Warren Buffett Is Buying Homebuilders While the Housing Market Freezes. Here’s What He Sees.

In September 2026, the 30-year fixed mortgage rate climbed to 7.03%. It rose for the fifth week in a row. In July, new home sales fell 10.5% from the month before. Single-family housing starts dropped 15.7% from a year earlier. The median price of a new home fell to a five-year low. It was down 9.7% from its December 2025 peak.

On paper, the US housing market looks frozen.

During the same period, Berkshire Hathaway raised its stake in Lennar from 6% to 12% in two weeks. The position was worth about $2.2 billion. Berkshire became the second-largest shareholder in one of America’s largest homebuilders. In May 2026, it agreed to buy another builder, Taylor Morrison, for $6.8 billion. That was a 24% premium.

Why does a country that finished urbanization long ago still need so many homes? Why is an investor known for caution buying builders while the market looks broken?

The answer sits in two forces. One is a national shortage. The other is a regional population shift.

The Missing Homes

The US housing problem is not total saturation. It is a long period of too little building.

After the 2008 financial crisis, residential construction stalled for more than a decade. Many small and mid-sized builders went out of business. Industry capacity shrank. Housing starts stayed below historical averages for years. The White House Council of Economic Advisers estimated in 2026 that if construction had kept normal historical growth after 2008, the country would have at least 10 million more single-family homes today. Zillow puts the current gap at about 4.74 million units. Congressional estimates put it near 5.5 million.

Any of those numbers points to the same conclusion. Even at 1.5 million homes a year, the country would need more than a decade to close the gap.

Harvard’s Joint Center for Housing Studies made a similar point in its 2026 “State of the Nation’s Housing” report. New construction is slowly adding supply. Building activity is still weak. Single-family housing starts fell 7% from a year earlier. The shortage has several causes. Credit tightened after the crisis, and smaller builders left. Construction costs rose. Regulatory delays, outdated zoning, and lending barriers held back new supply.

The Migration Map

A national shortage is one problem. Population movement creates another kind of demand. It concentrates in specific metros.

In recent years, Americans have moved from blue states to red states. California, New York, and Illinois keep losing residents. Texas, Florida, North Carolina, and South Carolina keep gaining them. In April 2026, a policy roundtable at Harvard Kennedy School asked a blunt question: Americans have left blue states. Can blue states win them back?

California is the clearest case. The state has recently seen about 230,000 residents leave for other states each year. Los Angeles County lost 54,000 people in 2025, the largest decline of any US county. International migration used to offset those losses. In 2026, California’s net migration fell from 361,000 to about 109,000. That was a drop of about 70%. The state slipped into negative population growth for the first time.

New York is in a similar spot. In 2025, the state had a net outflow of about 216,000 people. New York City shrank for the first time in three years.

The destinations are concentrated. Texas gained 391,000 residents in 2025, the most in the country. Florida gained about 197,000. North Carolina gained about 146,000. The U-Haul migration index ranked Texas first for population growth in 2025, Florida second, and North Carolina third.

Those numbers turn into metro growth. Dallas-Fort Worth, Houston, Phoenix, and Atlanta have all seen steady inflows of tens of thousands of people a year. The northern suburbs of Dallas grew from a sparse suburban area into a super-cluster metro with more than a dozen sub-centers between 2016 and 2026.

The migration is not mainly an ideological contest. It is a reallocation of economic opportunity. IRS data shows Florida gained more than $20 billion in net income from new residents in recent years. New York and California lost billions. People moved. Income moved with them.

Short-Term Pain

In 2026, the housing market is also dealing with short-term pain.

High interest rates are the main constraint. The 30-year mortgage rate reached 7.03%, far above the 2% to 3% range of the pandemic years. That created a lock-in effect. Many homeowners with cheap mortgages do not want to sell and take on a new loan at a higher rate. Existing-home supply stays tight. Harvard JCHS data showed existing-home sales in early 2026 still near the 30-year low set in 2023.

Demand is weaker too. Household formation has slowed for three years. Job growth fell from 1.5 million in 2024 to just 116,000 in 2025. Consumer confidence dropped more than 20 points in 2025. In April 2026, it hit a record low.

Builders feel the pressure. At D.R. Horton, homebuilding revenue fell 3% from a year earlier in the first three quarters of fiscal 2026. Pretax profit fell 19%. Lennar’s adjusted earnings per share in the third quarter of fiscal 2026 were $1.23. The market expected $1.30. Revenue was about $8.05 billion. The market expected $8.31 billion. Lennar stock fell more than 18% in 2026. It had already fallen more than 23% in 2025.

Discounting is common. In July 2026, the median new-home sales price fell to $393,800, a five-year low. New-home prices also fell below the median existing-home price of $434,100. That inversion is rare in US housing history. It shows builders are willing to give up profit to clear inventory.

Long-term investors see something else in that short-term weakness. The national shortage has not gone away. Migration keeps creating demand in specific metros. Construction capacity is still slow to recover. Those forces support housing construction over the long run.

What Buffett Sees

Berkshire’s buying pattern shows how a long-term investor behaves in a panic.

In August 2023, Berkshire took positions in three homebuilders at the same time. Lennar was the smallest, at about 153,000 shares. Over the next two years, Berkshire kept adding. By June 2026, it held more than 13.1 million shares, about 6.23% of the company. In September 2026, Lennar stock kept falling on weak results. Berkshire accelerated. From September 17 to 21, it bought nearly 2.7 million Class A shares and 75,000 Class B shares. From September 23 to 25, it bought about 1.659 million Class A shares and 20,700 Class B shares. In October, buying continued. Within two weeks, the stake jumped from 6% to 12%.

That was not the only move. In May 2026, Berkshire agreed to buy Taylor Morrison for $6.8 billion, a 24% premium. Berkshire also owns Clayton Homes, the largest US manufactured-housing producer. It owns paint maker Benjamin Moore and building materials company Johns Manville. Together, those holdings cover materials, construction, financing, and delivery.

Buffett is not only betting on the industry. He is betting on a change inside Lennar.

Traditional homebuilders hold large land reserves. That ties up capital and exposes them to land-price swings. In March 2026, Lennar finished a shift to an asset-light model. It moved from land ownership and related debt on the balance sheet to land-option platforms and option agreements.

Lennar now controls about 98% of its homebuilding land through options instead of owning it. CEO Stuart Miller explained the logic in an investor presentation. Optioning land lowers the worst-case loss and makes that loss predictable. If land prices fall, Lennar can walk away from options. It loses the option deposit, not the full decline in land value.

The effect is visible. Lennar’s land holdings on the balance sheet fell from about $20 billion two years ago to about $10.5 billion. Less than 5% of its land sits on the balance sheet. The freed capital went into a larger optioned pipeline. That pipeline grew from 69,000 homes to 486,600 homes. Asset turnover improved. Return on equity reached 21%.

Lennar is also choosing volume over margin. Management says it will accept lower margins in the short term to gain market share and scale. The company is working through land bought at higher prices. It is competing with existing homes and expanding the asset-light model.

The logic is simple. In a downturn, the builder with scale and the lowest cost structure tends to win. When the market recovers, Lennar can expand deliveries faster than peers. Its optioned land pipeline gives it room to move.

Buffett’s bet comes down to three points.

First, the national shortage is long-term. Tens of millions of missing homes mean the country will need to build for years, even if demand swings quarter to quarter. Market pessimism gives long-term investors an entry point.

Second, the asset-light model lowers risk. Lennar can stay flexible in a downturn and expand quickly in a recovery.

Third, Berkshire owns pieces across the housing chain. It has a builder, a manufactured-housing producer, building materials, and financial services. It can capture value at different points in the market.

What China Can Learn

The US housing market is split between shortage and regional boom. That split offers lessons for China’s cities and developers.

First, migration creates real demand. Americans moved from blue states to red states because of opportunity and cost. People will move when the gap is large enough. Recent Chinese data shows population moving from first-tier cities to second-tier provincial capitals, separately listed cities, and strong regional centers. From 2020 to 2024, second-tier cities gained a net inflow of 8.23 million people. They were the only city tier with net resident population inflow.

Second, housing demand will diverge by region. The US market has growth in the South and Sun Belt while parts of the Northeast and Midwest shrink. China may see a similar pattern. Policy cannot rely on one national approach for very different regional markets.

Third, builder business models matter. Lennar’s asset-light shift offers a model. It shows how to control risk while keeping expansion capacity, how to clean up a balance sheet in a downturn, and how to improve operating efficiency through platforms and standardization. Chinese developers facing similar pressure can study those moves.

The Signal

The US housing market is not simply overbuilt. It is mismatched. A national shortage meets a regional migration. Demand concentrates in some metros. Supply stays short in many places.

Buffett’s contrarian bet is a long-term wager on that mismatch. While the market worries about high rates, weak confidence, and bad quarterly results, he sees missing homes, a migration reshaping the country, and a builder that changed its business model.

That is the value-investing idea in plain form. Be greedy when others are fearful, but know what you are being greedy about. Buffett’s answer is clear. He is betting on an industry where short-term pain creates long-term opportunity, and on a company that moved first.

For everyone else, the US housing story is not just about stocks. It is about migration, city growth and decline, and business-model change. Aggregate numbers can hide the real story. The opportunity often sits in the parts that averages do not show.