Summary

Horizon Kinetics’ James Davolos frames just-IPO’d WaterBridge as “the leading water infrastructure company in Texas”—more like critical waste infrastructure than traditional midstream. In the Delaware Basin, about 3.7–4 barrels of hypersaline produced water come out with every barrel of oil and gas, and the produced-water infrastructure industry “was completely nonexistent before the shale boom”—it’s barely a decade old, yet now “absolutely critical infrastructure” for a basin representing roughly 10% of global production volume.
The biggest misperception, per Davolos: water escapes the shale treadmill. Oil declines 30%+ early, but the water cut rises as wells age, keeping total fluid volumes similar for decades—and moving into lower-tier Wolfcamp rock pushes cuts from 4:1 toward 5:1–6:1 or even over 10:1. His call: “If the Permian produces flat oil for the next twenty years, water volumes will grow almost certainly mid to high single digits, if not higher. I don’t think very many people appreciate that.”
Scarce pore space is the moat, and everyone now acknowledges the capacity problem. Deep injection began creating earthquakes, shallow injection can cause sinkholes and legacy-well blowouts, and “we are running out of pore space”—so switching providers “used to be very easy, now it is difficult, and I think in the future it’s getting closer to infeasible than just difficult.” Proof point: Devon, “in a first-of-its-kind transaction,” paid WaterBridge just to reserve future pore space.
Unit economics: ~78 cents per barrel, ~44 cents operating margin, and ~51% consolidated EBITDA margin on just shy of $400M run-rate EBITDA. Davolos cites an indicative, theoretical opportunity set of ~$3.5B of CapEx that could contribute ~$1B of EBITDA at full utilization—roughly a 30% unlevered return on incremental capital—spread over five to ten years and explicitly not guidance, with long-term fixed-fee contracts and CPI escalators on an 11-year weighted average.
The valuation arbitrage is the trade: the IPO was benchmarked against gathering & processing comps (~9x forward EBITDA, with a one-turn discount), but Davolos argues that’s the wrong peer set. “15% or better” organic growth for three to five years at eight times, with multi-decade CPI-escalated contracts and 50% margins—“Doesn’t sound right to you. It doesn’t sound right to me. So what does it sound like? It sounds like waste.” Waste comps (Casella, GFL, Waste Management, Clean Harbors) trade 14–18x+, and HK’s base case does not require a rerate.
The Five Point ecosystem is the meta-story: land (LandBridge), water (WaterBridge), sour gas (Northwind, sold to MPLX in what Davolos heard was a “hotly contested bidding war”), and power (PowerBridge, run by the former Talen CEO). LandBridge broke its IPO price at $17, ran to $80 by year-end, and sat in the mid-$50s at recording. Davolos thinks the market overrated data-center call options “happening now versus T+1, T+2, or T+3” while underrating water.
A structural quirk keeps incumbents entrenched: Texas mineral rights carry effective eminent domain, but water does not. “A rancher could literally just say, ‘No, you cannot cross my land, hard stop’"—so crossing 50 miles means negotiating with 50 landowners, favoring whoever already controls easements. Meanwhile disposal has migrated from junior marketing staff “up in the CFO’s office” as hypothetical $4–6/bbl water costs hit LOE against roughly $40 netbacks, driving E&Ps such as Devon to contribute systems and take equity rather than compete internally.

Deep dive

1. Horizon Kinetics’ thirty-year land thesis, from TPL to WaterBridge

Davolos traces the firm’s window into this niche to 1995, when Murray Stahl found “this funky liquidating land trust”—Texas Pacific Land, born from the 1880s bankruptcy of the Texas and Pacific Railway. The illustrative math: buying back stock at an implied $10/acre against $100 fair value made it “the most accretive compounding machine you’ve ever seen.” Stahl now sits on TPL’s board.
The framework distilled: versus capital-intensive, cyclical upstream/midstream/downstream and oil-field services, HK prefers land—“a pure margin business that’s perpetual with optionality.” Royalties are “mailbox money” in Texas parlance: no OPEX, no CapEx. “It’s a pretty good gig if you can get it.”
WaterBridge, public as of the day before recording, operates the water infrastructure; LandBridge is “the land company with basically triple-net leases that facilitate the activities of WaterBridge” plus next-generation power and data-center optionality.

2. Produced water: a decade-old industry born from earthquakes and sinkholes

The mechanism, as Davolos tells it: the Permian is an ancient seabed, so fracked rock yields water “many magnitudes more saline than seawater” with corrosive compounds—“in every sense of the word, it is a waste product.” The Delaware Basin runs ~3.7 barrels of it per barrel of oil and gas, within a Permian producing 11–12M boe/d against a little over 100M barrels globally.
Disposal evolved from calling the neighboring rancher, to deep injection below the shale—which began creating seismic events that caught the Railroad Commission’s attention—to today’s ~75% shallow disposal, which brings its own sinkholes and can cause blowouts through decades-capped legacy vertical wells. “Bookmark this concept of pressure and pore space.”
Enter the third party: WaterBridge can “guarantee your billion-dollar pad-drilling plan” for 50,000–200,000+ barrels of disposal, with injection wells, transport pipe, and long-haul offtake out of basin. In Davolos’s hypothetical, at $1/bbl and 4:1 cuts, that’s $4 of LOE when “you’re lucky if you’re netting 40 in the Permian.”

3. Water cuts break the treadmill—the episode’s key analytical claim

Davolos calls the shale-treadmill analogy “one of the biggest misperceptions”: water cuts are lowest at first production and climb as the well ages, so total fluid volumes stay remarkably stable even as oil declines—a big well means “decades, maybe 30 years, of water liabilities that you need to figure out.”
The second driver is moving into lower-tier acreage. Tier two isn’t more expensive mainly because it’s deeper—“it’s generally more expensive because there’s more water”—pushing water cuts from 4:1 toward 5:1–6:1 and, in some multi-decade scenarios, “over ten to one.” Hence his qualified call: flat Permian oil for twenty years still yields water-volume growth “almost certainly” in the mid-to-high single digits, if not higher.
Matt’s follow-up on whether this translates to the Eagle Ford or Bakken gets a flat “short answer is no”—they do not have nearly as high water cuts or production volumes; “the opportunity is very acute and the largest in the Permian, specifically the Delaware.”

4. Five Point: the sponsor whose Vulcan dismissal helped build the franchise

The origin story: David Capobianco, a traditional banker with midstream experience who worked at Paul Allen’s Vulcan and took Plains All American private, was, according to Davolos’s reading of filings and legal analysis, terminated and replaced by his team in an effort by Vulcan not to pay what was ultimately disclosed as a $20M settlement. Capobianco and his partner won when they litigated in court, and the episode helped lead to Five Point’s founding in 2012. He later acquired a water company run by Jason Long, now CEO of both WaterBridge and LandBridge.
Two pivotal moves: an area-of-mutual-interest JV with TPL to exploit the state-line core of northern Loving County (“I don’t think many people could have pulled this off other than David and Jason”), and buying the ~70,000-acre Hanging H Ranch in 2021 after a private-equity-backed buyer failed to close during COVID.
On keeping the vehicles separate, Davolos invokes Newmont buying and re-spinning Franco-Nevada: “you would never get the appropriate multiple of land within a broader portfolio,” and Aris “probably never got the multiple that it deserved”—investors must be able to isolate water as a pure play.
LandBridge’s IPO stumbled from a slated $19–22 to $17, ran to $80, and sat in the mid-$50s at recording. The narrative “got taken over” by six-to-eight plug-and-play multi-gigawatt data-center sites and beneficial water reuse—real call options, but the market “got a little too excited about them happening now versus T+1, T+2, T+3.”

5. Contracts and lock-in: acreage dedications beat MVCs, and Devon proves scarcity

The contract stack, at an ~11-year weighted average: acreage dedications (the preferable form, with penalties if the dedicated acreage’s water is not handled by WaterBridge, plus CPI-linked escalators), minimum volume commitments (useful to de-risk projects but “it doesn’t guarantee you all that flow”), and a small spot component for producers “in a jam” where “you can extract a lot of economics.”
The scarcity tell: Devon paid WaterBridge to reserve pore space it might not need for three or four years—behavior that, as Matt notes, “contradicts a lot of what the producers do in oil world.” Davolos says some debate how large the shortfall is and whether it bites now or in 2027–2028, but nobody denies the disposal-capacity problem.
On outsourcing, the shift is structural: Devon contributed an asset and became a large WaterBridge equity holder, while Conoco was a large Aris holder through the legacy Concho contribution. Decisions once handled by “some junior person in the marketing team” are “now up in the CFO’s office.” Davolos expects the industry to become oligopolistic, with large third-party share gains, part organic and part from consolidating E&P-owned systems.
The legal kicker: mineral extraction enjoys effective eminent domain in Texas; water egress does not. Cross 50 miles of 640-acre checkerboard sections and “deal with 50 different landholders asking you different rates”—which is why incumbency compounds.

6. Economics: 51% margins, 30% incremental returns, and the sour-gas option

Pro forma: ~78 cents/bbl for produced-water handling (85% of revenue, 8% skim oil, the rest mostly water-solutions work cleaning water for refracs), ~44 cents operating margin per barrel, and ~51% EBITDA margin on just shy of $400M run-rate EBITDA. Maintenance heuristic: 10–15% of undepreciated PP&E, “probably on the high side,” on 7–28-year useful lives.
The Speedway pipeline moves New Mexico water east across the state line into the Texas Panhandle—critical because New Mexico permitting averages over two years versus weeks in Texas, and Lea and Eddy Counties are “juggernauts of US energy growth.” The indicative opportunity set: ~$3.5B of CapEx yielding ~$1B EBITDA at full utilization, or ~30% unlevered incremental returns Davolos expects to “drift higher”—explicitly a case study, not guidance, over five to ten years.
The non-obvious growth leg: the Delaware’s eastern sour-gas shelf, laden with dangerous hydrogen sulfide, has long been ignored by many operators. Five Point’s Northwind built the acid-gas-injection infrastructure and, in what Davolos heard was a “very aggressive, hotly contested bidding war,” sold to MPLX. WaterBridge and LandBridge have pre-positioned surface and infrastructure “ready to go once those drill bits come,” whether from Devon, Oxy, or Coterra. Davolos does not call the opportunity de-risked, but sees visible growth if the window develops.

7. Valuation: priced like midstream, argued as waste

HK builds bottom-up DCFs but reconciles to comps: G&P peers (Western Midstream, MPLX, ONEOK) trade ~9x forward EBITDA, and the IPO targeted a one-turn discount—it priced at the top of the range and upsized. Davolos’ reframe: 15%+ organic growth at eight times with multi-decade CPI-escalated contracts and a 50% EBITDA margin “doesn’t sound right… It sounds like waste”—where comps trade 14–18x+ and, he argues, have worse returns and growth than WaterBridge.
The base case excludes multiple expansion: organic growth from Kraken, Speedway, and the sour-gas window, plus possible bolt-ons like Berkshire-owned Pilot’s legacy-Diamondback Midland business, debt paydown, and buybacks. Beneficial reuse—desalination for the Pecos River watershed, agriculture, and industrial cooling—is a call option, “definitely not within 12 to 18 months.”
Key sensitivities: Davolos’s KPI is whether Delaware volumes remain stable or grow, including whether production shifts farther west into the Delaware; the murkier second layer is “are these returns on incremental invested capital as high as we thought? Is the addressable market as big as we thought?”
Closing lessons from roughly a decade in the ecosystem: build core competency but keep “your head on a swivel” for adjacencies; demand capital-light models that would still be good businesses in maintenance-only mode; and “the ultimate capital-light real asset is land. It’s perpetual, there’s optionality, it’s finite.”

Macro Talk 92: China’s Economic Restructuring Is Halfway Through E208｜How P&G Managed to Mass-Produce CEOs