Latest / Energy Economics with Fexingo: Oil Prices, Renewables, and the Cost of Power / Why AI Data Centers Are Eating Rural Land Markets
Transcript
- Lucas: If you look at the headline from today about AI data centers transforming rural land markets, it sounds like a tech story, but it is actually an economics story about scarcity. We are seeing developers in states like Iowa and Kentucky bidding up acreage that has been in the same family for three generations, not to grow corn or soy, but to host servers that consume megawatts of power. Luna: And they are paying premiums that make no sense on a traditional agricultural yield basis, right? It is pure speculation on future power availability. Lucas: Exactly. And the kicker is that while these deals grab headlines, the underlying energy math is getting tighter. WTI crude is sitting at ninety-one dollars, which keeps baseline industrial energy costs elevated, and natural gas is flat at two dollars. But the real constraint isn't fuel price; it is physical grid capacity. You cannot just plug a hyperscale data center into a rural substation and expect it to work without a five-year upgrade cycle. Luna: So the land price reflects the bottleneck of the grid itself, not just the value of the dirt. Lucas: Precisely. When I dug into the recent SEC filings for some of the mid-sized utility providers in those regions, the capital expenditure budgets have shifted dramatically. They are allocating more toward transmission reinforcement than new generation, because the generation exists—it is often wind or solar—but the wires cannot move the electrons fast enough to the server farms. This creates a localized inflationary pressure on local economies that doesn't show up in the national CPI, which is currently at three hundred thirty-two. Luna: That is a fascinating disconnect. The national inflation number stays cool, but the local property tax base in these rural counties gets completely disrupted by commercial buyers. Lucas: Right. And it changes the political dynamic locally. Farmers are watching their neighbors sell out to tech firms for twenty thousand dollars an acre, which is triple the market rate for productive cropland in many areas. This isn't just about money; it is about the identity of the community. We saw similar friction with wind farms ten years ago, but this is different because the footprint is smaller but the power draw is massive. A single data center can consume as much electricity as a small city. Luna: Which means the utility has to decide: do we prioritize the residential grid or the industrial load? And usually, the answer is whatever brings the highest guaranteed revenue. Lucas: Usually. And that decision is being made behind closed doors in regulatory commissions. The Energy Secretary’s comments yesterday about Iran nuclear deals might sound unrelated, but if global oil supply tightens further, it puts upward pressure on diesel and gas, which makes operating backup generators for these data centers more expensive. That margin squeeze forces them to demand more firm, reliable power from the grid, which again, pushes against that capacity bottleneck I mentioned. Luna: So the geopolitical risk in the Middle East indirectly increases the cost of running AI models in rural America. Lucas: It does. It compounds the problem. When you combine high oil prices with a constrained grid, the cost of carbon-free power becomes less about environmental virtue and more about engineering reality. Nuclear is the obvious solution here, but nobody is building new plants quickly. So we are stuck with a patchwork of natural gas peaker plants and intermittent renewables, trying to serve loads that never sleep. Luna: And the land buyers know this. They are betting that the grid will be fixed before their leases expire. Lucas: They are betting that regulators will fast-track permits. But let’s look at the numbers. If a data center draws fifty megawatts, that is roughly the output of four large coal plants, all in one location. The existing rural substations were designed for agriculture and light commerce. Upgrading them requires transformers, switchgear, and miles of high-voltage line. That infrastructure takes time and money, and who pays for it? Often, it ends up being socialized across all ratepayers, including homeowners. Luna: That feels unfair. Why should suburban families pay for the grid upgrades that benefit a tech company's profit margin? Lucas: Because the utility companies argue that industrial customers provide stable, predictable revenue streams that stabilize the overall system. Without them, residential rates would be even higher due to volatility. But there is a growing backlash, as CNBC reported today, because communities feel they are losing control over their resources. The land is gone, the water usage for cooling is increasing, and the traffic from construction trucks is damaging local roads. Luna: It’s the classic extractive industry model, just with silicon instead of oil. Lucas: Spot on. And the financial returns for the land sellers are immediate, but the long-term tax base change is uncertain. Once the data center is built, it employs very few people compared to a modern farm. You lose the payroll taxes from farmworkers and gain corporate tax from the facility, but the net effect on local employment is often neutral or negative. So you have higher land values but less economic activity per square mile. Luna: Which makes the whole thing look like a balance sheet maneuver rather than genuine regional development. Lucas: Exactly. And this is where the market is starting to price in the risk. Look at XLE, the energy sector ETF, it’s barely moving, but individual utility stocks are seeing volatility as investors weigh these new liabilities. If a data center project gets delayed due to grid issues, the developer still owes the land lease fees, but they aren’t generating revenue yet. That creates cash flow stress for the builders. Luna: So the leverage is shifting back to the utilities. They hold the keys to the grid. Lucas: They do. And they are using that leverage to negotiate better terms. We are seeing contracts where the data center provider pays for the entire extension of the transmission line, not just the interconnection point. This shifts the capital burden off the utility and onto the tech firm, which improves the utility’s return on equity but raises the barrier to entry for smaller AI startups. Only the hyperscalers can afford this upfront cost. Luna: Which consolidates the market. Small players get priced out of rural expansion. Lucas: That is correct. It reinforces the oligopoly structure of the cloud computing market. Amazon, Microsoft, Google—they have the balance sheets to absorb these infrastructure costs. Everyone else has to rent space from them or find urban locations where grid congestion is even worse. This creates a moat around the big tech companies that is built on physical assets, not just software. Luna: Physical moats. I like that framing. It reminds me of why oil majors are so resilient; they own the pipelines. Lucas: Right. The parallel is strong. In the next five years, owning the power access might be worth more than owning the algorithm. Because if you can’t run the code, the code is worthless. We are seeing early signs of this in the bond markets too. Municipal bonds in counties hosting data centers are seeing lower yields because of the increased creditworthiness, but that stability comes at the cost of local autonomy. Luna: So the residents trade political voice for fiscal stability. Is that a fair exchange? Lucas: Depends on who you ask. The local government loves the tax revenue. The farmers love the sale price. The rest of the town wonders what happened to their quiet way of life. But economically, it is a rational transaction in the short term. The question is sustainability. Can these rural grids handle a second wave of expansion? Or will we see brownouts and blackouts that force the federal government to intervene? Luna: Intervention seems likely if consumer prices start rising due to unreliable power. Lucas: Absolutely. And remember, core PCE is already ticking up. If energy costs spike because of grid failures, that metric moves faster than anyone expects. The Fed watches that closely. So this isn’t just a local zoning issue; it is a macroeconomic variable that could influence interest rate policy if it escalates. Luna: I hadn’t connected the rural land deal to the Federal Reserve’s interest rate decisions. That’s a huge leap. Lucas: It feels like a leap until you trace the wire. Energy reliability drives inflation. Inflation drives monetary policy. Monetary policy drives asset prices. And asset prices drive land values. It is all one system. The reason we are talking about AI today is because it is the latest shock to that system, testing its limits in places we rarely look. Luna: So the real story isn't the technology; it's the fragility of the infrastructure supporting it. Lucas: Exactly. And once you see that fragility, you stop looking at tech stocks as purely software plays and start looking at them as utility-adjacent heavy industries. That changes your valuation models entirely. Luna: Do you think we will see more regulations coming out of Washington to manage this land use? Lucas: Probably. The Trump administration’s focus on deregulation might slow things down initially, but the sheer scale of the buildout will force federal standards on grid access and water usage. We are heading toward a new era of resource management where digital infrastructure competes directly with physical survival needs. Luna: Digital versus physical. That is going to be the defining tension of the decade. Lucas: It really is. And if you are investing or just trying to understand your local news, keep an eye on the utility filings. That is where the real drama is happening, long before it hits the evening news. Luna: Utility filings. Got it. I’ll be watching those instead of the tech press. Lucas: If walking through the economy with us has made something click, and you want to help keep these deep dives ad-free, honestly, if today was worth a coffee to you, that is the link — buy me a coffee dot com slash fexingo. It helps us cover the research time. Luna: Yeah, it makes a difference for the team, especially when we are digging into municipal bond data all week. Lucas: Thanks for that support. Now, back to the grid—what happens when the first major blackout hits a data center hub in the Midwest? Luna: That is the question hanging over every investor’s head right now. Lucas: One that we will have to watch closely in the coming quarters.