(Note: If you are just joining the seminar, please start by reading the Introduction)
One important lesson from Part 2 is that the investment approach needs to be heavily influenced by the distribution of returns of the securities in which one is investing. Consider the two extremes: high-grade bonds vs. venture capital.
The distribution of returns among high-grade bonds is as follows: out of 100 bonds you are likely to have a couple that will default and the rest will pay as promised. The ones that default, even with some recoveries, will result in a very large, 50%+ loss. The difference among the ones that don’t default will be very small, usually a percent or so in yield per year.
Your success will not depend on choosing among the bonds that don’t default but rather on eliminating the few that will. In other words, there is almost no penalty for incorrectly eliminating a security that you even remotely suspect could default.
Compare that with venture capital. Here you are likely to have a couple of huge winners with returns of 100x+, some far smaller winners and many large losers. Here, it doesn’t matter how many complete losses you avoid if you don’t catch the one or two 100x returning start-ups. The cost of a mistake of omission, unlike in high-grade bonds, can be huge.
Another lesson that Graham teaches us in Part 2 is that it is better to avoid trouble than to rely on assets to protect us. His concept is to focus on a company’s ability to pay rather than on the collateral supporting a bond. There are several reasons for this, including the time it would take to get your money back in bankruptcy, the frictional costs involved and the uncertainty of the market value of collateral when it needs to be liquidated.
Graham gives us several formulas/criteria to use as a checklist for a high-grade bond. This is probably the part that needs to be updated for modern times and analytical tools. Here is a brief overview of what is used in modern times:
Coverage Ratio
The benefit of the coverage ratio is that it measures ability to pay directly. The numerator should be a measure of pre-tax cashflow available to pay mandatory fixed charges and the denominator should be the fixed charges. The denominator is mostly interest expense, although it can include other items such as mandatory pension contributions or mandatory legal payments.
The numerator is where I think we need to update our approach from Graham’s. Graham used EBIT (Earnings before Interest and Taxes), but that’s too punitive. EBIT subtracts both depreciation and amortization. Most amortization is a non-cash accounting charge (more on this in a later reading). Depreciation is quite real, but here we need to think about the timing of payments.
When looking at interest coverage, we care about the following scenario: the company is in distress, its profitability is depressed and it needs to figure out if it can find enough cashflow to meet its interest payments until its business recovers. Depreciation has a corresponding cash outlay – Capital Expenditures. However, CapEx can be broken up into 2 components: growth and maintenance.
The maintenance CapEx component can mean different things depending on the question being asked. From the point of view of the equity investor, long-term maintenance CapEx is rarely below depreciation. That’s the reason for using Graham’s EBIT approach. However, when the company is in temporary distress and is just trying to make it through to the other side of whatever challenges it is facing, that’s not the relevant maintenance CapEx because much of it can be deferred. So, there is a lower, ‘bare-bones’ maintenance CapEx that is the right number to think about for the purposes of credit analysis.
The complete formula for the coverage ratio would be: (EBITDA – bare-bones maintenance CapEx) / (All Mandatory Fixed Charges). In practice, what we see in modern bank covenants is a simplified ratio of EBITDA/Interest. While it ignores some nuances, it is approximately right, especially if the threshold level is set higher than an EBIT/Interest ratio would be set to account for the numerator being an overestimate of the available cash flow.
Leverage Ratio
While Graham referred to Equity/Debt as the leverage ratio, there is another leverage ratio that is commonly used in credit analysis. While the coverage ratio measures ability to pay directly, its shortcoming is that it is relying on current interest rates.
For example, consider a company that has variable interest rate debt. It might have a coverage of 4x in a low-rate environment, but if interest rates were to spike that ratio would deteriorate substantially. The alternative is the leverage ratio defined as: Debt/EBITDA, which is one of the two ratios commonly seen in modern bank covenants. For context, a BBB-rated industrial company usually has Debt/EBITDA below 3.5x.
As a thought exercise, can you calculate the relationship between the coverage ratio (EBITDA/Interest) and the leverage ratio (Debt/EBITDA) at different levels of debt and different interest rates?
In Part 3, we are now moving away from high-grade bonds which have limited, defined upside and into a category of securities in between them and equities. While all of the securities mentioned here by Graham offer a wider range of upside than high-grade bonds, they differ in a number of ways.
Securities with speculative features usually combine the characteristics of a bond with that of a call option on the equity. In some cases, as with convertible bonds, the two are inseparable in a single package. In the case of bonds packaged with warrants the call option can be detached and sold separately.
Even though we are approaching Graham’s teachings as investors, we should pause here and put on our CFO hats. Why would we, as a CFO of a company raising capital, issue converts or offer a sweetener to a straight bond in the form of warrants? The answer is simple: because the market won’t let us issue straight debt as a high-grade borrower.
Perhaps it’s because the company has a limited history. Or it does not have the cashflow to support normal coupon payments. It could also be because it has exhausted its regular debt capacity subject to the constraints that would still put it within the high-grade category.
The reason matters less than the point that there is a negative selection bias that companies issuing hybrid securities, as a group, are weaker credits than those able to issue straight high-grade bonds.
To entice the market to overlook the weakness of the credit, the company then has to part ways with a portion of the equity upside. In other words, because the downside is greater, the upside must be as well.
With respect to high-yield or junk bonds, Graham has a very specific approach. He is looking for equity-like returns, which is why he sets the bar by only looking at distressed bonds trading at or below 70% of par value. That would translate into an annualized rate of return over 3-5 years well into the teens.
Finally, preferred stock when purchased near par value combines the worst characteristics of these securities: negative selection bias of weaker issuers with limited upside and few if any contractual protections. To Graham the answer is simple: just pass.
Part 3 also contains one of the most important passages in the whole book. I was confused the first couple of times that I read it, and it took me a bit of thinking to reason it out. I would suggest that if you were not at all confused upon reading it you are either already an expert investor or you haven’t thought about it deeply enough:
As between the two factors, it is undoubtedly true that it is more profitable to select the right company than to select the issue with the most desirable terms. There is certainly no mathematical basis on which the attractiveness of the enterprise may be offset against the terms of the privilege, and a balance struck between these two entirely dissociated elements of value. But in analyzing privileged issues of the investment grade, the terms of the privilege must receive the greater attention, not because they are more important but because they can be more definitely dealt with. It may seem a comparatively easy matter to determine that one enterprise is more promising than another. But it is by no means so easy to establish that one common stock at a given price is clearly preferable to another stock at its current price.
-Page 300 (6th Edition)
If you unpack what Graham is telling us here, I think you will reach the conclusion that his point is that two random stock whose securities you are comparing are very likely to be efficiently priced regardless of the differences between the underlying companies. For example, even if one business is demonstrably superior to the other, he is saying that the market likely already reflects this in the stock price, and that is the reason we should focus on analyzing the differences in the terms of the securities.
Graham’s point boils down to this: most stocks are approximately correctly priced by the market most of the time.
Also, bringing this back to dimensions of an investing style – this is a key dimension: do you want to be more of a business analyst or more of a security analyst? As we study different investors, keep this important dimension in mind.
There were far more insightful answers than I can fit in here, and over time I will try to highlight as many folks’ contributions as possible:
In response to Question 1 Andrei Tonkikh wrote:
Graham called the selection of high-grade bonds a “negative art” because the investor’s main task is to exclude unreliable issuers. His logic:
- A bond’s income is capped by its terms, while possible losses can be large.
- Common stocks have no predetermined ceiling on returns, so their analysis is aimed not only at spotting risks but also at finding opportunities.
I broadly agree with Graham: for high-grade bonds the reliability of payment comes first.
However, a bond is not only a debt obligation but also a security in its own right. Its price, yield, maturity, risks and the relationships among them matter too. I believe they should be assessed not in isolation but through a system of relationships and their dynamics. Developing such an approach is one of my future tasks.
In response to Question 2 Lattice Notes wrote:
Graham's main idea about safety is an organization's ability to pay its obligations. It is less about the claim against property or a legal claim. Although the underlying assets are important, but the value of those assets generally tend to also go down during times when the business is not operating. It also follows that it’s not easy to take legal action and the time and costs associated to get a claim. Hence, a borrower’s ability to pay is of primary importance. I agree this approach should be preferred than relying on recovering principal through legal proceedings.
In response to Question 3 Lattice Notes wrote:
The metrics that are important for assessing a borrower’s ability to pay include Debt/EBITDA, EBITDA/Interest, FFO/Debt, and RCF/Net Debt. They measure debt relative to operating earnings, the cushion available to pay interest, and cash generation relative to debt. This aligns with Graham’s focus on a company’s ability to meet fixed charges. For example, fixed-charge coverage can be approximated using EBITDA/Interest or FFO/cash interest, while Debt/EBITDA, FFO/Debt, and RCF/Net Debt assess the debt burden relative to the economic resources supporting it.
There is no universal cutoff. Moody’s uses sector-specific scorecards and combines several measures rather than relying on any one ratio. In its distribution-sector example, the boundary between Baa and Ba is approximately 3.0× Debt/EBITDA, 7× EBITDA/Interest, and 25% RCF/Net Debt. Baa is investment grade, while Ba is speculative grade. However, these ranges must be assessed based on the sector, earnings capabilities and variations, capital needs, and the specific business
In response to Question 6 Minh wrote:
Graham believes these securities are attractive in form because they appear to combine two desirable characteristics: the safety of a senior claim such as a bond’s fixed interest and priority over common shareholders and the possibility of unlimited appreciation through a conversion, participating, or subscription privilege.
I agree with Graham’s point only in theory. The speculative feature does allow investors to access potential future equity gains that a straight high-grade bond excludes. However, the privilege does not itself create safety. If the issuer is financially weak, the conversion right may simply make a risky bond more appealing without improving its credit quality.
Moreover, once a convertible rises substantially with the common stock, its market value becomes increasingly dependent on the stock price. At that stage, the investor no longer enjoys the original “safety plus upside” combination; holding it becomes more like holding a speculative equity position
In response to Question 7 dr. Solti-Tóth, László wrote:
On p. 300 (6th edition), under Terms of the Privilege vs. Prospects of the Enterprise:
“As between the two factors, it is undoubtedly true that it is more profitable to select the right company than to select the issue with the most desirable terms.” However, the authors continue that once it is established that an issue is “investment grade”, the terms of the privilege must receive greater attention. Additionally, their opinion is that “we doubt that a successful policy of buying privileged issues from the investment approach can be based primarily upon the purchaser’s view regarding the future expansion of the profits of the enterprise.”
As I understood it, “the prospects of the underlying enterprise” – can be translated to earnings and earnings growth (or cash-flow) [cf. examples of attractive issues on pp. 292-294, 6th edition]. As it is discussed in connection with the questions above, these are the real “security” (“collateral”) of a bond (the debt service).
Therefore, my conclusion is that though the future prospect does not have primacy in the case of bonds, the safe coverage ratio does.
In addition:
On p. 295 (6th edition) the authors derive the investment principle for these kinds of instruments: “A privileged senior issue, selling close to or above face value, must meet the requirements either of a straight fixed-value investment or of a straight common stock speculation, and it must be bought with one or the other qualification clearly in view.” I would like to add that in Chapters 6-8 they argue that a seniority of an issue does not make a security more attractive than the “junior” issues; if a company/ business/ enterprise is investable based on its fundamentals, then it may be worth to buying the junior issues (unless the spread between the interest rates of the two is so narrow that the premium to paid for the extra protection would not cut into the profit much).
Week 3 assignment is to read the following and answer the questions below:
Security Analysis Part 4
Security Analysis Part 5 chapters 31, 34, and 37 through 41
Question 1: What does Graham mean when he says that as far as a typical common stock is concerned, that analysis is unlikely to yield a dependable conclusion as to the attractiveness or real value of the stock? Do you agree/disagree?
Question 2: Why does Graham believe it is dangerous to project an earnings growth trend into the future? Do you agree/disagree? Is it also dangerous to project a company's level of average earnings into the future? Why/why not?
Question 3: What are the challenges that Graham identifies with Growth Investing? What are your thoughts on this topic?
Question 4: How does a company's dividend track record and current policy impact its attractiveness as an investment?
Question 5: When should a company be reinvesting earnings and when should it be paying them out as a dividend?
Question 6: What are the issues with relying solely on earnings to estimate the value of the business? Do you agree? How does this vary for different types of businesses (e.g. capital intensive vs. asset-light)?
Question 7: How should Qualitative considerations be used to supplement Quantitative considerations when assessing a company's past earnings record in order to estimate its earnings power? What is the problem with just using a historical average over a sufficiently long period?
Question 8: How does Graham suggest a common stock investor approach using the P/E ratio to decide if/at what price to invest in a stock? Which parts do you agree with? Which parts do you disagree with? How applicable do you think what he wrote is today? Why?
Question 9: How should the capitalization structure of the company be taken into account when analyzing the income statement? What investment implications does it have?
Question 10: Create an AI prompt based on the material covered by Graham in Part 4 and 5 that would aid you in your investing.
Now it’s your turn:
Submit your answers in the comments below this article with all your answers in a single comment. I will engage with some of the answers each week and highlight some of the ones I find most insightful in next week’s seminar assignment article. Please submit your answers by the evening of Thursday, October 29th.
Engage with the answers of some of your fellow seminar members in the comments below. Remember – the goal is to learn together. Be kind, be respectful and try to add to our learning as a community.
Feel free to ask any questions about the reading in your comment.
Until two weeks from now,
Gary
About the author
Gary Mishuris, CFA is the Managing Partner and Chief Investment Officer of Silver Ring Value Partners, an investment firm that seeks to apply its intrinsic value approach to safely compound capital over the long-term. He also teaches the Value Investing Seminar at the F.W. Olin Graduate School of Business.