THE BOND BRO DISPATCH · WEEKEND READ · NO. 1
The Same Liquidity, Spent Twice
Thursday, May 28, 2026 · The AI Capex Financing Collision · Premium
Executive Summary
Over the coming months, the largest private companies in the world come to public markets in a single cluster — SpaceX on June 12, with Anthropic and OpenAI both expected to follow later in 2026. The equity market is reading this as an IPO calendar. The bond market is reading it as a supply event. Both are right, and that is exactly the point: the artificial-intelligence buildout is now being financed through two channels at once — equity issuance and corporate debt — and both channels draw from the same finite pool of institutional capital in the same window. The collision is the story. The bond market is pricing it first.
THEME One AI capex cycle, financed simultaneously through the equity and bond markets, with both supply waves cresting in the same mid-to-late-2026 window.
RISK BIAS Constructive on duration into the supply collision; cautious on the AI / growth equity complex through each pricing window.
KEY WATCH Whether the mega-IPO calendar prices clean or soft. Soft pricing is the catalyst that converts equity indigestion into a duration-positive flight to quality.
Why is an IPO calendar a bond-market story?
Start with the scale. SpaceX is targeting a debut as early as June 12 on the Nasdaq under the ticker SPCX, at a reported valuation between $1.75 trillion and $2 trillion, raising somewhere in the range of $40 billion to $80 billion — the largest IPO in history by an order of magnitude. Pricing sits at roughly 58 to 65 times forward revenue, a multiple that prices flawless execution on Starlink monetization and eventual Starship commercialization. Behind it, Anthropic and OpenAI are both expected to tap public markets in the back half of the year. Stack the three together and you have, plausibly, $150 billion or more of mega-cap equity supply landing in a single concentrated window.
That number is the bridge to the bond market. An IPO is a claim on institutional capital. So is a corporate bond. So is a Treasury auction. The equity desk and the credit desk do not buy from separate wallets — they compete for allocations from the same pension funds, insurance general accounts, sovereign wealth vehicles, and asset managers. When the equity calendar gets this heavy, it does not stay contained to equities. It pulls on the same liquidity that has to clear the back end of the Treasury curve.
What is being financed on the other side of the ledger?
The same thing. The AI capital-expenditure cycle is one of the largest private investment programs in modern corporate history, and a growing share of it is being funded in the debt market rather than from internal cash flow. The five major hyperscalers — Amazon, Alphabet, Meta, Microsoft, and Oracle — issued roughly $121 billion in US corporate bonds in 2025, against a 2020–2024 average near $28 billion a year, per BofA, which expects that pace to run around $140 billion annually over the next three years. Citadel projects $70–100 billion of AI-related public corporate debt over the next twelve months; Morgan Stanley sees roughly $200 billion of incremental investment-grade tech supply through 2028. Barclays puts 2026 net investment-grade issuance up more than 30% year over year and names hyperscaler capex as the single biggest upside risk to that number. The names are familiar — Oracle’s capex-driven borrowing, the Meta Hyperion-style special-purpose-vehicle financings that ring-fence data-center assets off the parent balance sheet — but the scale is not.
The framing to hold: the IPO wave and the AI bond wave are not two stories. They are the same capital expenditure, mirrored across two financing channels. The equity market funds the equity slice; the bond market funds the debt slice; and through the second half of 2026, both slices hit the market at once.
Where do the two channels actually collide?
In the demand pool — and the collision surfaces at the back of the Treasury curve. Set the three supply sources side by side: mega-IPO equity supply, AI-capex corporate bond supply, and the Treasury’s own deficit financing, the quarterly refunding that never pauses. Three enormous, concurrent claims on one pool of institutional capital.
This is Force 2 of the Pressure Gauge — the supply-demand, mid-curve force — intensifying in real time. When supply outpaces the structural demand from pensions, insurers, and foreign central banks, the price of capital rises: yields go up, term premium widens, and the back end of the curve carries the strain. The 10Y sits near 4.49% in the upper band of the Q2 Pressure Gauge, with the 2Y at 4.04% and the 30Y at 5.02% — a 2s10s of roughly 45 basis points and a steep 10s30s near 53, the long end already carrying the supply signal. A collision is precisely the kind of pressure that keeps the 10Y pinned in that upper band — or pushes it higher into the issuance windows.
Why does forced index buying not save the picture?
Because it is the wrong kind of demand. Once SPCX lists at a multi-trillion-dollar market cap, it enters the Nasdaq 100 and every passive vehicle benchmarked to it — the QQQ complex alone runs into the hundreds of billions. Even the governance- and ESG-screened funds that would never hold a Musk-controlled entity by choice are forced to buy, or to carry tracking error against their benchmark — and tracking error, not the position itself, is what gets a passive manager fired.
That is real, mechanical, price-insensitive buying. But it is a reshuffle inside the equity bucket — passive funds fund the inclusion by selling the rest of the index pro rata — not net new demand for risk assets as a class. The primary issuance is the genuine cash drain. Anyone reading the forced-inclusion bid as a signal of institutional conviction is mistaking an index-construction rule for sentiment. They are different things, and the Dispatch read keeps them separate.
The Credit Map
The sector-level read on the debt side of the buildout, anchored against the on-the-run 10Y. Most hyperscaler IG paper prints in the single-A band at the 5–10Y part of the curve — the row to watch is single-A at 61 basis points and the 7–10Y bucket at 89, both still tight against the historical average. The full CUSIP-level Matrix — single-name dispersion across Oracle, the Meta Hyperion SPV, and the parent-level hyperscaler stack — ships in Wednesday’s Companion. The qualitative read for now: the cash-rich names print through demand at tight spreads, while the heaviest borrowers — Oracle most visibly — have cheapened as the market prices their funding dependence.
Source: Koyfin, US Corporate Credit dashboard, May 28, 2026 pre-market. UST 10Y from Global Yields.
What to watch
The calendar is the catalyst. The SpaceX roadshow opens around June 4, pricing lands around June 11, and trading begins June 12. Each step is a discrete liquidity event. Watch the pricing against the $1.75–$2 trillion range: a deal that prices at or above range and trades well says the pool absorbed it; a deal that prices soft or breaks issue says the pool is thinner than the headline money-market AUM suggests. Then watch the Anthropic and OpenAI timelines firm up behind it, and watch the next Quarterly Refunding Announcement for the Treasury’s supply intentions into the same window.
Hold the sequence in mind, because it runs in two stages. Heavy concurrent supply cheapens duration into the issuance windows — that is the first move. But a soft or failed equity clear is what flips the setup, turning supply pressure into a risk-off flight to quality that rallies the long end. Same supply, opposite duration outcomes, depending on whether the deals clear.
BASE CASE (~60%) Supply is absorbed with indigestion — air pockets and vol spikes around each pricing date, secondary softness in the AI / growth complex as managers raise cash to fund allocations, but no systemic break. Duration grinds modestly cheaper into the issuance, then firms.
RISK CASE (~40%) One of the mega-deals prices soft or trades poorly, the AI / growth complex sells off, and the risk-off bleed turns duration-positive — a flight to quality that rallies the long end while equity holders eat the air pocket.
How institutional money is positioned
The crowded trade is long the AI / growth complex and underweight duration. That is the vulnerable book into a supply collision. The late-cycle pattern is worth naming honestly: when the largest private companies on earth rush to public markets in a cluster, at peak multiples, it has historically been a monetization-into-strength signal — Aramco in 2019, the 2021 listing blowoff ahead of the 2022 drawdown. This is not a timing call; it is a fragility marker. SpaceX entering the Nasdaq 100 at roughly $2 trillion also makes an already historically concentrated index more concentrated, which raises every passive holder’s beta to a single name’s execution.
The bottom line
The AI buildout is being financed twice — once in equity, once in debt — and the bill for both comes due in the same window. The equity market is trading each IPO as its own event. The bond market is pricing the aggregate supply collision before equities have connected the deals to one another. If the calendar clears clean, duration cheapens modestly and the story is indigestion. If it clears soft, the bond market is the beneficiary of the risk-off it saw coming.
KEY TAKEAWAY One capex cycle, two financing channels, one liquidity pool — and the bond market sees the collision first.
Educational and macro commentary only. Not investment advice. Views are my own and do not represent any employer or affiliated entity. Subject to CFA Institute Standards of Professional Conduct.
© 2026 Positive Carry LLC, 6586 Atlantic Ave #115, Delray Beach, FL 33446