Kinder Morgan Retires the Shadow Backlog and Writes a $4.7 Billion Check With an A-Rated Hyperscaler's Name on It

The horses, finally, in the corral

For four consecutive calls, Kinder Morgan's management asked investors to believe in something invisible. There was the $10 billion shadow backlog introduced in late 2025 — projects "actively" under discussion, costed internally, no signatures. There were the assurances in July that "we expect to add significant projects in the back half of this year." And there was Rich Kinder's folksy plea a quarter ago that “we have to get the horses in the corral.” — Richard Kinder, Executive Chairman · 2026-07-22

On 21 October, the horses arrived, and the vocabulary changed with them. The word "shadow" has essentially vanished from the Q3 transcript. In its place: a backlog increase from $9.6 billion to $14.4 billion — $5.4 billion of new projects sanctioned in one quarter, against $700 million placed into service — the largest quarterly addition in the company's history, at a 5.7x multiple that management frames as $2.5 billion of eventual EBITDA. Chief Executive Kim Dang put the pivot plainly: “Before this quarter, most of our EBITDA from our backlog was expected to be online by the end of 2025. With Permian Link and Southeast Connector entering service between 2030 and 2032, we have meaningfully extended the period of high-quality cash flow growth.” — Kim Dang · 2026-10-21

That is the real story here, and it is not a growth-rate story. It is a duration story. A midstream name whose visible runway was about to run dry has just bought itself five more years of contracted growth — and, critically, the two headline projects are joint ventures with Kinder Morgan's share at roughly $4.7 billion, not $5.4 billion, which is why the balance sheet does not blow up on the announcement.

The uncomfortable detail: the AI trade is de-rating while KMI signs with an AI tenant

Here is the contrast worth chewing on. The two new projects are demand-pull, not producer-push. Southeast Connector moves about 2.6 Bcf/d into a region where interconnection requests exceed 200 gigawatts — roughly 8 Bcf a day of natural gas demand if it were all gas-fired — and is contracted with investment-grade utilities. Permian Link moves Permian gas north to Texas Panhandle power load, and its anchor shipper is, in management's phrasing, an A-rated hyperscaler.

That last phrase is a genuine evolution. On the October 2025 call, Kinder's chairman went out of his way to say the company preferred contracting with utility-grade counterparties because it “minimizes any risk that we have as opposed to contracting directly with AI developers, for example.” — Richard Kinder, Executive Chairman · 2026-01-21 Twelve months later, an AI developer — admittedly the investment-grade kind, with 20-year take-or-pay and storage-backed seven-nines reliability — is the anchor. The company didn't change its risk appetite; it found an AI customer that clears the same hurdle.

Timing matters. The global keyword tape shows AI data center up 53% over 360 days but down roughly 6% over the last 90, with 48 decliners against 6 advancers — hyperscaler-adjacent equities have been sold hard. Kinder Morgan, meanwhile, is up 8% over the same 90 days, an almost unbroken 17-week climb into this print. The market is paying for contracted, credit-wrapped gas demand and refusing to pay for speculative compute. That is a distinction KMI's structure now embodies, and it is the reason a midstream stock is outperforming the theme that supposedly drives it.

Returns, costs, and the discipline question

The obvious analyst objection — that a single-anchor-shipper project carries more risk — was answered with the same return threshold language management has used for years, with one nuance: risk is priced, not avoided. As Dang put it, “If a customer is willing to share in some of the cost, then we can move down the return threshold a little bit. If the customer's not willing to share in cost overrun, we need a higher return.” — Kim Dang · 2026-10-21 On Western Gateway, the $300 million clean-products JV, she noted “we do have certain cost protections against cost overruns.” — Kim Dang · 2026-10-21 Steel and compression inflation have lifted per-mile costs; customers, so far, have not balked.

Two numbers quantify the ambition. First, growth CapEx guidance has been reset again: from "at least $3 billion" in January, to roughly “$4 billion-ish for the next few years” — Kim Dang · 2026-10-21 today — with $1.5 billion more (Project 219 South, over 2 Bcf of open-season interest) waiting on signatures. Second, leverage is 3.5x, the bottom of the 3.5–4.5x range, expected to end the year at 3.6x versus a 3.8x budget. The balance sheet absorbed a mid-cycle step-up in capital spending and improved.

We've got a hell of a lot of those horses in the corral now with the saddles on, but we've got more horses just outside the corral, and I think you're going to see them.

The 2026 problem nobody wants to underwrite

The quarter itself was excellent and, in part, unrepeatable. Adjusted EBITDA rose 9% to $2.18 billion, adjusted EPS 21%, net income $764 million; every segment grew. But management was candid that roughly $225 million of this year's beat came from a stack of one-timers — Winter Storm Fern, a terminals contract buyout, a favorable FERC true-up, an unusually wide Waha-to-Houston spread — plus another $175 million from commodity prices against a $60 WTI budget. Dang's warning was unhedged: “2026, because it has some of these items, is going to be a difficult year to replicate.” — Kim Dang · 2026-10-21 The market will need to separate base-business growth from weather luck, and 2027 gets only 30% of Trident's contracted EBITDA.

Worth flagging on the fundamentals side: the latest 10-Q covers the June quarter, so it predates all of this. Still, the shape is consistent — total revenue of $4.5 billion, up 11% year over year with operating margin of 30.1%, and capex up 52% to $982 million. The company still carries net debt of about $29.6 billion, and trades at 22.5x free cash flow. That is a fuller price for a longer runway, not a cheap one.

What to watch

The retired word — shadow backlog — is the tell. When a company stops marketing a pipeline of maybes and starts booking take-or-pay, you should take the conversion seriously. The follow-on test is whether the demand center narrative keeps converting at this cadence, whether 219 South lands near $1.5 billion and then scales, and whether the middle system power corridor produces the next multi-billion-dollar sanction rather than another quarter of adjectives. On the current run rate, CapEx of $4 billion a year is still fundable from $5+ billion of cash flow plus modest leverage — right up until it isn't. Three more quarters like this one and the leverage question becomes the story.