Data center land prices are repricing parcels nobody wanted five years ago. What the buyers are actually buying, and what comes attached to it.
Ask a land broker what changed in pricing over the last few years and the conversation eventually arrives at data centers. It is worth understanding what those buyers are actually purchasing, because in most cases it is not the dirt.
What drives data center land prices: power, not land
A data center is a building that converts electricity into computation and heat. Enormous quantities of both. The limiting factor on where one can be built is almost never the site itself. It is whether the site can get power, how much of it, and how soon.
That makes two things the real subject of the transaction: proximity to transmission capacity, and a utility willing to commit to an interconnection timeline. A parcel that was agricultural or industrial-adjacent, with no particular development story attached, becomes valuable because of what runs past it rather than what sits on it.
Water for cooling, fiber routes, and a local jurisdiction that will actually approve the use round out the list. None of those are traditional land value drivers, which is why this bid surprises owners who thought they understood their own property.
For the owner, this is often a genuine repricing
An owner sitting on the right parcel may receive an unsolicited offer substantially above anything the local comparables would support. That is real, and it is not a trick.
It is also a tax event considerably larger than anything they had planned for, arriving without warning, frequently on land with a very low basis held for decades. The structure of the sale (and whether an exchange is contemplated at all) has to be settled before a contract is signed. After the fact, the options narrow sharply.
An unsolicited offer on land near transmission infrastructure is a tax conversation before it is a negotiation.
For the investor, the risk profile deserves more scrutiny than the enthusiasm allows
Capital is moving into this asset class quickly, and quick-moving capital tends to underwrite loosely. Five things deserve attention:
- Concentration. These are very large buildings with very few tenants, sometimes exactly one. The investment is substantially a credit position on that tenant, dressed as real estate. If the tenant’s business model changes, the real estate does not save you.
- Obsolescence. The equipment inside has a shorter useful life than the shell, and the power density and cooling specifications considered standard today were not standard five years ago. A building configured for one generation of hardware is not automatically suitable for the next.
- Development risk. Interconnection queues slip and utility commitments move. A project that pencils on a three-year timeline behaves very differently on a six-year one, and the carrying cost does not pause while you wait.
- Illiquidity. The buyer pool for a purpose-built facility is small, specialized, and gets smaller quickly if the sector cools. This is not an asset that trades its way out of trouble.
- Overbuild. A great deal of capacity is being built simultaneously, on demand forecasts that may or may not hold. Nobody in the middle of a boom enjoys discussing this one, which is itself a reason to discuss it.
A specialized asset, not a bad one
None of the above argues against the sector. It argues that the underwriting questions are different from the ones that apply to a warehouse or a shopping center. The relevant questions are about power availability, tenant credit, and construction timeline, not square footage and location in the ordinary sense.
Anyone putting capital in should be able to say clearly which of those they are underwriting, and what happens to the investment if that one assumption is wrong. If the answer is a general sense that the sector is growing, that is enthusiasm rather than analysis.
