Defining the sample

國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

4. Data and Analysis

4.1. Defining the sample

The study aims at analyzing US public companies having more than 100 million US dollars in revenues in 2011. This first criterion will allow us to avoid companies which could not access to debt because of their size. Also, big companies might be more aware of capital structure challenges and act according to it. Most of the time, smaller companies do not have a financial department and might not think of their financial strategy carefully.

Because they have an extremely specific structure which might not be analyzed the same way as other industries, financial companies will be removed from the study. Most financial ratios used in the study are not relevant for financial companies of require some adjustments.

Moreover, financial companies are well known for having opaque financials with complex network of subsidiaries. Those are the main reasons why financial companies are dismissed from the study.

Companies will have to show at least 2 years of financial history prior to the test period. Age is actually another key issue when seeking debt financing. We believe that companies with more than 2 years of financial history will have enough to show to financial institutions to easily subscribe debt.

After setting those first parameters on Bloomberg, we find 1907 US non financial public companies having more than $100 million revenue in 2011 and at least 2 years of financial history before 2011. Out of those 1907 companies, 313 (16.4%) are marginal-debt companies and 1594 (83.6%) are indebted companies.

‧

As mentioned earlier, Devos et al. (2012) find the evidence than most companies do not have debt in their balance sheet due to their inability to raise debt either from banks or financial institutions. It actually sounds normal that companies facing a high bankruptcy risk would have difficulties to convince banks or financial markets to lend them additional money. Then, most marginal-debt companies may be in this situation and have no other choice than not having debt.

This phenomenon is logical and well documented in the literature. Indeed, as this study aims at studying marginal-debt companies which use this strategy as a choice, we want to remove companies with high bankruptcy risk from the study. While in real life, each company is a specific case which involves many different parameters to be taken into account in order to assess its capacity to raise debt, this study will use a single variable, the Altman Z-score. Using a single and universal variable avoids any bias in the study. It also helps us avoiding too much complexity and restrictions in the data selection. Altman Z-score was introduced by Edward I.

Altman in 1968 in order to predict company bankruptcy. It is widely used in academic studies for analyzing bankruptcy risk and its impact on other financial indicators. In the scope of this study, we will use the Altman Z-score given under the VM001 (ALTMAN_Z_SCORE) mnemonic in Bloomberg and calculated with the following formula :

Z 1.2X 1.4X 3.3X 0.6X 1X

Where X1 = (Total Current Assets – Total Current Liabilities) / Total assets X2 = Retained Earnings / Total Assets

X3 = (Pretax Income + Interest Expense) / Total Assets X4 = Market Capitalization / Total Liabilities

X5 = Revenue / Total Assets

‧

Using the above formula, a Z-score below 1.81 indicates a high bankruptcy risk while a score above 3 indicates a company in a healthy situation. In this study, companies will be selected if they show a Z-score above 3 from 2011 to 2013. In Altman (2000), using a cutoff score of 2.675 the Altman Z-score proved to be between 82% and 94% accurate predicting bankruptcy one year in advance. Type II errors (declaring that a firm would bankrupt while it actually does not) is between 15% and 20%. Those results are reasonable with the purpose of the study and help eliminated a major proportion of companies which are not able to raise debt. Type II errors will only penalize the study by reducing the sample. This is not a key issue as it should not affect the results of the study. On the other hand, type I error would penalize the study as it would include companies with high bankruptcy risk in the study. This is why, as long as the test sample is not too small, it is better for this study to select a conservative Z-score as a selection parameter. Our sample will include companies which have a Z-score above 3 for each consecutive year between 2011 and 2013.

In addition to the Altman Z-score, we require companies to have positive net income for each consecutive year between 2011 and 2013. With this second parameter indicating that all companies in our sample should be profitable, there is a much higher probability that all marginal-debt companies in the sample could easily access debt.

Using those two financial health criteria, 718 companies are remaining. It represents 37.7% of the total number of US non financial public companies having more than $100 million revenue in 2011 and at least 2 years of financial history before 2011. The marginal-debt sample contains 167 (23.2%) companies against 551 (76.6%) for the indebted sample. We can notice that indebted companies were much more affected by the financial health criteria that marginal-debt companies. It tells us that, in average, marginal-debt companies are less risky than their

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

indebted peers.

Among those marginal-debt companies, 100 (60%) companies have strictly no debt from 2011 to 2013, and 74 (44%) have strictly no debt from 2009 to 2013. On the other hand, 43 companies had a debt-to-equity ratio above 5% in 2009, 2010, or both 2009 and 2010. 13 of them have a debt ratio above 20% in 2009, 2010, or both 2009 and 2010. The average debt-to-equity ratio for the period (e.g. from 2011 to 2013) is 0,36%.

On the other hand, we have 551 indebted companies having a 56,29% average debt-to-equity ratio. Sixty three of those indebted companies had a debt-to-equity ratio below 5% in both 2010 and 2009.

CONTROL GROUP Indebted companies

TEST GROUP Marginal-debt companies US non financial public companies

Available information for at least 2 years before the study period Total revenues > $100M in 2011

Altman Z-score > 3 from 2011 to 2013 Net income > 0 from 2011 to 2013 D/E > 0.05 for at least one year

between 2011 and 2013 D/E < 0.05 from 2011 to 2013 Figure 2: Sampling Summary

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

4.2. Analysis of the Sample Data