Oneok: A Good Business Standing Above Its Value

OKE · Oneok · Published Jul 16, 2026

Some companies are easy to admire and easy to decline. Oneok, the midstream operator of natural gas, NGL, refined products and crude pipelines, is one of them. The four gates below were applied on 2026-07-16, and revisited after the company’s Q2 2026 earnings in early August. The conclusion did not move: a fine business, run reasonably well, priced above what it is worth.

Gate 1 — Circle of Competence

Oneok owns roughly 60,000 miles of pipelines, fractionation, storage and terminal assets, moving natural gas, NGLs, refined products and crude from producing basins — Permian, Mid-Continent, Rockies and Bakken — to demand centers and export hubs like Conway and Mont Belvieu. It charges by volume. Around 90% of 2025 earnings were fee-based, and all four segments (NGL 35%, gas G&P 27%, refined products and crude 28%, gas pipelines 10%) run at 85–90% fee-based revenue. Residual commodity exposure in the G&P segment is roughly 75% hedged.

That structure matters for the competence question. This is not a direct commodity bet; it is a toll booth on volumes. The proof is in the record: twelve consecutive years of EBITDA growth through 2025, surviving the oil collapse and the pandemic. Volumes are, over the long run, a function of US shale production, so there is an indirect commodity exposure — but judging that US gas and NGL production and exports (LNG, petrochemical feedstock) remain structurally supported over a ten-year horizon, with new pipelines effectively blocked by permitting barriers, is analysis, not gambling. Pipeline networks are the same species of asset as a railroad: efficient scale and an irreplaceable physical network. This sits comfortably inside the circle.

The company became an integrated midstream major through the Magellan acquisition ($18.8 billion, 2023) and EnLink and Medallion ($5.9 billion, 2024).

Gate 2 — Moat

Three sources of advantage. First, efficient scale: laying a competing pipeline in the same corridor is, given permitting, environmental regulation and capital requirements, effectively impossible. Second, integration: a vertically connected NGL chain from wellhead to Mont Belvieu fractionation and export terminals carries bundled economics that segment-by-segment competitors lack. Third, switching costs: long-term fee-based contracts with minimum volume commitments, and gas processing physically tied to connected plants.

Pricing power has a regulatory spine. FERC index-based rates link interstate transport tariffs to inflation — cost pass-through is built into the system itself. The $500 million of realized Magellan synergies (with a further $150 million expected in 2026) attest to the negotiating strength of an integrated network.

Direction: steady to modestly expanding. The acquisitions deepened cross-basin, cross-product integration, though competition from Enterprise Products, Targa and Energy Transfer in the Permian-to-Mont Belvieu corridor is real and pressures fractionation margins. The moat is not narrowing. In an AI era, physical infrastructure is the kind of advantage that cannot be eroded by software; if anything, AI data center power demand may add a new source of gas throughput.

Gate 3 — Management

A serial acquirer: five large transactions over twenty years, most recently the Magellan and EnLink/Medallion deals. The record is disciplined — deals were consistently accretive to EPS and free cash flow, Magellan synergies were actually delivered, and Medallion was bought at a reasonable 6.3x 2025E EBITDA. This is not a pattern of buying at peaks and destroying value. The deductions: Magellan was paid partly with dilutive share issuance, and deal funding left the balance sheet structurally levered — debt-to-equity of about 1.5, improved from 2.36 but still high. A $2 billion buyback was authorized; the dividend, currently $4.28 annually (roughly 4.7% yield), has grown over the long term.

The retained-earnings test passes only conditionally. Adjusted EBITDA compounded at 17% annually from 2013 to 2025, but a substantial share of that growth rode on leverage and share issuance rather than purely retained profits. On candor: management did not hide the fact that 2026 guidance ($7.9–8.3 billion at the time) implied flat adjusted EBITDA — an admission that the twelve-year growth streak had ended. Synergy targets are stated numerically and tracked publicly. No dishonesty signals. Note that the 17.1% five-year ROE sits atop 1.5–2.3x leverage; unlevered returns are materially lower. That is structural for pipelines — but the cushion for equity holders if rising rates and volume shocks arrive together is thin.

Gate 4 — Price

The reported owner-earnings figure ($1,755 million, a 3.1% yield on a $57.35 billion market cap) deducts all growth capex and is likely too conservative; the reported D/E of 0.0 was an XBRL tagging error, with real debt around $33.7 billion. Normalizing — 2025 net income of $3.46 billion less maintenance capital — yields owner earnings of roughly $3.5–4.0 billion, a 6.1–7.0% adjusted yield.

With 2026 guidance signaling stagnation, no growth multiple is warranted. At 12x:

Against a market cap of $57.35 billion at $92.19 (2026-07-14), the base case implies roughly 37% overvaluation. No margin of safety. The 3.1% reported owner-earnings yield sits below the 4.3% ten-year Treasury, and even the normalized yield looks thin once $33.7 billion of debt is counted — EV of about $91 billion against $8+ billion of EBITDA is not a discount. A 30% discount would require a market cap of roughly $29 billion, a share price of about $47–52.

The August re-check reinforced rather than overturned this. Q2 2026 adjusted EBITDA of $2,121 million rose 7.1%, NGL throughput hit a record 1,630 MBbl/d, and full-year guidance was raised twice to $8.2–8.5 billion — the “stagnation” premise was partially wrong. But the raised midpoint of $8.35 billion is still only 3.3% above 2025, inflation-level growth that does not justify 15x. At the August 4 close of $87.65, the market cap of $55.2 billion still stood 15–31% above the intrinsic value range, with 40–46% further downside needed to reach the buy zone. Meanwhile the first hard evidence of Gate 5’s third risk — Permian corridor competition eroding pricing power — appeared: the NGL segment’s EBITDA fell 2.1% on record volumes, with the company citing lower average fee rates and narrower price differentials. Worth watching; not yet a trend.

Verdict

Watch. Gates 1 through 3 pass: a knowable, toll-based business with an efficient-scale moat and a candid, reasonably disciplined management. Gate 4 fails clearly — the market price sits well above intrinsic value with no margin of safety, and effective leverage near 4x net debt/EBITDA thins the cushion further. Quality and price were both reconfirmed by the quarter: the business runs well, and the price is still above what it is worth.

What would change the verdict: a price in the $47–52 range (a market cap near $29 billion), or evidence that structurally accelerating gas demand — AI data centers, LNG exports — justifies a durable growth multiple at today’s price. On the downside, a second consecutive quarter of NGL fee-rate compression, or net debt resuming its climb, would weaken the qualitative case itself.

This analysis is AI-generated, educational, and not investment advice. Figures may contain errors or be delayed. Disclaimer