Cost of Bankruptcy

Table 2: Corporate Tax Rate

Marginal-debt

mean Indebted mean Difference t-statistic Consumer

discretionary 33% 31% 2% 0,803 Consumer staple 35% 32% 3% 1,996

Energy 35% 30% 5% 2,098

Healthcare 30% 29% 1% 0,445 Industrials 34% 31% 3% 1,913 Information technology 28% 26% 3% 1,278

Materials 25% 31% -6% -0,783

Global 31% 30% 2% 1,661

Source: Bloomberg

4.2.2. Cost of Bankruptcy

As explained earlier in the sample selection process, only US public companies having more than $100 million in revenues, two years of financial history before the study period, an Altman Z-score above 3 between 2011 and 2013, and positive net income for the full study period are selected. We believe those criteria are sufficient to support the hypothesis that those companies could easily have access to debt. Because financial health and debt eligibility are key factor for our study, some other financial ratios will be studied to validate our hypothesis. The Altman Z-score provides a good indicator of the likelihood of default of the company. The net income filter ensures that all companies are profitable. In addition to those parameters, we want to have a deeper analysis of the asset structure and the cash flow.

‧

We compute the average Altman Z-score for both samples. Marginal-debt companies show a significantly higher Z-score in all GICS sector, ranging from 1.46 times higher in the consumer discretionary sector, to 5.72 times higher in the materials sector. Globally, the average Z-score for marginal-debt companies over the 3 years period is 9.92 against 5.74 for indebted companies. With a t-statistic of 7.46, we can say that the average Z-score of marginal-debt companies is significantly higher than the average Z-score of indebted companies. This holds at both 95% and 99% significance level. The result is coherent with our initial hypothesis as debt increases the bankruptcy risk. Then, in order to determine the impact of capital structure on the Altman Z-score, we compute a second test.

Table 3: Altman Z-Score

Marginal-debt

mean Indebted mean Difference t-statistic Consumer

‧

As mentioned in the previous section, X4 is equal to the market capitalization divided by the total liabilities. It is the main component affected by the capital structure in the Altman Z-score formula. We isolate this component and compare its average value for both the marginal-debt and the indebted sample. As a result, marginal-debt companies show a 6.72 average value for X4 against 2.26 for indebted companies. This leads to a 4.46 difference between both samples, while marginal-debt companies had a 4.18 higher Z-score.

Using the deleveraged X4, we find a 9.16 average deleveraged Z-score for indebted companies against 9.92 for marginal-debt companies. With a t-statistic of 1.28, the average deleveraged Z-score for both samples is not statistically significant at a 90%. Both samples show no difference in their average deleveraged Z-score.

Table 4: Deleveraged Altman Z-Score Marginal-debt

mean Indebted mean Difference t-statistic Consumer

discretionary 8,2 10,2 -2,0 -1,224 Consumer staple 11,6 9,1 2,5 1,455

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

The indebted company sample shows a 39.5% average operational cash flow volatility while the marginal-debt sample shows a 42.2% average operational cash flow volatility. With a t-statistic of 0.30, we find no statistical difference between the two global samples in terms of average operational cash flow volatility at both 95% and 90% significance level. The materials sector, the energy sector and the consumer staple samples are the only one showing a significantly higher volatility for the marginal-debt sample.

Table 5: Operational Cash Flow Volatility Marginal-debt

mean Indebted mean Difference t-statistic Consumer

discretionary 35% 45% -9% -0,712 Consumer staple 16% 29% -13% -2,142

Energy 16% 59% -44% -3,503 Healthcare 26% 28% -2% -0,452

Industrials 50% 41% 9% 0,517 Information

technology 33% 47% -14% -1,065 Materials 11% 27% -16% -4,742

Global 43% 40% 4% 0,383

Source: Bloomberg

‧

As mentioned in the previous section, we expect marginal-debt companies to have a lower asset tangibility ratio, asset tangibility being defined as tangible assets divided by total assets.

However, the data shows us that for all GICS sector, marginal-debt companies have an average assets tangibility ratio at least as high as their indebted peers. Globally, with a 0.86 asset tangibility ratio, marginal-debt companies´ assets are 19% more tangible than those indebted companies. The superior asset tangibility of the marginal-debt sample is statistically significant at both 95% and 99% significance level. This is contradictory to our first hypothesis and indicates that, instead of having a higher loss given default figure, marginal-debt companies might actually have a lower loss given default due to their higher asset tangibility.

Table 6: Asset Tangibility Marginal-debt

mean Indebted mean Difference t-statistic Consumer

discretionary 91% 79% 12% 4,172 Consumer staple 87% 72% 16% 3,192 Energy 87% 88% -1% -0,105 Healthcare 85% 64% 21% 5,538

Industrials 85% 68% 17% 5,702 Information

technology 81% 70% 11% 3,653 Materials 95% 83% 12% 3,727 Global 86% 72% 14% 9,681

Source: Bloomberg

‧

When analyzing the available collateral ratio, a higher ratio indicates that a company has more collateral available for new debt issuance. Results are again in favor of marginal-debt companies. Marginal-debt companies have on average 1.26 times their total equity in tangible assets net of financial debt, against 1.07 for indebted companies. With a t-stat of 3.37, we can say that marginal-debt companies have significantly higher ratio at both 95% and 99%

significance level. It reinforces the fact that marginal-debt companies might not suffer from higher loss given default.

Table 7: Available Collateral Ratio Marginal-debt

mean Indebted mean Difference t-statistic Consumer

Finally, by looking at the operational cash flow ratio, we find that the marginal-debt sample average ratio is 21.66% against 17.02% for the indebted sample. With a t-statistic of 3.43, the marginal-debt sample generates significantly higher operational cash flows compared to its total equity plus financial debt at both 95% and 99%.

‧

Table 8: Operational Cash Flow Ratio

Marginal-debt

mean Indebted mean Difference t-statistic Consumer