A short but interesting interview with David Capobianco, the managing partner of Five Point Energy. He discusses the business model of LandBridge, a publicly traded affiliate of Five Point Energy, which owns >270k surface acres of land in the Permian Basin in West Texas. He also talks about the re-rating opportunity given the particular suitability of West Texas land for data centers, which adds a new revenue source to acreage that was once used primarily for ranching or energy extraction.
For readers who are interested in learning more about the investment case for land and royalty companies, I highly recommend reading Murray Stahl’s letters as his Horizon Kinetics team has done a lot of work in this space.
Some of my notes from the interview are also included below the embedded video. These are for personal reference only and not meant as a comprehensive transcript; any mistakes are also my own.
On why the Delaware Basin is perfect for data centers: you have all of the elements necessary to build large, contiguous campuses for data centers. The two limiting factors for building out digital infrastructure are power access (with an attractive power price) and water. The surface in the Delaware Basin is uniquely configured with the lowest cost gas in North America. Very few grids can deliver the amount of power that is going to be necessary for the digital infrastructure of the future. In terms of water, the water production that comes along with frac wells in the Delaware is somewhere between 4-10x water versus the oil that comes out of a fracked well. In addition to all the produced water from the fracking resources, you also have the briny aquifers that can deliver ~1mn bpd of brackish water. By contrast, very few municipal locations or locations near population centers are going to be able to deliver enough water off their public resource to enable a data center to cool itself. You also have great connectivity for fiber that goes right through the Delaware Basin. The other factors are an attractive regulatory regime in Texas as well as carbon sequestration options.
On the optimal use of their land (e.g. energy extraction vs. digital infrastructure): their business model for surface acreage is the construction of infrastructure, which can be water, oil & gas, renewable, digital etc. These are not mutually exclusive, as they can function on the same location. You don’t need to worry about the oil & gas because you can configure your surface to have development corridors. That enables the energy participants to extract their oil & gas, send pipelines across, get the water they need and then deliver the oil & gas (and water) they produce. So LandBridge can set their land up to be able to handle the oil & gas but also to have contiguous big block areas for the power plants and data centers.
On the most effective renewable energy: the only way we will have 100% renewable-powered data centers will be with nuclear energy. That is the only renewable asset that runs with enough consistency and fail safes (with no down time) to be able to run a data center. If you look at one of the solar farms on their land, you can rely on it during hours of sun, but can’t rely on it when it is cloudy. So, for that 250MW, they expect an average of 25MW across a whole year. Nice to have, but not a must have. What you need for data centers is as source of baseload power. The only real solutions today are the locations where there are nuclear power plants and gas-fired power plants.
On LandBridge’s economics: they are strictly a surface play. They own a little bit of the mineral estate, meaning they get some royalties from production, but the lion’s share of their returns come from the surface. There are two things with the surface. One is the resources, e.g. caliche, which is a concrete-type substance used for construction, then the sand used for fracking, and also water (used for fracking, but also for data center cooling). Then you have the infrastructure piece. The first thing you do is to lease an area or location for development. The second piece is a sharing of the revenues generated from the land (e.g. the structure of their first agreement is a 5% share of the net power generated margin). So if you are trying to build out a data center with say 1GW of power, the spread between what you create it at and sell it for is for the owner/developer of the data center and power plant, but a 5% override goes to the them.
Importance of water management experience as a competitive advantage: as an example, a 1GW data center needs about 150-200k barrels a day of water necessary for cooling. The newer chips (e.g. Nvidia’s Blackwell chips) run hotter and hotter and get fit into smaller places, so cooling becomes increasingly important. A 1GW power plant requires about the same amount of water. So you effectively have a need for 350k barrels a day of water if you are going to build out a 1GW power plant and a 1GW data center. Five Point (LandBridge’s sponsor) manages approximately 4mn bpd of water throughout the whole of the Permian Basin. That is produced water. But they also have briny resources on their land that they currently exploit for fracking (send to companies to frack, who send them back produced water). This experience of managing that water in bi-directional pipeline flows is exactly what you need to manage the water around a power plant and data center.
On how long it takes to build and power a data center: if you already have the power ready to go, you have a connection to ERCOT, behind the meter power plant, you have a location etc. it will probably still take a couple of years to build a data center, and that is the fastest time. There are other options such as building out new power and creating a whole powered land for the data center. In the fastest possible scenario, you would sign a power purchase agreement with a hyperscaler, then immediately race to develop a power solution while they develop the data center, and you hope to have them meet at roughly the same time. There are other solutions which can be a little faster, if you are flexible based on how many MWs you want to bring on over what timeline.
On why Permian land prices will re-rate given a new revenue source from digital infrastructure: If you are in Langley, VA, Atlanta, GA, or these new targeted data center hubs, you are basically building around city centers or around neighborhoods. Nobody wants to see these giant data centers around them, or for their power grids and water tables to be drained by them. By contrast, take Loving County in West Texas. If it were a country by itself, it would be a top 10 or 15 oil producing nation. But it has a population of 130 people. There are massive swathes of land with very few people. That is unique because no one cares about seeing your data center, your water is not being taken from a population’s water sources and your power can be built and developed for you on your own site. West Texas is also very well configured for wind and solar, which will be a piece of the puzzle. But over the next decade, we will start talking about more nuclear reactors and SMRs that can be located right at your data center, built over a shorter time frame.
If you look at their most recent 46,000 acre acquisition (Wolf Bone Ranch), they purchased the land at US$~5k/acre. Their property trades at over US$~20k/acre in the public markets. In terms of their surface that they will put data center infrastructure on, the first 2,000 acres they have leased out at US$~4k/acre for 30-40 years. And then if you get a 5% power margin of a 1GW power plant, that could be US$~20-30mn. So you are talking about US$~10-16k/acre per year for very long-term contracts. As soon as the data center development happens, land values will re-rate completely.
Changing the paradigm around land: land is a unique resource. It doesn’t deplete and it has infinite optionality, particularly if you have it in scale. If you think about their ~270k acres of surface, their EBITDA for next year is projected to be US$~180mn. They can exploit 3-4x the amount of water, solar and wind infrastructure and still only get to 20-25% of their surface utilization. Historically, people didn’t understand why land was so attractive because the prevailing business model was like a mousetrap. You had ranchers who loved seeing their pristine ranch but running cattle didn’t pay so well, so you let pipelines through, and if there was a mineral estate you would lease drilling on your land etc. But it was a constant tension between rancher and industrialist. The industrialists paid the ranchers a lot but the ranchers couldn’t stand the industrialists. This caused development to really grind to a halt and they saw this happen as they developed their water business. What they recognized was they could change the whole paradigm around land. They could buy the land and be the best partner that exists for industrialists with respect to the infrastructure they want to build out. The other paradigm shift was having these operating businesses in water and data centers that can help drive the infrastructure build out on the surface, which also creates value.
His outlook for fossil fuels in the US: he thinks there has been a rapid evolution towards the recognition of 1) the growing demand for fossil fuels and 2) that fossil fuels can be exploited and utilized in an environmentally sensitive way. By doing so, you create an incredible strategic resource for the US and its allies (seeing what happened to Europe in getting its gas supply cut off from Russia was eye-opening). Under the Trump administration, with Chris Wright as the Energy Secretary, the dynamics are really supportive of the industry. We are now seeing this inflection point away from the narrative that demand has peaked and is going away. If you look at demand for crude oil over the last 40 years, it has risen every year for 40 years except for 3 recessions and 1 pandemic. Other than that, demand has increased at 1-1.5% every year, driven by population growth. The businesses involved in the energy ecosystem (e.g. in extraction, carbon sequestration, transportation, energy creation) are going to go from being an out of favor sector, which they have been over the last decade, to becoming a sector that has the only real secular growth outside of technology.