Buffett Did Not Buy Houses
In May 2026, Buffett's Berkshire Hathaway acquired Taylor Morrison, America's sixth-largest homebuilder, for $8.5 billion in cash. Almost simultaneously, Japan's Sumitomo Forestry paid $4.5 billion in cash for Tri Pointe.
Content accounts everywhere spun this as proof that Buffett expects home prices to rise. That reading is simply wrong.
Here is the conclusion first: Buffett did not buy a view on next year's prices. He bought the scarcest thing in American housing over the next decade — housing production capacity.
I am a developer myself, dealing with land, entitlement, construction and delivery every day, so from where I sit this transaction is not a bet on prices at all. Below I break it down four ways: why long-term capital enters when housing data looks worst; what makes Taylor Morrison valuable; what off-balance-sheet assets are and how large builders control enormous land positions with little capital; and what this means for understanding BTS and BTR.
Why Long-Term Capital Buys When the Data Looks Worst
The U.S. new-home market does not look good right now. New home sales are down sharply year over year, new home inventory sits at a 15-year high, and the 30-year fixed mortgage rate remains above 6%.
A short-term investor sees pressure and losses. Long-term capital sees something else entirely: over the long run, America is short of housing.
According to Freddie Mac's 2024 data, the U.S. housing shortfall is 3.7 million units. DR Horton, the largest builder in the country, can produce roughly 80,000 homes a year at full throttle. A 3.7 million unit gap is 93 times DR Horton's annual capacity.
The tension in U.S. housing is not that nobody needs homes. It is that high rates have locked many people out. Demand has not disappeared; it has been suppressed and deferred.
That is why short-term and long-term capital reach opposite conclusions. Short-term money asks: how are orders this year, what happens to gross margin next quarter, will inventory keep building? Long-term capital asks: will America still be short of housing in ten years, and who can keep producing it?
If the answer is that the shortage persists, then the worse today's data looks, the better the entry point for long-term capital. This is Berkshire's signature move: buy assets whose long-term logic is intact while short-term sentiment is bleak.
Sumitomo Forestry's acquisition of Tri Pointe at the same moment shows this is not an isolated event. Why would Japanese capital buy an American homebuilder? Because Japan's aging population leaves little room for housing growth at home, while the U.S. — despite high rates and cold transaction volumes today — still has population growth, job creation, household formation and housing replacement demand over the long run.
In one sentence: ordinary buyers see the short-term pressure of high rates; long-term capital sees the long-term opportunity created by a housing shortage.
What Makes Taylor Morrison Valuable
Berkshire bought Taylor Morrison because it has land. More precisely, it bought not raw land but a lot pipeline.
What is a lot pipeline? A proven housing production line: land control, planning and design, zoning and entitlement, civil engineering, permitting, construction, sales, and then the attached services of mortgage, title and insurance.
Land by itself is not valuable. Developable land is valuable. What is most valuable is a land supply chain already validated by a mature system that can reliably turn dirt into homes.
Based on 2025 figures, Taylor Morrison is the sixth-largest U.S. homebuilder, delivering close to 13,000 homes a year with home closings revenue of roughly $7.8 billion. It operates in 12 states and 21 markets, many with population growth, job growth, or strong long-term housing demand.
More importantly, it controls more than 75,000 homesites. At the delivery pace of the past twelve months, that is roughly 6.2 years of land supply.
How many homes a builder can deliver depends not only on its crews, subcontractors and sales staff, but on whether it has land supply for the years ahead. Without land, even the strongest builder is stuck.
In American development, the hardest part is usually not building the house. It is everything before it: Can this parcel be built on? Is the density sufficient? Is there water, power and sewer? Will the traffic study pass? How is stormwater handled? Will the community plan be accepted by the market? How long is permitting? Will neighbors object? How much uncertainty is there in municipal review? Only once those questions are resolved does a parcel on a map become a financeable, sellable, deliverable project.
So Taylor Morrison's most valuable asset is not its current standing inventory. It is a complete capability for converting land into housing.
That capability compounds with Berkshire's existing housing chain. Berkshire already owns Clayton Homes, a manufactured housing builder, along with many housing-adjacent businesses: building materials, paint, brick, brokerage, mortgage, title and homeowners insurance. Taylor Morrison was not an impulse purchase. It is Berkshire taking one more step along the housing supply chain.
Ordinary buyers make money from rising prices. Good builders make money converting land into housing. The most sophisticated capital makes money controlling the housing supply chain.
Off-Balance-Sheet Assets: Controlling Big Land With Small Capital
This is the most important technical point of the piece.
Most people evaluating a homebuilder look only at what is on the balance sheet: how much land, inventory, cash and debt. Anyone who truly understands builders looks at something else — off-balance-sheet controlled lots.
On-balance-sheet assets are what the company genuinely owns. If a company spends $100 million cash on a parcel, that parcel enters the balance sheet as inventory or a land asset.
The problem is that land is an extremely heavy asset. If a builder purchased outright every parcel it intends to develop, enormous capital would be frozen: money is committed before development begins, interest and carrying costs run before a single home sells, and if the market turns, all the risk sits with the builder.
So large builders do not put all land directly on their own balance sheet. They make heavy use of land options, land banking, and special purpose entities.
A simple example. A parcel can eventually be developed into 500 homes, with a total land price of $100 million. The traditional approach: the builder buys it outright for $100 million. It owns the land fully — and has $100 million of capital locked up, potentially for three to five years of development cycle.
The smarter approach: the builder does not buy the land directly, but signs an agreement with the landowner, a land banker, or an SPE. The builder might put down a 10% deposit — $10 million — to lock in the right to purchase those lots in phases over time. The land itself may be held by the SPE or land banker and never enters the builder's balance sheet. The builder does not necessarily own the land yet, but it owns the right to acquire lots on agreed terms.
With $10 million, the builder can control $100 million of land supply.
This structure has three advantages. First, capital efficiency: the builder does not tie up cash in land and can deploy it into projects under development, under construction, or in sales. Second, risk flexibility: if the market deteriorates, the builder can decline to exercise, losing at most the option deposit rather than being trapped in the entire parcel. Third, better financial optics: because the land is not fully on the balance sheet, reported assets are lighter, inventory turnover can be higher, and ROE looks more efficient.
Taylor Morrison is a textbook example. As of the first quarter of 2026 it controlled more than 75,000 homesites, 51% of them off balance sheet. In other words, more than half of the lots it needs for the coming years are controlled through options, contracts and land banking rather than sitting on its own books.
That is why I say Berkshire did not buy a pile of land. It bought a land control system. The value of that system is not "how much land" but "how much future supply is controlled with how little capital."
The Profit Logic Is Shifting From Demand to Supply
Back to the question ordinary investors care about: what does this have to do with us?
The real lesson is that the way money is made in U.S. real estate is shifting from the demand side to the supply side.
For the past decade-plus, ordinary buyers made money on three things: low rates, asset inflation, and leverage. Over the next decade that gets harder. Rates may not return to 2% or 3%. Insurance costs are rising. Labor costs are rising. Material costs are rising. Municipal review keeps getting more complex. Land costs keep climbing.
So excess returns in real estate will increasingly come from the supply side. Who can source cheaper land? Who can navigate more complex permitting? Who can reduce construction cost? Who can shorten delivery timelines? Who can pivot flexibly between selling and renting? Whoever can do those things wins in a high-rate, high-cost, housing-short environment.
This is why I keep arguing that modular development is the future of American real estate.
Modular is not simply "cheaper construction." Its core is higher housing production efficiency. The biggest problem with conventional construction is that it is uncontrollable: weather, labor scheduling, subcontractor coordination, material waste, site management, schedule slippage — every one of those can raise cost and stretch timelines. Modular logic moves part of on-site construction into a factory, converting uncontrollable field labor into a controllable manufacturing process. Standardized design, bulk procurement, factory production, on-site assembly — the goal is shorter schedules, less labor dependency, less waste, and higher delivery certainty.
That is the same direction as the logic behind Berkshire buying Taylor Morrison: the central problem in future American housing is production efficiency. The modular development projects I have delivered in Seattle and Friday Harbor follow exactly this path, and the cost structure and schedule data are documented there.
BTS and BTR: Two Roads, One Underlying Capability
This transaction also reframes the relationship between Build to Sale and Build to Rent.
Whether you ultimately sell homes for capital gain or hold them for rental cash flow, the prerequisite is identical: you must be able to produce housing efficiently in the first place. What Berkshire bought is that prerequisite capability, not one particular exit strategy.
So if you invest in U.S. real estate, or you are a builder, agent or developer yourself, it is worth asking again: is your edge built on which way the market moves, or on your ability to produce housing? In a market with a long-term shortage and short-term high rates, the answer is getting clearer.
