by examining the scale of the bankruptcy costs. Since researchers have shown a scale effect of bankruptcy costs, this should mean that bankruptcy costs are more significant for small firms than for larger ones. If bankruptcy costs are significant to small firms it could go some way to explaining why small firms have less debt than large firms. On the other hand, if bankruptcy costs can be shown to be relatively insignificant then they would not affect the optimal capital structure decision. The data used in this study comes from all firms who went into and completed receivership in the Hamilton region of New Zealand, between 1990 and The rate of failure, defined as going into receivership, was very low during this time period: from a high of approximately 0.12% in 1990 to a low of 0.02% in Background Researchers such as Kraus and Litzenburger (1973), Scott (1976) Baxter (1967) and Kim (1978) developed optimal capital structure models which incorporated bankruptcy costs. These models show that the value of a firm is maximised by increasing the level of debt financing to a point where the marginal present value benefit of the tax shield equals the marginal present value of the costs of bankruptcy. The actual cost of bankruptcy is difficult to determine. Bankruptcy costs include the direct or administrative costs, indirect costs, shortfall costs and tax credit losses. Haugen and Senbet (1978) argue that it is only the direct costs which are important when determining the capital structure of a firm. For this reason, and the relative ease in 1 See sample data for more specifics 2

4 determining the direct costs of bankruptcy, a lot of research has been done in this area. This research can be divided into two groups: those that have looked at bankruptcies or receiverships, and ; those that have examined liquidations of firms. Baxter (1967), Warner (1977), Ang, Chua and McConnell (1982) and Bradbury and Lloyd (1994) examined the direct costs of the bankruptcy process. The costs and firm sizes varied considerably. The largest firm sizes and lowest direct bankruptcy costs were in the Warner study (1977), the average firm size was $50 million at the time of bankruptcy with a direct cost of bankruptcy of 5.3% 3 of the market value of firm at the time of bankruptcy. The highest level of bankruptcy costs were 14.3% for the New Zealand small businesses of the Bradbury and Lloyd study. Both Stanley and Girth (1971) and Robertson and Tress (1988) looked at the direct costs of liquidations. The direct costs of liquidation in these studies were 31% and 38% respectively. Both groups showed a scale effect concluding that as the firm size increased so did the direct costs of bankruptcy, but at a decreasing rate. This indicates substantial fixed costs and economies of scale in the bankruptcy/receivership/liquidation process. The direct costs for the firms in bankruptcy and receivership relative to the value of the firm were lower than for those firms that were in liquidation. A possible reason for this is in some cases firms go through the bankruptcy or receivership process before entering into liquidation. The value of these firms have already been decreased by the direct and indirect costs of bankruptcy and possibly by the sale of assets. Sample source A sample of 43 Hamilton small business receiverships were examined. These firms came from a list of receiverships compiled by a local Hamilton, New Zealand receiver. It 2 In receiverships were performed by Hamilton receivers on companies registered with the Hamilton Companies Office. In 1994 there were only 5. These figures were divided by the number of registered companies in the Hamilton Companies office, being 21, The market value used was at the time of bankruptcy. When Warner looked at the value of the firm 5 years prior to the bankruptcy, the direct costs of bankruptcy were only 1.4%. The direct costs are even smaller if the likelihood of bankruptcy is also taken into account. 3

5 included all receiverships performed by Hamilton receivers on Hamilton registered companies between 1990 to By law when a company goes into receivership the directors must submit a Statement of Affairs to the Companies Office. This gives the financial position of the firm at the time of receivership. It includes the estimated realisable value of assets, a list of creditors, including the level of indebtedness and each creditors priority. Unfortunately, only 21 of the 43 firms filed this statement. The receiver must also file the Receivers receipts and payments upon cessation of their acting as receiver. This includes payments to the firm from debtors, from selling assets, GST refunds, and payments by the firm for administrative costs of the receivership, to preferential creditors, debenture holders and unsecured creditors. Sample Outcomes Table one gives an analysis of the outcomes of the receiverships. Table one Repayments to creditors Partly paid preferential creditors 2 Partly paid debenture holders 39 Fully paid debenture holders 2 5 Partly paid unsecured creditors 0 Total 43 Number of firms Table Two 4 These 43 firms are therefore the majority of receiverships performed over this period in the Hamilton region. It is possible that if the list had included receiverships performed by non Hamilton receivers for Hamilton companies the figure would have been higher. 5 In fact they more than repaid the debentureholder, this is because the figure in the Statement of Affairs is completed at the time the firm goes into receivership. During the receivership process the interest on the debt continues to accrue. This means that the total amount due to the creditors by the end of the receivership process is more than that stated in the Statement of Affairs. 4

6 Losses to the creditors (21 firm sub sample) mean median Preferential creditors 8% 0% Debenture holders 53% 60% Unsecured creditors 100% 100% Total creditors 79% 82% At the time the firms went into receivership the average age of 42 of the firms was 13 years with a median of 10 years, with 30 firms (71%) being five years or older. This is at odds with the general idea that small firms die young. Research by Williams (1982) showed that 69% of his sample went into bankruptcy by the age of 5 years, while the sample of Hall and Young had an average life expectancy of 6.9 years. Sample statistics Table three shows the summary statistics and dollar distributions of the direct costs and receivership proceeds of the full sample and the estimated asset values, total debt, short term and long term debt of the sub sample of 21 firms. Table three Summary statistics of the sample firms (number of firms) $ (000) Direct costs Receivership proceeds Estimated Asset Value Listed debt Short/t debt , million Long/t debt minimum maximum average median

7 # of firms Comparing the company directors estimated asset values in column four with the receivership proceeds of the 21 firm sub sample (average receivership proceeds $280,747 and a median of $139,017) shows that the directors estimates tend to be slightly biased upwards. This is also found in the Robertson and Tress (1988) study. However, the number of under and over estimations of the assets value were almost half: half indicating that, on the whole, directors do not intentionally overestimate the value of their assets. Column five shows the listed debt for the 21 firm sub sample ranged from $129,725 to $1,638,570 with an average of $677,545 and a median of $549,875. Using a leverage ratio of 0.41:1 the implied healthy book assets of the firm can be estimated. This is important as firms determine their optimal capital structure when they are healthy. This leverage ratio is a combination of two studies of accounting statistics 6 for healthy small businesses in the Waikato, New Zealand (Hamilton is the largest city in the Waikato region). The implied healthy book assets of the firm would range between $316,402 to $3,996,512, with an average of $1,652,549 and a median of 1,341,159. Column six and seven break listed debt into short term and long term debt. It shows that on average the sample has more long term debt, having on average 39% short term debt and 61% long term debt. Direct Receivership Costs The direct costs of receivership as a percentage of receivership proceeds and implied healthy book value were also calculated: Table four Direct costs of receivership mean median 6 This leverage ratio was calculated from a study by Vos(1992) and Vos and Forlong(1997) which gave leverage ratios of small businesses in the Waikato region of New Zealand. 6

8 As a percentage of receivership proceeds 23.5% 21.9% As a percentage of implied healthy book 3.64% 3.93% assets 7 This shows that a significant proportion of the market value of the firm at the time of receivership is consumed by the direct costs of receivership. However, the direct costs of receivership are only a small percentage of the firms implied healthy book assets. The costs in relation the implied healthy book value is more important for the determination of an optimal capital structure. This is because firms decide on their structure when they are healthy, not when they are on the brink of receivership. The scale effect Previous research has found a scale effect between firm size and the level of direct costs of receivership (see Warner 1977, for example). As shown in table five a Spearman rank test was used to find the level of correlation between receivership costs and 3 measures of the firms size; receivership proceeds, estimated asset value and book assets. The relationship between book equity and receivership costs was also calculated. The table shows that the level of receivership costs is most related to the level of receivership proceeds. Both book assets and estimated asset value were correlated to a significant level, but much smaller at α = 0.2. There is a significant negative relationship between book equity and receivership costs. This indicates that the more equity the firm has the easier it is for the receiver to complete the receivership process. Table five Spearman rank-order correlation coefficient of total receivership costs with: Receivership proceeds Book equity r s Significance, α 7 The direct costs of receivership as a percentage of receivership proceeds for the 21 firm sub sample averaged 27.3% and had a median of 24.7%. 7

9 Estimated Asset value Book assets To determine if there was a scale effect between the direct receivership costs (DRCs) and receivership proceeds, and between the DRCs and the implied healthy book assets of the firm, four regression equations were run. The relationships were tested using both a logarithmic and quadratic function regression. This would determine if the DRCs not only increase as the firm size increases (as shown in the Spearman rank) but if the costs increase at a decreasing rate. Direct costs of receivership was the explanatory dependant and firm size (using the receivership proceeds measure or book assets ) were the explanatory variables. The following models were tested: logarithmic log DC = a 0 + a 1 logrp where DC = the dollar amount of the direct costs of receivership, and RP = either the receivership proceeds or the implied healthy book assets quadratic DC = b 0 + b 1 RP + b 2 RP 2 If direct costs are a decreasing function of size then the expected coefficients of the logarithmic function regression are a 0 > 0 and 0< a 1 < 1. The hypothesised coefficients of the quadratic function are b 0 > b 1 >0 and b 2 < 0. Table six Logarithmic results Full sample, receivership proceeds and direct costs Intercept (a0) 1.32 (4.02) Sub sample, book assets and direct costs 4.87 (1.21) 8

10 Slope(a1) (10) (1.37) R sqr adj 70.2% 4.2% Table seven Quadratic equation Full sample, receivership proceeds and direct costs Sub sample, book assets and direct costs Intercept (bo) (2.61) (-0.21) b (10.41) (1.72) b *10 6 (-9.24) -0.01*10 6 (-1.22) r sqr adj 72.6% 19.7% The results for the DRC in relationship to the receivership proceeds were as hypothesised: a 1 = 1.32 which is more than 0, a 2 = which is between 0 and 1 and b 0 = is more than b 1 = which is more than 0 and b 2 is less than 0. Therefore, it seems that the direct costs of receivership increase at a decreasing rate as the receivership proceeds increase. The logarithmic function results for the relationship between DRC and the size of healthy firms were as hypothesised. However, the quadratic function results reject the hypothesis that the DRCs increase at a decreasing rate as the firm size (the implied healthy book assets) increases. The very low adjusted r squared variables of both equations show that the DRCs are poorly explained by the implied healthy book assets of the firm. The low Spearman rank correlation coefficient between the implied healthy book assets and receivership costs also shows that the book assets of the firm do not adequately explain the DRCs. 9

11 The regression results indicate that the level of DRCs are related to the market value of the assets at the time of the receivership, not the firms size when it was healthy. Since the optimal capital structure of a firm is determined when the firm is healthy, and since there seems to be no clear and definite relationship between DRCs and the implied healthy book assets, it would be unwise to assume that the smaller the firm the relatively more important the DRCs will be. Two main points have now been made. Firstly, there is a DRC scale effect but it relates to the receivership proceeds, not to the value of the healthy firm. Secondly, the overall level of direct bankruptcy costs as a percentage of the value of the healthy firm is very low. This would indicate that receivership costs, as with large firms (see Warner's (1977) conclusions), would not have a significant impact on the determination of a firm's optimal capital structure since the scale effect only relates to the unwell firm. Multi-study comparisons By comparing a number of studies it is possible to get a broader picture of direct bankruptcy costs. Table eight compares four direct bankruptcy cost studies. The size of the firms in the four studies were all noticeably different. The firms in the first three studies can be classed as small businesses, while the Warner firms can be classed as large firms. The book value or healthy value of the firms were all calculated slightly differently. As previously mentioned this study used a leverage ratio estimated from two studies on the financial statements of small firms in the Waikato region. The Bradbury and Lloyd study used a leverage ratio from publications of the New Zealand Corporate Financial Statistics, which compiles the financial statements for listed and unlisted public firms in New Zealand, i.e. large firms. This is a flaw in their methodology as statistics from large firms cannot be used to proxy those of small firms (Vos 1992). Their book value estimates are therefore doubtful. The A,C&M study used a leverage ratio for firms with total assets 10

12 between $250,000 and $3,000,000. As the firms in the Warner case were listed, the market value of the firm five years before the bankruptcy was used to indicate the value of the firm when it was healthy. 11

13 Table eight Direct bankruptcy cost study comparisons This ($NZ) study Bradbury and Lloyd ($NZ) Ang, Chua & McConnel ($US) Date Sample size 43 NZ firm 27 NZ firm 86 US firm receiverships (21 with information on listed debt) receiverships bankruptcies Market value 8 mean median min max Book/healthy value 10 mean median min max DRC mean median min max DRC as a % of market value DRC as a % of healthy value Scale effect for DRC and market value Duration in months 415, , ,990,532 1,652,549 1,341, ,402 3,996,512 52,647 45,583 4, , (mean) 21.9 (median) 3.64 (mean) 3.93 (median) 1,072, , ,070,309 3,927,096 1,623,305 17,278 40,318,112 49,943 34,311 2, , (mean) 8.1 (median) 4.4 (mean) 8.9 (median) 108, , ,136, ,516 n/a 22,980 4,540,620 2, , (mean) 1.7 (median) n/a Yes Yes Yes Yes 19 (mean) 13 (median) 42.6 (mean) 33 (median) Warner ($US) 11 US railroad bankruptcies ,000,000 n/a 10,400, ,700, ,700,000 n/a 29,300, ,500,000 1,880,000 2,000, ,000 2,890, (mean) 1.4 (mean) 14 (mean) 150 (mean) 156 (median) 8 This is determined as the receivership/ bankruptcy proceeds of this study, Bradbury and Lloyd and Ang, Chua & McConnell studies and the market value of the securities of the Warner case. 9 Both the mean and median do not include the firms which had zero asset values, (one third of the sample). 10 This is the estimated book value calculated by this study, B & L and A,C,&M studies and the market value of the Warner study 5 years before they went into bankruptcy. 12

14 Comparing the market value at the time of receivership with the healthy value of the firm shows that, regardless of the size of the firm, approximately 75% of the value of the firm is lost in the process of going from a healthy to a bankrupt firm. This indicates that debenture holders do not discriminate against small businesses when deciding when to put a firm into receivership. All four studies found a scale effect, showing that as the receivership proceeds increased so did the receivership costs, at a decreasing rate. There is a small difference between the direct costs of receivership as a proportion of the healthy firms' value for the small New Zealand firms and the larger US railroads (Warner, (1977)). This could indicate a scale effect. However, when the Warner study is closely investigated it shows that the maximum direct costs of bankruptcy for a firm was 3.2% for a firm worth $29.7 million, this percentage is almost the same as the average for this study, where the average book value of the firms assets was $1,652,549. The B&L study also has higher direct costs of receivership as a percentage of book assets than this study, which has smaller firms. These findings indicate that there is no scale effect for the direct costs of bankruptcy as a percentage of the healthy value of the firm (book assets). They also show that the costs are small (as a percentage of the healthy value) and relatively insignificant in the determination of the optimal capital structure for a firm. Conclusion The search to find a method of determining an optimal capital structure for firms has been going on for the past forty years. Firms, particularly small firms do not possess the optimal levels of debt of the Modigliani and Miller 1963 correction paper. The idea that the costs of bankruptcy may offset the tax advantage of debt is examined in this paper. The impact of direct bankruptcy costs on the determination of the optimal capital structure for small New Zealand businesses was examined. These results were 13

15 then compared with results from other studies to give a broader picture of the impact of bankruptcy costs on a firm optimal capital structure. The results of this paper show that on average a significant 23.5% of the market value of the firm at the time of receivership was consumed by the direct costs of receivership. The important measure of the firm's value is the implied healthy book assets or the value of the firm when it was healthy. This is because a decision as to the optimal capital structure of the firm is made when the firm is healthy. The direct costs of receivership to the value of the firm when it was healthy (implied healthy book assets) were not high, averaging 3.64%. This low cost of bankruptcy for a healthy firm is only slightly higher than the bankruptcy costs of Warner s firms when they were healthy. These results suggest that the direct costs of bankruptcy would have little impact on the firm's optimal capital structure decision. The scale effect of the direct bankruptcy costs as a percentage of the market value of the firm at the time of bankruptcy (bankruptcy proceeds) was confirmed in all four studies. It was also supported when three of the studies were compared, which indicates that as the market size of the firm increases the percentage of the firm consumed by the direct costs of bankruptcy decreases. These studies cover different samples, time periods and indeed different countries. Yet the overall pattern of findings in these various studies seems consistent. Only this study tested the scale effect of the direct bankruptcy costs and the book assets of the firm. It was concluded that there is not enough evidence to support a scale effect. The comparison between the studies showed that there was no scale effect between the direct costs of bankruptcy and the value of the firm when it was healthy. The results of this study have shown that the direct costs of bankruptcy are too small to impact on a small firms decision as to what its optimal capital structure will be. It also shows that there is no scale effect between the direct bankruptcy costs and the value of 14

16 the firm when it is healthy, implying that the direct costs of bankruptcy are not the reason why small firms have lower debt / equity ratios than their larger counterparts. Therefore, other factors must influence small businesses to have lower debt ratios than larger firms. One of these factors could be the negative agency advantage of debt for small firms as found by Vos and Forlong (1996). 15

17 Bibliography Ang J.S., J.H. Chua and J.J. McConnell, The administrative costs bankruptcy: a note Journal of Finance, 37, pp Ang J.S., On the theory of finance for privately held firms Journal of Small Business Finance, 1 (3), pp Baxter N., Leverage, risk of ruin and the cost of capital, Journal of Finance, 22, pp Bradbury M.E. and S. Lloyd, An estimate of the direct costs of bankruptcy in New Zealand, Asia Pacific Journal of Management, 11, 1, pp Hall G. And B. Young, Factors associated with insolvency amongst small firms, International Small Business Journal, 9,2, pp Haugen R.N. and L.W. Senbet, The insignificance of bankruptcy costs to the theory of optimal capital structure, Journal of Finance, 33, pp Hickman K.A. and M.J. Shrader, Economic issues in bankruptcy and reorganisation, Journal of Applied Business Research, 9, 3, pp Kim E.H., A mean-variance theory of optimal capital structure and corporate debt capacity, Journal of Finance, 33, pp Kraus A. and R. Litzenburger, A state preference model of optimal capital structure, Journal of Finance, 28, pp Miller M, Debt and taxes, Journal of Finance, 32, pp

18 Mogdilani F. and M. Miller, The cost of capital corporation finance and the theory of investment, American Economic Review, 48, pp Mogdilani F. and M. Miller, Corporate income taxes and the cost of capital: A correction. American Economic Review, 53, pp Robertson D.K. and R.B. Tress, Bankruptcy costs:evidence from small-firm liquidations, Australian Journal of Management, 10, pp Robichek A.A. and S. Myers, Problems in the theory of optimal capital structure, Journal of Financial and Quantitative Analysis, 1, pp Stanley D.T. and M. Girth, Bankruptcy: problem, process, reform, Washington D.C,: The Brookings Institute. Warner J.B., Bankruptcy costs: Some evidence, Journal of Finance, 32, pp Williams A., Patterns of small business failure, Institute of Industrial Economics, Newcastle, Australia, pp Vos E., Unlisted businesses are not financial clones of listed businesses Entrepeneurship Theory and Practice, Vol. 16, No. 4, Summer, pp Vos E. And C. Forlong, "The agency advantage of debt over the lifecycle of the firm, The Journal of Entrepreneurial and Small Business Finance, Vol. 5 No. 3, pp

19 18