Welcome back.

Every newsletter in your inbox this week is running some version of the AI-IPO-wave story — Anthropic’s S-1, the Nvidia financing web, the trillion-dollar valuations.

We’re going to touch it once, briefly, and then spend most of this issue somewhere more useful: the deals that actually closed in the lower middle market over the past two weeks, and what they say about financing discipline that a megadeal prospectus can’t.

On the Cirrus side, our recent deal flow has leaned into exactly that theme.

Over the past two weeks, we worked through a facility for a construction company — an $8 million asset-based revolver paired with an $800,000 term debt piece — structured to carry a sponsor through an accretive acquisition.

We also agented a $250,000 term loan for a consumer brands company in the beverage space, sized for exactly the kind of near-term inventory and growth need that doesn’t require (and shouldn’t carry) a nine-figure capital structure.

Both deals are instructive precisely because they’re small: the discipline is in the structure, not the size, and that’s the thread running through this entire issue.

$60 million term sheet. Renewable energy. Structured against fixed assets and project cash flows to support an energy operator’s expansion financing.

$30 million bank facility ($20 million revolving / $10 million term). Private credit lending. Supports scaling of the borrower’s loan book.

$15 million senior-secured credit facility. Growth-stage operating company. Up to $15 million total availability, $5 million funded at close, four-year term, 15% cash-pay interest, and a modest equity kicker (a penny warrant for 2% of common).

$8 million asset-based revolver with an $800,000 term debt piece. Construction company. Sized to a contractor’s borrowing base to smooth working capital across project cycles. (Closed and funded.)

$250,000 term loan. Consumer brands, beverage/ready-to-drink. Supporting near-term inventory and growth needs for a scaling beverage brand. (Closed and funded.)

Anthropic filed its S-1 this week, showing roughly $16.3 billion in first-half revenue, a loss figure so large it needed its own explanation in the coverage, and bankers lining up for what could be one of the largest technology listings on record.

Somewhere in the same two weeks, Cirrus closed a term loan for a beverage company that would round to a rounding error on Anthropic’s balance sheet.

I want to talk about the second deal, not the first — because the beverage loan, and the construction facility that closed alongside it, tell you more about how financing actually works than anything in a 400-page prospectus.

Here’s the conventional wisdom circulating right now: the most sophisticated capital structures are the biggest ones, with the most exotic collateral — GPU-backed paper, data-center guarantees, hyperscaler credit support flowing through circular financing arrangements that read more like reinsurance than lending.

Founders read these stories and absorb an implicit lesson: bigger, more aggressive, more structurally creative is the direction smart capital is moving. That’s a reasonable inference if you only read the headlines.

It’s the wrong lesson if you’re running a $4 million revenue construction business or a regional beverage brand.

The structures underpinning the AI financing wave work because of who is standing behind them. A hyperscaler’s contracted, investment-grade cash flows can support financing structures that would be reckless anywhere else — long tenors, thin equity cushions, collateral that’s genuinely hard to value independently.

That’s not a template; it’s a function of counterparty quality that almost no company outside a handful of trillion-dollar balance sheets can replicate.

The strongest version of the counterargument here is real: some of these structural innovations — longer-duration asset-backed paper, more creative collateral packages — may well filter down to other capital-intensive sectors over time, including renewables and infrastructure, where Cirrus is already active.

I don’t dismiss that. But filtering down takes years, and it arrives wrapped in covenants and collateral discipline suited to the borrower, not copied wholesale from a megadeal.

What the construction and beverage deals actually demonstrate is the opposite instinct: match the facility to the borrowing base, not to the ambition.

The construction company’s $8 million revolver is sized and structured against real, monitorable collateral — receivables and work-in-progress that a lender can actually underwrite month to month.

The beverage brand’s $250,000 term loan is proportionate to an inventory cycle, not stretched to fund three years of growth on year-one assumptions. Neither deal is exciting.

Both are survivable in a bad quarter, which is the only test that actually matters once the financing closes and the business has to run on it.

What this means in practice:

The trillion-dollar filing will get the attention this week.

The construction loan and the beverage term loan are the ones worth studying if you’re actually running a business that has to make payroll in a bad quarter.

Anthropic filed a Form S-1 for a planned Nasdaq listing (ticker: ANT), disclosing roughly $16.3 billion in first-half 2026 revenue — reported growth of more than tenfold year over year — alongside a net loss figure large enough that multiple outlets flagged it needs context (a meaningful portion is tied to non-cash items rather than cash burn).

Why It Matters: This is the first time the market gets audited-adjacent numbers behind the AI infrastructure spending story, rather than estimates.

For the Altitude Audience: The real signal isn’t the valuation chatter — it’s the gap between revenue growth and reported losses, which is exactly the kind of scrutiny that’s pushing AI infrastructure financing toward structured credit rather than pure equity. We’re intentionally not turning this into a deeper Nvidia/chip-financing story — that ground has been thoroughly covered elsewhere this week.

DoorDash announced a deal to acquire Instacart for about $65 per share (in cash, with a capped stock election option), a roughly 26% premium to Instacart’s pre-announcement price.

The deal is expected to close in late 2027 or the first half of 2028, pending shareholder and regulatory approval.

Why It Matters: This is a bet that the next phase of delivery-platform economics is grocery and retail fulfillment, not just restaurant orders — and it’s being financed through a mix of cash on hand and a bank bridge facility rather than a pure stock deal, which tells you something about how acquirers are thinking about capital structure even on enormous transactions.

PepsiCo lowered its projected payout under the earnout tied to its roughly $1.95 billion acquisition of Poppi, the prebiotic soda brand, after softer category sales. (Note: this development was reported October 14, 2026 — just outside our formal coverage window — but it’s included here because it’s directly relevant to this issue’s theme and to the beverage sector specifically.)

Why It Matters: Earnouts look like free upside when a deal is signed and look like a liability dispute the moment performance misses plan. Any founder negotiating an earnout-based exit, or any lender underwriting a business with earnout obligations on its balance sheet, should read this as a reminder that earnout structures need the same underwriting discipline as the deal itself.

Short-end rates are drifting lower, while the long end has remained unusually volatile — the 10-year swung from near 4.83% to roughly 5.0% over two weeks.

The Fed held rates on October 28, with two dissents favoring a hike; equities ripped higher following the November 3 midterm results; and credit spreads remain historically tight even as a government shutdown-driven data blackout leaves the labor market and inflation picture genuinely unclear heading into the December Fed meeting.

Market Summary:

A real caveat belongs here: the extended government shutdown forced the BLS to cancel the October household survey outright, so there is no October unemployment rate, and the November payrolls report is delayed to December 16.

October CPI data was still working through standard reporting timelines as of this writing. That’s a genuine blind spot heading into the Fed’s December 8–9 meeting — tight credit spreads and a strong equity rally are easier to read as “all clear” when you’re missing two of the three data series the Fed itself is watching most closely.

For anyone locking in financing terms over the next month, that gap is worth pricing in as uncertainty rather than resolving it in either direction.

Serval, which builds AI agents to automate internal IT and employee-support workflows, raised $150 million in a Series C led by Sequoia Capital, with participation from Elad Gil, Evantic, First Round, General Catalyst, Redpoint, Conviction, and Headline — tripling its valuation to $3 billion. (Growth equity/venture.)

The round is notable less for its size than for its positioning: Serval is explicitly being framed as a credible AI-native challenger to ServiceNow’s enterprise service-management franchise, which is a different kind of competitive claim than most AI-wrapper startups can credibly make. Announced October 29, 2026.

DoorDash agreed to acquire Instacart in a cash-and-stock deal valued at roughly $17 billion (at least 75% cash, with stock capped at 25%), at approximately $65 per share — a 26% premium to Instacart’s pre-announcement price.

The deal, announced on October 29, 2026, is expected to close in late 2027 or the first half of 2028, pending shareholder and regulatory approval, and is being financed with cash on hand and a bank bridge facility.

Strategically, it hands DoorDash Instacart’s grocery fulfillment infrastructure, retail technology, and advertising business — deepening its push beyond restaurant delivery into broader retail commerce. (Announced, not completed.)

Altitude is the #1 newsletter for founders, operators, dealmakers, and capital allocators aiming to reach their highest potential. 🏔️

To your growth,

Thanks for reading Altitude!

If you're financing growth, evaluating a refinancing, or just want a quick check on a term sheet, connect with me directly on LinkedIn. If you know someone who'd get value from this, forward it their way.