Some organisations (Banque de France, BIS, etc.) regularly publish the ratio of non-financial companies' debt to gross domestic product (GDP), and make international comparisons. For example, the Banque de France calculates that of the six major OECD countries, French companies have increased their debt as a percentage of GDP the most, rising from 56% in 2001 to 87.6% 20 years later, just behind Japanese groups which are at 104.6%, and well ahead of German companies at 50.7%.

Should we be concerned?

We don't think so, for three reasons.

The first reason is that these ratios do not, as they should, take into account the cash held by companies when calculating debt. However, companies hold a considerable amount of cash and cash equivalents, and the impact thereof is very significant. At 31 December 2020, the gross debt of French companies was estimated by the Banque de France at €1,888bn, but cash and cash equivalents amounted to €886bn, or 47% of the previous amount. In other words, the net debt ratio as a percentage of GDP is 55.5%, and not 87.6% as calculated in gross terms, which is nevertheless the aggregate that is highlighted, wrongly in our opinion.

The second reason is that these debt figures are calculated based on companies' corporate accounts, not their consolidated accounts. This means that when a parent company takes on debt to finance its foreign subsidiaries, the debt is included in the ratio, but not the assets that are the counterpart to it and that constitute a (financial) asset in the accounts. For tax reasons, since the French corporate tax rate has been at the top of the tax rate bracket in the main OECD countries for years, most French groups house the group's debt in France, and not abroad, even if it means capitalising the foreign subsidiary, whose dividends will be paid out under the parent company-daughter regime, with taxation at 1.4%.

The third reason can be illustrated by the acquisition of Tiffany by LVMH for €13.4bn. This is a lot of extra debt for French companies, but as Tiffany has marginal activities in France, it has no impact on GDP, and therefore a deterioration in the debt/GDP ratio, while there is no instant impoverishment, nor any doubt as to LVMH's ability to repay its acquisition debt when the time comes.

*         *            *

While GDP is the favourite indicator of macro-economists, because it measures wealth created during a year as the sum of added value, a company does not repay its debts out of its added value, but out of its free cash flow. With value added as a starting point, the company first pays its employees and production taxes before arriving at its EBITDA, which is a much better approximation of its capacity to repay its debts than value added can be.

Overall, it should be noted that while the gross debt of French companies as a percentage of GDP rose from 56% to 88% between 2001 and 2020, it fell from 56% to 47% of corporate equity, from a high of 75% in the first quarter of 2009.

If we wanted to play the macro-economist, we'd note that the equity of the companies under review thus rose from 100% of GDP in 2001 to 187% in 2020, i.e. from €1,500bn to €4,028bn (gross debts in the meantime rose from €840bn to €1,888bn). Since equity is the measure of a company's ability to withstand a crisis, we can only be reassured by this development.

While not the best measure of the size of corporate debt, the gross debt-to-equity ratio seems to us to be a much better measure than the gross debt-to-GDP ratio.

And if we take it as debt net of cash and cash equivalents and not as gross debt, we can even estimate it at 25% of equity at the end of 2020.

And since the EBITDA of French companies is estimated at €420bn in 2019 and €371bn in 2020, we can calculate the ratio of net banking and financial debt to EBITDA, which in 2020 is 2.7 for 2020 EBITDA or 2.4 for 2019 EBITDA.

All in all, it seems to us that the Banque de France figures are nothing to write home about, or at least if you do, it would be to recommend to your parents that they needn't spend too much time worrying about corporate debt/GDP ratios.

These 2 graphs, published in Option Finance and based on data from CACIB and Dealogic, highlights several elements.

Firstly, the current lack of standardisation of the terms used to call "impact loans", those that others (like us) call "sustainable loans". Then, the well-documented dominance of Europe in this field, but the US is not so far behind.

Finally, within Europe, the disparities are striking, because if Germany and the UK together do as much as France or Italy separately, Belgium does as much as Germany or the UK, even though its GDP is 5 to 7 times lower.

In France, 30% of syndicated loans have an interest rate that is indexed, at the margin, to non-financial criteria. And 40% of large groups and SMEs have already taken out such loans.

With the collaboration of Simon Gueguen, lecturer-researcher at CY Cergy Paris University

The most common business valuation method used by investors (and, in general, the most efficient) is to discount expected free cash flows, possibly with a probability distribution, at a rate determined by the market risk associated with those flows. In this area, there is one factor that makes it technically difficult to correctly anticipate flows: investment options.

When developing a new product or entering a new market, companies often have the opportunity, depending on the success of the first year, to increase or decrease the investment made. Their decision will therefore depend on new information, which is not available (from the entrepreneur or from the market) when the first investment phase is launched. It is an option, in the financial sense of the term, and the fact that the company can choose to adjust the amount invested at a later date implies additional value for the company, similar to a time value. In particular, the greater the uncertainty, the greater the value of having a choice later on^[1].

It has long been known that these investment options represent a substantial portion of the value of companies, especially in risky sectors^[2]. It is widely believed that investors tend to overlook the value of these options in their valuation process, with the result that companies with these options are systematically undervalued. This month's paper^[3] confirms that the presence of investment options often leads to an inaccurate valuation of the company, but shows that this is not a simple negative bias: overvaluations are about as common as undervaluations.

To estimate these valuation differences, Lyandres et al constructed a valuation model based on the use of available capital. In the model, firms have the opportunity to raise additional funds (at some cost) when they observe that ongoing investments yield high profits. Like any valuation model, its construction is debatable; but it is a necessary step since it involves estimating the differences between an observed market price and a theoretical value taking into account the investment options.

The next step is to calibrate the model, i.e., to measure its parameters for a sample of companies (and in particular according to their sector, a crucial variable when it comes to investment options). The data covered by the empirical study covers the period from 1980 to 2014 in the United States. Once the model is calibrated, it is used to determine a theoretical value for the companies in the sample. Companies whose observed value is higher than the theoretical value are considered overvalued, and undervalued in the opposite case.

The first result of the study is to show that valuation gaps (both positive and negative) are more frequent for small, young, high R&D spending firms with volatile results: all these characteristics are a priori indicative of the presence of strong investment options. Lyandres et al emphasise that this observation applies to both overvalued and undervalued companies. In other words, the market misprices investment options, but it does not systematically undervalue them.

In order to confirm that the observed effects correspond to mispricing, Lyandres et al compare the stock market performance of undervalued firms with that of overvalued firms. Once risk factors are taken into account according to standard models, the former return an average of 1.05% in the month following the observed mispricing, and the latter only 0.15%, i.e., an annualised performance difference of 10.8%. This is a very high rate, suggesting that the model constructed is able to identify valuation gaps that can be used by investors.

Finally, Lyandres et al show that the observed financial policies are consistent with the mispricing hypothesis. Firms identified as overvalued carry out more capital increases, while undervalued firms carry out more share buybacks (and their managers' share purchases are also more numerous).

The interest of the paper is twofold. On the one hand, the proposed valuation model seems to provide results that can be used by investors. In this area, however, it is advisable to exercise caution and to check that the profitability/risk ratio obtained remains favourable on other samples. On the other hand, and this is what we will focus on, the results contradict the short-termism hypothesis according to which the market systematically underestimates investment options. Investment options are indeed a headache for business valuation teams, but they lead to errors on the upside as well as the downside.

Problem 1

A young M&A analyst values an unleveraged company at 1,000 using a DCF. Its share capital is composed of 40 common shares and 60 non-voting preferred shares. The value of the share is therefore 1000/(60+40) = 10 

After analysis, she considers that the absence of voting rights requires a discount of 20%. She therefore has 40 ordinary shares valued at 10 euros and 60 preference shares at 8 euros. She notices that the equity value is equal to 10 * 40 + 8 * 60 = 880 and not 1000. She wonders if this is not a nonsense since the cash flows have not moved.

Can you help her?

Problem 2

Why is Berkshire Hathaway, Warren Buffett's company, currently not included in the Dow Jones, which includes 30 of the largest US companies by market capitalisations, and is unlikely to be included despite being the 7th largest US market capitalisation with a market capitalisation of $638bn?

Problem 3

What is the beta of the cash?

Problem 4

A company pays an exceptional dividend representing 40% of its equity value financed by new debt (a leverage recapitalisation). What is the impact on its P/E ratio?

Solution to problem 1

Her error in reasoning is to consider that the discount is a complete loss of value, like evaporation. In fact, it benefits the voting shares and their holders. She should thus reason:

40 x P + 60 x P x (1 - 20%) = 1000, where P = 11.36 and the value of the preferred shares at 9.09. 

Solution to problem 2

Because the Dow Jones weights its components, not by their market capitalisation or free float, but by their share price[1] . As Warren Buffet has always refused to proceed with a stock split, his shares are currently priced at around $424,000, and if they were included in the Dow Jones, the Berkshire Hathaway share would then represent 98.9% of the index, depriving it of any representativeness, except for that of the fluctuations of the Berkshire Hathaway share!

Solution to problem 3

This question, which is classic in recruitment interviews for trainees or analysts in investment banking, is easily solved if you remember that the beta measures the relative volatility of an asset compared to the volatility of the market as a whole. 

However, the volatility of one euro of cash is zero, since one euro of cash will always be worth one euro, in one day, in one month, in one year or in ten years (be careful not to confuse this with the value today of one euro received in ten years, which is another question where you need to discount cash flows). So, the beta of the cash is 0 since its “price" is one euro whatever the market fluctuations.

Solution to problem 4

Too many people try to do the calculation in their head, which leads nowhere because not all the data are available, such as the cost of debt or the corporate tax rate, or they make the mistake of forgetting the additional financial costs of the debt.

The correct reasoning is much simpler. As every reader of the Vernimmen or this newsletter knows, the P/E ratio and risk vary in opposite directions. After this exceptional dividend, the share has of course become riskier because it bears a much higher amount of debt. So the P/E ratio will go down.