According to a common idea, increased financial leverage of a business entity endangers its financial stability (defined as stability of earnings and cash flows or creditworthiness) as it increases fixed cash outlays by the amount used to service the debt. By doing so, increased leverage shifts the break-even quantity of sales/production (minimal quantity, at which a company is able to cover all of its expenses, both operating and financial) and adds volatility to (i.e., reduces the stability of) earnings and cash flow fluctuations.
Theoretically, in times of macroeconomic distress, highly leveraged companies should appear more prone to severe declines of stock prices and consecutive defaults, since under recessionary conditions they would encounter more serious declines in cash flows and earnings than otherwise identical entities.
But in practice, one can argue that not all companies are operating in the same economic environment. Some operationally stable companies may have access to cheaper credit resources and, thus, can leverage their equity without imposing significant volatility on their bottom line. If the majority of highly leveraged companies belong to this category, the amount of debt per dollar of equity would be a poor predictor of a company’s financial stability.
In order to assess whether financial leverage of the companies traded in the U.S. was indeed a significant contributor to the deterioration of their financial risk profiles in the years 2008-2009, I conducted the following test:
Using the Value Line database for January 2008, we take a snapshot of the financial leverage and stock prices of companies traded in the U.S.
With stock price being commonly viewed as an aggregate opinion of market participants about the financing, investing and dividend strategy of a company (from which the company’s risk profile is derived), I believe the relative change in the stock price to be, in general, the best available indicator of relative deterioration of a company’s financial stability. This seems especially true during recessions, when characteristics of investing strategies, which are often put on hold, play a lesser role in the determination of stock prices. (Although in individual cases, other numerous factors might also contribute to this relative change.) That being said, in order to estimate the degree of financial risk profiles’ deterioration, I compare the stock prices from the database for January 2010 with those for January 2008.
I break the sample down by four categories, based on the financial leverage criterion (expressed as the total debt divided by the assets' total book value) and test whether any conclusion can be inferred.
Before discussing the results it’s necessary to recognize the imperfection of the databases used. For the purposes of analysis, companies with zero stock prices and zero or negative equity were excluded, as most of those companies seemed to be reported with errors. Aside from that, companies available in only one database were excluded, since there were no means to find out whether they weren’t present in the other database by a mistake, due to merger and acquisition or due to delisting.
This will probably lead to survivorship bias and an underestimation of the number of severely distressed companies. Another alteration was made to exclude companies with the market capitalization of less than $500 million. For most of such companies, no liquid stock market exists. This results in inadequate stock pricing, with prices failing to properly reflect changes of the issuer’s fundamentals. That being said, the final variant of the database is not error-free, but with a sample of 1734 companies, I see no reason to believe that errors will have significant effect on our conclusions.
Debt-to-assets ratio is used to break the sample down by four categories:
<0.2: not leveraged
0.2-0.5: mildly leveraged
0.5-0.8: significantly leveraged
>0.8: extremely leveraged
For each group, I calculate the average rate of stock price change and compare it to the average of the whole sample in order to evaluate the statistical significance of the deviation.
Results of the grouping are presented below.
Table 1. Distribution of companies by debt-to-assets ratio and the rate of stock price change (Click to enlarge)
Chart 1. Distribution of companies by debt-to-assets ratio and the rate of stock price change (Click to enlarge)
Chart 2. Distribution of companies by debt-to-assets ratio and average rate of stock price change (Click to enlarge)
Although the pattern in the last diagram is not entirely consistent, it seems sufficiently convincing to conclude in favor of the thesis under consideration: The more leveraged, the more susceptibility to the economic cycle.
Interestingly, the average rate of stock price change for the first group appears significantly higher than that of the second group. An explanation that comes to mind is that a considerable portion of the companies in the first group are free of leverage; not because they strive for higher financial stability, but because their poor creditworthiness (brought about by high operating risk or other factors) precludes them from borrowing in significant amounts. Lack of debt wouldn’t provide much of a relief for such companies and doesn’t lead to increased financial stability.
Before making final conclusions, a formal hypothesis testing needs to be conducted. Results of a two-tailed t-test are presented below.
Table 2. Results of hypothesis testing (Click to enlarge)
As seen above, stock price changes for the third and fourth groups fall within their confidence intervals. Thus, these changes are not statistically different from the average rate of stock price change of all groups.
The average stock price change of the first group lies within 1% outside the confidence interval. Such proximity, as well as the group’s inconsistency with the pattern exhibited by other groups (see the second diagram), warrants higher scrutiny of the underlying data. As mentioned above, companies with zero or negative equity were excluded from the analysis. But companies with zero debt were not, although it’s quite likely that some of them were recorded with zero debt by a mistake. If the number of companies with underestimated debt was significant, it must have lead to misclassification of the companies in favor of the first group and consecutive overestimation of the first group’s average stock price decline. With that in mind, and considering the proximity to the confidence interval, it seems reasonable to deem the findings for this group also inconclusive.
That leaves us with only one group that we can make valid conclusions about: The second group. Average stock price decline within the group was milder than that of other groups, and this deviation from the sample average was statistically significant. This finding is consistent with the thesis under consideration and suggests that susceptibility to financial stability of a business entity is inversely correlated with the business’s financial leverage measured as debt per dollar of assets (or equity).
In view of the above, we can conclude that the initial thesis, stated as presence of a positive relationship between the susceptibility to the economic cycle and the degree of financial leverage (measured as debt-to-assets ratio), remains largely unconfirmed. The sample was broken down in four groups. With small sizes of samples and high variance, three of four groups were found to exhibit only statistically insignificant differences in the degree of risk profile deterioration, (measured as the relative decline in stock prices) over the period between January 2008 and January 2010.
At a 10% significance level, only the findings regarding the group of mildly leveraged companies (20 to 50 cents of debt per dollar of assets) were conclusive and consistent with the thesis. With corrections made to account for possible errors in the underlying data, no evidence contrary to the thesis appears to be present. Combined, these findings suggest the presence of only a weak link between the susceptibility to economic cycle and the financial leverage. Inconclusiveness of the findings regarding three of four groups warrants further research of the matter and will require the use of a database less prone to errors, than the one applied above.
Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.