# Debt capital markets: bonds, covenants & rating agencies
In May 2020, with the world locked down and credit spreads still wobbling, Apple's then-Treasurer Gary Wipfler walked into the market and printed $8.5 billion in bonds across five tranches. The 30-year piece priced at 2.65%. A few months later, in August 2020, Apple returned for $5.5 billion more, and the 40-year tranche landed at 2.55%, with one tranche carrying a coupon of just 1.65%. A separate $1 billion 10-year piece was placed at roughly 1.5%. Apple did not need the cash. It was sitting on nearly $200 billion of liquidity. What Apple needed was *duration at a generational low*.
Fast-forward to 2026. The 10-year Treasury is anchored around 4.2%, investment-grade spreads have normalized in the 95-110 bps range after the 2022-2023 dislocation, and CFOs refinancing 2020-vintage paper are watching all-in coupons triple. Apple's treasury team is being studied in business schools for a reason: they understood that DCM strategy is not about minimizing this quarter's interest expense, it is about *positioningpositioningThe mental space you want your brand to occupy in your target customer's mind relative to alternatives.Voir la définition complète → the liability stack across rate regimes you cannot forecast*.
This lesson will walk you through the architecture of the debt capital markets the way a senior CFO must see it: the IG/HY divide, the covenant package as a negotiated instrument, the rating agencies as quasi-regulators, and the strategic playbook you should be running on Monday morning.
The market splits at BBB-/Baa3. Above it: investment grade, where pension funds, insurance companies, and global aggregate index buyers are forced buyers. Below it: high yield, where the buyer base shrinks to dedicated HY funds, CLOs, and credit hedge funds, and where covenant packages get materially more restrictive.
The pricing differential is not linear. In the current 2026 market, the jump from BBB- to BB+, a single notch, typically costs 130-180 bps of spread, depending on sector. That is not a credit story. That is an *index inclusion* story. When Ford was downgraded to junk by S&P in March 2020, roughly $36 billion of its debt fell out of IG indices overnight, forcing mechanical selling regardless of fundamentals. Ford's CFO at the time, Tim Stone (and his successor John Lawler), spent the next three years executing one of the most disciplined "rising star" campaigns in corporate finance, regaining IG status from Moody's in November 2023 and S&P in late 2024.
Investment grade. Buyers are duration-matchers. They want size ($500M minimum for benchmark status, $1B+ for true liquidity), simple structures (senior unsecured, bullet maturity), and minimal covenants, typically just a change-of-control put at 101 and a limitation on liens. They will not negotiate hard on covenants because they are buying a curve, not a credit story.
High yield. Buyers are credit pickers. They want yield, structural protections, and a story. The covenant package, what bankers call the "incurrence covenants", becomes a heavily negotiated instrument. Restricted payments baskets, debt incurrence ratios, asset sale sweeps, and the infamous "J.Crew" and "Chewy" trapdoors that allow IP and subsidiary transfers out of the credit group are all on the table.
Roughly $3.5 trillion of US corporate debt sits in the BBB tier as of early 2026, about half of the entire IG market. This is the "BBB bulge" that regulators have been worried about since 2018. For a CFO running a BBB- credit, the strategic question is binary: *do you defend IG at the cost of growth optionality, or do you accept HY and use the covenant flexibility to run a more aggressive playbook?*
AT&T's John Stankey and CFO Pascal Desroches chose the former path post-WarnerMedia spin in 2022, explicitly committing to deleverage to 2.5x net debt/EBITDAEBITDAEBITDA (Earnings Before Interest, Taxes, Depreciation, and Amortization) measures a company's operating profitability before financing and accounting decisions, used to compare core performance across firms.Voir la définition complète → to protect BBB ratings. Kraft Heinz, by contrast, accepted a BB+ rating in 2020 after the Buffett-3G writedowns and used the flexibility to restructure without IG-buyer pressure.
Junior bankers think the coupon is the price. Senior CFOs know the covenant package is at least half the price. In 2026, with private credit funds offering covenant-lite structures at premium pricing, the negotiation has shifted, but for public bond issuers, the indenture remains the document that will govern your strategic flexibility for the next 10-30 years.
Maintenance covenants are the bank loan world: you must maintain a leverage ratio below X at every quarter-end, or you are in default. They force quarterly conversation with lenders and create real cliff risk.
Incurrence covenants dominate the bond world: you must meet the test *at the moment you take an action* (issue more debt, pay a dividend, sell an asset). If you do nothing, you cannot trip them. This is why HY bonds are structurally more flexible than term loans, and why the leveraged loan market's shift to "cov-lite" since 2015 has made loans look more bond-like.
When your treasurer brings you a draft indenture, three baskets will determine your operational flexibility:
1. Restricted Payments (RP) basket. Governs dividends, buybacks, and investments in non-guarantor subsidiaries. Built as a "builder basket", 50% of cumulative net income since issuance, plus a starter amount. *This is the basket that gets you sued.* Revlon's 2020 transfer of IP to an unrestricted subsidiary used RP capacity creatively and triggered years of litigation.
2. Debt Incurrence basket. Typically a 2.0x fixed charge coverage test, plus general baskets. Determines how much secured and structurally senior debt you can layer on top.
3. Permitted Investments / Unrestricted Subsidiary capacity. The "trapdoor." Allows assets to move *outside* the credit group, where existing bondholders have no claim. Post-J.Crew (2016) and PetSmart/Chewy (2018), sophisticated bondholders demand explicit blockers, but mid-market issuers often still have the flexibility.
In a 2025 BB rated bond I'd advise a client to focus negotiation on three points: (1) a larger general RP basket (push from $50M to $150M for a $2B EBITDAEBITDAEBITDA (Earnings Before Interest, Taxes, Depreciation, and Amortization) measures a company's operating profitability before financing and accounting decisions, used to compare core performance across firms.Voir la définition complète → company), (2) portability, allowing the bonds to survive a change of control without a put, which is worth 15-25 bps of spread to a PE-owned issuer, and (3) the "available amount" definition, which should include equity contributions and asset sale proceeds not otherwise required for reinvestment.
Vérification des acquis
1. In August 2020, what was the coupon on Apple's 40-year bond tranche that has since become a case study in DCM strategy?
2. According to the lesson, what is the approximate spread cost in the current 2026 market of dropping a single notch from BBB- to BB+?
3. Which investor type is typically a primary buyer in the high-yield market but NOT in the investment-grade market?
4. Select ALL correct statements about Apple's 2020 bond issuances as described in the lesson.
Sélectionnez toutes les réponses correctes.
5. Select ALL correct statements about the structural differences between investment grade and high yield markets.
Sélectionnez toutes les réponses correctes.
Moody's, S&P, and Fitch are not your friends. They are not your enemies. They are quasi-regulators with published methodologies and analysts who are evaluated on the accuracy of their ratings. Your job as CFO is to understand the methodology better than the analyst does.
Each agency publishes its corporate methodology, and they are publicly available, dense, and remarkably specific. S&P's framework starts with a Business Risk Profile (industry risk, country risk, competitive position) crossed with a Financial Risk Profile (cash flow/leverage metrics) to produce an "anchor" rating, which is then adjusted by modifiers (diversification, capital structure, liquidity, financial policy, management/governance, comparable rating analysis).
The financial risk profile is where the math lives. For a "Significant" financial risk profile, typically BBB territory for a standard industrial, S&P wants FFO/Debt of 20-30% and Debt/EBITDAEBITDAEBITDA (Earnings Before Interest, Taxes, Depreciation, and Amortization) measures a company's operating profitability before financing and accounting decisions, used to compare core performance across firms.Voir la définition complète → of 3.0-4.0x. Drop FFO/Debt below 20% on a sustained basis and you are negotiating against a downgrade.
Moody's uses a similar scorecard with explicit weightings, for a typical manufacturing company, leverage (Debt/EBITDAEBITDAEBITDA (Earnings Before Interest, Taxes, Depreciation, and Amortization) measures a company's operating profitability before financing and accounting decisions, used to compare core performance across firms.Voir la définition complète →) and coverage (EBITA/Interest) together account for roughly 30-35% of the scorecard.
Twice a year, your CFO and Treasurer sit across a table (or Teams call) from a lead analyst and a backup. The analyst has read your 10-KKThe average number of new users each existing user generates through referrals. Above 1.0, growth compounds on itself and becomes exponential.Voir la définition complète →, your competitors' 10-Ks, and your transcripts. The conversation that matters is *forward-looking*: capital allocation policy, M&A appetite, willingness to issue equity in stress, hedge book composition.
The single most powerful thing a CFO can do is articulate a financial policy framework with explicit numerical commitments. When Microsoft's Amy Hood committed to maintaining AAA-equivalent metrics through the Activision acquisition, S&P's reaffirmation at AAA in 2023 was not a gift, it was a response to a credible policy framework backed by $80B+ of annual operating cash flow.
Every CFO should know the answer to: *what specifically would cause a one-notch downgrade, and what would cause two notches?* If you cannot recite the agency's downgrade triggers, "sustained Debt/EBITDAEBITDAEBITDA (Earnings Before Interest, Taxes, Depreciation, and Amortization) measures a company's operating profitability before financing and accounting decisions, used to compare core performance across firms.Voir la définition complète → above 3.5x" or "FFO/Debt below 18%", you are not running the credit. The credit is running you.
Boeing's saga is instructive. Following the 737 MAX grounding (2019), pandemic (2020), and Alaska Air door-plug incident (January 2024), Moody's downgraded Boeing to Baa3 in April 2024, one notch above junk. CFO Brian West and the team executed a $24 billion equity raise in October 2024 specifically to defend IG status, the largest equity issuance by a non-financial corporate in over a decade. The dilution was painful. The cost of falling out of IG indices on roughly $50 billion of debt would have been catastrophically worse.
Tactical issuance, when to print, what tenor, what coupon, is your Treasurer's job. *Strategic* DCM is your job. Here is the framework.
Pull your debt maturity schedule. If more than 20% of your debt matures in any single year, you have unhedged refinancing risk. The 2026 refinancing wall, roughly $1.8 trillion of US corporate debt maturing in 2026-2028, much of it issued at 2020-2021 coupons, is the defining DCM event of this decade. Companies that termed out in 2020-2021 (Apple, Microsoft, Alphabet, Verizon) are running 200-300 bps below current market rates on their long end. Companies that relied on revolvers and 3-5 year notes are now refinancing into a 5-6% all-in cost environment.
Tender offers, exchange offers, and open-market repurchases are not crisis tools, they are continuous portfolio management. When your bonds trade at 85 cents on the dollar because rates have risen, a tender at 90 creates accounting gain *and* reduces gross debt. Verizon executed roughly $