Look at 30-year hyperscaler debt (which has widened more than 3 standard deviations this month). First it was Nvidia which brought $25bn of new supply. Then SpaceX with $25bn the week after. Amazon also issued $25bn after that, and those are trading roughly 50bp back in secondary.

AI-related issuance across IG, high yield and loans now sits at a whopping $489bn YTD vs. $322bn for ALL of 2025. It’s also 23% of USD IG gross supply this year. The 6 biggest tech names carry more risk in the US high-grade index than the 6 biggest banks (the first time that has ever happened).

Base rates are clearly doing damage underneath it. 30 year treasuries are above 5% with real yields approaching 3%, so long-duration credit reprices regardless of what any individual issuer does. Both the governments AND hyperscalers are competing for the same long-end buyer. This is about absorption, not fundamental deterioration.

The more interesting part worth thinking about is what is actually happening underneath.

Project finance paper will not necessarily price on its own. A data center bond leased to Meta gets quoted off Meta’s curve + a premium for structure, tenor and complexity. So when the tenant’s curve widens on supply, the project paper widens with it, whether or not anything about the project changed. Take Hyperion for example, the $27bn Meta package done with Blue Owl last October, priced at par, traded above 109, and sits around 95 now.

Of course some of the move is just duration, some is supply, and the price alone tells you almost nothing about how much of it reflects the structure.

Which is the problem. Everything is widening together but very little is being separated in the move. When rates and supply are pushing the whole complex around, price stops telling you what you own. The terms are the only thing left that can, and the premium over the tenant’s curve is supposed to compensate you for whatever sits between you and the tenant.

That is why the recent move is more interesting. Project-finance AI bonds are widening mechanically with their tenants, but the market is still not distinguishing between structures that place completion risk on the hyperscaler and those that leave it with the bondholder.

Besides tenant counterparty risk, what bondholders should be asking is what happens if completion slips and who has to fund the gap? Right now, the market is not paying nearly enough attention to the difference.

Free Issue Sponsored by 9Fin:

New report: Where is risk building across BDC portfolios?

9fin has analysed filings from 157 BDCs to produce an industry-first watchlist of potentially at-risk loans.

The report identifies:

• 468 individual loan positions showing signs of potential stress

• $5.7bn of loans at fair value

• $1.2bn of value erosion from $6.9bn of par value - roughly 17%

See which borrowers and lenders are exposed, where value is deteriorating and where pressure may emerge next.

Request Access

The Same Word Has Different Meanings

Every lender deck in this sector says they have a lease that is “contracted” but “contracted” means very different things depending on who the issuer is.

Take the Meta (Sopaipilla) deal from last week because it shows what the strong version looks like. This is a $12.3bn senior secured note due November 2048, funding a 960MW campus in El Paso. Construction started October 2025 with rent commencing in December 2028. It’s about 3 years and change before the first dollar of revenue actually hits.

Here is what the documents actually say. Meta begins to pay rent from December 2028 whether or not the campus is finished. An abatement applies ONLY for force majeure, a landlord-caused delay, or if landlord tells the contractor to stop. The whole thing is capped at $218mm against the delay start-up insurance. Meta also agrees to cover cost overruns past 105% of the fixed budget. This is why the transaction is not exposed to typical construction risk. The tenant is taking that.

Beignet, the earlier Meta deal, worked the same way. Date-certain rent commencement.

Now the other end of the same cohort. The Hut 8 transactions, River Bend and Beacon Pointe, were the first where cash flows don’t commence until the data center is actually complete. Rent starts on delivery, not on a date. Completion risk moved from the tenant to the bondholder, in named deals, inside twelve months.

That isn’t just a slightly weaker version of the same bond. In the first structure you’re primarily exposed to the tenant’s contractual payment obligation, subject to whatever structural and enforcement risk sits above it. In the second you’re also exposed to construction execution, interconnection, equipment delivery, customer acceptance, and the project’s ability to stay liquid until rent begins.

What hasn’t happened is the same distinction being remade in the recent move. Tenant curves widened on supply, project paper followed mechanically, and structurally opposite deals moved together. A broad supply-driven selloff won’t separate them for you.

Why The Difference Matters

Because a signed lease is not revenue. Just like how a planned megawatt is not capacity or a hyperscaler tenant is not a powered site. Until a facility is fully energized, accepted AND billing, you’re lending against the conversion of a development project into an operating asset.

The reason this all matters is because the delivery dates will inevitably slip. Goldman expects only around half of scheduled capacity to arrive on time over the next two years.

From a credit perspective, a delay is not inherently a credit default. Nothing in the indenture trips because the interconnection date moves by 6 months.

The consequences rest with what’s been negotiated in the lease agreement. Commencement mechanics, delay penalties, rent deferrals, acceptance tests, customer remedies, termination rights, etc.

A delayed project can push out when rent starts, trigger customer remedies, and force amendments without ever having missed a coupon. The project doesn’t need to fail either for the bond economics to deteriorate. They can arrive late enough that the original cash-flow assumptions stop working. The difference is a hyperscaler can absorb that on its balance sheet; while the others have potentially limited liquidity.

There’s currently about $45bn of datacenter high yield paper that will mature between 2030 and 2031. If we take a typical 5-year bond issued against an asset expected to energize only in 2028, this leaves only 2 years of operating history before actual maturity. In practice, refinancing discussions typically start 12 to 18 months early, so the next lender will have only 6 to 12 months of stabilized performance in their underwriting. Now if the delivery date slips by 1 year, almost none of that remains. The delay ate into any refinancing cushion.

What Completion Risk Looks Like

Completion risk is easiest to understand by example. This one is worth watching precisely because it is not a marginal project.

Project Jupiter sits on 1,400 acres in Doña Ana County, New Mexico comprising 4 buildings, up to 2.45GW, and up to $165bn of announced investment. STACK Infrastructure and BorderPlex Digital Assets are building it, Blue Owl is the equity, and Oracle is the tenant.

The campus is designed to be fully powered by an on-site microgrid rather than relying on the grid. In late April, Oracle and BorderPlex said the energy plan had changed so that Bloom Energy fuel cells would replace the gas turbines and diesel generators originally proposed. A cleaner permitting story.

This is where it gets complicated. Energy Transfer’s Green Chile Project, a 17-mile extension, needs rights-of-way across state trust land. In March the State Land Office canceled applications for 2 crossings and a meter-station lease. Energy Transfer requested reconsideration and the Land Office denied that on July 14.

This is not a pipeline that has been rejected outright admittedly. The current route still lacks two state-land crossings and a meter-station lease, and the stated August 15 in-service date is now weeks away. Oracle’s own FERC submission said that missing that date would substantially increase costs and could jeopardize the campus operating timeline.

That does not prove the project misses its customer delivery date. However, the original delivery schedule appears increasingly hard to reconcile with the public record, which is a different and more useful observation.

Notice what is not failing. The construction is moving and the buildings are going up. What has run into trouble is a second, separate piece of infrastructure the first one cannot operate without. Project-on-project risk, where permitting, land access and construction all have to land twice. The building can be ready while the power is not.

What this project shows is that delivery risk can be visible in permitting and land records long before it appears in an issuer’s reporting. Importantly, sponsor quality and deep capital do not eliminate it. Instead, they just determine how much room a project has to absorb it, and who ultimately bears the cost.

I do think it’s important to highlight that the current issuance cohort is too young to offer much realized loss history. Most of the paper was issued last year against assets expected to energize in ‘27/’28 so the first test is on the come. That doesn’t mean you shouldn’t differentiate now. It just makes the documents more important, because the lease mechanics, the reserves and the completion schedule are the only evidence that exists before operating performance does.

In fact, industry participants are saying this out loud. CoreWeave’s VP of strategy made essentially the same point: compare the announced delivery date with how long the project should actually take to build, then mark the date on a calendar. Before there is loss history, the calendar is the early warning system.

New Mexico is also not the only jurisdiction where this happens, and in places the slowdown is deliberate. New York’s temporary pause covers any facility or group of facilities on the same or contiguous sites that can consume at least 50MW, so splitting a development into phases does not necessarily get you around the threshold. Fourteen states have considered statewide moratoria. New York is the first to impose one.

For a hyperscaler, a delayed project may be simply a portfolio management problem. Capacity can move to another site or another market. For a levered developer however, the same delay can consume interest reserves, push revenue to the right and become a liquidity problem. The same physical event can produce a completely different credit consequence depending on who owns the delay.

What Bondholders Should Underwrite

When does rent begin? A fixed contractual date provides very different protection from a lease where rent is contingent on completion, energization or customer acceptance. That nuance separates Sopaipilla from River Bend. In one structure the tenant starts paying even if the project is late. In the other, the project has to keep funding itself until the facility is delivered.

Next is who funds the delay. Look for the rent-abatement cap, the cost-overrun threshold, delay insurance, reserve sizing and any right the tenant has to terminate. An IG hyperscaler does not protect the bondholder from schedule risk if the project company still has to carry the interest, construction costs and lost revenue until completion.

Some additional non-obvious items you should be sure to ask when doing your diligence:

• How much of the long-lead equipment has actually been ordered, and how much is on site?

• Where does the project sit in the interconnection queue, or if it runs behind the meter, where does the on-site generation stand in its own permitting?

• What still has to be permitted, financed and built that is not the data center itself?

• What remains outstanding on the construction payment schedule?

• What triggers the tenant’s termination rights, and how much cushion is there

• How much room is left between the announced delivery date and the first refinancing conversation?

A campus that is energized and operating is a different credit from one carrying 2 more years of construction and an unresolved power dependency, even when the same name appears on both leases.

That brings this back to where it started. Spreads across the AI complex are wider, and much of the move can be explained by supply, duration and project paper following its tenant’s curve. Those forces reprice the entire complex, but they do not tell you which bond is carrying the better structure.

That main difference still has to come from the documents. A bond where the tenant has a fixed date rent payment and a bond where the project funds itself until delivery may share the same tenant and sector but they are not the same credit risk profile.

Microsoft, Meta, Amazon and Apple all report in the coming week, and everyone will be watching the capex line. But capex only tells you how much money is being committed. It does not tell you when the capacity turns on, how much liquidity gets consumed before it does, or who owns the delay when it does not.

JBI Bulletin Board

1) Now Hiring: Credit Analysts

I’m hiring credit analysts to add capacity. If that’s you, or someone you’d vouch for, reply with a short note on your background.

If you liked this, you might also like:

Multi-baggers don’t really exist in this asset class. The math fights you: your upside is capped at par plus coupons, and anything trading far enough below par has usually earned the discount by dying. I’ve seen it happen a handful of times.

This is one of them. A bond I wrote up three years ago at the lows is now a five-bagger on price, and counting coupons, roughly 6x your money.

The only question remaining: is there juice still left in the trade?

Putting this behind paywall early as it is illiquid despite being exchange traded and every cent matters for this trade.

Read more

Baby bonds are a retail product: sold in $25 increments, listed on an exchange and often bought by investors screening for coupons. This one pays 7.50%. I’d bet most buyers never opened CION’s schedule of investments, and a security priced without one gets mispriced in both directions.

Sometimes the panic is wrong and the paper is cheap. Sometimes the coupon does the selling and the paper is rich.

Which one is this? I thought it would be interesting to find out.

Background

CION Investment Corporation (“CION”) is an externally-managed BDC lending to mid-market companies in the US. Mostly senior secured loans plus a collection of equity and restructuring positions accumulated through prior workouts.

The past year has not been kind to the portfolio. Fair value went from ~$1.79bn in 1Q’25 to $1.70bn in 1Q’26, the # of companies from 104 to 89, net assets from $757mm to $660mm, and NAV per share from $14.28 to $13.11.

And lately it’s been getting worse faster: $59mm of unrealized losses in 4Q’25, another $36mm in 1Q’26, gross debt-to-equity now at 1.78x. Asset coverage sits at 156% against a 150% requirement.

Credit quality tells the same thing. During the quarter, non-accruals rose from 4.32% to 5.35% of cost, the internally rated 3-through-5 bucket reached 15.0% of fair value, and borrower interest coverage slipped from 2.26x to 2.08x. At fair value, non-accruals are just 1.53%, and that low number should not reassure anyone. It’s low because the weakest positions have already been written down substantially.

Earnings quality is deteriorating too. PIK interest ran ~22% of total investment income in the first quarter, and after financing and operating costs, most of what remained of earnings arrived as accruals rather than cash.

Unsurprisingly, the reduced dividend is not covered by current quarterly NII, and recurring cash earnings fall far shorter. 1Q’26 NII of $0.25 vs. a $0.30 payout, and if you exclude one unusually strong quarter (3Q’25), coverage has been at or below 1.0x in 4 of the past 5.

Management has bought back some stock though. $9.7mm during the quarter at an average of $8.71, accretive to NAV at reported marks, but eating into liquidity. This comes at a time when the book is shrinking, with ~$265mm funded against ~$356mm of repayments/sales over the LTM, netting to $91mm of runoff.

Against that backdrop, CION sold $135mm of 7.50% unsecured notes due 2031 in February 2026 and used part of the proceeds to repay $100mm on its JPM facility. Longer duration, less secured debt, fine as far as it goes. But it doesn’t go very far, because $115mm still matures in 2026 and more than $500mm in 2027.

Portfolio Overview

Before the numbers, it’s worth knowing what CION actually owns.

The book comprises 89 companies across 23 industries, mid-market borrowers with $25mm to $75mm of EBITDA (median of $35mm).

Sector exposure is diversified but not featureless. Business services leads at 14.0%, healthcare at 11.8%, then retail at 10.7%, electricity at 8.4%, and diversified media at 6.8%, with everything else below 6.7% each. Business services and healthcare are standard direct-lending. Double-digit retail is less standard, and given where the marks sit, worth keeping in mind when we get to the near-term maturity cohort.

The origination model matters for reading the marks too. Alongside directly originated and club loans, CION runs an opportunistic strategy: buying illiquid, lightly syndicated 1L paper at discounts created by ratings or technical pressure. There’s also a joint venture with EagleTree Capital, formed in late 2021, that pursues higher-yielding and junior-capital deals.

One more piece of context on the situation: the market has already voted. CION trades at 0.47x NAV against 0.74x for the BDC universe, carries 162% net leverage against a peer average of 119%, and shows non-accruals at cost of 5.35% against 3.50% for peers. The trailing distribution yield is 20%+. A yield like that is the market telling you it doesn’t believe the payout, the NAV, or both.

Estimating the Portfolio’s Fair Value

Management reports $1.70bn of fair value (excluding $97mm of short-term Treasuries) against ~$1.87bn of cost, for NAV of $660mm, or $13.11 per share. I treat reported fair value as the management case, not as an independently underwritten recovery value.

Here’s how I go about re-marking the current portfolio:

Read more

That split is starting to show up across the market. Shareholders want the hyperscalers to move fast, defend their position, and spend before a competitor does. Bondholders are financing the same strategy, but their upside stops at the coupon while the funding burden keeps rising.

AI is creating different winners and losers inside the same capital structure: the company can become more valuable while the bonds get cheaper.

The Former Tech Credit Bargain

For most of the past decade, the largest technology companies were almost perfect bond issuers. They generated enormous FCF, held large cash balances and did not need much physical capital to grow. Debt came to market occasionally, usually for opportunistic reasons, and investors paid up for the scarcity.

The equity and credit stories moved together. Stronger growth produced more cash, the cash strengthened the balance sheet and the balance sheet supported tighter spreads. Management could invest heavily without forcing creditors to absorb a completely different funding profile.

AI is breaking that arrangement.

The buildouts need data centers, chips, networking equipment, power generation, etc. The companies that once sat at the asset-light end of the corporate credit universe are becoming some of its largest capital spenders. Current capex estimates at 3.2% of US GDP in 2027 and direction is clear. These businesses are becoming structurally capital intensive.

The financing has followed. AI infrastructure credit has grown into a roughly half-trillion-dollar market in less than 2 years across hyperscaler bonds, data-center financings, and leveraged credit (I mapped the market here). This issue is about a narrower question: who earns the return on all that capital, and who is simply providing it?

The Same Decision, Different Payoffs

Suppose Meta announces another $100 billion of AI spending. Equity investors may like the decision because it signals that management sees a large market and plans to move before competitors close the gap.

If the spending works, shareholders get the benefit. Revenue grows, the competitive position improves and the equity may re-rate. The bondholder receives the same coupon that was already promised.

If the returns take longer, the bondholder still financed the build. FCF falls, issuance rises and the existing curve has to absorb more paper. Meta can remain one of the safest companies in the market while its bonds widen anyway.

Traditional credit analysts are trained to ask whether the company can pay. With these issuers, the answer is usually obvious. Repayment does not settle whether the bond is attractive at its current price. A larger opportunity gives management a reason to spend faster, which pushes FCF lower and brings forward the next financing. The announcement that improves the equity story can weaken the bond technical on the same day.

Management may be doing exactly what it should do. A hyperscaler that protects near-term FCF by refusing to invest may lose its position. A hyperscaler that spends aggressively may build a much stronger company and still leave existing creditors worse off.

This is different from the traditional shareholder-creditor conflict. Historically, the transfer was easy to spot: a sponsor extracted a dividend, leverage rose and cash left the company. Here, the cash stays inside the business and may create real long-term value. Existing creditors can still underperform because most of that value belongs to the equity.

Good management is normally treated as an uncomplicated positive in credit analysis. This cycle makes that harder. The strongest management team may be the one most willing to tolerate years of lower FCF and repeated issuance to protect the franchise.

We can see that in recent performance. Meta and Nvidia equities rallied while their bonds widened. SpaceX’s long bonds fell roughly ~8 points shortly after issuance despite a massive equity cushion and IG ratings.

Amazon completed a $25 billion bond package while investors reassessed the rest of the hyperscaler complex around it. Those moves fit the capital structure: equity owns the payoff if AI works, while credit funds the route to get there.

The Capital Allocation Arms Race

Credit investors now have to underwrite several years of capital allocation, not only the current business. Revenue, margins, leverage and coverage still matter, but they can be overtaken by a spending plan that changes quarter by quarter.

The next few years of capex need to be treated almost like a separate business line.

• How much spending is required to stay competitive?

• How long before it produces cash?

• How much debt comes first?

The current leverage ratio tells you very little if the ratio is about to change.

And the competitive dynamic makes the analysis harder. Amazon does not set its datacenter budget in isolation. Microsoft, Alphabet, Meta and Oracle are all reacting to one another, and no management team wants to be the first to decide it has enough capacity.

Each company may be making a sensible decision, while the group spends and borrows more than any one participant would have chosen in a calmer market. No management team wants to learn 3 years from now that the competitor who overspent was right. Rational behavior at the company level can still produce a poor outcome for the sector’s existing creditors.

Credit has seen this pattern before. Telecom demand in the late 1990s was real, and the internet became more important than even the bulls expected. Plenty of bonds financing the buildout still performed badly because capital arrived faster than the economics. AI does not need to fail for creditors to lose money; they may simply be funding several years of expansion at spreads that assumed the cash returns would show up sooner.

The Frequent-Issuer Discount

Large banks provide the familiar precedent. JPMorgan or Bank of America can be excellent credits and still trade wider than similarly rated industrial companies because investors know another bank deal is coming after earnings, followed by another one later in the year.

Recurring supply removes scarcity value. Buyers do not need to chase today’s bond when a newer, more liquid issue may arrive with a concession a few months later.

Hyperscalers are beginning to develop the same problem. They do not need to issue continuously to survive, but they may need to issue continuously to compete.

Take an Amazon bond trading at T+65. Amazon announces another large financing, and the new bond needs to clear at T+80. The old bond is immediately expensive against the new one and reprices even if Amazon’s earnings, ratings and leverage are unchanged.

The next deal then hangs over the curve. Investors know another concession may be coming, so the outstanding bonds do not tighten back to the prior level. Future issuance creates a higher spread floor.

This is no longer a hypothetical future calendar. 7 companies are already responsible for a remarkable share of the market’s new supply, and the spending estimates imply they are only getting started. Scarcity was part of the old technology-credit bargain. Recurring benchmark issuance is replacing it.

An exceptional balance sheet protects repayment, but it also gives management the capacity to borrow far more than a weaker company could. That capacity can become a technical problem for existing creditors.

Each new transaction makes the old bonds less scarce, and default risk is only one source of loss. If Treasury yields rise by 30bps and the credit spread widens another 50, a bond with 14 years of duration can lose more than 11 points. The analyst can be right about the company and wrong about the investment, while the issuer remains overwhelmingly money-good.

The Relative-Value Reset

This supply wave is changing the high-yield hurdle too. Amazon recently borrowed at an all-in yield of 5.3% for 10-year. This compares to BBs which yield at ~6% at the index level.

A long-duration BB issuer may offer only a modest pickup over Amazon while introducing real downgrade, default and recovery risk. Buying the hyperscaler looks obvious until the supply overhang is included.

Moving up in quality does not solve the trade automatically. High yield looks expensive vs. the all-in yield available in IG. Hyperscaler debt has stronger fundamentals, but the bonds may remain cheap because another large deal is always coming.

The comparison also affects unrelated borrowers. A portfolio manager offered Amazon, Meta, bank paper and data-center project finance at yields above 5% has less reason to stretch into a marginal BB credit.

That BB company does not need to deteriorate for its refinancing cost to rise. The alternative simply became more competitive. AI financing can tighten conditions elsewhere in credit by consuming the same portfolio capacity.

Hiring in Credit?

What the Week Said

HY tightened 4bps to 263bps last week. Meta and Nvidia shares rallied as their bonds weakened, and long-duration paper from SpaceX, Meta and Oracle continued to reprice wider. The pressure was concentrated in issuers asking the market to finance the AI buildout rather than reflecting a broad flight from corporate credit.

8 of the 12 largest drags on the IG index this year are companies financing the AI buildout. Meta alone has detracted more from index returns than any other issuer. The sector financing the most exciting trade in markets is the sector losing the most money in credit.

According to PitchBook LCD, AI-related IG issuance reached $218 billion through July 8, vs. $80.5 billion during all of 2025. AI-related HY issuance added another $31.9 billion.

Amazon’s $25 billion deal cleared, but its size became part of the valuation problem for the rest of the sector. The buildout remains intact, but the market is charging more to finance it.

What to Watch

The next phase depends on whether cash returns begin catching up with capital deployed. Upcoming earnings calls at the end of this month should be read for incremental AI revenue, margins,and FCF rather than another increase in capex alone.

New-issue concessions will be equally important. If each large transaction needs wider pricing than the previous one, the market is building a structural frequent-issuer discount rather than working through a temporary supply glut. If concessions compress over the next two quarters and the curves grind back toward their old levels, this was digestion and the thesis was early.

Concentration may become the harder constraint. An investor can like Amazon, Meta, Microsoft and a portfolio of data-center bonds individually, then discover that all of them depend on the same capex cycle, power buildout and demand assumptions.

The real ratio worth tracking is incremental AI cash flow against incremental AI financing. Until the first number starts catching the second, every increase in strategic confidence gives equity another catalyst and credit another supply estimate. Hyperscalers can afford the buildout; the open question is whether their existing bondholders are being paid enough to finance it.

JBI Bulletin Board

1) Something New is Coming

The institutional version of JunkBondInvestor is coming. The list gets in first and locks the founding rate before it closes. Add your name here.

2) Now Hiring: Credit Analysts

I’m hiring credit analysts to add capacity. If that’s you, or someone you’d vouch for, reply with a short note on your background.

If you liked this, you might also like:

JunkBondInvestor is an independent financial research publication providing general market commentary, corporate credit analysis, and educational content. The information provided is for informational purposes only and does not constitute investment advice, financial planning, or personalized recommendations of any kind. Nothing published constitutes a solicitation, offer, or recommendation to buy, sell, or hold any securities, nor does it guarantee any financial outcome. Any credit opinions or outlooks expressed are solely the independent views of the authors and are intended to reflect general credit trends, not specific investment recommendations. Investing in high-yield bonds, distressed debt, and leveraged loans carries significant risk including potential loss of principal. JunkBondInvestor operates under the Publisher’s Exemption of the Investment Advisers Act of 1940 and applicable state laws.

Numbers like that usually mean one of two things: the market is missing something, or the calculation is hiding something. In this case it’s a third thing, and it’s more interesting than either.

One housekeeping note. The security is exchange-listed but thinly traded, so this is mainly a PA trade idea. Because the situation is sensitive to size, I’m putting this one behind the paywall early. I’ll open it up in a few weeks.

Read more

The lease is the clean part. The tenant is IG, the rent is contracted, and the cash flows are visible. The residual is harder. It is the building and what that building is worth once the contracted rent runs out.

That first piece was about measuring how much of the bond depends on the residual. This one is about whether that residual is money-good when the bond actually needs it. A building can look valuable on day one, carry an appraisal, and sit behind a flawless tenant, but still fail the refinancing test if the next generation of compute needs a different physical asset.

Density Requirements Are Increasing

Much of today’s AI data center capacity is being designed around the density requirements of the current generation. The rough underwriting case is a facility that can support something like 200kW per rack, which works for today’s hardware but may not work for the next refresh cycle. The next turn points much higher, with density moving toward 1.2MW to 1.5MW per rack, and the ceiling keeps moving.

That is not a change you absorb with a maintenance budget. Carrying a building across that gap can mean heavier foundations to bear the load, a different cooling architecture, and a rebuilt power distribution system to push several times the power onto the same floor. The cost can approach a new build, and the timeline can run to years.

The cadence is as much of a problem as the magnitude. A bond financed over 10 or more years has to outlast not one hardware generation but several, each denser and hotter than the last. The chips turn over every few years, while the building that houses them is being asked to stay current across the entire life of the debt, on a capex budget that assumed it would not have to.

Some of what is being financed today may prove to be single-generation infrastructure. It can house the hardware it was designed for, but the hardware after that may not fit without a rebuild. The lease covers the first generation, while the residual quietly assumes a second.

The Retrofit Is A Timeline Risk

Obsolescence does not always mean the building is worthless. Sometimes it means the building needs a second construction project before it can be leased or financed again. That distinction matters for credit, because a retrofit may be technically possible and still fail the bondholder if it takes too long, costs too much, or requires inputs that are not available when the refinancing has to happen.

A facility that needs new cooling, new power distribution, heavier floors, tenant downtime, utility coordination, and scarce contractors is not simply an old building with a capex item. It is a construction project with a maturity date. If the balloon comes due in year 10 and the building needs years of work to support the next density profile, the fact that the asset is “upgradeable” does not solve the credit problem. It becomes the credit problem.

The density cliff reads like an engineering problem until you put a maturity date next to it. Then it becomes a refinancing problem. The building may have a path to relevance, but the debt may not have the time to wait for it. The market underwrites the residual as if it is already there, while the creditor should ask whether the residual has to be built again.

Buildings Designed for One Job

Not every building is equally exposed, and that is where the absence of a clear quality convention bites. There is no broadly traded Class A, B, or C shorthand for data centers the way there is in office or industrial. A facility engineered to be re-densified, or de-densified, at modest cost is a genuinely different asset from one packed into the tightest footprint its economics allowed, and the market often charges roughly the same for both.

The people who design these buildings know the difference. The spread often does not, because the spread is set by the tenant and the lease, and the lease says nothing about whether the shell survives the technology cycle. A long-term lease can make two buildings look identical on a screen while one has flexibility and the other is a single-purpose machine.

Workload cuts the same way. A building running training has different requirements from one running inference, and a site optimized for one is not automatically useful for the other. Training can tolerate lower redundancy and a more bare-bones envelope, while inference often wants to sit close to the data and the users. That makes location part of the asset rather than incidental to it.

A campus chosen purely for cheap power, far from anywhere, can be fine for the workload it was built for and stranded if that workload moves. The building does not become impaired because demand for compute disappears. It becomes impaired because the demand that remains needs a different building.

Put the two frameworks together and the sector starts to sort itself. The riskiest credit is not just the highest residual basis. It is high residual basis attached to a building with low adaptability.

Power Is Part of the Residual

The same problem shows up in power. Capital is not the binding constraint in digital infrastructure anymore, capacity is. Power, interconnection, substations, transmission upgrades, turbines, permitting, electricians, utility approvals, and local politics now decide when a site can become useful.

That turns residual risk into timeline risk, because the building cannot be refinanced on theoretical power. It needs delivered power. A building that needs more density also needs more power, and more power means more utility coordination on a schedule the lender does not control.

If the upgrade depends on interconnection, grid upgrades, switchyards, or generation capacity, then the refinancing case depends on institutions and equipment queues outside the credit agreement. Every month spent waiting on power is a month in which the maturity date gets closer. That is a very different risk from a rent roll, and it is much harder to solve with spread.

Behind-the-meter (BTM) generation is often presented as the fix, but it is better understood as a bridge with its own cost and execution risk. If a site needs a gigawatt, the sponsor may need to overbuild generation, absorb massive upfront capex, and pay for power as if the plant runs at full utilization even when the data center does not. What used to be 6 or 7 cent power can become 12 cent power before utilization, and something much higher once actual utilization is reflected.

The same is true of utility upgrades. A developer may have to fund substations, switchyards, transmission upgrades, generation interconnections, or system improvements before a site can receive the capacity it needs. What used to be a manageable contribution can become hundreds of millions of dollars, or more, paid upfront before the building produces the cash flow the financing assumes.

This is the part spreads struggle to price. You cannot price a power delay like carry, and a local moratorium is not something 25bps of spread absorbs. These outcomes are binary. Either the site gets delivered, the power shows up, and the tenant can refresh into the building, or the whole refinancing case is waiting on something outside the credit agreement.

Is the Appraisal Worth Anything?

In a normal real estate credit, the appraisal at least has a transaction history behind it. Office, industrial, multifamily, hotels, and retail all have flawed but visible comps, while data centers have much less history, especially for assets that are technologically stale.

In rated CMBS, at least one agency effectively answers that question before assigning recovery. In stress, it marks certain data center assets 30% to 50% below current appraised values, and it is careful about how much value it will credit above the rating of the tenant. That is not a universal market price, but it is a useful admission from the structured-credit process: the appraisal is not the same thing as realizable value, especially when the value rests on the building rather than the lease.

In a securitized deal you can at least see the discount. Investors can see the property value, the debt, the amortization, the residual exposure, and the rating agency stress. In corporate format, the appraisal disappears into the issuer’s balance sheet and the haircut is never struck. The same building carries the same obsolescence risk in both, but in one wrapper the markdown is explicit, while in the other it is easier for the market to ignore.

The residual is most overstated exactly where it is least visible. Plenty of corporate-format deals are priced fine, and plenty of securitized ones are not safe just because the haircut is printed. The point is narrower. The wrapper decides whether the market sees the building risk clearly or lets it disappear behind the tenant name.

Scarcity Buys Time, Not Adaptability

There is a serious argument that obsolescence is overstated, and it deserves a hearing. Power is the binding constraint in this sector, and it is choking new supply. A site that already exists, already energized, with interconnection that would take years to replicate, carries scarcity value. If you cannot build the new building, the old building keeps earning, and on that logic today’s stabilized assets may grow more valuable as the queue for new capacity lengthens.

That argument is real, but it only protects assets the market can still use. Scarcity does not turn an inflexible shell into a flexible one. It does not make an air-cooled facility liquid-cooling ready, or a low-density floor plate able to support the next rack profile, or a remote training campus useful for inference workloads that want to sit closer to users and data.

The tenant is making the same calculation. If the technology improves far enough, the hyperscaler on the lease may find itself holding more capacity than it needs, in buildings that no longer match the workload. If the lease permits assignment, the tenant may be able to hand the weakest sites to someone else. Scarcity can make the residual valuable. It cannot make every residual equal.

When There Is No Second Life

The deepest version of the problem is that these buildings often have limited alternative use. When an office tower empties, there is a conversion market, however painful, into residential or mixed use or something. A purpose-built data center campus on cheap land far from population, designed around one kind of compute, has far fewer paths. If the workload it was built for moves or shrinks, the asset does not step down in value the way a flexible building does. It can gap lower, because the only natural buyer is another operator who wants precisely what it already is.

Location decides where on that spectrum an asset sits. A facility inside a major metro, near the data and the users, has paths to a second tenant and a second use. A gigawatt campus chosen purely for cheap power, an hour from anything, has the residual of a specialized industrial site whose industry may not need it in 10 years.

The lease does not tell you which one you own. Residual basis tells you how much it matters. This is why the market needs a better vocabulary than tenant, lease term, and appraised value. Two facilities can have the same tenant and the same lease and still be radically different collateral. One is a flexible shell with power, location, and upgrade paths. The other is a purpose-built machine whose second life depends on the first workload still mattering.

When the Rent Runs Out

Residual basis told you how much of the bond depends on the building. The debate this week is whether the building will be there to depend on, in any economically useful sense, by the time the contracted rent runs out. The harder version is whether the building is still useful as-is, or whether the residual the bondholder is counting on is really another construction project waiting for funding.

No one underwriting this paper can tell you what compute looks like in five years, let alone the 10 or 15 the residual reaches for. None of that uncertainty means you should avoid the sector. It means you should know how much of your principal is riding on it, because the spread certainly is not paying you to find out.

The lease can pay perfectly and the credit can still disappoint. That is the problem with residual basis. Nobody tests it at signing. The test comes years later, when the rent runs out, the refinancing has to happen, and the building has to prove it still matters.

JBI Bulletin Board

1) Something New is Coming

The institutional version of JunkBondInvestor is coming. The list gets in first and locks the founding rate before it closes. Add your name here.

2) Now Hiring: Credit Analysts

I’m hiring credit analysts to add capacity. If that’s you, or someone you’d vouch for, reply with a short note on your background.

If you liked this, you might also like:

JunkBondInvestor is an independent financial research publication providing general market commentary, corporate credit analysis, and educational content. The information provided is for informational purposes only and does not constitute investment advice, financial planning, or personalized recommendations of any kind. Nothing published constitutes a solicitation, offer, or recommendation to buy, sell, or hold any securities, nor does it guarantee any financial outcome. Any credit opinions or outlooks expressed are solely the independent views of the authors and are intended to reflect general credit trends, not specific investment recommendations. Investing in high-yield bonds, distressed debt, and leveraged loans carries significant risk including potential loss of principal. JunkBondInvestor operates under the Publisher’s Exemption of the Investment Advisers Act of 1940 and applicable state laws. This issue includes sponsored content and partnership links.

Shutterstock shareholders gave back 30% in a session. Getty’s sub-$1 stock is fighting to stay listed. Eighteen months of deal work gone just like that.

The $628mm of bonds sold to fund the deal? Up more than 14 points.

That’s not backwards if you’ve been following the situation and the deal dying is what could send this paper to par.

The redemption is potentially days away yet the bonds still trade two points below it. The only questions are why is there a discount and is there a chance these bonds remain outstanding?

Situation Background

In January 2025, Getty announced its merger with Shutterstock, the two largest stock content platforms. $3.7bn of combined value, $150-200mm of promised annual cost synergies.

The rationale was obvious. AI’s crushing stock photography from the bottom up. Combining meant scale, cost cuts, and a bigger rights-cleared content library to license to the AI labs, the one thing the models cannot generate for themselves.

In February 2026, the DOJ cleared the deal. That left the UK’s Competition and Markets Authority (“CMA”) as the final practical gate. The CMA conditionally cleared the transaction on June 10, but only if Getty divested Shutterstock’s full Editorial segment. The problem market was UK editorial, the news and celebrity photography business, and the remedy meant selling Rex Features plus Backgrid and Splash News. Getty proposed a narrower fix, selling Backgrid and Splash while keeping Rex. The CMA said no.

As recently as May, commentary pointed to an active buyer pipeline for the divestiture assets, with both companies guiding to a 2026 close. Then nothing sold. The simplest explanation: the editorial business was unprofitable and no acceptable buyer showed up at an acceptable price.

Rather than execute a forced sale on bad terms, Getty announced on June 30 that it will let the merger agreement terminate when the Second Extended End Date passes on July 6.

The Security and the Escrow Structure

On October 21, 2025, Getty issued $628mm of 10.5% senior secured notes due 2030 to fund the cash consideration to Shutterstock shareholders, refinance Shutterstock debt, and pay fees.

Here’s the defining structural feature.

Read more

Issuance ran at a few billion dollars a month in early 2025, then inflected in the autumn and has since run at $50bn+ a month. That is too large to treat as an equity side debate or a venture financing story. The buildout now depends on credit investors continuing to fund it, roll it, and eventually refinance it.

This issue is a data pack on the AI infrastructure credit market: what it is made of, how it prices, and what the spread dispersion says about where investors think the risk actually sits.

Free Issue Sponsored By 9fin:

AI Chat — Ask complex questions across 9fin’s full data set (covenants, cap tables, financials, filings, news, and 9fin analysis) and get a fully cited answer in seconds. Effortlessly screen for opportunities, deep-dive into a credit and track evolving situations in real-time.

Research Grid — Compare credits across metrics, outlooks and key risks using 9fin’s frameworks or your own questions. Produce consistent, side-by-side analysis in minutes so you can spot outliers instantly and track changes in fundamentals.

Learn More

Structure drives the spread

The spread range in AI infrastructure is less about one asset pricing differently and more about where the creditor sits in the stack.

At the tight end, senior datacenter ABS can clear around 120bps, supported by lease cash flows, tenant credit, collateral, and tranche subordination. At the wide end, neocloud corporate debt can trade closer to 600bps, where the lender is taking enterprise risk tied to compute demand, customer concentration, GPU residual value, and refinancing access.

Both sit inside the same AI infrastructure buildout, but they are not the same exposure. One is closer to a structured claim on contracted datacenter cash flow. The other is closer to a corporate claim on the economics of selling compute. The spread range is the market separating those claims.

That distinction matters because the labels can obscure the risk. “AI debt” includes hyperscaler bonds, leased datacenter paper, ABS tranches, private placements, construction loans, and merchant compute debt. The underlying theme is similar, but the creditor claim is not.

The bands overlap, which is where the relative value sits. The widest paper in one structure can price near or through the tightest paper in another. That is the useful part of the data: not that every instrument is comparable, but that the market is still drawing the boundaries between structure, borrower, tenor, and demand exposure.

High yield separates the operators from the compute bet

Inside the high-yield and frontier part of the market, the ordering is cleaner if you separate three groups: datacenter and colocation names with contracted tenants, power and equipment names tied to the buildout, and neocloud or merchant compute names taking direct demand risk.

The contracted datacenter names generally trade through the merchant compute names. Credits backed by leased capacity, customer contracts, or infrastructure-like assets cluster well inside CoreWeave, while power and equipment names sit in their own part of the stack. CoreWeave, the only neocloud credit in the set and the name most directly exposed to selling compute, trades north of 500bps, roughly 110bps wide of the next datacenter name below it.

One way to read the ordering is that the market is more comfortable underwriting contracted infrastructure exposure than merchant compute demand. A leased datacenter credit depends on tenant quality, lease durability, asset location, power access, and residual value. A merchant compute credit depends more directly on GPU utilization, customer demand, pricing, refinancing access, and the residual value of the hardware.

That distinction matters because the collateral is different. A claim on a datacenter or colocation name may run through contracted customers and powered infrastructure. A claim on a merchant GPU cloud is much more exposed to chips that depreciate quickly, are worth less off the stack they were deployed on, and may be difficult to control if the borrower is also dependent on third-party space, power, and cooling.

The move since pricing reinforces the split. Over the period in this data, most of the datacenter and infrastructure-linked high-yield names rallied: Applied Digital tightened about 390bps, from 630bps at issue to 240bps, with WULF in 228bps, VoltaGrid 181bps, and Cipher 148bps. CoreWeave was the exception, with its longer 2032 paper widening slightly since recent issuance.

IG curves are separating

The IG part of the market is less volatile, but the curves are not moving as one. Oracle and Meta sit wide of the tighter large-cap cluster, while Apple, Alphabet, Amazon, and Nvidia remain the lower-spread group.

Part of that is structural. Oracle is BBB, longer-dated, supply-heavy, and more levered than the AA large-cap issuers, so it should not trade on top of Apple, Amazon, or Alphabet. Meta is a cleaner comparison but still has its own capex and duration profile. The point is not that the bond market has made a clean verdict on hyperscaler AI spending. It is that the widest IG curves are also the names where investors have more to underwrite: capex intensity, leverage trajectory, supply, and long-end duration.

Oracle’s long end trades around 214bps to 245bps, triple-digit wide of Amazon and Alphabet at comparable tenors, roughly 140bps over Amazon and 90bps over Meta at the long end. Meta’s long bonds trade around 120bps to 160bps. The tighter cluster, Apple, Alphabet, Amazon, and Nvidia, sits in a narrow band.

The useful observation is the dispersion, not a single-cause explanation. Oracle has widened in secondary rather than merely pricing cheap at issue, which suggests the market is continuing to reprice the name after the deal. But the spread gap likely reflects a mix of rating, tenor, leverage, supply, and capex exposure. I wrote about Oracle back in December of last year. Check it out here.

Dispersion within ratings is wide

Mapping bonds to ratings gives the expected downward-sloping relationship, but the dispersion inside each rating bucket is large. BBB spans roughly 75bps to 245bps at the Oracle long end. That is a 170bps range inside one rating category. The AA cluster is tighter.

Some of that range is tenor. Once maturities are matched, the rating signal still matters, but it does not explain the whole spread. In the 5-to-10-year part of the curve, the wide end of IG and the tight end of HY begin to overlap, especially around the power, equipment, and infrastructure-linked names. The seam between rating categories is therefore not a clean line; it is a zone where issuer, tenor, structure, and sector exposure matter as much as the rating label.

The same point shows up inside IG. The median IG curve flattens past 10 years, while the range around that median widens. The long end does not simply pay more because it is longer. It pays more in the names where the market has more to underwrite.

In this market, rating is the starting point. Issuer, tenor, structure, and exposure explain the rest.

Spread gap versus all-in yield gap

Measured by all-in yield, the HY-IG gap is about 200bps: median YTW is around 4.5% in IG and 6.5% in HY.

The difference comes from duration and curve exposure. IG contains more long-dated paper, where a larger share of the all-in yield comes from the Treasury curve. HY is concentrated in shorter 4-to-6-year tenors, where the spread is doing more of the work. The spread comparison and the yield comparison therefore tell different stories about how much extra income investors are actually receiving.

Total return shows the same split. Year to date, HY and datacenter names led with double-digit returns as spreads compressed on top of carry, while the worst performers were long-duration IG, including the Oracle and Vertiv long bonds, where rates and spread widening compounded. A single sector return figure averages those opposite ends into a number that fits neither.

The absolute yields are not especially high though. The widest name, CoreWeave, tops out near 10%. The point is that the spread gap makes the risk premium look larger than the actual income pickup once duration and curve exposure are taken into account.

Datacenter ABS

Datacenter ABS is one of the less-covered corners of the market, and it has a clear structure to it. The enhancement curve is steep and fairly consistent across sponsors: senior AAA tranches clear around 120bps to 145bps, while the pickup to the BBB mezzanine is roughly 100bps to 175bps. QTS runs from 160bps to 300bps, and Switch from 145bps to 260bps.

On an all-in basis, the sector yields roughly 6% to 8%, not much below HY. The difference is the claim. ABS investors are taking long-duration exposure to contracted datacenter cash flows with tranche subordination, cash-flow controls, and collateral backing. HY investors are generally taking corporate credit risk, operating risk, refinancing risk, and, in the case of neoclouds, more direct compute-demand risk.

That does not make ABS risk-free or even directly comparable to HY. The tenors are long, which brings extension and refinancing risk. Tenants roll, collateral values can change, and the residual is still connected to the same AI infrastructure demand ecosystem, just filtered through lease and tenant performance. The point is not that ABS avoids the risk but that it transforms the risk into a different claim.

Maturities cluster in 2030-2031

The maturity profile is uneven. Roughly $45bn of HY paper comes due across 2030 and 2031, most of it issued in 2025. IG has no comparable cluster; it is laddered along the curve, with its largest single bucket out in 2036 and sizeable amounts beyond. The first concentrated refinancing need sits in the frontier part of the market. How spreads, demand, and market access look then will say a lot about how durable the current financing bid really is.

Four risks under one label

AI infrastructure is not a single trade. At least four distinct risks sit under the label:

1. Hyperscaler duration

2. Leased datacenter residual

3. Securitized tenant cash flow

4. Merchant compute demand

They are related, but they do not move together. Treating them as one exposure blends very different creditor claims and risk profiles.

More broadly, this looks like an asset class still forming, the way CLOs and datacenter REITs once did. Half a trillion dollars of debt has formed in roughly 18 months, and the taxonomy is still unsettled: neocloud vs. operator, property vs. securitization, corporate claim vs. asset-backed claim. The spread dispersion reflects that. The market is still deciding which labels matter and how much each claim should be paid.

The funding base is also concentrated. USD is 87% of the market, and of the roughly $66bn issued offshore, almost all is Alphabet, across six currencies, and Amazon, across four. Everyone else funds in dollars.

Pulled together, the data shows a few consistent patterns.

Lease-backed and structured paper generally prices through merchant compute exposure. Inside IG, the wider curves are also the names where investors have more to underwrite, including capex intensity, leverage, supply, and duration. Ratings set the broad neighborhood, but issuer, tenor, structure, and exposure drive the trade. The spread gap overstates the income pickup once duration and curve exposure are taken into account. And the first concentrated refinancing cluster sits in 2030-2031.

The main message is simple: this is not one credit market yet. It is a set of claims being priced apart in real time.

High Yield Weekly Market Performance

High Yield Primary Market

High Yield Secondary Market

JBI Bulletin Board

1) Something New is Coming

The institutional version of JunkBondInvestor is coming. The list gets in first and locks the founding rate before it closes. Add your name here.

2) Now Hiring: Credit Analysts

I’m hiring credit analysts to add capacity. If that’s you, or someone you’d vouch for, reply with a short note on your background.

If you liked this, you might also like:

JunkBondInvestor is an independent financial research publication providing general market commentary, corporate credit analysis, and educational content. The information provided is for informational purposes only and does not constitute investment advice, financial planning, or personalized recommendations of any kind. Nothing published constitutes a solicitation, offer, or recommendation to buy, sell, or hold any securities, nor does it guarantee any financial outcome. Any credit opinions or outlooks expressed are solely the independent views of the authors and are intended to reflect general credit trends, not specific investment recommendations. Investing in high-yield bonds, distressed debt, and leveraged loans carries significant risk including potential loss of principal. JunkBondInvestor operates under the Publisher’s Exemption of the Investment Advisers Act of 1940 and applicable state laws. This issue includes sponsored content and partnership links. JunkBondInvestor receives compensation from sponsors for featured placements and may earn referral fees from partner programs.

The company’s market cap sits around $2 trillion. The bonds are under 2% of that. Cash alone covers the debt more than three times over.

You would have to wipe out something like 98% of the enterprise value before a bondholder lost a dollar. On the one question that usually decides a credit, whether you get paid back, the answer is obvious.

An equity analyst would’ve stopped there. Money-good, take the spread, move on.

That is the wrong question. With a cash pile this size almost everyone gets paid, so default is not what you are pricing. You are pricing what the bond does over the years you hold it, and that looks very different at the five-year than at the thirty. Same company, five bonds, and they do not trade as one.

The gap between how this deal got bought and how it should be priced is the reason for the write-up. I’ll walk through how a credit investor looks at and prices SpaceX, tranche by tranche. Everything presented here is based on public information.

What SpaceX Actually Is

For a creditor, the relevant fact is that this is 3 businesses with different economics.

• Space is the launch arm, Falcon 9 and Starship, with a near-lock on cheap access to orbit and the contracts to fly NASA and national-security payloads.

• Connectivity is Starlink, around 12 million subscribers, recurring revenue, and the only part that reliably makes money.

• AI is xAI, Grok, the Colossus data centers, and X; it brought in about $3.2 billion last year and lost roughly $6.4 billion doing it, per public filings, with the loss widening every year.

Revenue across all three ran about $18.7 billion in 2025. Starlink generates almost all the usable cash, AI burns it and reaches for more, and Space underneath supplies the launch capacity that makes the rest cheap. A lender is really backing the satellite business and betting the AI spend doesn’t drain the company before it pays off.

What This Deal Actually Is

This is a refinancing first, and growth capital second. The main job is to term out a $20 billion bridge loan. The practical effect is to move Musk’s AI debt stack out of expensive short-term funding and into the IG market.

For context, the debt dates back to Twitter. Musk funded the buyout in 2022 with about $13 billion the banks couldn’t offload, so they held it for two years. Twitter became X, X merged into xAI, and xAI borrowed another $5 billion along the way. SpaceX bought xAI this February and inherited the pile.

SpaceX took it out with the $20 billion bridge and these bonds repay the bridge at coupons of 5.35% to 6.65%, about $1.5 billion a year. The IG rating also buys access to the high-grade market, roughly $8 trillion deep.

Current Capitalization

What’s Good About the Credit

• Tiny loan-to-value. $25 billion of bonds under a company worth around $2 trillion, so gross LTV of about 1.6%. Even after the $600 billion drop, even haircutting the equity hard, the cushion is enormous; the bonds don’t take a loss unless the whole enterprise falls something like 98% first. That is the main reason this clears as IG.

• Starlink makes money. Most of the $18.7 billion of 2025 revenue runs through connectivity, and it’s profitable and growing, with 12 million subscribers, recurring billing, and margins that widen as the network fills in. You’re lending against a real broadband franchise with runway left in enterprise, government, and direct-to-cell.

• A mountain of cash. Around $114 billion against $25 billion of bonds and about $9 billion of asset financings, so roughly $80 billion net. Near-term liquidity is not the worry.

• A launch near-monopoly the government needs. SpaceX puts more mass into orbit than everyone else combined, at costs nobody has matched, and flies NASA and national-security payloads. Not cash flow today, but a moat under the structure and a reason it doesn’t get competed away.

• Management pulls the right levers, so far. Roughly half the capex is discretionary, so it can slow the AI buildout if cash tightens. It just raised $86 billion of equity in the IPO and pays no dividend. The agencies credit it with conservative financial policy, balance sheet ahead of payouts, which isn’t something you assume with this management.

What Worries Bondholders

• It burns cash, and will for years. No positive FCF until 2030 on S&P’s view, with the burn worsening next year and again in 2028. The cushion covers the early going. After that the company needs revenue to ramp fast and the debt market to stay open, and it needs both.

• A wall of supply is coming, and the long bonds wear it. S&P’s base case takes debt to about $132 billion by 2028, and this deal landed in a record month for high-grade issuance, more than $175 billion as AI borrowers flood the market at once. Every future deal reprices existing paper wider. The long bonds are also not underwriting today’s balance sheet but the balance sheet SpaceX is trying to become, roughly five times the debt it carries now, with the AI plan still unproven.

• The AI business loses billions with no proven payoff. The IG case assumes monetization arrives, through compute deals and enterprise and consumer revenue, before it drains too much cushion. You fund it the whole way and don’t get the upside if it works.

• Starship has to deliver. The economics that make Starlink cheap to scale depend on Starship reaching full reusability roughly on schedule. Slip badly and the connectivity growth story, which the credit rests on, weakens.

• The equity owns the upside, and the cushion moves. Your best case is getting paid back; the Mars and AI and space-data-center optionality goes to the stock. And the cushion under you is a market price that fell $600 billion in three days and bounced. It can shrink fast, right when you’d want it.

• One person runs everything. Governance is weak across the board: controlled company, concentrated voting, thin oversight. Fitch was blunt about it, calling the Musk dependence a “key rating constraint.” This is someone running a $60 billion all-stock acquisition of a coding startup while the AI arm bleeds. The IG case assumes he keeps choosing the balance sheet. So far he has.

• The long end has an ESG problem. Moody’s flagged emissions and on-site power around the data centers, MSCI scores the company at the bottom of its ESG scale, and desks tied some of the weak long-end demand to mandates that can’t hold the name for thirty years. Fewer natural buyers for the bonds already under the most pressure.

• And it’s hard to model. Three businesses with different economics, almost no segment guidance, and agency capex and debt forecasts that span huge ranges because even they are guessing.

The credit setup is unusual: a fortress balance sheet attached to a business that still needs years of external funding. The real question is where the market is paying you for that risk, and where it is not. Across the five bonds, the answer is not the same, and that is the whole trade.

How the Bonds Are Currently Trading

Read more

Buy one and you are not buying Microsoft. You are buying Microsoft for the contracted years of the lease, and after that a building, its value resting on whether anyone still wants it once the contracted cash flows run out.

The market prices the lease and treats the building as a pass-through, handing you the residual for free on the assumption it is worth close to par. A growing share of the risk in this paper sits in exactly that residual, and most people are not pricing it.

Start with the cleanest version of the problem: what you actually own when you own one of these bonds.

The Tenant Is Not the Entire Asset

There is no Class A, Class B, Class C convention in data centers the way there is in office or industrial. No premium for a facility that can be re-densified cheaply, no discount for one that cannot. The credit curve is flat where it ought to be steep, because the tenant and the lease carry the entire analysis.

That holds only while the lease term and the debt term line up. A 20-year lease behind 10-year paper is not the same risk as a 10-year lease behind 10-year paper, and neither is the same as a structure where the contracted lease is shorter than the life being financed. The market treats all of them as one credit because the name on the lease is the same.

You can see the same indifference in leverage, with public deals now clearing at 95% to 98% LTC and the oversubscription used to pull equity back out of the project. That matters more than the thin cushion suggests. Residual value is an equity risk by design: the lender is meant to amortize out before the building’s worth is ever in question, leaving the bet on what the asset fetches in 10 years to the equity underneath. Strip the equity down and back-load the amortization, and that bet quietly migrates onto the bondholder, unpriced.

The tenant is not even fixed for the term the analysis assumes. Some leases permit assignment, and unless equivalent credit support remains in place, the IG name underwritten at closing may not be the credit support investors think they own for the full term. Even the credit doing the work in the spread is contingent on terms most buyers never read.

Residual Basis Is the Real Credit Line

Last week I argued the market was not differentiating hyperscaler-leased paper from neocloud-leased paper, applying one spread across a real gap in tenant quality. Hold the tenant constant and a second, more important gap appears.

The number that decides it is residual basis: how much of your principal comes back from contracted rent you can model, and how much rides on re-leasing or refinancing the asset later. The first is a real estate loan with a defined exit. The second is a bet on a building, or on a tenant who does not yet exist.

Same tenant, same rating and coupon. Deal A amortizes almost all of its principal from rent under contract. Deal B runs out of lease the day the bond matures. Deal C looks identical to A on the page, the same tenant and the same 20-year lease behind 10-year paper, but it is sized as a partial amortizer with a large balloon, so 40% of recovery depends on refinancing against the back half of the lease and the building underneath it.

Three bonds the market prices within a few basis points of each other. Three different credits. The headline terms, the ones that set the spread, are the ones that hide the difference.

Estimating residual basis is not complicated, which is part of what makes the omission strange. Take the bond’s amortization schedule and ask how much principal retires from contracted rent over the life of the paper. Whatever is left at maturity is the balloon, and the balloon is the residual basis. A bond that pays down steadily over a 10-year term against a 20-year lease is mostly contracted. A bond that IOs for 9 years and bullets in year 10 is almost entirely residual, however IG the tenant looks on the front page.

The lease term remaining at the refinancing date is the other half of the calculation. A balloon that comes due with 15 years of contracted hyperscaler rent behind it is a financeable asset. The same balloon coming due with 3 years of lease left is a building looking for a tenant, in a market that may by then have repriced the whole sector. Residual basis tells you how much you are exposed. The remaining lease at the refinancing point tells you how dangerous the exposure is.

Run the screen this way and the sector reorders around the only number that matters: what fraction of par is contracted, and what fraction is a balloon. One is a credit you can underwrite. The other is a forecast about a building and a market a decade out.

The rating agencies understand this and notch hard when debt does not amortize inside the first lease term, and they notch faster when repayment leans on an extension or a re-lease. The structured market reflects it in the gap between fully amortizing deals and those built around an anticipated repayment date, with calls stacked at years 8, 12, and 16 to manage exactly this exposure. The $29bn Meta-risk project financing that cleared last year was built on that logic, the call schedule doing work the lease alone could not.

This is the normal case, not the edge case. Most of these structures do not fully amortize from contracted rent, because the cash flows are not long enough, so a slug of principal is always left to refinance. The paper is not the clean credit pass-through it is sold as. Bullet structures, single-asset deals, and thinner amortization are pushing the residual share higher, not lower. The part of the credit nobody prices is the part that keeps growing.

Hiring in Credit?

The Refinancing Leg

The residual is not a vague unknown. It is a specific event: a refinancing, on a date you can read off the structure, into conditions you cannot.

When a balloon comes due, three things have to hold at once.

• The lease has to have enough term left to support new debt.

• The market has to still want data center paper at a price that clears.

• The asset has to still support a loan.

The structures pricing today are underwritten as if those 3 are independent and all 3 hold. They are not independent, and they tend to fail together.

The structures do push back on this. A master trust pools many assets and blends their maturities, so a single balloon refinances against a longer weighted-average stream rather than one expiring lease, which genuinely softens the timing risk. What it does not do is change what sits underneath. A pool of single-tenant hyperscaler buildings refinanced into a cooler market is still a hyperscaler real estate bet, just a diversified-looking one. Pooling the residual is not the same as pricing it.

The refinancing window is also the moment the holder has the least leverage. A fully amortizing bond never faces it, because there is nothing left to refinance. A balloon hands the borrower an option and the lender a deadline. If the sector has cooled, if rates have moved, if the tenant has signaled it will not renew, the refinancing the residual basis quietly assumed becomes the hardest financing in the structure to execute, at exactly the moment it has to happen.

One bond is repaid by a lease. The other has to find a new lender, a willing market, and a usable building, all on one date years from now.

Format Can Hide the Residual

The corporate market makes none of these distinctions, and that is where the residual disappears most completely.

Earlier this year a large single-tenant data center, leased long term to a hyperscaler, was financed in the IG corporate market rather than as an ABS. In corporate form it read as clean IG paper, and within days other developers were asking their bankers to put the same trade in corporate clothing, because the format quietly lowers the apparent cost of the residual by hiding it. Comparable assets in the securitized market were clearing at better ratings, shorter maturities, and wider spreads, and once investors put the two side by side the corporate bond repriced wider.

Some of this paper now clears in a day or two with little diligence done, because demand is deep enough that the format prices regardless. Speed is not scrutiny.

A long lease wrapped as shorter corporate paper presents as a tenant credit, and the residual exposure an ABS structure would isolate and amortize against vanishes into the format. That gap corrected quickly because it was visible the moment someone compared the two. The residual mispricing inside a single bond does not correct that way, because nothing forces the comparison.

The Case for the Other Side

There is a real bull case here, and the relative-value argument is stronger for stating it.

These are not ordinary commercial leases. Many are long-dated and structured hell-or-high-water, so the tenant pays through almost anything, and the tenant is among the best credits in the world, funding capex measured in the hundreds of billions out of net cash. The asset underneath is scarce in a way that matters: an energized site with interconnection that would take years to replicate does not sit empty for long, so if one hyperscaler walks, replacement demand for the power and the shell can be robust. This is why project finance lenders like the paper. Long lease, elite tenant, hard asset behind a queue no one else can jump.

Grant all of it. None of it touches the residual. A hell-or-high-water lease is bulletproof for exactly as long as it runs and silent on the day after, and the residual is the part beyond the lease. The strongest form of the bull case, that scarcity makes the re-lease easy, is itself a bet on the residual rather than the lease, placed years out into a market no one can price today. It may well pay. The point is that it is a different bet from the contracted rent beside it, carrying different risk, and the spread charges the same for both. Taken seriously, the bull case is an argument for pricing the residual, not ignoring it.

High Yield Market Performance

Primary Market

Secondary Market

JBI Bulletin Board

1) Something New is Coming

The institutional version of JunkBondInvestor is coming. The list gets in first and locks the founding rate before it closes. Add your name here.

2) Now Hiring: Credit Analysts

I’m hiring credit analysts to add capacity. If that’s you, or someone you’d vouch for, reply with a short note on your background.

If you liked this, you might also like:

JunkBondInvestor is an independent financial research publication providing general market commentary, corporate credit analysis, and educational content. The information provided is for informational purposes only and does not constitute investment advice, financial planning, or personalized recommendations of any kind. Nothing published constitutes a solicitation, offer, or recommendation to buy, sell, or hold any securities, nor does it guarantee any financial outcome. Any credit opinions or outlooks expressed are solely the independent views of the authors and are intended to reflect general credit trends, not specific investment recommendations. Investing in high-yield bonds, distressed debt, and leveraged loans carries significant risk including potential loss of principal. JunkBondInvestor operates under the Publisher’s Exemption of the Investment Advisers Act of 1940 and applicable state laws. This issue includes sponsored content and partnership links. JunkBondInvestor receives compensation from sponsors for featured placements and may earn referral fees from partner programs.

You’ve probably never heard of them. You’re not supposed to. You’re not the customer. Ask any woman in your life, at least one has scrolled past their Instagram ads. This:

For a couple of years, Oddity looked like it had solved the one thing nobody in consumer solves. Paid acquisition that actually pays for itself. Land the customer at breakeven, make it back when they reorder at 70%-plus gross margins, do it forever. Lifetime value over acquisition cost. The infinite money glitch.

Then the glitch closed.

83% more to acquire a customer. Not in one geography. All of them, at the same time. Revenue down 26% in 1Q’26. EBITDA negative. The stock trades near $10, off almost 90% from the highs last year.

But back when this was all working, Oddity sold $600 million of zero-coupon exchangeable notes due 2030. They go for 69 today. 9.5% YTM. Busted convert sitting under a broken stock.

So who’s buying it at 69, and what do they think they know?

Situation Overview

Oddity is an Israeli company that sells beauty and wellness products online, and almost nowhere else. Two brands carry it: IL MAKIAGE in color cosmetics and SpoiledChild in anti-aging and hair. A third, METHODIQ, just launched.

It’s 97% direct to consumer. No real wholesale, department store counters, or shelf space worth mentioning. You take a quiz, the AI reads your answers and points you at product, and Oddity buys the traffic that feeds the quiz through paid social. Nearly all of that spend goes through one platform: Meta.

The company went public in July 2023 at $35 on a simple premise. Acquire a customer at roughly first-order breakeven, then earn it back on repeat purchases at gross margins above 70%. For two years, it ran the way it was drawn up. Revenue compounded above 25% a year, first orders and repeats came in about even, and management guided to a long-run 20% growth at 20% EBITDA margins. People bought the algorithm because the results kept showing up.

Then in late February 2026, management got on a call and described a “dislocation” with its largest advertising partner.

The cost of that word was specific. Customer acquisition at IL MAKIAGE rose 83% year over year in the first half of 2026, after three straight years of mid-teens CPA inflation, the boring kind you can budget around. And it hit the US, Canada, the UK, Australia and Israel at the same time, which is not how failing brands behave. Management’s read is that the problem is technical, an algorithm and signal issue on Meta’s end, not fading demand for the product.

For a business that acquires at breakeven, the math flips the moment a new customer costs more than the first order brings in. Then every acquisition burns cash. The first quarter is what that looks like.

Net sales fell 26%. First orders were halved. Repeats, the customers already through the door, fell only 15%, the one number that held. The mix did the rest: gross margin gave up 520bps to 69.7% on lower order value, and Adj. EBITDA swung from positive $52 million a year earlier to negative $7 million. Management guided Q2 to sales down 25 to 30%.

A quarter in, the engine still isn’t fixed, but the signals are turning. The ad partner thinks 40 to 60% of the inflation is recoverable through its own adjustments. IL MAKIAGE’s acquisition costs fell 28% from April to May, the first sequential relief since the break. SpoiledChild is running a milder version of the same problem, and METHODIQ is tracking toward roughly $25 million in its first year. None of it has restarted the core yet.

Through the collapse, the company repurchased its stock even as EBITDA turned negative. At ~$10/share, the stock is priced as if the engine is broken rather than stalled. But the equity is the obvious security to fixate on, and the obvious one is rarely where the trade is. The better question sits a rung up the capital structure: what Oddity owes, and what it owns against it.

Current Capitalization:

In mid-2025, with the stock in the $60s and $70s and the model still working, Oddity sold $600 million of 0% exchangeable notes due 2030. The conversion price was set far above where the shares trade now, so the equity option attached to the notes is deep out of the money and what changes hands today is essentially unsecured credit. The notes trade around 69 implying a 9.5% YTM.

Read more

OpenAI put a version of that question on the table this week. It is reportedly weighing a steep cut to what it charges for tokens, to pull users from Anthropic and stay ahead of the cuts it expects Anthropic to make in return. Two companies whose businesses still depend on investors subsidizing enormous compute spend, lining up to cut price.

Price cuts do not prove weak demand. They can also reflect falling unit costs. But they do prove something important for credit: the model layer may be more competitive, more substitutable, and less able to hold price than the infrastructure bid assumes. That is what the start of a price war looks like, and a price war is how a market begins to find out whether a product has any pricing power at all. A product that has none, sold at a loss, is one that never earns back the capital behind it.

It matters far beyond the labs.

Roughly a fifth of this year’s high yield issuance has gone to fund AI data centers, sold as infrastructure: long-lived buildings, contracted cash flows, a long runway.

But a data center bond is only as good as the company leasing the building, and those tenants are not all the same animal. Some are hyperscalers with broad, diversified demand, cloud and enterprise and advertising and their own internal workloads, that will pay the rent whatever happens to the price of AI. Others exist mainly to resell compute, and their ability to pay traces straight back to whether the model layer can charge enough for tokens to survive.

If the models never earn their cost of capital, the first group might be fine. The second might not be. The question is whether the market is pricing that difference.

We Have Seen This Before

You know what other industry doesn’t have any pricing power?

Airlines. Or close enough. They have loyalty programs and premium cabins and consolidation, but set against their capital intensity and a fuel bill someone else controls, none of it has ever been enough to earn the cost of capital. Indispensable product, demand compounding for decades, and a balance sheet that turns all of that growth into creditor losses, over and over. One of the most reliable destroyers of invested capital in modern markets.

The record is almost unmatched. Nearly every major US carrier has been through bankruptcy, several of them more than once. Warren Buffett used to hold airlines up as the purest example of a business that grows forever and rewards no one who funds it, joking that investors would have been spared a fortune if someone had grounded the Wright brothers at Kitty Hawk. Passenger traffic rose almost every decade since then. The capital lit itself on fire anyway.

Why Airlines Break Their Lenders

Start with why the original was such a dependable way to lose money.

An airline carries every feature that makes a business hard to lend to, all at once. The product is a commodity; a seat is a seat, and the carrier that charges more loses the passenger. The assets are enormous, debt-financed, and constantly depreciating. The largest single cost, fuel, is set by someone else. Put those together and you get a business that can grow for fifty years and still not cover its cost of capital, with credit that gets impaired every cycle.

AI did not build one airline. The closest analogy splits that model across three layers, and they are not the same credit. The mapping is not clean, and it is worth saying so. A token is not quite a seat, a GPU is not quite a plane, and the building that holds the GPUs is barely an aviation asset at all. The point is not that the parallel is exact. It is that the same dangerous combination, heavy capital set against a product that may not command a price, can be assembled out of these parts.

The labs sell tokens to end users. A token is not a pure commodity the way a seat is; models differ on quality, ecosystem, and the cost of switching, and a dominant one could hold real pricing power that no airline ever had. But this week’s news is a sign that the pricing power may be thinner than the bull case assumes. If it is, the labs are the airlines.

The neoclouds own the GPUs and rent compute to the labs and the hyperscalers. They are the closest thing to aircraft lessors. But a lessor only ever worked because planes held their value and the customer base was diversified, and the neoclouds tend to have neither. The fleet loses value far faster than any airframe; a three-year-old GPU is already being written down by the next generation on the truck. And the book is often concentrated in a handful of customers. It is a lessor stripped of the two things that made lessors safe.

The powered shells, the buildings and the power and the interconnect, are the one piece that is not really airline-shaped at all. If they map to anything in aviation, it is the scarce, location-bound infrastructure no operator can do without, the gate or the takeoff slot rather than the carrier or the plane. And the asset that actually matters is not the concrete. It is the power connection, which is genuinely scarce and, where it is scarce, genuinely valuable. That is what the AI infrastructure bonds being raised right now are mostly secured against, and it is the most durable thing in the stack.

Here is the chain that matters, and it only applies to the second kind of tenant. A shell bond leased to a neocloud is a claim on real estate, leased to a lessor whose collateral is melting and whose customers may be airlines. The rent comes from the neocloud. The neocloud’s revenue comes from the labs. The labs are funded, for now, by investors willing to subsidize a business that does not yet cover its costs. That bond is not infrastructure. It is a bet on how long that patience lasts.

The Market This Week

Take a look at JetBlue, whose bonds now yield north of 13%.

S&P just cut the company to CCC+ on a billion-dollar FCF deficit this year and another expected next. Spirit is in bankruptcy. Allegiant priced a secured deal wide, broke it to a 100 handle, and watched it slip to 99.25 within days.

Record numbers of people are flying, and the credit still looks like that. The sector has produced bankruptcy after bankruptcy across fifty years of rising demand. That is the warning the analogy carries, and it is not theoretical.

Now look at how AI infrastructure priced into the same week.

• Cipher’s Stingray deal, $810 million at a 6% yield, is secured against a data center in Andrews, Texas leased to Amazon for fifteen years.

• Meanwhile, Applied Digital priced $1.59 billion at 7% against a campus in Ellendale, North Dakota leased to CoreWeave.

Amazon pays the rent off cloud, retail, advertising, and its own workloads, whatever happens to compute prices. CoreWeave is a levered single-B neocloud whose ability to pay depends on its own customers continuing to pay it. One landlord rents to a blue chip. The other rents to a lessor that rents to airlines.

The market priced both as infrastructure, and the tenant distinction did not show up with nearly enough force. The bid is discriminating at the margin, but not nearly enough for how different the tenant risk is. It is mostly buying the theme, and the theme is infrastructure.

What It Would Mean for the Credit

So suppose it does not get there. Credit has owned something like this shape before, and learned the rules the expensive way. The asset was rarely the problem. The tenant was. Three lessons to learn from the airline experience.

Structure is most of the outcome. The investors who came through aviation whole owned the secured claim on the asset, not the airline’s unsecured promise. That is the split sitting in this week’s supply. The shell-backed secured paper is the senior secured claim. CoreWeave’s corporate notes are the unsecured lessor bond, a direct claim on the operator. Its existing 9.25% notes gave back a point on the supply while the lease-backed paper held. The market may already be sorting these, even if the spread between them says it has not gone far.

The collateral question starts after the tenant is gone, and the powered shell is better than the bears claim and trickier than the bulls assume. It is not a box of melting silicon; the silicon belongs to the lessor. It is a building, a power connection, and an interconnect, and where power is scarce it can re-lease well. But a shell built to one tenant’s spec in a remote location is not a fungible 737 with a deep global lease market behind it. Recovery turns on whether the next tenant wants that capacity at that site.

The lease is only as good as the tenant’s decision to honor it. Airline creditors lived inside Section 1110 of the bankruptcy code, a short window in which the airline affirmed and cured or handed the asset back. There is no Section 1110 for a data center. If CoreWeave ever restructured, would it affirm the Ellendale lease at that rent, or reject and renegotiate? The bondholders would be downstream of that decision, and the decision would turn on whether its own customers were still paying.

One pattern worth noting, not as an alarm. CoreWeave took a $1 billion equity check and a $6 billion compute commitment from Jane Street, which now reportedly wants to build a data center of its own. Customers funding suppliers so those suppliers can build more capacity for the same customers is not automatically bad. But it is a late-cycle shape credit investors have seen before, in the late-1990s telecom buildout, with vendors financing the carriers that bought their gear.

The Question Is the Tenant

Demand was never the question for airlines. They flew more passengers in 2025 than in any year in history, and the sector had already buried multiple flag carriers. Compute demand can be every bit as real and still leave the credit question open, because what strains an airline is not weak demand but an indispensable product it cannot monetize above its cost of capital.

None of this is a short, which is where the reflexive bears get it wrong. The shell may hold its value, and the Amazon-leased paper may be exactly the utility it is priced as. The point is not to avoid the sector. It is that a landlord renting to a blue chip and a landlord renting to a weak lessor are wearing the same lease at spreads far closer than the gap in tenant risk justifies, and only one of them is a utility.

If there is a discipline here, it is to underwrite the tenant, and the tenant’s tenant, at least as hard as the building. Three questions, which you can ask on any of this week’s prints.

• Is the company paying this lease a hyperscaler that pays regardless, or a neocloud whose survival depends on customers holding compute prices that may be commoditizing?

• Do you own the asset, or just a claim on that tenant?

• And if the lessor failed, would the powered shell re-lease to someone else, or is it stranded at that site for that spec?

The analogy may break exactly where it should. Compute may stay scarce, the labs may find their pricing power, and the economics may close. That is a real possibility, and this is not a call that they won’t. But the question worth keeping in front of you is the one the price war just raised. If the model layer has no pricing power, the pain does not stay with the models. It runs down the lease, through the lessor that can no longer collect, and into the debt that financed the building they all run on.

JBI Bulletin Board

1) Something New is Coming

The institutional version of JunkBondInvestor is coming. The list gets in first and locks the founding rate before it closes. Add your name here.

2) Now Hiring: Credit Analysts

I’m hiring credit analysts to add capacity. If that’s you, or someone you’d vouch for, reply with a short note on your background.

If you liked this, you might also like:

JunkBondInvestor is an independent financial research publication providing general market commentary, corporate credit analysis, and educational content. The information provided is for informational purposes only and does not constitute investment advice, financial planning, or personalized recommendations of any kind. Nothing published constitutes a solicitation, offer, or recommendation to buy, sell, or hold any securities, nor does it guarantee any financial outcome. Any credit opinions or outlooks expressed are solely the independent views of the authors and are intended to reflect general credit trends, not specific investment recommendations. Investing in high-yield bonds, distressed debt, and leveraged loans carries significant risk including potential loss of principal. JunkBondInvestor operates under the Publisher’s Exemption of the Investment Advisers Act of 1940 and applicable state laws. This issue includes sponsored content and partnership links. JunkBondInvestor receives compensation from sponsors for featured placements and may earn referral fees from partner programs. All sponsored content is clearly marked. Editorial views and recommendations are independent and not influenced by sponsor relationships.

That’s the whole pitch. People have to eat. Recession comes, they trade down to canned vegetables, the cash keeps flowing, the bonds pay like clockwork. Boring is the point. Boring is why the money is cheap.

B&G Foods just borrowed money at 11%.

Not a lot of companies issue debt with double digit coupons. The ones that do tend to be in trouble, or close enough that the market wants to be paid for the ambiguity.

Either the market got this one wrong, or the boring food company isn’t boring at all. Which is it?

Situation Overview

B&G Foods is one of those companies nobody has heard of but everyone has bought from. Green Giant, Crisco, Ortega, plus roughly 50 other brands of shelf-stable and frozen food across the US, Canada, and Puerto Rico. All of it parked in the middle of the grocery store, which is exactly the problem.

For two decades, the model was a debt-funded roll-up of orphan brands. Big food companies wanted mature, slow-growth names off their books. B&G took them, plugged them into a shared distribution backbone, spent nothing on innovation, and milked the cash. The cash went to the next deal and to one of the fattest dividends in the food group.

Green Giant came from General Mills in 2015 for $765mm at ~7.8x. Spice Islands and Tone’s came in 2016 for $365mm at ~9.4x. Crisco came from Smucker in 2020 for $550mm at ~8.1x. All of it levered.

The model needed flat categories and cheap debt. It got neither. Consumers drifted out of center-store and frozen toward fresh, private label, and better-for-you, and brands that had been starved of investment for years had nothing to fight back with. Then interest rates went up.

Now the roll-up runs in reverse. Green Giant US frozen went to Seneca in March for $63mm against roughly $200mm of revenue. Le Sueur US fetched $59mm from McCall Farms. Green Giant Canada closes this quarter. Brands bought at 8-9x EBITDA are walking out the door at a fraction of revenue.

When inflation hit in 2022, B&G pushed price hard, +11% for the year, peaking at +17% in Q4. Volume fell 6%, the classic elasticity trade. Then the pricing rolled off through 2023 and the volumes never came back.

By 2024-25, both levers were dead: price flat to slightly negative, volume still eroding 3-5% a year. That’s the signature of a portfolio losing relevance rather than riding a cycle. Shoppers who traded down to private label didn’t trade back up, and brands with no marketing budget have no way to win them back.

Meanwhile, GLP-1s are the newer risk factor. Roughly 1 in 8 US adults is now on the drug, and the data shows treated households cutting grocery spend MSD, with the cuts concentrated in calorie-dense categories. For BGS, the exposure is milder than for the snack and candy names, since most of the portfolio is cooking ingredients. But Crisco sits squarely in the crosshairs, fats and baking, the same SKU already carrying the soybean-oil problem.

What’s left splits four ways, and the segments have almost nothing in common.

Spices & Flavor Solutions is the crown jewel: ~30% EBITDA margins, +9% sales in Q1. The growth is real and it’s volume-led, which makes it the exception in this portfolio.

The rest of the portfolio is what it is. Meals (Ortega) is stable. Specialty, which holds Crisco, hot cereal, and baking staples, is mature and grinding lower. Frozen & Vegetables runs ~6% margins and is mostly sold. Blend it all and you get ~$1.75bn of revenue, ~$280mm of adjusted EBITDA, ~16% margins.

Then the cost problem. Soybean oil, Crisco’s main input, is up more than 50% year over year and sitting near 2022 highs on biofuel demand. Crisco runs a quarterly pass-through that protects gross profit dollars, but management has already said that if oil stays here, pricing goes portfolio-wide.

Which brings us to the balance sheet. Net debt is about $2.0bn, a touch over 7x, under an equity stub worth around $330mm. Cash interest runs ~$163mm. Capex is $30-35mm. After taxes and what’s left of the dividend, FCF for actual debt paydown is maybe $30-50mm a year. On $2bn of debt, that barely registers.

The company also just showed everyone its real cost of capital. Two weeks ago B&G priced $475mm of senior unsecured notes at 11%, issued at ~97, to retire the $509mm of 5.25% notes going current in September. The deal cleared and the 2027 wall is now gone.

The price was steep though. Cash interest on that tranche roughly doubled, from ~$27mm to ~$52mm a year, which consumed almost everything the dividend cut had freed up. The new notes broke below issue and sit in the mid-90s.

So the near-term cliff is handled, but the structure left behind is unforgiving. Three quarters of the debt is first lien, and the next maturity wall is bigger than the one they just paid 11% to clear.

The question now is what a perpetually levered, slowly declining roll-up is worth, with one genuinely good asset buried inside it and a commodity problem bolted to the side.

Here’s what it means for the existing capital structure.

Read more

Soft data prints, everyone reaches for the cut, credit grinds tighter. Bad payrolls meant the Fed was coming. Weak retail sales meant the Fed was coming. Every disappointment got bought because the disappointment was the catalyst for rescue. Good news was good news and bad news was good news too.

Friday broke that.

172,000 jobs against an 85,000 estimate. Roughly double, with the prior month revised up to 179k. The 10-year ripped to 4.53%. Equities had their worst week since October, ending a winning streak that had run essentially uninterrupted, on a payroll number that would have been celebrated in any normal cycle.

The pain trade for 2026 was never recession; it was reacceleration.

A labor market that inflects up instead of rolling over. An AI capex cycle that keeps getting supersized. Energy prices held firm by a conflict that won’t resolve. And a Fed that suddenly can’t ease into any of it. Strong growth stops being a comfort and becomes the risk.

We’re there now. The next move from the Fed may not be a cut anymore. It may be a hike. Maybe soon.

The macro people have been writing “more growth, more inflation” for weeks. The credit market spent those same weeks pricing the opposite. That gap closed on Friday, and it closed in the one way the leveraged credit complex is least equipped to handle: through rates, not through spreads.

Rates First, Spreads Second

The S&P fell 3%, snapping a nine-week win streak. The Nasdaq dropped nearly 5%, the Russell 3%. Crude closed higher on the week. And HY spreads?

They widened just seven basis points, to 265. But HY still returned -0.42% on the week, because the yield moved even if the spread didn’t.

The bigger point may be that the equity move was largely mechanical, not fundamental: a Google $80 billion-plus equity secondary, index providers keeping inclusion rules unchanged ahead of major IPOs, and tech concentration unwinding. A rotation, not a credit event, and spreads treated it as one.

Then look at the dispersion, because it confirms this was rates, not fear.

The worst total-return cohort on the week wasn’t the junk tail. It was BBB, down 0.57%, the longest-duration, lowest-spread paper in the stack, punished precisely because it has the most rate sensitivity and the least spread to absorb it. In a genuine credit panic you’d see the reverse, the CCC tail blowing out while quality holds and everyone hides in BBs. This was the opposite.

The forward curve spent the back half of last year pricing meaningful cuts. By Friday it was pricing a real chance of a hike by year-end. Seven FOMC participants already see zero cuts in 2026. Waller has gone from advocating cuts to arguing the Fed should drop the easing bias entirely; Logan has openly floated that rates may need to go higher.

The Two Economies, One Rate

None of this resolves, because the thing driving the reacceleration is structurally insensitive to the tool the Fed would use to slow it.

The US is running two economies on one policy rate.

Economy One is AI infrastructure, hyperscaler capex, and the power buildout underneath it. Commitments to spend north of $750 billion on capex in 2026 alone. The Dallas Fed’s high-frequency tracker has GDP running at 3.23%.

And critically, this is no longer a pure capex story. Dell’s AI server revenue is up more than 750% year over year. Enterprise adoption is translating into real, booked demand. This part of the economy does not care what Powell does with the funds rate. It is spending regardless.

And here’s the part the soft-landing crowd keeps getting wrong: this boom is stoking employment and inflation at the same time. There’s still no real evidence of AI-driven job losses in the data. It’s Jevons paradox playing out in real time, cheaper compute creating more demand, more buildout, more hiring. The capex that was supposed to be deflationary is, right now, inflationary. That’s why a 172k payroll print isn’t a fluke. It’s the mechanism.

Economy Two is everyone who borrowed at 6-8x assuming SOFR normalizes lower. Housing frozen. Building products cyclically impaired. Consumer-facing credits grinding. Every quarter rates stay higher than the LBO model assumed is another quarter of cash flow eaten by debt service instead of equity cushion.

The Fed has one rate for both. Too low for the first, too high for the second. Cut, and you pour fuel on an AI-driven nominal growth fire while energy is already firm. Hold or hike, and you keep grinding down the levered cohort. The “Fed put” that underwrote credit since 2008 is functionally suspended, and a lot of spreads still price it as active.

And the inflation side is getting stickier, not softer. Even if the Strait reopens tomorrow, the effects persist. Inventories drawn down during the conflict have to be rebuilt, which is incremental demand after the disruption ends. Firms and governments are rebuilding around energy security: bigger buffers, redundant supply chains, diversified sourcing. All of it embeds cost.

Layer the AI power-demand curve on top, with data-center bottlenecks bidding up transformers, grid connections, construction labor, and cooling, and the pressure spills well beyond a narrow set of sectors. Structurally firmer, not transitory.

The AI Narrative Is Turning Quietly

Here’s the part that complicates my own software thesis: Software credit bounced.

The loans had their best month in nearly a year in March, and the re-rating has continued. The market is revising down the speed of AI disruption while continuing to reward the infrastructure layer. Semiconductors and data-center beneficiaries keep getting paid. Software incumbents are clawing back.

I don’t think this is the resolution of the terminal value question. I think it’s the market repricing the adoption timeline, and there’s a real argument underneath it.

The whole infrastructure trade assumes more compute is unambiguously good. But more capable models also consume more compute, and unlike traditional software, AI carries a meaningful, ongoing cost to serve each user. The economics depend on whether productivity gains scale faster than the cost of generating them.

There are already tentative signs that cost is becoming the constraint: enterprises scrutinizing AI budgets, surprise token bills, agentic workflows that don’t always survive contact with the unit economics.

If AI proves more expensive to deploy than the bulls assume, adoption gets slower and more selective. That’s bad for the infrastructure credits financing buildings against shifting revenue timelines, and paradoxically good for the software incumbents everyone spent the winter selling, because the disruption clock extends too. Both sides of the trade are betting the buildout happens on schedule. It might not.

So my framing on software hasn’t changed, it’s just gotten more textured. The terminal value question isn’t resolved. It’s deferred. And a deferred problem on a 2027-2028 maturity wall is still a problem.

Two Markets, One Screen

New issue came roaring back. More than $13 billion in the first three days. On volume alone, a wide-open market.

The composition and the aftermarket explain the actual situation.

QXO tapped the market for its roughly $17 billion acquisition of TopBuild. The bond piece cleared as an evenly split $3 billion deal: 6.50% five-year and 6.875% eight-year, both at par, both at the tight end of talk. Both tranches held a ~101 handle after the break.

Now hold that against the rest of the sector. The names I’ve been tracking on the distressed side, Cornerstone Building Brands, Jeld-Wen, US LBM, Cabinetworks, are a different animal. PE-owned, financed in leveraged loans, levered for a rate environment that never showed up and earnings exposed to exactly the commodity inflation we’re now staring at. Fuel and input costs will pressure margins across that group into the second half.

Two building products markets, running in parallel. The well-capitalized acquirer prints at par. The over-levered 2021 vintage sits in the 70s waiting for a spring selling season the macro just made harder. The difference isn’t the sector. It’s the capital structure and who owns the equity.

If QXO is the market open and cheap for the right borrower, B&G Foods is the other side of it. The food company paid 11.625% to refinance a 2027 maturity and still watched the bonds fall to the mid-90s, and that coupon came after it had already halved its dividend. That’s the point. The market is open, but for the wrong borrower it’s open after the dividend cut, after the deleveraging promise, and still at an 11-handle.

What to Watch: CPI Wednesday

The verdict on all of this lands Wednesday.

The desks are penciling May CPI at +0.52% on the month, which pushes the year-over-year rate to roughly 4.2%, a three-year high. Energy goods, mostly gasoline, up around 8%. Core a more contained +0.23%, but airfares set to jump on the jet-fuel surge and the Spirit Airlines bankruptcy. Gas is back above $4 a gallon for the first time since 2022.

A 4.2% headline doesn’t give the Fed room to validate the cuts still half-embedded in the curve. It does the opposite. And it’s not just a US issue: the ECB looks set to start hiking, 25bp with a hawkish bias, and the Bank of Canada is holding at 2.25% with the path tilted to hikes. This is a Treasury-supply and term-premium story before it’s a credit story.

If CPI prints hot, the rates move that drove this week extends, and the BB complex takes the next leg of damage while the short-duration, high-carry tail keeps outperforming. If it surprises soft, you get a relief squeeze into a heavily short book. I lean to the hot side on the data, but I hold the squeeze risk in mind.

And the oil side just got worse, not better. Crude finished the week higher, and the comment that matters came from Trump: the blockade may not lift until Labor Day. That pushes any reopening of the Strait well past the point where it would help.

One thought I keep coming back to.

For two years, this market was conditioned to believe that weakness would always be rescued, because it always had been. The thing that breaks that conditioning isn’t a recession. It’s strength, the kind the Fed can’t lean against without making the inflation worse.

That’s the uncomfortable shape of this cycle. The economy is fine. That’s the problem. The good news is the bad news now.

JBI Bulletin Board

1) Something New is Coming

The institutional version of JunkBondInvestor is coming. The list gets in first and locks the founding rate before it closes. Add your name here.

2) Now Hiring: Credit Analysts

I’m hiring credit analysts to add capacity. Buy-side experience required, ideally software or building-products coverage. You know the names, the docs, and the playbooks. If that’s you, or someone you’d vouch for, reply with a short note on your background.

If you liked this, you might also like:

JunkBondInvestor is an independent financial research publication providing general market commentary, corporate credit analysis, and educational content. The information provided is for informational purposes only and does not constitute investment advice, financial planning, or personalized recommendations of any kind. Nothing published constitutes a solicitation, offer, or recommendation to buy, sell, or hold any securities, nor does it guarantee any financial outcome. Any credit opinions or outlooks expressed are solely the independent views of the authors and are intended to reflect general credit trends, not specific investment recommendations. Investing in high-yield bonds, distressed debt, and leveraged loans carries significant risk including potential loss of principal. JunkBondInvestor operates under the Publisher’s Exemption of the Investment Advisers Act of 1940 and applicable state laws. This issue includes sponsored content and partnership links. JunkBondInvestor receives compensation from sponsors for featured placements and may earn referral fees from partner programs.

Why can’t someone just vibe code this? Why would anyone pay for it?

And you’ve got the rebuttal ready. The switching costs. The integrations with a hundred counterparties nobody outside the company can see. The years of embedded process knowledge no language model reconstructs over a weekend. You’ve given that speech more than once.

Here’s the first test case.

Avis has been a Verra Mobility customer for twenty years. In late May, it terminated the contract, effective September, and decided to run its tolling itself. Vibe coded, insourced, partnered out, nobody’s confirmed which. Doesn’t really matter. The stock fell 70%. The bonds gave up a dozen points, near par into the mid-80s.

And the bigger question isn’t even Avis. It’s the two customers still in the building. Hertz and Enterprise renew in 2027. If Avis can walk and do this itself, why can’t they?

The equity has already picked a side. It’s pricing all three gone. The credit hasn’t moved nearly as far, which means the two markets are looking at the same company and disagreeing about how this ends.

So is the selloff overdone, and if there’s a trade here, where in the cap stack does it sit?

Situation Overview

Verra Mobility is one of those businesses most people have interacted with without realizing it.

If you rent a car, drive through a toll, and later see a toll-related charge on your rental bill, there is a decent chance Verra was somewhere in the middle. The company helps rental car companies and fleet operators manage tolls, violations, billing, payments, and the administrative mess that comes with moving millions of vehicles across different toll roads and jurisdictions.

That is Commercial Services, the crown jewel. Verra matches a toll event to the right vehicle, the vehicle to the rental agreement, and the agreement to the renter, then runs the billing, collections, and dispute handling behind it. Unglamorous work, and harder to run than it sounds.

Government Solutions is the other real business: automated traffic enforcement for municipalities, the red-light and speed cameras, school-zone and bus-lane programs, anchored by New York City. Lower margin than Commercial Services, more exposed to politics and regulation, but sticky once the cameras are in the ground. Parking Solutions is small and not why anyone owns the stock, so set it aside.

Before the recent drama, the Verra story was that this was a boring, high-margin, infrastructure-like business. Commercial Services was not the largest segment by revenue, but it was the profit engine. In FY’25, Commercial Services represented roughly 45% of revenue but nearly 70% of EBITDA. That margin profile is what made the company look decent.

The pitch was that Verra had built a platform too complex to replicate: a fragmented tolling landscape, consumer-level billing layered on top of fleet tolling, and years of plumbing into authorities nobody else had wired up. Investors treated that complexity as a moat.

Then Avis Budget terminated its contract.

In late May, Verra announced that Avis had sent a termination notice effective September 2026, ending a relationship of almost twenty years. Management put the hit at roughly $135mm to $145mm of annualized Commercial Services revenue and $120mm to $125mm of segment profit, before any cost reductions.

The problem was never just the lost revenue. It was what Avis was saying about the business. If one of the three largest rental car companies thinks it can insource this, the next question writes itself: why not Hertz, why not Enterprise, both up for renewal in 2027.

The reaction was brutal. The stock fell about 70%, the CEO stepped down a week later, and the credit stopped trading like a simple multiple reset.

The first-lien term loan dropped into the mid-80s. The 2029 notes sold off too. What the credit is pricing now is not the Avis earnings hit. It is the possibility that the highest-margin segment has been structurally impaired.

Two ways to read it. The bull read is that Avis is underestimating the work, hits billing errors and authority friction, and either crawls back or quietly proves the platform was worth paying for, in which case Hertz and Enterprise stay and the worst you get is harder pricing. The bear read is that Avis does not need a perfect clone, just a good-enough internal solution that keeps more of the economics for itself, and once that exists Hertz and Enterprise hold a much stronger hand and the old Commercial Services margin is gone for good.

Which one you believe is the whole call, and it is a credit question now, not an equity one. If the remaining customers stay, Verra is still a cash machine and the debt is money-good. If both eventually follow Avis, you are no longer valuing a premium tolling platform, you are valuing a lower-margin government enforcement business with a damaged commercial segment. That range of outcomes is wide enough that the equity and the credit have drifted apart, and that gap is where the trade lives.

Can Avis just vibe-code this?

This is the question the whole selloff hinges on, so it is worth being precise about what Verra actually does.

Read more

This week the sort broke, in both directions.

Verra, the kind of recurring-service credit everyone had in the safe bin, got bombed this week. Rackspace, a services name the market had left for dead, has spent the spring getting saved. Same force, opposite outcomes.

The bins do not hold. The mechanism underneath them does, and it is the better cut anyway.

The Safe Name That Broke

Until now, AI disruption in credit has been a slow story. Terminal value, net revenue retention, who loses pricing power over the next three years as the models improve. Erosion you could see coming.

Verra is something faster.

Avis terminated its tolling contract, a high-margin slice of Verra’s commercial services and more than a tenth of total revenue, effective September. The equity was cut in half on the day. The bonds had traded near par. They fell into the high 80s. Suddenly and violently.

The headline 2026 numbers barely moved, because the loss only lands for one quarter of the year. The real damage is a 2027 event, and 2027 is also when Verra’s other two big rental-car customers, Hertz and Enterprise, renew. The three contracts together are something like 80% of commercial services and a third of the company. So the market did not reprice this year’s print. It repriced whether Avis is one-off or the first of three, a year before the income statement will say.

The market’s shorthand for the fear is vibe coding. But nobody has confirmed Avis is even building this in-house rather than just switching to a competitor, and the real issue was never whether an AI tool can write the software. It is what one termination did to belief. A workflow the market had underwritten as mission-critical plumbing got re-underwritten, overnight, as something a determined customer might rebuild. The cause of the cancellation barely matters. The repricing of substitutability is the event.

Verra’s defense is that the moat was never the code. It is the plumbing. Millions of vehicles, hundreds of toll authorities, and the violations, billing, disputes and registration machinery that took years to wire together. AI writes code. It does not stand up that network over a summer.

The bulls will tell you the government and parking businesses carry the debt even if every rental-car dollar walks. Maybe. The point is not who wins that argument. It is that the market repriced the credit to stressed before anyone could have it.

The Doomed Name That Didn’t

The names that screen as most exposed did the opposite of Verra this week. The most-active paper in high yield was CoreWeave and the rest of the AI-infrastructure complex, which has printed oversubscribed in primary all year. For those credits AI is not a threat to the revenue line. It is the revenue line. The more AI there is, the more compute, memory and power the customer buys, and the better the credit gets.

That is the easy half of the pole. The harder half is the inversion, and Rackspace is the cleanest version of it I have seen in credit.

Rackspace ran an LME two years ago to cut debt and buy liquidity. Earlier this year the debt traded around 30 cents, and it fell further in February in the same AI selloff I was writing about then, priced as one more business about to be disrupted into the ground. The company used the weakness to buy back nearly 100 million of its own paper.

Then the story turned. A non-binding MOU with AMD in May to build an enterprise AI cloud, the companies explicit that talks were preliminary, on top of earlier tie-ups with Palantir and Uniphore. The secured debt ran from the 50s to north of 80. Look at what moved it. Not contracted revenue. A framework. The bonds repriced optionality, not cash flow, which is exactly the convexity you want in a name the market had buried.

The driver is the cleanest consumed mechanic there is. AI moving from training to inference broadens demand for exactly the GPU and CPU workload management Rackspace sells. Roughly 2.5 billion of 2028 maturities still hangs over it, so this is not finished. But it is a credit the disruption tape left for dead and the consumption tape brought back.

Two credits, opposite outcomes. The sector label called neither. The mechanism called both.

Consumed or Replaceable

Here is why the AI-exposed versus AI-safe binary fails. It sorts on sector. The thing that actually determines the credit outcome is mechanism.

The real axis is whether AI makes your product more consumed or more replaceable.

Consumed means your product is the surface the models run on, or the thing AI creates more demand for. Compute. Memory. The data layer. The fraud defense. The more AI there is, the more of you the customer needs.

Replaceable means your product is a workflow, an output, or a labor arbitrage that a motivated customer can now reconstruct. The proprietary report. The back-office process. The seat of a credit analyst ;)

This axis cuts straight across the old line. Rackspace sits on the consumed side. Verra is being accused of the replaceable one. NIQ, Kantar, Gartner, Accenture, the running joke about advisors replaced by Claude, none of them software, all of them priced as replaceable.

The diligence question is no longer how sticky is the revenue. It is which side of the axis the business sits on.

What the Week Said

The headline index did almost nothing. High yield returned about 0.55% on the week. Spreads ended flat around 257 bps, the ten-year hovered near 4.44%, and the all-in yield came in 16 bps to 6.96%, with effectively all of that move coming from rates rather than spread.

Underneath, the dispersion that the index always hides. Flight to quality within high yield, the B-versus-BB differential pushing toward the top of its historical range. The rating buckets are pulling apart while the headline sits still. The index is lying to you, again.

Equities went the other way and went hard. The Nasdaq 100 cleared 30,000, Micron crossed a trillion dollars for the first time with its high-bandwidth memory already sold out for the year, and May closed up double digits for a second straight month.

The two poles sat on the same leaderboard. Avis’s bonds were among the most-traded names in high yield this week, right next to the AI-infrastructure complex. The tape that bid compute to the top of the volume list took the name AI was used to threaten from near par into the high 80s.

Which is the problem with owning high yield as a number right now. You are being paid less than you have all cycle for a tail that is quietly getting fatter.

The Refi Tax

Four billion priced across five deals, bottom of the syndicate range, a holiday week pinched around the GS leveraged finance conference.

The theme was the cost of capital, and it cut both ways. The low-coupon maturities got repriced sharply higher. Navient paid 9.375% to take out 6.75s. PBF came 7.25% to refinance 6s. Avis printed its 8% add-on at 7.77% to chip at the 5.75% 2027s.

And the names refinancing down still landed expensive. Borr Drilling upsized to 2.035 billion of secured to retire 10% and 10.375% paper, and the best it could print was 8.75% and 9%.

Whichever direction the refi ran, everyone cleared at eight to nine and up. The maturity wall is not a wall. It is a repricing, and it lands hardest on exactly the levered names the feature says AI is already pressuring.

And before anyone reads consumed as safe. The most consumed credits alive, the AI-infrastructure names, are also the most capital-hungry, rolling enormous capex into this same eight-handle market. In the third quarter the largest tech firms spent more on capital investment than they earned from operations.

The revenue tailwind sits on top of a balance-sheet hole. Consumed protects the top line. It does nothing for the refi. It is a second tail, not the absence of one. And the rate relief that would close either tail may not be coming, for the reasons I laid out last week.

The Screen Worth Running

Stop sorting the book by AI-exposed and AI-safe. Re-sort it by consumed versus replaceable. The mislabeled names are where the edge lives: a credit whose real moat is operational plumbing, priced by the tape as if it were replaceable code.

The replacement-fear overshoot is a buy list, not a short. When a credit gaps on a replacement headline and the customer that pulled the contract is itself running negative EBITDA, that is convexity, not confirmation.

Fade nothing in AI infrastructure on valuation alone, but respect that the most-traded paper in high yield is now the most consensus. Crowded is not the same as wrong. It is just crowded, and it gets repriced faster when the story wobbles.

And do not confuse a melt-up with a healthy credit tape. The equity index is green because ten names are green. Underneath it the buckets are separating, and the headline is the last place you will see it.

JBI Bulletin Board

1) Built for the Desk

The institutional version of JunkBondInvestor is coming. The list gets in first and locks the founding rate before it closes. Add your name here.

2) Now Hiring: Credit Analysts

I’m hiring credit analysts to add capacity. Buy-side experience required, ideally software or building-products coverage. You know the names, the docs, and the playbooks. If that’s you, or someone you’d vouch for, reply with a short note on your background.

If you liked this, you might also like:

JunkBondInvestor is an independent financial research publication providing general market commentary, corporate credit analysis, and educational content. The information provided is for informational purposes only and does not constitute investment advice, financial planning, or personalized recommendations of any kind. Nothing published constitutes a solicitation, offer, or recommendation to buy, sell, or hold any securities, nor does it guarantee any financial outcome. Any credit opinions or outlooks expressed are solely the independent views of the authors and are intended to reflect general credit trends, not specific investment recommendations. Investing in high-yield bonds, distressed debt, and leveraged loans carries significant risk including potential loss of principal. JunkBondInvestor operates under the Publisher’s Exemption of the Investment Advisers Act of 1940 and applicable state laws. This issue includes sponsored content and partnership links. JunkBondInvestor receives compensation from sponsors for featured placements and may earn referral fees from partner programs. All sponsored content is clearly marked. Editorial views and recommendations are independent and not influenced by sponsor relationships.

The category has been a graveyard for a decade. E-commerce hollowed out anything you could buy without touching it, store leases turned into a fixed cost that crushes you the moment comps go negative, margins are thin enough that one bad season tips the model, and the list of names that went from investment grade to liquidation is long enough that the reflex to pass is rational.

So you skip it. You see a fragrance-and-candle retailer with declining sales and you move on.

Bath & Body Works gives you every reason to do exactly that. Comps have been negative for quarters. Body Care is bleeding share. The CFO is walking out mid-turnaround. On the surface, this is the same retail story you’ve passed on a hundred times.

But the debt might not need the turnaround to work. And it’s trading wide. So the question is whether that’s a mispricing or the market getting it right.

Situation Overview

Bath & Body Works (Ba2/BB+) is a specialty retailer of personal care and home fragrance. You know the brand: the body lotions and mists, the hand soaps and sanitizers, the candles and Wallflowers that anchor the Home Fragrance line.

came out of the old L Brands in the 2021 split that also created Victoria’s Secret. The result left a clean, pure-play operator with ~1,900 company-owned stores across the US and Canada, an e-commerce channel it calls Direct, a franchised international fleet that costs almost nothing to run, and, since February, wholesale distribution on Amazon.

Three categories carry the business.

• Body Care is the core, the franchise the brand is known for.

• Home Fragrance is the candle business, weighted to the holidays.

• Soaps and Sanitizers is smaller and steadier.

From a lender’s perspective, the credit appeal is these are consumable, replenishment categories, so the revenue base holds up better than apparel. Customers come back on a cadence, gross margins sit in the low-to-mid-40s%, and the business throws off $600-800mm of FCF in a normal year.

Fragrance is bought after sampling, which is why these stores never got hollowed out the way commodity apparel did. That cash flow profile, not the growth, is what keeps this retailer at BB-ratings instead of single-B.

What’s been happening

For most of the past year the story hasn’t changed: comps down, and getting worse. 1Q’26 kept the trend going. Net sales dropped 3.2% to $1.38bn, and Adj. EBITDA declined about 22% to $212mm. Gross margin gave up nearly 270bps, pressured by tariffs, inflation, and the deleverage of spreading a fixed cost base over a smaller top line.

The market actually read it as a relief print. Expectations were on the floor, guidance held, the 2Q’26 guide came in no worse than feared, and the equity rallied.

Taking a step back, nothing in the trajectory actually moved. Sales are still falling, and the recovery is still something management is describing rather than delivering. They have pointed to a second-half inflection for several quarters now, and the inflection has not arrived.

What’s actually broken

Almost all of the weakness sits in one place. Body Care, the heart of the brand, was down “mid-teens” and has now decelerated three quarters running while the rest of the business held roughly flat. The core is the part that is bleeding.

Read more

That is the narrative investors want to believe. Software agents replace back-office labor. Models automate workflows. Productivity rises. Services inflation eventually rolls over because expensive humans get swapped for cheaper code.

Maybe. But before AI becomes deflationary, it has to become extremely capital intensive.

It needs data centers. Data centers need land, power, transmission, substations, transformers, cooling, backup generation, fiber, gas turbines, construction labor, and years of utility planning. The model may live in the cloud, but the cloud still needs a grid connection.

That is the part the Fed cannot solve.

Rates do not fix transformer lead times. They do not clear transmission queues or build gas turbines faster. They do not unwind hyperscaler capex commitments. The Fed can slow demand at the margin, but it cannot print power infrastructure.

That is why the AI trade is starting to become a credit one.

Equity Sees Growth. Credit Should See Capex

Equity investors have treated the AI buildout as a clean growth story. Nvidia gets the multiple. Hyperscalers get the strategic narrative. Power assets get re-rated. Data center landlords get the secular demand story. The entire complex has moved from “infrastructure” to “AI infrastructure,” which is the kind of label that usually makes capital cheaper until everyone remembers capital still has to be repaid.

Credit should be less comfortable with this.

The first stage of the AI buildout is not software margin expansion. It is an industrial capex boom running into a constrained physical system. Power is scarce. Equipment is scarce. Interconnection is slow. Permitting is political. Utility bills are already a cost-of-living issue.

The numbers back this up. Consensus 2026 AI capex estimates are around $750 billion (I’ll take the over on this). Power demand is rising materially for the first time in decades. Grid operators in Texas, Virginia, Georgia, Arizona, and parts of the Midwest are warning that hyperscaler buildouts are reshaping capacity forecasts. Utilities are filing rate cases citing data center demand. Transmission investment requirements are accelerating.

This is not some abstract long-term problem. It is the exact type of supply-side constraint that keeps inflation sticky and makes the Fed’s job harder.

I wrote about this a couple weeks ago through the 1990s telecom analog. Supply doesn’t widen spreads mechanically. It builds the vulnerability. The cracks show up in the marginal cohort first.

Credit Is Still Behaving Like the Rate Environment Doesn’t Matter

You have seen the front-end repricing.

Front-end pricing for 2026 has flipped from cuts to hikes in a matter of weeks. The move in the implied policy path is on the order of 80 bps. The rates market has done most of the work. Credit has barely moved. High yield spreads are actually tighter over the same period.

Those two things do not sit comfortably next to each other. The rates market is saying the Fed may not have room to cut. Credit spreads are still behaving like the Fed path does not matter much.

The incoming data is easier to square with the rates move than the spread move. April PPI ran hot at 1.4% m/m and 6.0% y/y, and the pressure was not isolated to energy. Goods passthrough showed up across transportation, metals, building materials, and industrial chemicals. ISM prices indices stayed elevated on both the manufacturing and services side, with services prices back near the highs of the current cycle. The ADP weekly pulse firmed into a monthly run-rate well above what the Fed itself characterizes as breakeven payroll growth.

None of these are individually decisive. Together they do not describe an economy giving the Fed cover to cut. None of this is obvious, especially in BB spreads at 152 bps.

HY index spreads now sit 10bps wide of their 52 week lows. Meanwhile, the 10-year Treasury is up nearly 50 bps YTD at 4.56% vs. the HY index yield at 7.12%. The attribution is not flattering: returns have been carried by coupon, while rates have been a drag and spread compression has done very little.

Carry is still printing. That part is true. The issue is whether the conditions that made carry easy survive another quarter where AI capex keeps pulling on the same physical inputs the Fed cannot conjure.

The Market Is Open, But the Price Discrimination Is Real

A Fed stuck between sticky inflation and political pressure does not shut the new issue market. It splits it.

RR Donnelley priced an upsized $900 million five-year senior PIK note. 11% coupon, 95 OID. Fitch moved the outlook negative on higher pro forma leverage in the low-5x area.

Same week, Granite Construction priced $600 million of 6.375% eight-year senior notes at par to refinance old converts.

Granite at 6.4%. RRD at 12.4% PIK. Same week. Roughly 600 bps apart on clearing yield.

The market is not closed. Strong credits refinance. Quality borrowers extend maturities. HY spreads stay tight. Weaker borrowers can still get deals done, but the price starts to look like punishment.

The interesting part is that AI is now on the other side of that triage. The market has infinite imagination for the capital needs of the AI buildout. It has much less patience for old-economy borrowers trying to refinance levered balance sheets at the wrong point in the rate cycle.

AI gets the growth multiple. Legacy refi risk gets the coupon.

Power Is Where the Politics Starts

Certain credits have treated the AI power complex as a clean tailwind.

Talen, Vistra, NRG, and Constellation traded as utility or power names two years ago and now trade as AI infrastructure plays. CoreWeave moved from novelty borrower to somewhat accepted institutional credit. Bitcoin miners, electrical equipment names, power generators, utilities, and cooling assets all sit somewhere on the same narrative map.

The demand story is real. The supply response is slow. That is why the trade worked. The next phase is harder.

US electricity demand is rising for the first time in decades and ratepayers are being asked to absorb more of the bill. Iran has pushed gas prices higher. Power bills are climbing too, partly because hyperscaler buildouts are driving demand growth that utilities pass through to customers.

The cleaner AI story gets harder when the power bill leaves the data center and lands in the regulated rate base. A hyperscaler PPA may be contracted. The household bill is political.

The cost-of-living backdrop is already strained. Real incomes have not kept pace with prices for most households, inflation approval is in the high 20s for the administration, and energy is back in the headlines as a household-budget item rather than a macro abstraction. Higher power bills hit the same working-class base through a different channel than the pump.

The political coalition that has tolerated the AI buildout starts to look less stable when the buildout shows up as utility bills, transmission fights, land-use battles, and local permitting pressure.

The issue is whether the local cost allocation case becomes politically intolerable before the productivity gains show up in household incomes. That is a credit problem before it is a macro problem, because the names exposed to local cost allocation fights are levered, regulated, and trading like the tailwind is one-directional.

That is not an anti-AI argument. It is the politics of who pays.

The PUC Is Where the AI Trade Meets the Customer Bill

State Public Utility Commissions (“PUCs”) are political animals. They feel pressure faster than federal regulators.

In states like Virginia, Georgia, and Ohio, data center cost allocation is already becoming a live regulatory issue. The direction of travel is toward less socialization, more scrutiny of large-load tariff structures, and higher customer contributions to interconnection costs. Virginia rate cases over the next two quarters are worth watching closely.

If the politics turns, the IPPs and utilities are probably easier pressure points than the hyperscalers. Hyperscalers bring jobs, capex, lobbying capacity, and federal influence. Utilities and IPPs sit closer to the customer bill, which makes them easier to drag into the state-level fight over who pays for the grid.

That does not mean the demand story is fake. It means the easy “AI infrastructure is a one-way tailwind” phase is probably over. The next phase has more regulators, more politics, and more attention to who absorbs the cost.

The question is whether these AI-credits should trade like a clean infrastructure credit, or whether the market eventually demands a larger concession once power, permitting, and regulatory risk are priced more explicitly. Same question across the IPPs, data center names, and power-linked credits. Some have contracted demand. Some have merchant exposure. Some have regulatory protection. Some have regulatory risk.

The market is still learning the difference.

What This Means for Credit

The first AI trade was Nvidia and the direct GPU ecosystem. The second was power. The third may be higher funding costs for everyone not attached to the buildout, and a political ceiling on some of the names that are.

That matters for broader credit because the AI capex cycle is not happening in a vacuum. It is happening while inflation is sticky, the Fed path has repriced hawkishly, and spreads are still trading like carry is the only thing that matters.

There are three implications:

Read more

Most are pretending the question doesn’t exist. Wix decided to do something about it. Went out and spent on the answer. Bought Base44. Built Harmony on a proprietary LLM. Pushed distribution into ChatGPT and Claude. Ran a Super Bowl ad.

You can see the spending in the numbers. You can also see it in the share price. Down 48% this year alone.

The convert has moved but not to the same degree. Busted and trading on yield. High-70s, ~6% YTM, priced like IG-adjacent software paper. The question now: is 6% enough?

Situation Overview

Wix sells subscription software that lets SMBs, freelancers, and agencies build and run websites without writing code, plus a layer of business apps on top (payments, commerce, bookings, domains, email, paid ads). Mostly recurring, paid upfront, low capex. Until recently it ran at low-to-mid 20s% FCF margins with a clean balance sheet.

Three things changed that.

The first is AI. The customer question shifted from “I need a website but can’t code” to “I should be able to describe what I want and have software build it.” Wix’s attempt to answer that is two products:

• Wix Harmony is an AI site builder integrated into the core platform that launched in late Jan’26.

• Base44 is a “vibe coding” tool that lets non-technical users describe an app or workflow in plain English and have software build it. Wix acquired Base44 in mid-2025.

Base44 has scaled fast: $59mm at YE’25, $100mm by early Mar’26, $150mm by mid-May’26. Harmony now runs on a proprietary in-house LLM, which management has described on the Q1’26 call as meaningfully lower cost than 3P alternatives. Wix is also pushing distribution into AI platforms, including Base44 inside ChatGPT and MCP integrations with tools like Claude and Cursor.

The pitch is that these products move Wix out of website builder land and into a bigger AI-powered web, app, and workflow market.

The second change is that the cost of all this showed up in Q1’26.

Revenue grew 14% y/y to $541mm. Bookings grew 15% to $585mm. New user cohort bookings up ~46%, with the Q1’26 cohort adding ~6mm new users. Headline looked fine. Go down the P&L. Adj. EBITDA was $34mm (6% margin) vs. $106mm (22% margin) in Q1’25, well below consensus.

S&M nearly doubled y/y to $191mm, or 35% of revenue vs. 21% a year ago. That includes $20mm+ of Super Bowl spend. Per management commentary, the other ~$90mm went to acquisition marketing across core Wix and Base44, with blended TROI extended to 7-9 months. FCF was $75mm, $112mm ex-acquisition costs, against $142mm in Q1’25.

The stock fell 27% on the print.

Guidance moved with the print. FY’26 revenue and bookings stayed at mid-teens, but FCF margin guidance (ex-acquisition costs) was cut from low-to-mid 20s% to high-teens. Management attributed ~$100mm of that to foregone interest income and new interest expense from the tender, plus a ~$64mm FX headwind from a stronger Israeli shekel weighted to 2H’26. None of those explanations are wrong. But taken together, they describe a business that now requires more marketing spend and more capital.

The third change is the balance sheet.

In Apr’26, Wix completed a modified Dutch auction tender, buying back ~17.5mm shares at $92 for ~$1.6bn. Almost 30% of shares outstanding. Funding came from cash on hand and a new $500mm credit facility from Bank Hapoalim. The transaction left Wix with roughly $1.65bn of gross debt, including the new $500mm bank facility and $1.15bn of 2030 converts, against the net cash position the company carried for years. Two months later the stock is at $55.

The facility terms are more interesting than the headline borrowing capacity. Hapoalim got liquid assets sitting with the lender, contractual and statutory lien and setoff rights, a negative pledge against a general floating charge, and a financial covenant. Specifically, Wix has to maintain ~1bn NIS of Israeli Treasury Bills (Makam) at the bank and keep its existing $120mm cash deposit there, with a Bank Debt / FCF covenant capped at 2.0x.

Read more

The answer often matters more than leverage, price, or all-in yield, and almost nobody focuses on it.

Credit markets have always worked this way. Different cohorts hold paper under different rules. Different rules mean different prices when paper has to move between them. The handoff between buyer bases has always been where price discovery happens on stressed names.

What’s different right now is the degree. The buyer base is more fragmented than it used to be. Private credit and HY ETFs are both meaningfully larger than they were five years ago. Active managers are smaller. CLO reinvestment cohorts have shifted. The crossover IG bid has gotten more conditional. Several absorbers in the chain thinned at the same time, and the price gaps at each handoff have widened as a result.

The credit may have changed, but the degree to which that price discrimination gets expressed is much wider now. Dispersion grows the smaller and shallower the next buyer universe is.

Most market commentary works off the spread because that’s what’s visible. The headline gives you the index, the rating bucket, the sector. It doesn’t give you who actually owns the paper and what would happen if they had to sell. Spread levels are an output. The buyer base is the input.

Today, HY spreads remain inside 300bps. The headline gives you nothing. What’s actually happening sits below it.

Where the price gap opens

Take a single-B loan trading in the mid-90s, owned mostly by CLOs in their reinvestment period. The CLO can hold it, add to it, reprice it. Spread reflects the underlying credit risk plus a modest liquidity premium. The buyer base is deep and sticky.

Downgrade it one notch. The company is still the business it was yesterday. Inside the CLO, the math changes. OC tests start mattering more. If the loan slips below par triggers or migrates closer to the CCC bucket, it begins to weigh on the structure rather than support it. Managers tight on their tests cannot easily add to it. Some need to trim. The largest natural buyer just got smaller.

The next bid is no longer the CLO at the same intensity. It is a credit hedge fund, a multi-strategy book, or a newer-vintage CLO with cushion to spare. Each requires a different price. The loan slips several points. The catalyst was the downgrade. The price gap was created by the owner base changing.

Downgrade it again. The CLO universe is largely out of the conversation. The natural buyers are distressed funds, opportunity funds, and the small set of credit hedge funds that specialize in this part of the market. The loan slips again. The business hasn’t always changed. The mandate that holds the paper has.

A handoff can produce significant price gaps that underlying fundamentals don’t dictate. The new owner base requires a different price to clear, and the gap between those prices is the dislocation. Sometimes that creates opportunity for buyers who can underwrite the next mandate before the old one fully exits. Sometimes it produces lumpy, choppy price action because there isn’t a natural buyer base at all.

Take broken converts for example. Once a convertible breaks, the equity option is too far out of the money for convert funds and the credit isn’t clean enough for HY managers. Neither side has a real reason to own it. Each fund type that might pick it up operates under its own restrictions, and many of the more flexible vehicles are too small to absorb size.

The depth of each buyer matters as much as whether one exists. A broken convert that nobody is forced to own and nobody is structurally set up to buy will trade where the most opportunistic capital wants it to trade, which is can be below where fundamentals would suggest. I wrote about this dynamic at over the weekend:

The same logic runs in bonds. A BB owned by crossover IG buyers, total return funds, and indexed ETFs has a deep bid. Downgrade it and the crossover bid weakens. Total return funds may still hold it but the marginal buyer is a dedicated HY fund with different concentration limits. Another downgrade and dedicated HY funds become the only natural holders.

Each step is a handoff. Each handoff has a price.

Why the handoffs are harder now

The slack in the system has gotten thinner at every step.

CLOs in reinvestment used to absorb deteriorating paper as a matter of course. Active HY funds had inflows to deploy. Crossover IG could lean in when yields were attractive. Insurance had room. Private credit was growing fast enough to absorb risk the BSL market wasn’t pricing.

Most of that slack is gone.

The class of 2020 to 2022 CLOs has been rolling out of reinvestment. Once a CLO exits reinvestment, it stops being a buyer at the margin. It can sell discretionarily but it cannot grow into deteriorating risk the way it could before. The cohort that absorbed the last wave is no longer the cohort that can absorb the next one.

Active HY fund flows have been one-directional for months. The asset class has seen consecutive weekly inflows, but the composition matters. The incremental capital is heavily ETF-driven. Active loan funds continued the pattern of sustained outflows that has defined the year. ETFs buy the index in proportion to the index and don’t differentiate between credits that should be owned at current prices and credits that shouldn’t be. Active managers do, and active managers are losing capital. A growing share of the incremental HY bid is a passive vehicle that doesn’t care about handoffs.

The crossover IG bid is more conditional than it was a year ago. Conditional on rate cuts arriving, on the curve flattening, on IG yields staying low enough that reaching into BBs makes sense. With the curve now pricing more hike risk than cut risk, that condition is breaking.

Insurance demand for CLO senior tranches remains solid, but spread compression has limited how much new equity formation it can support. CLO equity returns have come down. New CLO issuance ex-refi has been running below last year’s pace. The supply of new CLO vehicles to absorb deteriorating loan paper has slowed.

Most large private credit platforms have hit their quarterly redemption caps this year. The platforms managing through it are doing so by selling their most liquid paper, often the BSL exposure held alongside the private loans. That selling lands back in the loan market as incremental supply at the wrong moment.

Every absorber in the chain has gotten thinner. The depth of each remaining buyer matters more in that setup, not less.

The rates story is the trigger

I’ve been saying for weeks that the macro setup is quietly building its own vulnerability and that the long duration universe is the wrong place to be sitting. Financial conditions easier than pre-conflict despite oil 50% higher. One year inflation swaps trading 75bps higher. The market pricing more growth and more inflation simultaneously and treating it as benign. Labor stabilizing into the same impulse. The Fed not pushing back. The reflation trade leaking into credit while most accounts hadn't repositioned for it.

This week the leak became visible.

CPI came in hot midweek. A global bond selloff hit Friday on UK gilt concerns and renewed inflation worries. OIS forwards now price 60% odds the Fed delivers a hike rather than a cut by December, a complete reversal from where the curve sat two months ago. Even with spreads flat, high yield returned -0.49% for the week driven by the move in base rates.

Crossover IG buyers are the most rate-sensitive cohort in the HY bid and have been a structural absorber for BBs through this cycle. Their willingness to reach into HY for yield depends on a forward curve that suggests rates are coming down. When the curve flips, the math changes. Crossover allocations get pulled back. The deep bid behind BBs gets thinner.

BBs at current spreads do not compensate for the crossover bid stepping back. The credit doesn’t have to deteriorate for BB spreads to widen. The owner base just has to change.

The same logic works down the stack. Higher base rates compress interest coverage on floating-rate borrowers. Loans that were comfortably in CLO portfolios start consuming OC test cushion. The structures that absorbed the original financing become marginal sellers. The handoff cascade starts at the top of the rating stack and works lower.

A higher-for-longer regime functions as a buyer base thesis. The cohorts that absorbed the last cycle were built around an assumption that rates would come down. They didn’t. They aren’t. The owner base will have to reorganize around that reality, and the price will reset at each step.

What this looks like in real names

ZoomInfo is the cleanest example from last week. The company reported Q1, beat on both revenue and earnings, and cut full-year guidance. The bonds went from the mid-80s to a 78 context. Roughly 7 points lower on a quarter that beat. The existing holder base reassessed the terminal value and the next buyer required a different price. The handoff happened in one print. The 3.875% 2029s now yield close to 14%. The loans repriced to the mid-80s in parallel.

The same pattern shows up across building products, consumer cyclicals, business services, and the longer tail of sponsor-backed LBOs from the peak vintage. The single-B paper financed into the BSL market in 2020 to 2022 has been drifting lower for two years. Companies still current. Maturities still in the future. The loans trade where they trade because the buyer base that absorbed the original financing is no longer the buyer base that has to clear the secondary market.

The primary market reinforces the divide. Clean deals for clean issuers cleared without trouble this week. DIRECTV priced $1.4 billion of secured notes at par to yield 9.25%. Wayfair priced $400 million of eight-year secured notes at 7.125%, tight to talk, and the deal triggered a Fitch upgrade on both the secured debt and corporate rating. Kennedy-Wilson placed $1.8 billion across two tranches to fund a take-private. Travel + Leisure cleared $900 million of secured notes at 6.25%.

Every one of those deals priced inside guidance or at the tight end. The natural buyer base for clean BB and strong single-B paper is intact. The handoff problem isn’t in primary. It’s in the names already outstanding where the existing holder base is being asked to keep holding through a setup it didn’t underwrite.

Things to be mindful of

A few questions worth running on the names you already own.

Who owns this paper today, and what are their constraints. What would have to happen for them to step back. Who is the next buyer if they do. What price does that buyer require to show up. How wide is the gap between the current holder and the next one. How deep is each successive buyer.

Work on the company itself still matters. Cash flow, leverage, maturity profile, business model, all of it. The marginal source of return increasingly sits in the handoffs between buyer bases. A credit priced for stability where the current holder is one bad print away from leaning back is the more dangerous setup than one already priced for a handoff.

The harder question is when the handoff happens without a rating change. A CLO manager looking at a weakening loan can choose to trim before the downgrade arrives. The buyer is another holder who hasn’t done the work yet. The price moves before the rating moves. By the time the agencies catch up, the loan has already changed hands and the price already reflects it.

The valuable work is spotting credits where the market will reorganize the ownership before the agencies do, where the current holder is fragile, and where the next mandate has different math than the current one.

A growing share of the incremental HY bid is ETF-driven, which means the bid in the index is firmer than the bid in any individual name. The index holds up while specific credits get punished. Dispersion keeps widening regardless of where the index trades.

The single-B handoff inside the loan market is where the largest price gaps live right now, and the post-reinvestment CLO cohort is worth watching as a driver of softer secondary bids in the back half of this year. The same credit can have a different secondary price depending on which CLOs own it.

JBI Bulletin Board

1) Now Hiring: Credit Analyst

I’m looking to bring on an experienced full-time credit analyst (ideally distressed background). At least a couple of years covering the space, buy-side experience strongly preferred. Software or building products coverage. You should know the names, know the docs, and the playbooks. If you’re interested or know someone who fits, reach out with a brief note on your background.

If you liked this, you might also like:

JunkBondInvestor is an independent financial research publication providing general market commentary, corporate credit analysis, and educational content. The information provided is for informational purposes only and does not constitute investment advice, financial planning, or personalized recommendations of any kind. Nothing published constitutes a solicitation, offer, or recommendation to buy, sell, or hold any securities, nor does it guarantee any financial outcome. Any credit opinions or outlooks expressed are solely the independent views of the authors and are intended to reflect general credit trends, not specific investment recommendations. Investing in high-yield bonds, distressed debt, and leveraged loans carries significant risk including potential loss of principal. JunkBondInvestor operates under the Publisher’s Exemption of the Investment Advisers Act of 1940 and applicable state laws. This issue includes sponsored content and partnership links. JunkBondInvestor receives compensation from sponsors for featured placements and may earn referral fees from partner programs. All sponsored content is clearly marked. Editorial views and recommendations are independent and not influenced by sponsor relationships.