When Reston, VA–based Stanley Martin Homes sold a 190-acre Devlin Tech Park tract in Prince William County to Amazon Data Services for $700 million on Oct. 31, 2025, the deal set off tremors that reached far beyond Northern Virginia.

The transaction — first reported by The Washington Business Journal — ranks as the second-highest price ever paid for a single-use, single-parcel of raw land on record in the United States.

It also stands as a vivid case study in how a counter-strategic, “land-heavy” homebuilder can out-maneuver even the most efficient “land-light” rivals when the value of expertise meets the momentum of an AI-powered business era.

A Contrarian in a Land-Light World

The industry has shifted toward land-light,” said Steve Alloy, president of Stanley Martin Homes and executive chairman of Daiwa House USA Holdings. “But it’s not really how we’re wired.”

While most national builders have spent the past decade purging their balance sheets of owned land — choosing to option rather than hold — Alloy and his team doubled down on precisely the kind of ground most of his peers would avoid.

If there’s a flat, 60-foot-wide by 100-foot-deep lot, they’ll be the highest bidder,” he said of public homebuilding organizations, including NVR, Stanley Martin’s next-door neighbor and perpetual rival. “So for 30 years, our strategy has been to pursue the sites that are harder to do off-balance-sheet, that require more complicated engineering or zoning. We invested in really skilled land development managers and engineering managers. It’s part of our DNA.”

That contrarian DNA just produced a once-in-a-career payoff.

Stanley Martin assembled the Devlin Road property for $51 million between 2021 and 2022, rezoned it for 516 single-family homes, then — recognizing that “cloud plus AI” had made data-center land exponentially more valuable — petitioned for a second rezoning.

After a two-year court fight and a community backlash that reached the Virginia Court of Appeals, Stanley Martin’s entitlement victory cleared the way for Amazon’s purchase, closing last month.

“Aha” Moment in the Age of AI

When we first had people approach us to buy or partner with us to change it to data centers,” Alloy recalled, “we literally said, ‘Wait—it was so hard to get our 516 single-family homes there. We’re not selling this. We need to build houses.’ We were wrong in that line of thinking. The cloud impact on the data-center industry, followed by the AI impact, caused the valuations to explode.”

The $700 million sale, roughly 1,272% above Stanley Martin’s original investment to assemble the larger 270-acre parcel, now joins an elite handful of U.S. land transactions topping half a billion dollars.

We could only find four others that were larger,” Alloy said. “Ours was second at $709 million.” (He and his team’s own online sleuthing, he added with a laugh, used “the same ChatGPT, Gemini, and Google searches” that The Builder’s Daily employs.)

The deal’s economics underscore the new gravitational pull of artificial-intelligence infrastructure. According to UBS, global spending on AI-related data centers is projected to surge from $375 billion in 2025 to $500 billion in 2026. Already, AI infrastructure spending has outpaced investment in traditional office construction.

The expectations of very high returns in this industry are trumping the high interest rates we’re facing today,” Raymond James chief economist Eugenio Alemán told The New York Times in August.

Across the U.S., hyperscale data-center campuses are remaking the development map. Related Companies’ $7 billion Oracle–OpenAI “Stargate” complex in Michigan, Microsoft’s $9.7 billion IREN partnership, and Amazon’s accelerating footprint in Northern Virginia have pushed the value of entitled industrial land into the stratosphere. CNBC recently described the phenomenon as “a gold rush for invisible real estate” — a market where parcels once penciled for housing now trade hands as energy-hungry digital factories.

When “Hairy” Land Becomes a Hidden Asset

Few builders are equipped to exploit that pivot. It demands not just timing, but a decade-long investment in capability.

In some of our counties, we’re well-known for being able to get a difficult zoning case to the finish line more easily,” Alloy said. “Lots of people can do it, but we’ve built deep entitlement, county-relationship, and engineering expertise in-house.”

That expertise, Alloy explained, grew from necessity.

When I took over as CEO in 1998, we were a small niche player. We asked ourselves: if we’re up against giants like NVR, how do we compete? The answer was to develop the skills they didn’t need—zoning, engineering, local relationships. Over the years, that’s how we scaled.”

In a market obsessed with capital efficiency, Stanley Martin’s approach might seem antiquated. Yet its success exposes an overlooked truth: land-light models can’t capture upside from residual land appreciation—especially when use-case volatility creates windfalls that only control, patience, and know-how can unlock.

Buying something for $51.5 million and selling it for $709 million,” Alloy said with a wry smile, “is not something we’ll see again. That’s a once-in-a-career backflip.”

Daiwa House’s Expanding U.S. Blueprint

Behind Stanley Martin’s contrarian play stands its parent company, Daiwa House Group, one of Japan’s largest public enterprises.

We’re over 650 group companies around the world now, with more than 75,000 people,” Alloy said. “If you think of Berkshire Hathaway as a conglomerate that owns tons of businesses, Daiwa House is that—but only in real estate, construction, and development.”

The scale is staggering. Daiwa’s global revenues exceed $36 billion (down from $50 billion due to currency shifts), spanning single-family and multifamily housing, commercial and industrial development, hotel and senior housing operations, and building-materials distribution. Alloy now serves as executive chairman of Daiwa House USA, overseeing its portfolio of U.S. operations—Stanley Martin in the East, Trumark in the West, CastleRock in Texas, and a growing stake in multifamily giant Alliance Residential.

Our goal,” Alloy said, “is to replicate what Daiwa House has created in Japan and elsewhere—to enter a variety of real-estate segments through strong U.S. operating businesses.”

The model: a network of self-sustaining “mini-Daiwa Houses” anchored in regional expertise, cultural alignment, and local leadership. Each platform—Stanley Martin, Trumark, CastleRock—acts as both homebuilder and acquirer, nesting smaller operators into a patient-capital ecosystem.

That structure has already begun to differentiate Daiwa’s U.S. presence from both domestic publics and private-equity-backed builders.

It’s kind of at a different size than everybody else,” Alloy noted. “And I do think, over time, it’s only a matter of time before Daiwa becomes one of the very top U.S. players.”

The Broader Market Signal

The Devlin sale may have been singular in scale, but it reveals a broader shift in value creation across American real estate. As the AI revolution consumes land, power, and water at an unprecedented pace, entitlements themselves have become a new form of currency. The builders best positioned to capitalize are those with the technical fluency—and risk tolerance—to navigate the rezoning maze.

At the same time, the sale underscores a paradox for U.S. housing. Every acre diverted to data centers is one less for housing in markets already short on supply. Prince William County’s rezoning battle encapsulates the tension between digital and physical infrastructure—between the need for data storage and the need for homes. The winners will be nimble operators, able to read both markets at once.

For Alloy, that dual literacy is now part of Stanley Martin’s competitive identity.

We pursue different sites than our national competitors do,” he said. “Our strategy gives us a longer horizon and more flexibility. And when something extraordinary comes along — like this — we’re ready.”

The Takeaway for Homebuilding Leaders

The lesson is not that every builder should start hoarding land in hopes of a billion-dollar data-center payday. Instead, it’s that capability that continues to compound. In an era defined by capital scarcity and technological upheaval, the builders who invest in zoning intelligence, engineering depth, and community trust may find new ways to monetize their expertise—even in places they never expected.

Stanley Martin’s deal with Amazon may mark the crest of one wave, but the strategic insight beneath it endures: the harder the land, the richer the learning curve — and the longer the return tail.

As the cloud and AI boom rewrite what land is worth, homebuilders who know how to make the impossible permissible will own the next frontier.