ONEOK In-Depth Value Investing Analysis

Hunting ten-year five-baggers among great growth stocks — pressing the upside question: "Can it get much bigger?"

• QHow high is its market ceiling? Is it enlarging an existing pie, or creating an entirely new market?5/10

  A

  The ceiling is not low, but in essence it is "grabbing a larger share of a mature, slowly expanding existing pie," not creating an entirely new market—measured against Baillie Gifford's LTGG yardstick for finding new-market creators, this is its least sexy side.

  ONEOK does North American midstream energy infrastructure: moving natural gas, NGLs (natural gas liquid products), crude oil, and refined products from the point of production—through processing, fractionation, storage, and pipelines—to industrial, export, and power-generation demand. The report says it owns about 60,000 miles of pipeline network, in a "mature, capital-intensive, fairly heavily regulated" industry. This arena has existed for decades; ONEOK is not inventing a new kind of demand but, within an already-existing oil-and-gas logistics network, collecting more "tolls" through better-connected assets.

  The pie itself is indeed growing modestly, giving it real incremental room. The U.S. Energy Information Administration (EIA) May 2026 Short-Term Energy Outlook projects U.S. Lower-48 natural gas production to rise from about 118.9 Bcf/d in 2026 to about 124.0 Bcf/d in 2027, net natural gas exports to grow about 18% in 2026, and LNG exports to exceed 18.1 Bcf/d by 2027. Add AI-data-center-driven gas-fired power, and the industry estimates data centers could add 3–20 Bcf/d of demand by 2030 (base case about 6 Bcf/d). ONEOK itself discloses it is in discussions with more than 40 data-center and power-generation counterparties, with potential demand exceeding 5 Bcf/d. These are all structural tailwinds.

  But we must distinguish "high ceiling" from "creating a new market":

  First, ONEOK is grabbing share within an existing pie, not opening up virgin territory. Its rivals—Williams (market cap about $88.6 billion), Kinder Morgan, and Energy Transfer—long ago blanketed North America's trunk pipeline network, and the competition is "whose assets are more connected and better located," not "who invented this business first." The report itself characterizes it as "a high-quality asset holder in an ordinary industry, not a super company in a great industry."

  Second, its incremental ceiling is firmly boxed in by physical and capital limits. Building a new pipeline segment or a processing plant routinely takes several years and several billion dollars (the report cites ONEOK's roughly $1 billion share in the Texas City terminal JV). In this "trade steel and concrete for growth" business, the scale ceiling rises asymptotically; there will be no software-style exponential market expansion.

  Third, although demand is steady, it is subject to the long-term overhang of energy-transition policy. The upper edge of the ceiling is not fully decided by ONEOK but by how long fossil energy can hold its share of the U.S. energy mix.

  Conclusion: by Baillie Gifford's LTGG standard, ONEOK's market ceiling is "enough to keep building it into a decent cash-flow platform," but it is by no means "the infinitely extendable new market required for a fivefold return in ten years." It is enlarging its own portion of an existing pie, not creating a new pie—this is the first watershed that separates it from a truly great growth stock.

  Jun 11, 2026
• QCan its revenue at least double over the next five years? Is growth driven mainly by volume, price, or new business?3/10

  A

  It most likely cannot double organically within five years; reported revenue did multiply several times over the past five years, but almost entirely on M&A consolidation and commodity-price swings, not the kind of organic volume-and-price growth Baillie Gifford wants. This is where this question is most easily misread.

  First the facts. ONEOK's revenue grew from about $8.54 billion in 2020 to about $33.63 billion in 2025, roughly a 4x increase in five years, with 2024→2025 alone up 55%. At first glance it looks like a super-growth stock, but break down the drivers and it is all "accounting optics" rather than "business quality":

  • M&A consolidation: the 2023 acquisition of Magellan brought a refined-products/crude network, and in 2025 it completed the EnLink deal (the report notes 41 million shares issued at a fair value of about $4 billion) and integrated Medallion. EnLink was consolidated in 2025 and was the main reason revenue jumped 55% that year—this is "bought revenue," not something the company grew on its own.
  • Commodity-price pass-through: midstream revenue contains large amounts of "buy NGL/natural gas at market and resell" purchase-and-sale volume (POP contracts, spread optimization), so when commodity prices swing, the nominal revenue figure swings wildly—which is why revenue was $16.5 billion in 2021, $22.4 billion in 2022, and back down to $17.7 billion in 2023, like a roller coaster. This kind of revenue has a lot of "water" in it and does not map directly to profit.

  Precisely for this reason, revenue is simply not a good yardstick for this company. The report never uses revenue to tell the growth story; it uses adjusted EBITDA instead: 2025 adjusted EBITDA of about $8.02 billion, up 18% year over year, with the 2026 guidance midpoint at about $8.1 billion. Only EBITDA strips out commodity pass-through and reflects true earning power.

  So using EBITDA as the yardstick, can it double over the next five years? The answer is: almost impossible. Growing the 2026 guidance midpoint of $8.1 billion EBITDA to about $16 billion within five years would require a CAGR of about 15%—whereas ONEOK's own disclosed organic-growth baseline is "12–13 consecutive years of single-digit to low-double-digit EBITDA growth," and the 2026 guidance midpoint is up only about +1% year over year ($8.02 billion→$8.1 billion). This is a typical mature-infrastructure growth rate, not a doubling curve.

  Now look at what drives growth, item by item:

  • Volume: real but modest. Permian gathering, Gulf Coast exports, and data-center/power-generation gas hookups (potential >5 Bcf/d) will bring incremental throughput, but only through new pipelines and processing plants that take years to come online, released in stages.
  • Price: very weak. The report makes clear its price increases rely on contract CPI/PPI adjustments and FERC annual tariffs, not brand-style pricing power—basically only enough to slightly outpace inflation.
  • New business: essentially "adjacent expansion" (refined products, Permian NGL, export terminals), not a genuine second revenue curve.
  • M&A: this is the real engine behind the past doubling, but growing revenue by continuing to issue stock to buy companies dilutes per-share value, and Baillie Gifford would not count this as "good growth."

  Conclusion: the "doubling" of reported revenue is an accounting illusion, mixed together from M&A and commodity prices; on a true operating basis (EBITDA, per-share cash flow) the next five years will most likely be single-digit compound growth, nowhere near doubling. This is a "dividend + moderate growth" name, not a "revenue doubles in five years" growth stock—honestly, on this question it fails.

  Jun 11, 2026
• QFive years out, what will take over as the next growth engine? Does this "second curve" exist today?4/10

  A

  Frankly, ONEOK has no clear "second curve"—its growth engine over the next five years is an extension of the same main curve as today's engine (more pipelines, more processing volume, more M&A synergies), not a new growth pole of a different nature that can take over after the core business peaks. By Baillie Gifford's yardstick of "does the second curve already exist today," this is precisely its shortcoming.

  First let me lay out what in the report could count as a "growth baton" and evaluate each as to whether it is a genuine second curve:

  One, realization of M&A synergies (closest to a near-term engine, but not a new curve). The report notes the 2026 guidance includes about $150 million of annual synergies, with additional synergies still expected after 2027, from integrating Magellan, EnLink, and Medallion. This is real incremental EBITDA, but it is essentially "squeezing efficiency out of already-bought assets"—one-time and convergent, unable to contribute year after year, let alone deliver a fivefold return in ten years. It is "wiping the existing pie clean," not a new pie.

  Two, organic project startups (a natural extension of the main curve). Eiger Express (Permian gas to the Gulf Coast), the Medford fractionation plant rebuild, the Bighorn processing plant, the Texas City terminal JV, and more are all named in the report. These bring incremental throughput, but they are identical in nature to what ONEOK does today—still gathering, processing, pipelining, and charging tolls. This is a thickening of the main curve, not a second curve.

  Three, data-center/power-generation gas hookups (the only item with a whiff of a "new story," but still new customers for an old business). The report discloses ONEOK is in discussions with more than 40 data-center and power-generation counterparties, with potential demand exceeding 5 Bcf/d; the industry estimates data centers could add several Bcf/d of gas-fired power demand by 2030. This is indeed a structural tailwind, and AI electricity demand forcing gas-fired power is a real trend. But be clear-eyed: ONEOK's role here is still "move the gas over and charge freight," just with the customer switching from a traditional power plant to a hyperscale data center. It does not create a new profit model; it just plugs a few more big customers into the old network. Calling it a "second curve" flatters it—it is a "high-quality new order for the first curve."

  So does a true second curve (a growth pole of a different nature) exist today? The answer is: basically not. A company with a real second curve would let you see a new business that is small today but can grow independently into a towering tree in the future, with economics superior to the main business. ONEOK has nothing like that—it has no energy-transition new platform (no large-scale hydrogen, CCS, or standalone power assets), no asset-light high-return new model; all its "growth" shares the same "heavy-asset + toll-collecting" DNA and the same ceiling.

  This raises a key concern from Baillie Gifford's angle: when U.S. oil-and-gas production growth slows, or energy-transition policy tightens, and the main curve's volume growth peaks, ONEOK has nothing ready to take the baton. Its returns would then collapse into "dividend + inflation-level price increases," which is exactly the "quasi-bond equity" the report's bear case describes.

  Conclusion: ONEOK's growth is a single-engine linear extrapolation, not a two-engine relay. The second curve does not exist today—this is one of the core reasons it does not deserve the "great growth stock" label.

  Jun 11, 2026
• QWhat is its core competitive advantage? Will this moat widen or narrow over the next three to five years?6/10

  A

  Its core competitive advantage is a hard-to-replicate asset network + location + regulatory/permitting barriers + customer connectivity—a "moderately-to-fairly strong, real but capped" moat; over the next three to five years it most likely "widens slowly," but it is by no means the kind of super-moat that deepens automatically—and its vulnerability happens to hide in that same "regulation/permitting" block.

  Where the moat comes from (item by item, by strength):

  • Replication difficulty / licensing-and-regulatory barriers (strongest). ONEOK's assets are regulated by FERC and constrained by multi-tier permitting and environmental rules. The report notes: a new competitor "even with money may not be able to replicate an equivalent-scale network of routes within a reasonable time." Building a new segment of assets takes years and enormous capital (ONEOK's share in the Texas City terminal JV is about $1 billion); this triple gate of time + capital + approval is a real barrier.

  • Scale and system connectivity (moderately strong). The report notes about 60,000 miles of pipeline network, NGL fractionation capacity over 1.2 million barrels/day, natural gas processing capacity over 7 Bcf/d, and a refined-products system of about 14,000 miles / 53 terminals / 4 marine terminals / about 115 million barrels of storage. The more complete the system, the more it can aggregate volume, spread unit costs, and cross-sell.

  • Switching costs / location (moderate). Once upstream producers, refineries, and exporters connect to a given processing-storage-transport-terminal chain, switching is hard, because alternative routes do not always exist and often mean longer distances and higher losses. ONEOK's positioning in the Williston, Permian, Mid-Continent, and Gulf Coast makes it "supply-chain infrastructure" to customers rather than a freely replaceable service provider.

  • Physical network effect (weak to moderate). Not an internet-style network effect, but the "the more complete the system, the more volume it attracts" self-reinforcement does exist (the report's cited discussions with 40+ data-center/power-generation counterparties are an example).

  • Brand (weak). The report bluntly says "there is no Coca-Cola-style brand wall here"; its competitiveness is not in consumer mindshare.

  The next three to five years: the forces widening it vs. the risks narrowing it.

  The basis for widening is real: the platform has extended from a single NGL company to a balanced four-segment platform of NGL / natural gas G&P / refined products and crude / natural gas pipelines (2026 guidance EBITDA mix about 35% NGL, 27% G&P, 27% refined products and crude, 10% natural gas pipelines), and it keeps expanding in Gulf Coast, Permian, exports, and data-center/power-generation connections. The more connected the assets and the more customer interfaces, the harder it is to bypass.

  But the narrowing risk is equally concrete, and it is two sides of the same coin as its strongest moat: the report lists permitting, regulatory compliance, litigation/activism, environmental disputes, interest rates, and policy change as major risks. "Hard to permit" blocks competitors but may also block ONEOK's own core growth projects; if energy-transition policy tightens faster, the fossil-energy throughput it lives on is exactly what gets structurally eroded. Rising cost of capital would also weaken its "trade capital for growth" moat logic.

  Against Baillie Gifford's "forward-looking" discipline—moat strength should be judged on the margin of the next 4–8 quarters, not on historical metrics. Historically, 60,000 miles of pipeline only proves it "was once" strong; the marginal signals that truly determine the moat's trajectory going forward are: whether new Permian/Gulf Coast pipelines come online on schedule, whether data-center gas contracts can be firmed up, and whether net leverage can be pushed from 3.8x back to the 3.5x target (if leverage won't come down, it will limit continued investment in the moat). These forward-looking variables say more about whether the moat is wide or narrow three-to-five years out than those 60,000 miles do.

  Conclusion: the moat is real, moderately-to-fairly strong, coming from steel and permits rather than brand, and directionally widening slowly but with a clear policy ceiling and capital constraint. It is sound enough, but not the kind of top-tier moat that keeps getting deeper automatically and leaves rivals permanently behind.

  Jun 11, 2026
• QIf its core business were disrupted, does it have the DNA to reinvent itself? How does it treat mistakes and bad news?4/10

  A

  To be blunt: ONEOK is a heavy-asset infrastructure company that "eats off physical assets," and it does not have, and does not need, the kind of "self-reinvention DNA" that lets a company completely transform after being disrupted—the thicker its moat, the slower its turn. Its attitude toward mistakes and bad news, by contrast, is relatively candid and well-disclosed. These two points should be viewed separately.

  First, the implicit premise of "self-reinvention DNA"—does it even apply to this company?

  Baillie Gifford uses this question to test companies like Amazon (from bookstore to cloud) or Nvidia (from gaming to AI) that "grow a new body when the core is disrupted." ONEOK is not that species, for structural reasons: all its value is locked into 60,000 miles of steel pipeline network and underground storage, which are sunk, specialized, and non-relocatable assets. In the liquidation-value section the report is very clear—pipelines, terminals, storage tanks, and interconnects are "worth far more in normal operation than in a break-up sale," and "if it truly entered forced disposal, specialized assets could also be realized below book value." In other words, once the core business (fossil-energy logistics) is truly disrupted by the energy transition, ONEOK cannot convert its pipelines into some other money machine; it can only watch its assets strand. This is exactly the source of the "permanent capital loss risk" and "valuation multiple contraction" the report repeatedly stresses.

  So does it have any adaptability at all? Yes, but at a very low level: it can do category migration and adjacent expansion within fossil energy—from a pure NGL company, growing via M&A into a four-segment platform of NGL / natural gas / refined products / crude, and plugging new pipelines into new demand outlets like LNG exports and data-center power (the report notes discussions with 40+ data-center/power-generation counterparties, potential >5 Bcf/d). This is the flexibility of "switching trees within the same forest," not the reinvention DNA of "swimming from land back to the sea." When the whole forest (hydrocarbon energy) is cut down, this kind of flexibility cannot save it. So the honest conclusion is: low disruption resistance, and self-reinvention DNA essentially absent—this is the essential mark of it being a "quasi-bond equity" rather than a growth stock.

  Second, "how it treats mistakes and bad news"—here it does well.

  • Candid disclosure. The report repeatedly cites negative/sensitive information ONEOK proactively discloses: in its forward-looking statements it "lists permitting, regulatory compliance, litigation/activism, interest rates, ratings, and external market changes as major risk sources"; it candidly admits that winter weather, Bakken/Permian volume swings, refined-products seasonality, and spread changes affect quarterly results; and it clearly states current net leverage of about 3.8x is still short of its self-set 3.5x target—putting its own "not-good-enough" on the table rather than papering over it.

  • Incentives that restrain it from just chasing size. The report notes that in the 2025 short-term incentive, EPS carries 40%, ROIC 40%, and safety (TRIR) and environment (AREER) 10% each, with long-term equity incentives tied to relative TSR percentile (0%–200%). Writing ROIC, safety, and environment into the metrics means management is institutionally required to answer for "whether capital is used well and whether operations have accidents," not just chase scale—an ex-ante discipline against the "cost of mistakes."

  • A reserved doubt. The report also cautions that M&A synergies are "the thing most easily overestimated by the market," and that capital allocation over the past two years has bet heavily on M&A; if future ROIC cannot hold in the double digits and per-share value does not improve in step with the share expansion, then one should watch whether management would downplay the bad news of integration falling short. The final answer on this item depends on whether synergies over the next 2–3 years genuinely land in per-share cash flow.

  Conclusion: as an infrastructure company, ONEOK's attitude toward mistakes and bad news is acceptable and even preferable (transparent disclosure, metrics including ROIC/safety/environment); but it innately lacks the DNA to reinvent itself after being disrupted—this is not a matter of management being good or bad but the fate of the heavy-asset specialized-network business form. If Baillie Gifford is looking for a "rise-from-the-ashes" growth stock, this is not the answer.

  Jun 11, 2026
• QDo management (especially the founders) have a long-term vision, with interests deeply tied to the company? Are they willing to sacrifice today's profit for five-to-ten years out?4/10

  A

  Management is broadly rational, with interests "aligned but not deeply" with shareholders; the long-term vision shows in a willingness to invest in multi-year projects at the cost of periodically compressing free cash flow—but it is not a founder-type owner-operator, and insider ownership is under 1% of shares outstanding, so on "deep alignment" it only rates moderate.

  First, the degree of alignment (what Baillie Gifford values most—whether founders/management put their net worth on the line)—this is a clear weakness.

  The report discloses it specifically: the CEO's ownership guideline is 6x base salary and other executives 2–4x, and all executives met it in 2025; CEO Pierce Norton II holds about 127,782 shares, and all directors and officers together about 1,228,173 shares (including director phantom stock), still under 1% of shares outstanding. Measured against the current ~$90.57 share price and ~$57 billion market cap (about 630 million shares), the absolute value of the CEO's modest holding, set inside a nearly $60 billion company, hardly qualifies as "net worth on the line." This is not remotely the same order of magnitude as the growth stocks Baillie Gifford prefers, where founders hold heavily (often double-digit percentages). ONEOK is a textbook case of professional-manager governance, not owner-operator—executives have "some economic alignment," but nowhere near "deep" interest binding with the company.

  Second, the long-term vision and "willingness to sacrifice today's profit for five-to-ten years out"—this is actually a relative bright spot.

  The nature of the midstream infrastructure business forces management to make long-cycle decisions, and ONEOK is serious about this:

  • Investing real money in the future, periodically sacrificing current FCF. The report notes 2025 total capex of about $3.152 billion (above $2.021 billion in 2024 and $1.595 billion in 2023), which pushed reported FCF down from $2.867 billion in 2024 to $2.447 billion in 2025. Knowing it would compress current free cash flow, management still keeps building Eiger Express, the Medford rebuild, the Bighorn processing plant, the Texas City terminal, and other projects that only come online years later—this is precisely the concrete action of "sacrificing today for five-to-ten years out."

  • A clear and self-disciplined capital-allocation framework. The report notes the company's explicit priorities: high-return organic projects → maintain and grow the dividend → keep an investment-grade balance sheet → repurchase under authorization. The board authorized a $2 billion buyback, but actual 2025 repurchases were small—showing it did not spend recklessly to prop up the short-term share price but put money into long-duration assets. 2025 net income up 12% to $3.39 billion, adjusted EBITDA up 18% to $8.02 billion, a continuation of 12–13 consecutive years of growth, reflecting long-term discipline.

  • Incentive design that folds in the long term. Short-term incentives include ROIC 40% and safety/environment 10% each, with long-term equity incentives tied to 3-year relative TSR—not tying incentives entirely to a single "get bigger" point, which is relatively reasonable for the midstream industry.

  But the reservations must be stated clearly (the deductions from Baillie Gifford's angle):

  • Heavy bets on M&A over the past two years, with strong aggression. In 2025 it completed EnLink (41 million shares issued, about $4 billion) and the Delaware Basin JV (about 4.9 million shares issued, about $391 million). The report bluntly says "M&A does not create value automatically" and "synergy is the thing most easily overestimated by the market." Trading stock issuance for growth, if per-share intrinsic value does not rise in step, dilutes long-term shareholders—this is the part of the "long-term vision" that needs continued verification by results rather than being already proven.

  • No founder's "this is my life's work" irrational long-term commitment. Professional managers have limited tenures and are inevitably more swayed by within-tenure TSR metrics than a founder would be.

  Conclusion: management's capital allocation is rational, long-term-project-oriented, and willing to sacrifice current FCF for the distant future—all true; but "interests deeply tied to the company" only rates moderate because holdings are under 1% and there is no founder, and the heavy-M&A approach still needs the next 2–3 years to prove its sincerity by "turning synergies genuinely into per-share cash flow." It is a reliable professional steward, not a founder with net worth on the line.

  Jun 11, 2026
• QIf it disappeared tomorrow, how much would customers miss it? Is its way of growing sustainable and not reliant on harming society and regulation?5/10

  A

  Answer the two dimensions separately: indispensability—very high (if ONEOK's network stopped tomorrow, the refineries, power plants, exporters, and gas utilities directly connected to it would be immediately hit hard); the social/regulatory sustainability of its way of growing—moderate with a long-term open question (it serves fossil energy, itself a target of energy-transition policy, but natural gas as a "transition fuel" still enjoys high social acceptance over the short-to-medium term). What this Baillie Gifford question wants is a double answer of "both indispensable and able to be tolerated by society and regulators for the long haul"; ONEOK passes the first, and gets a question mark on the second.

  First: indispensability—high.

  ONEOK sells not an optional consumer good but the "toll passage" embedded in the physical backbone of the energy value chain. The report discloses that its assets connect directly to utilities, industrial customers, petrochemical facilities, refineries, exporters, local gas distribution companies, and power facilities; its natural gas pipelines connect directly to LNG, industrial, and power-generation demand. Once these pipelines, processing plants, and terminals stop:

  • upstream producers' gas and NGLs cannot be moved out or sold;
  • downstream refineries, power plants, and gas companies lose their feedstock and fuel supply;
  • alternative routes do not always exist—the report makes clear that switching "often involves longer distances, higher losses, and lower netbacks."

  This "cut it off and everything stops" property is exactly where it most resembles infrastructure, and the source of its moat. Customers would miss it dearly, because there is simply no equivalent substitute in the short term. Layered with structural incremental demand—ONEOK says it is in discussions with more than 40 data-center and power-generation counterparties, potential demand exceeding 5 Bcf/d, with AI electricity demand forcing gas-fired power—its indispensability only grows over the visible next few years.

  But an honest caveat that knocks "indispensability" down a notch: its indispensability is regional and nodal, not nationally unique. In many regions it coexists and competes with the networks of Williams, Kinder Morgan, and Energy Transfer. Customers cannot do without "a given pipeline," but at the industry level, leaving ONEOK does not mean there is no gas to be had. So it is "high," not "sole-source monopoly."

  Second: whether its way of growing is sustainable and not reliant on harming society and regulation—this is Baillie Gifford's real test, and the answer is "moderate, with a long-term open question."

  This has two sides:

  The relatively positive side: natural gas is widely viewed as the "transition/peaking fuel" during coal retirement and intermittent renewable supply, with higher social acceptance than coal and oil. ONEOK serves LNG exports (helping allies replace Russian gas) and supplies gas for data-center power—narratives that are encouraged rather than suppressed in the current policy context, and the EIA projects U.S. natural gas production and exports to keep growing through 2027. Its growth does not rely on deceiving users or regulatory arbitrage, but on providing real logistics services and charging reasonable tolls, and its price increases are constrained by FERC tariffs and contract CPI/PPI, so it cannot charge sky-high prices unilaterally—meaning its way of making money is tolerated by society at present.

  The unresolved side: its entire business is built on fossil energy continuing to be extracted, transported, and consumed. The report lists permitting, environmental disputes, community/activist resistance, and policy change as major risks, and points out that "the moat of these assets is precisely their vulnerability: permits are hard, environmental disputes are many, and policy changes are fast." That is, the regulatory barrier that protects it from replication is also the rope that could throttle it: if the carbon-neutral path accelerates, if new pipeline permits are blocked, or if environmental litigation escalates, its growth will be directly limited by "society and regulation." This is the real question mark on "whether growth is sustainable and not reliant on fighting regulation"—it does not harm users, but it is in inherent tension with the long-term direction of decarbonization.

  Conclusion: ONEOK is highly indispensable to customers (cut it off and everything stops, with data-center/LNG tailwinds on the demand side), passing this dimension solidly; but "the long-term social/regulatory sustainability of its way of growing" only rates moderate—natural gas as a transition fuel holds up over the short-to-medium term, but the long term is under the overhang of energy-transition policy, so on this double question half of it is solid and half is in doubt.

  Jun 11, 2026
• QWhat are the unit economics of this business (gross margin, incremental returns)? Do they improve or worsen as it scales? Where does the money it earns go?5/10

  A

  The unit economics are "a top student within heavy-asset infrastructure, but only middling once placed in Baillie Gifford's growth-stock pool": the fee-based model keeps margins stable and scale brings a modest unit-cost improvement, but its growth is bought with constant capital outlays, and incremental return on capital is only "double-digit ROIC, good for midstream but far from outstanding." The money it earns goes mainly to growth capex and M&A, then to the dividend, with buybacks minimal.

  Gross margin / contribution margin: the headline gross margin is distorted by commodity pass-through; the real number to watch is EBITDA margin, and the quality is good.

  A midstream company's "revenue" contains large amounts of buy-at-market, sell-at-market purchase-and-sale volume (POP contracts, spread optimization), so accounting gross margin is distorted and meaningless. To see true earning quality, use EBITDA against revenue: 2025 adjusted EBITDA of about $8.02 billion on revenue of about $33.63 billion, an EBITDA margin of about 24%; but the cleaner measure is fee-based cash flow—the report notes about 90% fee-based overall (by segment: NGL about 85%, G&P about 90%, refined products and crude about 90%, natural gas pipelines about 85%), meaning 90% of revenue is recurring, predictable "tolls," with only about 10%–15% exposed to commodity spreads. The stability of this margin is its true strength—far more resilient than a pure upstream cyclical stock.

  Incremental returns / scale effect: it "improves slightly" as it grows, but not software-style "the bigger the sweeter."

  • The improving side: the report notes about 60,000 miles of pipeline, over 1.2 million bbl/d of fractionation, and over 7 Bcf/d of processing capacity bring "better unit costs, volume-aggregation capability, and cross-selling synergies"; M&A synergies (the 2026 guidance includes about $150 million of annual synergies) are also incremental profit from scale. The more complete the system, the lower the marginal cost of adding another customer.
  • The capping side: but nearly every unit of new growth must be preceded by an enormous capital outlay—the report cites ONEOK's roughly $1 billion share in the Texas City terminal JV, and new pipelines and processing plants routinely cost several years and several billion dollars. This is "capital for growth," not software's "marginal cost trending to zero" exponential leverage. So the benefits of scale are gradual, diluted by capital intensity, without a sudden leap in unit economics.

  Incremental return on capital (what Baillie Gifford cares about most—does reinvested money earn?): double-digit ROIC, good in the industry, but not outstanding.

  The report notes the company defines ROIC as EBIT/Invested Capital, set its 2025 ROIC target at 11.58%, and that the actual figure was roughly "still around the double digits." The report's characterization is apt: "an acceptable, even preferable, level for heavy-asset midstream, but far from 'outstanding return on capital.'" Compared with the growth stocks Baillie Gifford truly wants (incremental returns on capital often 20%–40%+, without needing proportional capital outlays), ONEOK's roughly 11%–12% ROIC is only "above the passing line"—it can create value on reinvestment, but not much and not fast.

  Where does the money it earns go? (capital-allocation flows, in the priority order the report discloses)

  The report notes the company's explicit capital-allocation priorities: ① high-return organic projects → ② maintain and grow the dividend → ③ keep an investment-grade balance sheet → ④ repurchase under authorization. In practice:

  • The bulk is growth capex: 2025 total capex of about $3.152 billion, 2026 guidance $2.7–3.2 billion (of which maintenance is only $550–600 million, showing the vast majority is growth-oriented).
  • Second is M&A: the core moves over the past two years were integrating EnLink, Medallion, and Magellan, expanding the platform via stock issuance.
  • Third is the dividend: annualized dividend of $4.28, yield about 4.7%, with 2025 operating cash flow about $3 billion higher than dividend payments—ample coverage.
  • Buybacks are minimal: the board authorized $2 billion, but actual 2025 repurchases were very small.

  This "earn money, mainly plow it into build-outs + M&A, then pay out a big dividend" approach is itself rational, but it brings out a key judgment from Baillie Gifford's angle: it needs continued large reinvestment just to sustain growth, and the return on that reinvestment is only double-digit—this makes it "a high-quality tool that steadily spits out cash," not a "self-snowballing, ever-faster" compounding machine.

  Conclusion: the unit economics are top-tier among midstream peers (high fee-based, stable margins, modest scale improvement), but incremental return on capital is only double-digit and growth is highly dependent on capital outlays, so within Baillie Gifford's growth-stock standard it is only middling—which is exactly the financial backdrop of the report's repeated "solid but not perfect infrastructure business."

  Jun 11, 2026
• QFor it to 5x in ten years, what conditions must all hold at once? Are these conditions realistic? What expectations does today's share price imply?3/10

  A

  Straight to the conclusion: ONEOK rising fivefold in ten years is almost impossible—it would take a whole stack of unrealistic conditions all landing together—and today's share price implies nothing like a "5x expectation" but the pricing of a "reasonable-to-slightly-low, yield + moderate growth" bond substitute. On ONEOK, this Baillie Gifford question is basically a counterexample.

  First the hurdle: what does 5x in ten years mean?

  Fivefold corresponds to about 17.5% annualized total return (including dividends) for ten straight years. It would take today's ONEOK—about $90.57, market cap about $57 billion—to roughly a $285 share price / a market cap of about $285 billion in ten years (ignoring share issuance; otherwise the bar is higher). Yet the report's own expected annualized returns are: conservative 6%–8%, base 9%–11%, optimistic 12%–14%—even the most optimistic scenario of 14% falls short of the ~17.5% needed for a fivefold in ten years. This is the ceiling the report's own author computed by DCF, not pessimism I added on top.

  Which conditions must hold at once? Listed one by one, with realism flagged:

  1. Owner Earnings/EBITDA compounding rapidly for ten straight years (needs about 12%–15%/year). Realism: very low. The report notes the 2026 guidance midpoint EBITDA of about $8.1 billion, up only about +1% year over year, with the company's organic-growth baseline being 12–13 consecutive years of single-digit to low-double-digit growth. Multiplying EBITDA several times in ten years would require an endless stream of large acquisitions, and M&A requires stock issuance that dilutes per-share value, which instead drags on per-share returns.
  2. Valuation multiple expanding sharply (P/E re-rating from about 16x to 25x+, converging toward the broad market). Realism: low. The report's bear case clearly says "the market is already willing to give it a not-low utility-style valuation," and its net leverage of 3.8x is above its self-set 3.5x target and interest-rate sensitive, so it deserves a discount to pure long-haul assets. The multiple is more likely to stay flat or even contract than to double.
  3. Leverage keeps falling while it also expands rapidly (having it both ways). Realism: low. Deleveraging requires cash to repay debt, and rapid expansion requires cash to build/acquire; the two compete for the same money and cannot both be delivered to the extreme.
  4. Data-center/LNG/industrial demand realized at the top of the range, with core projects free of regulatory delay. Realism: medium. The demand tailwind is real (potential >5 Bcf/d in discussion), but the report lists permitting, environmental, and policy issues as major risks, and project delays are the norm.
  5. Oil and gas not significantly displaced in the energy mix for ten years. Realism: medium, but with a long-term policy overhang.

  Key judgment: these five mustallhold at once for even a shadow of fivefold, and they even fight one another (deleveraging vs. expansion, M&A-driven scale vs. per-share non-dilution). If any one deviates, the return collapses back to the report's "conservative 6%–8%." Multiplying five low-to-medium-probability events together, the actual probability of fivefold in ten years is vanishingly small. This is not that ONEOK is bad, but that a heavy-asset, regulated, single-digit-organic-growth, nearly $60 billion market-cap infrastructure company mathematically cannot grow fivefold—Baillie Gifford's firepower should be trained on the blue-sky imagination of years 3–10, and ONEOK's years-3–10 imaginative room is firmly sealed off by its business form.

  So what does today's share price imply?

  It implies a "reasonable-to-slightly-undervalued bond substitute," not a growth stock:

  • Priced at a TTM P/E of about 16.2x and a dividend yield of about 4.7%, the market assumes "single-digit EBITDA growth + a stable high dividend + gradually falling leverage," i.e., the report's conservative-to-base scenario.
  • The report's DCF shows the current price is only "near fair or slightly discounted" to conservative intrinsic value ($80–95/share), and carries about a 14%–30% discount to base value ($105–130). That is, the market has already priced it to a "not-cheap, not-expensive, slowly rewarding through dividends and moderate growth" position, without pre-paying for any rapid growth.
  • Analyst consensus confirms it too: an average price target of about $95.48, only about 5% upside, rated Buy—the market expects "a bit more upside," not "another fivefold."

  Conclusion: the conditions needed for fivefold in ten years are both unrealistic and mutually conflicting, and even the report's most optimistic scenario falls short of the bar; today's share price implies a reasonable-to-slightly-undervalued high-yield infrastructure-equity expectation, not a growth expectation. By Baillie Gifford's yardstick, ONEOK clearly fails on this most core question—it is a "yield + moderate growth" tool, not a "fivefold in ten years" candidate.

  Jun 11, 2026
• QWhy has the market not yet realized all this? Is it that it cannot understand, looks down on it, or cannot see far enough? What would become the "narrative inflection point"?3/10

  A

  The most honest answer is: on the matter of "a fivefold-in-ten-years growth stock," the market has not "failed to realize" anything—because there simply is no buried growth story waiting to be discovered. ONEOK is already fairly fully, even fairly, priced as a "reasonable-to-slightly-undervalued high-yield infrastructure equity," and Baillie Gifford's "why can't the market understand it" question essentially comes up empty on ONEOK. The places where there is a genuine, faint perception gap are very limited and not large in direction.

  Why the market has basically "got it right," rather than "cannot understand / looks down on it / cannot see far":

  • Not "cannot understand." Although the midstream business model is more complex than consumer goods, ONEOK is a large-cap with a market cap of nearly $57 billion, covered by 11-plus sell-side firms, rated Buy with an average price target of about $95.48, with ample institutional research. It is not an obscure small-cap; there is no question of no one understanding it.
  • Not "looks down on it." On the contrary, the report's bear case points out "the market is already willing to give it a not-low utility-style valuation" and "its good days are already partly priced in." The current TTM P/E of about 16.2x, though below the broad market's roughly 27x, is the midstream industry's reasonable structural discount (heavy-asset, regulated, single-digit growth, elevated leverage), not a sentimental disdain by the market. Over the past year it rose from a 52-week low of $64 to near a high of $96, already re-rated with considerable enthusiasm.
  • Not "cannot see far." The long-term tailwinds of data centers / LNG / gas-fired power have long been in the market's telling—by early 2026 there were plenty of sell-side "midstream renaissance" and "AI forcing natural gas" narratives. This "far" story is not being ignored; rather it is the current market consensus, and even the main reason supporting its valuation.

  So where is the only thing that qualifies as a "perception gap"? Small in direction, and neutral-leaning:

  What the report identifies as the genuine possible market mispricing is the difference between Owner Earnings and reported FCF: reported-basis FCF (about $2.447 billion in 2025) looks unattractive because of the large growth capex, whereas conservative Owner Earnings after stripping out growth capex is about $4.3 billion, corresponding to about 13.3x and an Owner Earnings yield of about 7.5%. That is, investors fixated only on GAAP free cash flow may systematically understate its true distributable earning power, because they mistake "growth investment" for "the capex needed to sustain the business." This is a real but mild perception gap—it can explain why ONEOK carries about a 14%–30% discount to base intrinsic value ($105–130/share), but it is far from enough to support fivefold, at most a limited re-rating "from reasonable-to-slightly-low back to fair."

  What would become the "narrative inflection point"? (both directions listed)

  The upside inflection (nudging the market to re-rate it a bit from "bond substitute" toward "growth stock"):

  • Data-center/power-generation gas hookups going from "in discussions with 40+ counterparties, potential >5 Bcf/d" to firmed long-term contracts that begin coming online, turning the "story" into "incremental EBITDA on the statements";
  • Net leverage clearly pushed from 3.8x back to the 3.5x target, removing the biggest discount rationale of "quasi-bond equity";
  • EnLink/Medallion/Magellan synergies genuinely landing in per-share cash flow (not just looking good on the EBITDA line), proving M&A created per-share value.

  The downside inflection (breaking the narrative and contracting the valuation):

  • Synergies failing to materialize for 2–3 consecutive years, ROIC falling below cost of capital;
  • Interest rates rising or the market re-rating it to a lower "utility/pipeline" multiple;
  • Core growth projects delayed long-term by permitting/environmental issues, falsifying the growth logic;
  • Dividend growth forced to slow for reasons of a deteriorating balance sheet.

  Bringing it back to Baillie Gifford's angle: the standard use of this question is to find a wrongly-sold-off growth stock where "the market is short-sighted but you see further"; ONEOK is not that name—the market understands it quite well, its discount is "reasonable rather than mistaken," and the only perception gap (Owner Earnings vs. reported FCF) is mild in direction and limited in magnitude. In other words, the reason to like ONEOK is "it is fairly priced as a decent high-yield infrastructure tool, with a slight discount," not "the market cannot understand a great growth stock." The latter does not exist.

  Jun 11, 2026