Welcome to Episode 130 of Special Situation Investing.
Before we jump into today’s write up, it’s worth acknowledging the fact that we’re covering another land-based royalty company. Because of this fact, it might appear we’ve moved from special situation investing to simply royalty investing, but I can assure you, we’re still looking for opportunity wherever we can find it. It just so happens that we’re finding a lot of value in the royalty space these days, Therefore, we continue to share our research with all of you, even if we’re covering the same ground more than once.
LandBridge Overview
With that disclaimer out of the way, let’s jump into today’s wave-top analysis of LandBridge Company (LB).
LandBridge owns about 220,000 acres of land in the Permian Basin acquired through the purchase of several ranches beginning with the acquisition of Hanging H Ranch in October of 2021. The company went public via an IPO on 28 June 2024, and as of this writing, had the following financials:
Market Cap - $2,430,000,000
Long Term Debt - $360,845,000
Revenue (2023 Pro Forma) - $92,902,000
Free Cash Flow (2023) - $50,259,000
Free Cash Flow Margin (Q1 2024) – 90%
To anyone who’s read our past material or listened to the podcast, they will immediately draw a parallel between Texas Pacific Land (TPL) and LandBridge. They would be correct to make the connection. Both companies are large land owners in the Permian Basin, both have royalty business models, and both profit from the energy industry. But the companies have significant differences as well.
For starters, LandBridge markets itself as an infrastructure development company and downplays its oil and gas business segment. In fact, LandBridge is largely focused on two specific infrastructure problems that its land is uniquely positioned to offer solutions to. The first issue relates to produced water processing along the Texas and New Mexico boarder and the second issue relates to digital infrastructure in the form of data centers.
Produced Water
For every barrel of oil pumped out of the ground in the Permian as many as four barrels of water are also produced and must be processed by oil and gas operators. Some of the produced water is used as sourced water in the vertical drilling of other wells and some of the water is recycled or filtered for other uses but most of the water is simply re-injected into the ground. In the state of Texas, this re-injected water has regulatory approval but in New Mexico regulations surrounding produced water disposal are much more restrictive.
Because the Permian sits underneath both states, any landowner with property along the New Mexico/Texas boarder that is also located inside the Permian Basin has a massive opportunity to profit from regulatory arbitrage. As it turns out, the Texas and New Mexico border is exactly where the majority of LandBridge’s acreage sits and the company, through its partner company WaterBridge, has already set up 980 miles of pipes to move 1.8 million barrels of produced water over the boarder every day.
LandBridge’s chairman David Capobianco (we mistakenly titled him CEO in the recording) put it this way in a June 2024 interview with Hart Energy that I highly recommend to our readers:
So right now in the Delaware, we're about 4:1 water/oil ratio in the wells. As those wells age, yes, you do have a slightly higher [percentage] of water.
But even more important than that, some of the newer zones—the deeper zones—have higher water ratios: 5:1 and 6:1 and even 7:1.
So yes, I would anticipate that the Delaware Basin’s water/oil ratio will rise over time.
But the long-term opportunity is that New Mexico alone will grow its water production between 750,000 and 1 bbl/d per year. That means over the next three to five years, there'll be 3 to 5 million more bbl/d of water produced in New Mexico.
Some of that will be recycled. The lion's share of that water will move to Texas, south and east.
And because of our contiguous location across the state line, it will likely either come to our ranch for management or come through to our ranch. That 3 [MMbbl/d] to 5 MMbbl/d creates a $200-300 million annual free cash flow opportunity for the ranch.
Data Centers
The second issue that LandBridge is well positioned to solve focuses on digital infrastructure, specifically data centers. We’re not experts in digital infrastructure but based on the research we have done, the situation can be summed up by reviewing the requirements for construction of both a traditional and an AI data center.
Imagine you’ve been tasked to select the construction site for a new data center. What are the top items on your checklist, the things you absolutely must have in order to begin construction and sustain operations? The first, and likely most important, item on your checklist is power. In fact, data center power consumption is projected to reach mid single digit demand of all power produced in the world in the coming decade.
In data-center-saturated markets like Langley VA, where more than 60% of global data centers are located, utilities companies like Dominion Energy are unable to provide any additional power to data centers because they don’t have any more to give. Massive data center demand recently drove Amazon to purchase a nuclear powered 960-megawatt campus from Talen Energy for $650 million, which again speaks to the current demand for both energy and facilities. Given today’s data center power requirements coupled with the massive demand for more capacity in the future, your first requirement for a new facility would be power.
Next, you need cooling to ensure that all of those warehoused computers don’t overheat and shut down. This item is linked to power in that many facilities cool the data centers with standard air conditioning. But increasingly both crypto miners and data centers are moving to direct-liquid cooling of the equipment. In light of this fact, you would look for a location with lots of power and plenty of water so you have the option to cool the facility using either standard air conditioning, direct-liquid cooling, or both.
Furthermore, you would want access to fiber optic cables so you could network your data center to the internet. Interestingly, much of the United States fiber optic network is collocated with railroad tracks. This is because laying down a fiber optic network in the 1990s across a country with countless distinct land owners would have been an approval nightmare and probably could not have been accomplished. But obtaining an easement from a single landowner, a railroad, who’s property already traversed the nation was relatively straight forward and profitable for both parties.
Finally you would want a sight with plenty of space to build the data center and relatively few people in the surrounding area to complain about noise and all of the other issues that go along with large-scale infrastructure projects. Not in my back yard or NIMBY issues are already being raised by concerned citizens in areas with high data center saturation like Langley VA.
AI data centers share the same basic requirements as traditional data centers but with one meaningful difference. AI data centers require vast amounts of raw data to be collocated so that it can be accessed and processed quickly by whatever program the data center is running. As fast as data can be transmitted today there are still limitations on how much can be moved how fast.
With traditional data centers, great efforts were made to minimize latency. In other words, you wanted your data center as close to the end user as possible so when a Facebook user pulled up their favorite cat picture, the picture could be accessed with minimal delay. When, on the other hand, an AI data center is tasked to generate a picture of a cat having a birthday party, the AI program might access all of the stored information it has relating to cats and birthday parties before generating a response. Accessing all cat and all birthday party related information instead of one specific cat picture takes a very long time if the information is stored in thousands of different locations across the globe instead of a single collocated facility. So with an AI data center you want a very large facility with lots of collocated information so that the program can generate responses internally first and then send them to the end user. For this reason, AI data centers will favor the availability of cheap and abundant energy, water for cooling and remote locations above proximity to end users.
If cheap and abundant energy, water for cooling, access to fiber optic cables, and remote land are highly prized by data center operators then the case for building them on LandBridge acreage should be obvious. In fact the case could not have been articulated better than by Mr. Capobianco himself in the same Hart Energy interview cited previously. He stated:
DC: And then the last piece, which will be an important part of the future also, is data-center development: The digital infrastructure that will be developed in West Texas on contiguous pieces of land.
ND: To power that, are y'all talking to any small modular [nuclear] reactor companies like Oklo [Inc.]?
DC: Small modular reactors are the future of data centers, [but] we look at that as a decade or so out. The need is today; it is imminent. We need to build five to seven times as much data-center capacity than has been built in our history.
And to do that, we're going to need to focus on places with low-cost fuel. In the Permian, where you have $0 to negative-$7 gas.
It’s pretty attractive. And you have an incredible regulatory environment.
We, because of our sister company [WaterBridge], have unlimited water for cooling.
We have very good fiber. Not the kind of fiber hub you have around Langley, Virginia, where 61% of worldwide data centers are. But we have very good fiber through our surface. And over time that will improve.
Over time, the connectivity between the data centers developed in the Permian Basin will also turn it into a fiber hub.
ND: Are y'all already talking to some data-center developers?
DC: We are. At Five Point, we’ve created a partnership with a developer with the goal of building out a 1-gigawatt data center on our surface.
To add to that, building out a 1-gigawatt data center will be driven by a 250-megawatt solar farm that we have on the same contiguous acreage and a gigawatt power plant.
ND: Data centers seem very appropriate for West Texas for lots of reasons, but among them they don’t require many people on the ground to support operations versus, say, a chip manufacturer.
DC: It's more of a management/monitoring program.
But you appropriately recognize that there aren’t many people who live out there. But we also expect that, over time, all the development that I'm talking about will develop West Texas as well. And I'm talking about modular condominiums, restaurants, places for people to live—more than just the man-camp program that we have today.
ND: Eagle Ford, Barnett and Haynesville gas producers will be glad to see more Permian gas get used in-basin and not sent to the Gulf Coast. How quickly might the data centers come online?
DC: Construction would begin in two years. Operations could be possible in four years. We've signed a 100-year lease agreement.
ND: Can you reveal who is the data center operator?
DC: I'll be making a press release about that in the next few months.
ND: Some use for that gas will be appreciated out there.
DC: Well, it's stranded and more than that, [using it in-basin] will massively change the economics of the Permian.
To put it into context, a gigawatt data center may take about 300 MMcf/d. We have six locations for data centers. You're talking about almost 2 Bcf of in-basin gas need. It would dramatically change the dynamics of prices at Waha.
In a separate article the Houston Business Journal stated that for LandBridge “a single gigawatt data center could literally be $30 million to $40 million [in revenue] annually for 100 years.” LandBridge already has locations for six data centers and plans to begin construction on the first project within two years.
Valuation
There are many ways to model the future growth and valuation of LandBridge but for the sake of simplicity let’s just compare the revenue projections offered by Mr. Capobianco to Texas Pacific Land’s current revenue and market cap. Among his estimates for the next three to five years he predicted the following: revenue from water royalties between 200 and 300 million per year, revenue per data center between 30 and 40 million per year, along with continuing, but less significant, revenue from oil & gas, sand, and other infrastructure projects. If we assume 250 million in revenue from water, 30 million from a single data center and another 70 million from oil & gas and other infrastructure projects we come to 350 million in total revenue.
Texas Pacific Land generated just over 600 million in revenue for the last several years and commands an impressive 20 billion market cap, This suggests that with revenue of 350 million per year, LandBridge could easily justify half of that market cap or 10 billion dollars. A 10 billion dollar valuation is four times higher than today’s market cap and is based on a lower P/E multiple than TPL’s as well as a low end estimate of what LandBridge could produce if Mr. Capobianco’s estimates are fully realized. Additionally, the margin of safety with a company like Landbridge is significant given that operating costs are minimal, the debt is manageable, and the land underlying the company could always be sold if needed. With all of these considerations taken into account, a bet on LandBridge hardly seems like a speculative undertaking.
Summary
If this high level overview of LandBridge peaks your interest in the company, then I would highly recommend that you read the company’s full S-1. The document is much more informative than other similar companies S-1 filings and it goes into great detail about the companies operations. I would also recommend that you research water production in the Permian basin along with the regulatory issues surrounding water disposal in New Mexico to better understand the company’s unique ability to serve that market. Finally, it’s worth reading up on the future of data center development. The data center market is rapidly growing and it’s worth understanding the industry dynamics even if you decide not to invest in the space.
LandBridge has incredible potential but it will take time to fully develop and for that reason I’m dollar cost averaging into the position over time. This is not a spin-off or restructuring that will double or triple in the near term based on a specific catalyst but rather a great company at a fair price with a very long runway ahead of it. Again, just reading more about the company and the industries surrounding it will be a great benefit to you even if you decide not to invest and I highly recommend you spend some time studying LandBridge on your own.
With that, we’ve wrapped up another episode of Special Situation investing. As always, we thank you so much for your support and your interest. We’ll be back again in two-week’s time with another write-up for ya.
LandBridge Company (LB)