Private capital no longer buys mature assets — it finances balance sheets

Two announcements, one pattern

On 30 August 2026 ONEOK, Inc. signed a definitive agreement to buy Brazos Midstream's natural gas gathering and processing assets in the Permian's Midland Basin for US$4.425bn in cash, with closing expected in the fourth quarter of 2026 and about 1.2 Bcf/d of processing capacity once Cassidy II is completed in 2027 FACT. The same announcement carried the more consequential number: funds managed by Apollo will invest US$9bn for a nonvoting Class B interest in a newly formed subsidiary, ONEOK Holdings, L.L.C., structurally subordinate to ONEOK's debt FACT. The Form 8-K filed the next day puts names and dates to it: the investor is AP Falcon Holdings LLC, an Apollo affiliate, contributing US$9bn in cash for 900 million Class B units, with closing set for the later of the satisfaction of conditions or 10 September 2026 FACT.

Three days later, on 2 September, Diversified Energy Company agreed to acquire Birch Permian Holdings for about US$1.8bn, funded mainly by an asset-backed securitization of approximately US$1.5bn — TCG Capital Markets, Carlyle, as sole structuring and placement agent — plus available liquidity FACT. In the same release, Carlyle and Diversified expanded the partnership they had announced on 24 June 2025 from a US$2bn framework to a collaboration under which they may pursue up to US$10bn of proved developed producing (PDP) acquisitions over time, subject to mutual agreement and deal-by-deal approvals FACT.

Read together: US$9bn of committed capital and a framework of up to US$10bn went into United States oil and gas in four days, and in neither case did the fund buy the asset FACT. Apollo owns a subordinated claim on a holding company; Carlyle places amortizing debt secured on producing wells. The manager's product is no longer the field but the balance sheet of the company that already owns it INTERPRETATION.

What Apollo actually bought

The mechanics are what the headlines omit. The Class B receives 15% of ONEOK's quarterly cash flow. Its return is capped at a 7.0% IRR for the first nine years, stepping to 7.35% in year ten and to a final ceiling of 7.85% in year fifteen; any distribution above the capped return reduces the capital balance. From the eighth anniversary of closing — or earlier, if the capital account falls to US$200m — ONEOK may acquire the remaining interest at the same 7.0% IRR FACT.

ONEOK will apply US$5bn of the proceeds to debt reduction, including a US$1.2bn term loan, taking expected pro forma 2027 leverage to about 3.25x debt-to-EBITDA without issuing a single common share. The company states that its rating agencies consider the transaction credit-enhancing and that it expects to receive full equity credit; that is ONEOK's expectation, not a published agency decision FACT. Cash tender offers for up to US$2bn of senior notes across twenty series maturing between 2029 and 2064 carry an early tender deadline of 14 September 2026 and expire on 29 September FACT.

What Apollo has bought is quasi-equity with a ceiling: equity-like subordination, senior-like predictability, no share of whatever the assets earn beyond 7.0% INTERPRETATION. Set against the 7.28% blended coupon Diversified paid in May 2024 on its ABS VIII tranches, rated A and BBB+ by Fitch, the cap is a price for structure, not for upside FACT.

It is also, apparently, not the end of the chain. The 8-K provides that before the fifteenth anniversary the Class B member may not transfer its units without the Class A member's consent, except for certain permitted transfers, with a right of first offer to ONEOK thereafter FACT. Within that perimeter, Bloomberg reported on 31 August, citing a person familiar with the matter, that Apollo intends to slice the stake into levels of seniority so that securities backed by the holding can obtain investment-grade ratings, with the deal placed partly with Apollo's insurer Athene, third-party insurers and other Apollo and client funds FACT. That is an attributed report, not a verified transaction: no tranche sizes, ratings or coupons are public, no Apollo, Athene or ONEOK release confirms it, and whether such a repackaging falls within the "permitted transfers" of the LLC agreement cannot be determined from the filing itself. If it is completed as described, the ultimate funder of ONEOK's deleveraging would be insurance balance sheets rather than a private-equity fund — which sharpens, rather than softens, the point INTERPRETATION.

What Carlyle actually financed

The Birch transaction is the other template. The assets comprise 480 net wells on about 46,000 net acres, producing around 68 Mboepd, with proved reserves of roughly 1,168 Bcfe and a PV-10 of about US$2.0bn; closing is expected in the fourth quarter with a US$50m break fee FACT. A US$1.5bn securitization against a US$1.8bn price means roughly 83% of the purchase is debt secured on wells already producing; the consideration mix for the balance, and the tranche ratings, coupon and amortization of the new notes, are not yet disclosed FACT.

The scale matters for Carlyle too. Its asset-backed finance group had deployed about US$8bn since 2021 and managed about US$9bn as of 31 March 2025, so a single US$1.5bn PDP deal is a material share of the platform FACT. A framework that has grown fivefold in fourteen months is a statement of intent — though "up to US$10bn" is a ceiling on opportunities, not committed capital FACT.

Data point. According to Jefferies, PDP securitization has reached roughly US$20bn of cumulative issuance since 2019 across more than 15 issuers, with senior tranches pricing at about 200–250 bp over Treasuries and structures supporting up to about 70% loan-to-value. This is an arranger's estimate; no rating-agency aggregate for 2024–2026 was located ESTIMATE.

The reader should be careful with that figure: Diversified has been the near-exclusive issuer since 2019, and the total comes from a bank with a commercial interest in the product. What is verifiable is narrower and more useful: investment-grade ratings on mature-well debt, a coupon in the low sevens, and a manager willing to underwrite ten billion dollars of it.

The mirror image: a sponsor exits in paper

The same week supplied the other half of the cycle. On 3 September Williams announced the closing of its acquisition of 100% of Momentum Midstream from EnCap Flatrock Midstream, a financial sponsor, in a transaction valued at about US$5.5bn: roughly US$3.5bn in cash and debt consideration and roughly US$2bn in Williams equity, about 36% of the total, at an implied valuation of about 8.5x projected 2027 EBITDA — the multiple is Williams' own figure from the 3 August announcement, not restated at closing FACT. The asset is a Haynesville system whose three take-or-pay pipelines can move 4.05 Bcf/d toward Gulf Coast LNG, power and industrial demand FACT.

Here private capital did sell a whole asset, and a strategic did buy it — but read the consideration. The sponsor built the system with its own capital and exits into a listed company's balance sheet, paid partly in shares rather than entirely in cash; who among EnCap Flatrock's funds, management or co-investors receives that equity, how many shares were issued and whether they are locked up is not disclosed INTERPRETATION. Entry and exit rhyme: Apollo goes in through a subordinated claim on a holding company, EnCap Flatrock comes out paid partly in a strategic's paper. In both directions the manager's exposure is to a balance sheet, not to a field INTERPRETATION.

The Mexican contrast

On 22 May 2026 Moody's Ratings affirmed Pemex at B1 with a stable outlook and a "ca" baseline credit assessment, citing funding needs averaging about US$14.9bn a year over 2026–2028, persistent negative free cash flow, and its view that "current development schemes are unlikely to materially increase output" FACT. On 2 September, at the CCE energy-infrastructure forum, Pemex's director general said ten mixed contracts had been signed and that the company needs to go beyond what the Hydrocarbons Law establishes, using the room its own corporate statute allows, to define new private-collaboration schemes — deep water, seismic, refinery and petrochemical maintenance, logistics and storage — without amounts or production targets FACT.

That statement is about collaboration schemes, not about preferred equity or securitization; the gap is the inference, and it deserves stating precisely. The mixed contracts one year on have produced ten signatures and no incremental barrels; more to the point, they carry no balance-sheet financing layer of the kind Apollo and Carlyle have just supplied twice. Under a mixed contract Pemex keeps title, the partner recovers costs through a waterfall, and no bankruptcy-remote vehicle holds investment-grade PDP collateral. No comparable transaction — fund-financed mature fields via capped preferred equity or PDP securitization — was located in Mexico or Latin America with a citable source INTERPRETATION. Whether the mixed-contract framework or Pemex's statute could accommodate such a structure is an untested regulatory question that this memo does not attempt to answer.

What changes for capital

The strongest objection is that two deals in one week do not make a market. ONEOK used the Class B because it is investment-grade, wanted to avoid dilution and had a US$5bn debt problem; Diversified has run the PDP playbook since 2019; and private equity has not stopped buying — Brazos is being sold by a sponsor-backed company and Birch by its owners. Williams–Momentum partly supports the objection: a strategic bought a whole asset for US$5.5bn. But the seller was a sponsor exiting, not a fund buying, and roughly US$2bn of the price was paper. The objection narrows the argument without dissolving it: these are templates that scale, sized at US$9bn and US$10bn by two of the largest managers in the world in the same week — and, if Bloomberg's report holds, re-sold onward to insurers who want rated paper, not wells OPINION.

For investors, the trade is duration and structure rather than asset upside: a contractually capped return with equity-like subordination, deployed through frameworks rather than auctions INTERPRETATION. For an operator of mature fields in Mexico or Latin America, the reading is more demanding. Both templates require three things — audited proved developed reserves the operator actually owns, a hedging programme, and a creditor-friendly jurisdiction in which a special-purpose vehicle can hold the collateral. Where the state company holds title and the partner holds a cost-recovery claim, the first condition fails before the other two are tested INTERPRETATION.

That is the contrast worth putting in front of anyone thinking about private capital in Latin American energy or the Mexican upstream this month. The United States has just shown, twice, that the cheapest way to fund a mature asset is not to sell it but to borrow against what it already produces. Mexico's debate on schemes "beyond the Hydrocarbons Law" opens the question of whether that layer could ever sit under a Mexican contract; it does not answer it. Nothing here is a recommendation on any security: it is a description of how the money moved, and of where, for now, it cannot.