Bitcoin Portfolio Insurance: Bond Risks And Contagion

As bond risks grow and contagion appear more likely than ever, every investor needs to consider bitcoin as portfolio insurance.

Editor’s note: This article is the second in a three-part series. Plain text represents the writing of Greg Foss, while italicized copy represents the writing of Jason Sansone.

In part one of this series, I reviewed my history in the credit markets and covered the basics of bonds and bond math in order to provide context for our thesis. The intent was to lay the groundwork for our “Fulcrum Index,” an index which calculates the cumulative value of credit default swap (CDS) insurance contracts on a basket of G20 sovereign nations multiplied by their respective funded and unfunded obligations. This dynamic calculation could form the basis of a current valuation for bitcoin (the “anti-fiat”).

The first part was dry, detailed and academic. Hopefully, there was some interesting information. At the end of the day, though, math is typically not a strong subject for most. And, as for bond math, most people would rather chew glass. Too bad. Bond and credit markets make the capitalist world function. However, when we socialize losses, and reward the risk takers with government funded bailouts, the self-correcting mechanism of capitalism (creative destruction) is jeopardized. This topic is important: Our leaders and children need to understand the implications of credit, how to price credit, and ultimately, the cost of crony-capitalism.

Heretofore, we will continue our discussion of bonds, focusing on the risks inherent to owning them, the mechanics of credit crises, what is meant by contagion and the implications these risks have for individual investors and the credit markets in general. Buckle up.

Bond Risks: An Overview

The main risks inherent to investing in bonds are listed below:

  1. Price*: rRsk that the interest rates on U.S. treasuries rise, which then increases the yield the market requires on all debt contracts, thereby lowering the price of all outstanding bonds (this is also referred to as interest rate risk, or market risk)
  2. Default*: Risk that the issuer will be unable to meet their contractual obligation to pay either coupon or principal
  3. Credit*: Risk that the issuer’s “creditworthiness” (e.g., credit rating) decreases, thereby rendering the return on the bond inadequate for the risk to the investor
  4. Liquidity*: Risk that bond holder will need to either sell the bond contract below original market value or mark it to market below original market value in the future
  5. Reinvestment: Risk that interest rates on U.S. treasuries fall, causing the yield made on any reinvested future coupon payments to decrease
  6. Inflation: Risk that the yield on a bond does not keep pace with inflation, thereby causing the real yield to be negative, despite having a positive nominal yield

*Given their importance, these risks will each be covered separately below.

Bond Risk One: Price/Interest, Rate/Market Risk

Historically, investors have primarily been concerned with interest rate risk on government bonds. That is because over the last 40 years, the general level of interest rates (their yield to maturity, or YTM) have declined globally, from a level in the early 1980s of 16% in the U.S., to today’s rates which approach zero (or even negative in some countries).

A negative yielding bond is no longer an investment. In fact, if you buy a bond with a negative yield, and hold it until maturity, it will have cost you money to store your “value.” At last count, there was close to $19 trillion of negative yielding debt globally. Most was “manipulated” government debt, due to quantitative easing (QE) by central banks, but there is negative-yielding corporate debt, too. Imagine having the luxury of being a corporation and issuing bonds where you borrowed money and someone paid you for the privilege of lending it to you.

Going forward, interest rate risk due to inflation will be one directional: higher. And due to bond math, as you now know, when interest rates rise, bond prices fall. But there is a bigger risk than this interest rate/market risk that is brewing for government bonds: credit risk. Heretofore, credit risk for governments of developed G20 nations has been minimal. However, that is starting to change…

Bond Risk Two: Credit Risk

Credit risk is the implicit risk of owning a credit obligation that has the risk of defaulting. When G20 government balance sheets were in decent shape (operating budgets were balanced and accumulated deficits were reasonable) the implied risk of default by a government was almost zero. That is for two reasons: First, their ability to tax to raise funds to pay their debts and, secondly and more importantly, their ability to print fiat money. How could a federal government default if it could just print money to pay down its outstanding debt? In the past, that argument made sense, but eventually printing money will (and has) become a credit “boogie man,” as you will see.

For the purpose of setting a “risk-free rate,” though, let’s continue to assume that benchmark is set by the federal government. In markets, credit risk is measured by calculating a “credit spread” for a given entity, relative to the risk-free government rate of the same maturity. Credit spreads are impacted by the relative credit riskiness of the borrower, the term to maturity of the obligation and the liquidity of the obligation.

State, provincial and municipal debt tends to be the next step as you ascend the credit risk ladder, just above federal government debt, thereby demonstrating the lowest credit spread above the risk-free rate. Since none of the entities have equity in their capital structure, much of the implied credit protection in these entities flows from assumed federal government backstops. These are certainly not guaranteed backstops, so there is some degree of free market pricing, but generally these markets are for high grade borrowers and low risk tolerance investors, many of whom assume “implied” federal support.

Corporates are the last step(s) on the credit risk ladder. Banks are quasi-corporates and typically have low credit risk because they are assumed to have a government backstop, all else being equal. Most corporations do not have the luxury of a government backstop (although lately, airlines and car makers have been granted some special status). But in the absence of government lobbying, most corporations have an implied credit risk that will translate into a credit spread.

“Investment grade” (IG) corporations in the U.S. market (as of February 17, 2022) trade at a yield of 3.09%, and an “option adjusted” credit spread (OAS) to U.S. treasuries of 1.18% (118 basis points, or bps), according to any Bloomberg Terminal where you might care to look. “High-yield” (HY) corporations, on the other hand, trade at a yield of 5.56% and an OAS of 3.74% (374 bps), also per data available through any Bloomberg Terminal. Over the past year, spreads have remained fairly stable, but since bond prices in general have fallen, the yield (on HY debt) has increased from 4.33%… Indeed, HY debt has been a horrible risk-adjusted return of late.

When I started trading HY 25 years ago, the yield was actually “high,” generally better than 10% YTM with spreads of 500 bps (basis points) and higher. However, because of a 20-year “yield chase” and, more recently, the Federal Reserve interfering in the credit markets, HY looks pretty low yield to me these days… but I digress.

Subjective Ratings

From the above, it follows that spreads are largely a function of credit risk gradations above the baseline “risk-free” rate. To help investors evaluate credit risk, and thus price credit on new issue debt, there are rating agencies who perform the “art” of applying their knowledge and intellect to rating a given credit. Note that this is a subjective rating that qualifies credit risk. Said differently: The rating does not quantify risk.

The two largest rating agencies are S&P and Moody’s. In general, these entities get the relative levels of credit risk correct. In other words, they correctly differentiate a poor credit from a decent credit. Notwithstanding their bungling of the credit evaluations of most structured products in the Great Financial Crisis (GFC), investors continue to look to them not only for advice, but also for investment guidelines as to what determines an “investment grade” credit versus a “non-investment grade/high yield” credit. Many pension fund guidelines are set using these subjective ratings, which can lead to lazy and dangerous behavior such as forced selling when a credit rating is breached.

For the life of me, I cannot figure out how someone determines the investment merits of a credit instrument without considering the price (or contractual return) of that instrument. However, somehow, they have built a business around their “credit expertise.” It is quite disappointing and opens the door for some serious conflicts of interest since they are paid by the issuer in order to obtain a rating.

I worked very briefly on a contract basis for Dominion Bond Rating Service (DBRS), Canada’s largest rating agency. I heard a story among the analysts of a Japanese bank who came in for a rating because they wanted access to Canada’s commercial paper (CP) market, and a DBRS rating was a prerequisite for a new issue. The Japanese manager, upon being given his ratingm, inquired, “If I pay more money, do I get a higher rating?” Sort of makes you think…

Regardless, rating scales are as follows, with S&P/Moody’s highest to lowest rating: AAA/Aaa, AA/Aa, A/A, BBB/Baa, BB/Ba, CCC/Caa and D for “default.” Within each category there are positive (+) and negative (-) adjustments of opinion. Any credit rating of BB+/Ba+ or lower is deemed “non-investment grade.” Again, no price is considered and thus I always say, if you give me that debt for free, I promise it would be “investment grade” to me.

Poor math skills are one thing, but adhering to subjective evaluations of credit risk are another. There are also subjective evaluations such as “business risk” and “staying power,” inherently built into these ratings. Business risk can be defined as volatility of cash flows due to pricing power (or lack thereof). Cyclical businesses with commodity exposure such as miners, steel companies and chemical companies have a high degree of cash flow volatility and therefore, their maximum credit rating is restricted due to their “business risk.” Even if they had low debt levels, they would likely be capped at a BBB rating due to the uncertainty of their earnings before interest. tax, depreciation, and amortization (EBITDA). “Staying power” is reflected in the industry dominance of the entity. There is no rule that says big companies last longer than small ones, but there is certainly a rating bias that reflects that belief.

The respective ratings for governments are also very, if not completely, subjective. While total debt/GDP metrics are a good starting point, it ends there. In many cases, if you were to line up the operating cash flows of the government and its debt/leverage statistics compared to a BB-rated corporation, the corporate debt would look better. The ability to raise taxes and print money is paramount. Since it is arguable that we have reached the point of diminishing returns in taxation, the ability to print fiat is the only saving grace. That is until investors refuse to take freshly printed and debased fiat as payment.

Objective Measures Of Credit Risk: Fundamental Analysis

In the case of corporate debt, there are some well-defined metrics which help to provide guidance for objectively evaluating credit risk. EBITDA/interest coverage, total debt/EBITDA and enterprise value (EV)/EBITDA are great starting points. EBITDA is essentially pre-tax cash flow. Since interest is a pre-tax expense, the number of times EBITDA covers the pro-forma interest obligation makes sense as a measure of credit risk. In fact, it was this metric that I had determined to be the most relevant in quantifying the credit risk for a given issuer, a finding I published in “Financial Analysis Journal” (FAJ) in March 1995. As I mentioned in part one, I had worked for Royal Bank of Canada (RBC), and I was well aware that all banks needed to better understand and price credit risk.

The article was titled “Quantifying Risk In The Corporate Bond Markets.” It was based on an exhaustive study of 23 years of data (18,000 data points) that I painfully accumulated at the McGill Library in Montreal. For our younger readers out there, this was before electronic data of corporate bond prices was available, and the data was compiled manually from a history of phonebook-like publications that McGill Library had kept as records. In it, I showed a nice pictorial of risk in the corporate markets. The dispersions of the credit spread distributions measures this risk. Notice, as the credit quality decreases the dispersion of the credit spread distributions increases. You can measure the standard deviations of these distributions to get a relative measure of credit risk as a function of the credit rating.

The data and results were awesome and unique, and I was able to sell this data to the RBC to help with its capital allocation methodology for credit risk exposure. The article was also cited by a research group at JPMorgan, and by the Bank for International Settlements (BIS).

It should be obvious by now that anyone who is investing in a fixed income instrument should be keenly aware of the ability of the debt issuer to honor their contractual obligation (i.e., creditworthiness). But what should the investor use to quantitatively evaluate the creditworthiness of the debt issuer?

One could extrapolate the creditworthiness of a corporation by assessing various financial metrics related to its core business. It is not worth a deep dive into the calculation of EBITDA or interest coverage ratios in this article. Yet, we could all agree that comparing a corporation’s periodic cash flow (i.e., EBIT or EBITDA) to its periodic interest expense would help to quantify its ability to repay its debt obligations. Intuitively, a higher interest coverage ratio implies greater creditworthiness.

Referencing the aforementioned article, the data proves our intuition:

EBITDA interest coverage ratio

Indeed, one could convert the above data into specific relative risk multiples, but for the purposes of this exercise, simply understanding the concept is sufficient.

Similarly, one can use some basic math to convert subjective ratings into relative credit risk. But first, realize that risk is related to both standard deviation and volatility as follows:

Risk is related to standard deviation and volatility

A glance at market data provides the standard deviation of the risk premium/yield spread for various credit rating categories, which then allows the calculation of relative risk.

The standard deviation of risk premium/yield for various credit ratings allows for the calculation of relative risk.

Therefore, as an example, if an investor wishes to purchase the debt of corporation XYZ, which has a credit rating of BB, that investor should expect a risk premium/yield spread of 4.25 times the current market yield for AAA-rated investment-grade debt (all other factors being equal).

Objective Measures Of Credit Risk: Credit Default Swaps

CDS are a relatively new financial engineering tool. They can be thought of as default insurance contracts where you can own the insurance on an entity’s credit. Each CDS contract has a reference obligation that trades in a credit market so there is a natural link to the underlying name. In other words, if CDS spreads are widening on a name, credit/bond spreads are widening in lock step. As risk increases, insurance premiums do, too.

Allow me to get into the weeds a bit on CDS. For those less inclined to do so, feel free to skip down to the italicized section… CDS contracts start with a five-year term. Every 90 days, a new contract is issued and the prior contract is four-and-three-quarters-years old, etc. As such, five-year contracts eventually become one-year contracts that also trade. When a credit becomes very distressed, many buyers of protection will focus on the shorter contracts in a practice that is referred to as “jump to default” protection.

The spread or premium is paid by the owner of the contract to the seller of the contract. There can be, and usually is, much higher notional value of CDS contracts among sophisticated institutional accounts, than the amount of debt outstanding on the company. The CDS contracts can thus drive the price of the bonds, not the other way around.

There is no limit to the notional value of CDS contracts outstanding on any name, but each contract has an offsetting buyer and seller. This opens the door for important counterparty risk considerations. Imagine if you owned CDS on Lehman Brothers in 2008 but the counterparty was Bear Stearns? You may have to run out and purchase protection on Bear Stearns, thereby pouring gas on the credit contagion fire.

I believe it was Warren Buffet who famously referred to CDS as a “financial weapon of mass destruction.” That is a little harsh, but it is not altogether untrue. The sellers of CDS can use hedging techniques where they buy equity put options on the same name to manage their exposure. This is another reason that if CDS and credit spreads widen, the equity markets can get punched around like a toy clown.

Many readers may have heard of the CDS. Although technically not an insurance contract, it essentially functions the exact same way: “insuring” creditors against a credit event. Prices of CDS contracts are quoted in basis points. For example, the CDS on ABC, Inc. is 13 bps (meaning, the annual premium to insure $10 million of ABC, Inc. debt would be 0.13%, or $13,000). One can think of the premium paid on a CDS contract as a measure of the credit risk of the entity the CDS is insuring.

In other words, applying the logic from Foss’ FAJ article described above, let’s estimate the relative CDS premiums of two corporate entities:

  1. ABC, Inc.: Credit rating AA+, EBITDA interest coverage ratio 8.00
  2. XYZ, Inc.: Credit rating BBB, EBITDA interest coverage ratio 4.25

For which entity would you expect the CDS premium to be higher? That’s right: XYZ, Inc.

It turns out that the difference between CDS premiums and risk premiums/yield spreads is typically quite small. In other words, if the market’s perception is that the credit risk of a given entity is increasing, both the CDS premium and the required yield on its debt will increase. Two examples from recent events highlight this point:

  1. Look at the recent fluctuations in CDS pricing on HSBC (a bank). It turns out HSBC is one of the main creditors of Evergrande (of Chinese real estate fame). According to my interpretation of historical CDS data, five-year CDS pricing on September 1, 2021 was 32.75 bps. Just over a month later, it had increased nearly 36% to 44.5 bps on October 11, 2021. Note: This was during the month of September that news of Evergrande’s impending collapse circulated.
  2. Turkey has been experiencing a well-publicized currency collapse of late. The one-month and s-month variance on five-year CDS pricing of Turkey’s sovereign debt is +22.09% and +37.89%, respectively. Note: The yield on the Turkish 10-year government bond currently sits at 21.62% (up from 18.7% six months ago).

One could argue that the most accurate way to assess credit risk is via tracking CDS premiums. They are neither subjective, nor are they an abstraction from financial data. Rather, they are the result of an objective and efficient market. As the saying goes: “Price is truth.”

This dynamic interplay between CDS premia and credit spreads is extremely important for corporate credit and it is a well-worn path. What is not so well worn, though, is CDS on sovereigns. This is relatively new, and in my opinion, could be the most dangerous component of sovereign debt going forward.

I believe inflation risk considerations for sovereigns will become overwhelmed by credit risk concerns. Taking an example from the corporate world, two years prior to the GFC, you could purchase a CDS contract on Lehman Brothers for 0.09% (9 bps), per historical CDS data. Two years later, that same contract was worth millions of dollars. Are we headed down the same path with sovereigns?

Think of the potential for long-dated sovereign bonds to get smoked if credit spreads widen by hundreds of basis points. The resultant decrease in bond value would be huge. This will cause many bond managers (and many economists) indigestion. Most sovereign bond fund managers and economists are still focused on interest rate risk rather than the brewing credit focus.

Moreover, the price of sovereign CDS premia effectively set the base credit spread for which all other credits will be bound. In other words, it is unlikely that the spreads of any institution or entity higher up the credit ladder will trade inside the credit spread of the jurisdictional sovereign. Therefore, a widening of sovereign CDS premia/credit spreads leads to a cascading effect across the credit spectrum. This is referred to as “contagion.”

So, I ask the reader again: Is the U.S. treasury rate really “risk free”? This would imply that the inherent credit risk is zero… yet, at present, the CDS premium on U.S. sovereign debt costs 16 bps. To my knowledge, 16 bps is greater than zero. You can look up CDS premia (and thus the implied default risk) for many sovereigns at WorldGovernmentBonds.com. Remember, price is truth…

Bond Risk Three: Liquidity Risk

What exactly is liquidity, anyway? It’s a term that gets thrown around all the time: “a highly liquid market,” or “a liquidity crunch,” as though we are all just supposed to know what it means… yet most of us have no idea.

The academic definition of liquidity is as follows: The ability to buy and sell assets quickly and in volume without moving the price.

OK, fine. But how is liquidity achieved? Enter stage left: Dealers…

Let’s imagine you own 100 shares of ABC, Inc. You would like to sell these 100 shares and buy 50 shares of XYZ, Inc. What do you do? You log into your brokerage account and place the orders… within a matter of seconds each trade is executed. But what actually happened? Did your broker instantly find a willing counterparty to purchase your 100 shares of ABC, Inc. and sell you 50 shares of XYZ, Inc.?

Of course they didn’t. Instead, the broker (i.e., “broker-dealer”) served as the counterparty in this transaction with you. The dealer “knows” that eventually (in minutes, hours or days) they will find a counterparty who desires to own ABC, Inc. and sell XYZ, Inc., thereby completing the opposite leg of the trade.

Make no mistake, though. Dealers do not do this for free. Instead, they buy your shares of ABC, Inc. for $x and then sell those shares for $x + $y. In the business, $x is termed the “bid” and $x + $y is termed the “ask.” Note: The difference..