Chadd Garcia drills into LandBridge's value

Summary

• Chad Garcia’s core call: LandBridge is a higher-quality Permian land/surface-use royalty play, trading ~20–22x EBITDA versus TPL’s ~28–31x — and the gap widens once you strip minerals. Put a Viper-style 10x on LandBridge’s 6% mineral royalty stream and the remainder trades ~22–24x; do the same for TPL with a PrairieSky-style 16x and TPL’s residual produced-water, source-water and surface-easement business is ~50x — “you almost have a 20-turn spread between the two.”
• The load-bearing number is pore space: LandBridge earns a royalty on 1.7 million barrels/day of produced-water injection today and holds over 7.5 million barrels/day of incremental capacity (5M+ permitted, ~2.5M likely permittable). Management — “while not giving long-term guidance” — repeatedly said 5 million barrels/day can be absorbed within five years at 15 cents a barrel, “which is all profit,” implying a ~25% free-cash-flow CAGR before any data centers, surface royalties, or frac-sand growth. Andrew says he thinks they cited ~$300M of incremental FCF repeatedly at the investor day.
• Why the stock is stuck near ~$70: Garcia blames a TMT-driven short that zeroed the data-center thesis, comped the rest to a 10x royalty business, and bet sponsor Five Point would “hit the bid” on any rally. He thinks the market reaction to an “extremely bullish” investor day has been muted, and expects Five Point’s future selling to be greater at WaterBridge than at LandBridge — “LandBridge is the vehicle that the sponsor is going to want to own for as long as possible,” with the sponsor separately putting $1B into PowerBridge.
• The land strategy is a blocking-position game, not passive royalty collection: 2024-vintage acquisitions grew free cash flow 150% year-over-year. LandBridge filled in TPL’s checkerboard in Loving County to unlock New Mexico-to-Texas water crossings, locked up the panhandle-corner pore space just as the over-pressurized area to the east will probably lose ~2M bbl/d of capacity over 3–5 years, and bought the 1918 Ranch (October 2025) for $375M against $20M current EBITDA. Garcia said its pore space was worth about $75M and described the purchase as 3.3x EBITDA, “because if you’re a competitor… guess what’s between it? LandBridge land.”
• The data-center option is a free call option but hasn’t hit yet — no hyperscaler is definitively announced in the Permian. Glimpses: Chevron’s contemplated 2.5GW (expandable to 5GW) power-generation facility filing for tax abatements in Reeves County, abutting LandBridge acreage, with Microsoft offtake contemplated; Eric Schmidt’s BOLT targeting 10GW of Permian data centers; and a LandBridge–NRG power-generation agreement from September. PowerBridge’s leadership previously built Cumulus in Pennsylvania — “nobody wanted anything to do with this project until it was complete and then… there was insatiable demand” (sold to Amazon, with ~$100B of capital attracted). The PowerBridge project is not expected to require LandBridge capital.
• On fair value Garcia won’t pin a number but says “triple digits is pretty reasonable… you start applying TPL multiples to this thing and you get to $150 pretty quick”; forced to choose, he picks LandBridge over WaterBridge for the highest FCF/share growth, the TPL re-rate gap, and the free data-center call — though he owns both.
• Postscript on Secure: GFL’s ~11x takeout validates the “waste, not energy services” thesis Garcia discussed in November 2024 (stock +77% in a year, +465% over five), but he calls the price “too cheap.” Abrams Capital crossed 10% and publicly opposed the deal; Andrew’s counter — with support agreements locked and the asset likely fully shopped, “isn’t this just worth taking the bid?” — versus Garcia’s view that the GICS code, gross-to-net reporting fix, and sell-side education were only just kicking in.

Deep dive

1. Land royalty 101: TPL’s history explains why surface economics differ from minerals

• Garcia’s setup: TPL was formed in 1888 from the bankruptcy of Jay Gould’s Texas and Pacific Railroad, which had received 3 million acres of land grants; the trust’s job was to sell acres and return capital, so by now it’s down to ~882,000 acres — and having long ago sold anything with agricultural or development value, what’s left is “the most desolate part of the country,” West Texas land that “looks like you’re on Mars.”
• His provocation — that this desolate real estate “is probably worth the real estate value of Manhattan” — draws a laugh from Andrew, but the revenue stack backs the point: just under half mineral royalties, ~30% source water (fracking intake, built by ex-EOG hires around 2016–18 into “probably the leading source water business in the basin”), ~16% produced-water royalties on pore space, and ~10% gravel and sand pits used for frack sand. Surface easements are separately described as recurring, high-quality revenue but are not assigned a percentage here.
• The pushback on lazy comps: investors see pure oil-and-gas royalty plays like Venom at ~10x EBITDA or PrairieSky at ~16x, look at TPL at ~30x, “and they don’t really get it” because they are equating the businesses. Andrew frames surface rights as potentially even more valuable than mineral rights; Garcia’s explanation is that surface land can support infrastructure, water and easement revenue in addition to minerals.

2. LandBridge’s mix is deliberately built around the highest-quality streams

• LandBridge exists “to help WaterBridge grow out its pipeline infrastructure and provide pore space” — WaterBridge is ~30% of the water currently sourced on its land, with other operators providing the rest. Revenue splits: surface-use royalties ~73% (mostly pore-space royalties plus royalties for fracking water), resource sales (frac sand, gravel, source water) ~20%, and minerals just 6%.
• Garcia’s quality ranking is explicit: mineral royalties, despite gushing cash at TPL, are “the lowest quality revenue stream because once you extract a mineral and the royalty’s paid on it, it’s gone forever,” and they’re directly tied to oil and gas prices. LandBridge acquires minerals only incidentally with land it wants, and he’d expect the percentage to shrink over time.
• The physical driver: Delaware Basin wells produce four to six barrels of water per barrel of oil — water six to seven times saltier than the ocean, mixed with hydrocarbons, “nasty stuff” that must be separated, processed and injected into salt-water-disposal wells. E&Ps need reliable takeaway and “want to sign 10–15 year deals that guarantee the water’s gone and never think about it again.”

3. The valuation spread — and Garcia’s read on who’s short and why

• The sum-of-parts math: LandBridge trades ~20–22x EBITDA, TPL ~28–31x. Strip LandBridge’s minerals at a Viper-like 10x and the rest is ~22–24x; strip TPL’s “vast” mineral package at PrairieSky’s 16x and the residual produced-water, source-water and easement business sits near 50x — a spread of roughly 20 turns for arguably similar assets.
• Andrew’s challenge: the company itself flags the comp gap on slide six of its investor day — isn’t the market already “on to it”? Garcia’s answer: no, because the investor day’s “extremely bullish” content drew a muted reaction. His explanation is a heavy short from TMT investors who “said, ‘I don’t believe the data center thesis. That’s a zero,’” comped the rest at a 10x royalty-business multiple, and assumed Five Point would sell into any squeeze.
• Why TPL keeps its premium: Munger’s dictum (“don’t ever sell your royalty checks”), multi-decade family holders, and Horizon Kinetics’ ~15% stake — Garcia credits the late Murray Stahl as “the real godfather” of the whole space, with TPL a 20-bagger since his firm bought it in 2016–17.

4. The pore-space runway: 25% FCF compounding with no incremental capex

• The numbers Garcia leans on: 1.7M bbl/day currently injected and paying royalties; 5M+ bbl/day of incremental permitted pore space; another ~2.5M permittable — call it over 7.5M bbl/day total. Management and the chairman said “several times” they could “pretty easily” fill 5M bbl/day within five years at the current 15 cents a barrel, “which is all profit… not going to be any incremental expense to that.”
• That alone implies free cash flow compounding at a ~25% CAGR — “without any data centers… without any royalties for surface use or frac sand or any of that growth.” Andrew’s needle: “I’m laughing where you say without giving long-term guidance — they said this would generate $300 million of incremental free cash flow. And I think they mentioned that like 15 times.”
• Andrew’s valuation pushback remains: at 20–25x EBITDA “you need that [growth] just to justify that multiple.” Garcia’s response is that the announced water projects provide ample growth even without data centers, while the data-center thesis remains additional optionality.

5. Governance: Andrew’s hair stands up; Garcia mostly shrugs

• Andrew’s discomfort, laid out in full: Five Point did secondaries in 2025, he hasn’t seen aggressive insider buying or “aggressive PSU targets or stock-option grants,” 10% of the proxy details related-party transactions, and the structure — PowerBridge for data centers and, Andrew believes, fiber; WaterBridge for water; LandBridge as the royalty layer — “seems like either too much financial engineering or rife with conflicts of interest… it just makes the hair on the back of my neck stand up.”
• Garcia’s rebuttal: the LPs of the three vehicles are different funds with diverging monetization timelines; each entity has conflicts committees; and Devon’s stake in WaterBridge is a real check — “Devon is not going to want to see WaterBridge being taken advantage of for the sake of LandBridge.” He says he sees nothing untoward, while acknowledging the skeptical case that some WaterBridge economics flow to LandBridge.
• On insiders, a partial concession: management owns ~13% excluding the Five Point holdings, and Garcia thinks they’ve bought some shares recently (“I’ll have to check on that one”) — a hedge he leaves intact. His structural read: sponsors “see more than we do,” and LandBridge benefits from whatever’s coming “without having to put capital up.”

6. Active land management: the 150% uplift and the blocking-position playbook

• The metric — awkwardly named “surface use economic efficiency” — shows the 2024 acquisition vintage growing FCF 150% year-over-year. Andrew ties it to the old TPL bear case (“they just wouldn’t pick up the phone”): “I am 100% a believer that if you’re actively managing these things and calling the oil and gas companies, you can create a lot of value.”
• The playbook, as Garcia tells it: first, fill in TPL’s Loving County checkerboard so both sides get free crossing — because “every landowner’s land you crossed with infrastructure is an opportunity for somebody to take a toll” — unlocking New Mexico-to-Texas water flows. New Mexico has more water than it can handle, and E&Ps are reluctant to rely on its regulatory and disposal regime.
• Second, the 2024 deal locked up the pore space and access at the southeast corner of New Mexico just as the over-pressurized shallow-disposal area to the east will probably lose ~2M bbl/day of capacity in 3–5 years. A competitor pipeline, which Garcia tentatively identified as Engeol, probably handles ~300k bbl/day; once it reaches a likely agreed cap, competitors will have to come to LandBridge and negotiate.
• Third, the 1918 Ranch (October 2025): $375M for land with $20M of current EBITDA plus 900k bbl/day of contiguous incremental pore space. Garcia said the pore space was worth about $75M and described the purchase as 3.3x EBITDA; “really they’re the only ones that could do that,” since any competitor routing New Mexico water there has to cross LandBridge land. Andrew asks why TPL wouldn’t compete; Garcia: TPL has bought some pore space but lacks the infrastructure, and he doesn’t think it wanted the produced-water headache.

7. Data centers: no Permian hyperscaler yet, but the glimpses are stacking up

• Garcia is careful with the state of play: “there hasn’t been a large hyperscaler data center definitively announced in the Permian Basin yet” (Facebook is looking at Ector County land; its announced big one is El Paso). The nearest thing: Chevron’s mid-November analyst-day plan for a 2.5GW power-generation facility expandable to 5GW, $5–7B for phase one against $16B of LTM capex — with December tax-abatement filings and air-quality permits surfacing in Reeves County, on ~6,000 acres that include TPL land and abut LandBridge’s. Offtake is “being contemplated with Microsoft.”
• The believers are named: Eric Schmidt’s BOLT (TPL invested in December) with “a stated goal of bringing 10 gigawatts of data centers to the Permian Basin,” and LandBridge’s own September agreement with NRG for another power-generation facility. Garcia’s hedge stays: “before I’m going to say it’s going to happen… you need the offtake agreement with a hyperscaler.”
• Why it’s slow: the contracts are brutal — hyperscalers want 10–15 years and guaranteed uptime, power providers want 20–25, “the damages both ways could be immense.” The precedent that anchors his optimism: PowerBridge’s leadership includes a former Talen CEO/CFO who formed Cumulus around two Pennsylvania nuclear facilities when the state had zero data centers — no takers “until it was complete and then there was insatiable demand,” Amazon committed $20B, and with Blackstone/QTS the campus attracted ~$100B of capital. Crucially, the PowerBridge project is not expected to require LandBridge capital.

8. Fair value, the sibling choice, and the Secure/GFL endgame

• On valuation Garcia demurs, then relents: a low-single-digit FCF yield plus a 25–35% five-year FCF CAGR “isn’t scream[ing] too expensive,” the data centers “don’t need to hit for this to work,” and “triple digits is pretty reasonable — you start applying TPL multiples to this thing and you get to 150 pretty quick” (stock $67.58 at recording). Forced to pick between his children: LandBridge over WaterBridge — most FCF/share growth, the TPL re-rate gap, and the free data-center option, though WaterBridge gets its own re-rate if the municipal-waste thesis migrates to the Permian.
• Secure Energy — Garcia’s November 2024 thesis — was taken out by GFL at ~11x, versus ~15x for municipal-waste comps. His verdict is genuinely mixed: “a little disappointed with the price,” but the stock is +77% in a year, +465% over five, “probably a crowning achievement” for a management team that deliberately rebuilt the business (75% waste, 80% of that recurring; volumes held through a 20% Canadian rig-count drop) — and Waste Connections buying the tribunal-forced 20% divestiture had already “validated” the comp.
• The live disagreement: Abrams Capital crossed 10% — unusual for them — and publicly said it would vote against the deal. Andrew’s take-the-bid logic: management (2%) and a 20% holder are in support agreements, the vote may exclude locked shares, and with Andrew estimating “three strategic players and seven private equity firms” to call, “this was probably fully shopped… isn’t this just worth taking the bid?” Garcia’s counter: the thesis was mid-inflection — GICS code still reads energy, the gross-vs-net pass-through-oil reporting only changed in Q4 (“the company screened horribly”), his Raymond James room went from 12 people to 30–40 in a year, and sell-side notes fretting GFL is “not a pure play” ignore Waste Connections’ 6% E&P exposure. “There was a lot of education left to do.”