Hadrian’s $1.37B Raise: One Building Is 77% of It

On August 6, 2026, Hadrian said it raised $1.37 billion in new equity financing to accelerate the buildout of domestic manufacturing capacity for defense, aerospace and industrial systems. The release, datelined Torrance, California, says the raise values the company at $7.87 billion, and it does not say whether that figure is pre-money or post-money. Read as post-money, $1.37 billion against $7.87 billion computes to roughly 17 percent of the company, which is PSV arithmetic on the released numbers rather than a company statement. The round is co-led by WCM Investment Management, Washington Harbour Partners, Valor Equity Partners, 137 Ventures and Baillie Gifford. JPMorganChase’s Strategic Investment Group joined as anchor co-lead, investing through the bank’s Security and Resiliency Initiative, the $1.5 trillion, ten-year program JPMorganChase announced on October 13, 2025 that carries up to $10 billion of direct equity and venture capital across supply chain and advanced manufacturing, defense and aerospace, energy, and frontier technologies. 1789 Capital took major participation, and the release also lists Morgan Stanley Wealth Management, funds managed by Apollo, accounts advised by T. Rowe Price Associates, CapitalG, Andreessen Horowitz, Founders Fund, Lux Capital, Altimeter and Construct Capital.

Hadrian says the money funds new factories, expanded research and development, and additional production lines over the next year, including munitions and autonomous systems, under a model it calls Factories-as-a-Service. Founder and chief executive Chris Power said that “production is now the frontline of deterrence” and framed the raise as funding the technicians and engineers who will rebuild the industrial base. The company also says it is broadening access to technician equity. Those are Hadrian’s characterizations of its own raise. The release names no customer contract value, no revenue figure and no delivery volume, and nothing in the record attributes a dollar of the round to any specific program.

The real estate is in the numbers the company itself published, in two releases five months apart. The August release says that since its Series C twelve months earlier, Hadrian opened factories in Mesa, Arizona and Muscle Shoals, Alabama, bringing its footprint to just under 3 million square feet across four sites, two of them in Torrance. Its March 20, 2026 release opening the Alabama plant is more specific: four facilities totaling approximately 2.85 million square feet, of which the Alabama site alone is 2.2 million square feet. Do that subtraction and the other three sites together come to roughly 650,000 square feet, close to the approximately 600,000 square feet the boilerplate at the bottom of that same March release still ascribes to three facilities, the two figures differing by about 50,000 square feet on rounded numbers. Hadrian’s January 29, 2026 release puts Mesa at 290,000 square feet, which leaves the two Torrance sites somewhere around 310,000 to 360,000 square feet combined. On the company’s own published figures, 2.2 million of 2.85 million square feet computes to roughly 77 percent of Hadrian’s entire physical footprint on a single site.

Run the capital against the space and this tenant stops looking like industrial. Hadrian’s Alabama plant is a public-private partnership with the Navy that the company puts at more than $2.4 billion of total investment, combining more than $1.5 billion of private capital with $900 million of government funding through Navy appropriations, which computes to the public side carrying roughly 38 percent. Against a 2.2 million square foot site, $2.4 billion computes to roughly $1,090 per square foot. Mesa is the same shape at smaller scale: $200 million of investment against 290,000 square feet computes to roughly $690 per square foot. Nothing speculative gets built at those numbers, and neither figure is shell cost. Most of it is process: machine tools, robotics, power distribution, environmental systems, and the software layer Hadrian calls Opus. When a tenant’s improvements run several times the value of the box they sit in, that tenant is not moving at lease expiry. Put it in your renewal probability, your rollover assumptions and what you are willing to fund on day one.

The Alabama site is the more useful signal for anyone who owns older industrial. The March release says the Cherokee plant was formerly the largest railcar manufacturing facility in the country before production was outsourced to Mexico. That is the profile worth screening for: a stranded heavy-industrial building with the power service, the rail, the crane capacity and the floor loading already in it, in a rural county, reactivated by a venture-funded operator with a federal program behind it. Note also who holds it. At the ribbon cutting, Representative Robert Aderholt thanked AE Shoals for providing the facility that makes the project possible, which places Hadrian as an occupier at the site that is roughly three quarters of its footprint rather than the owner of it. Hadrian has not published a lease term, a rent or an ownership position at any of the four sites, so tenure at the other three is simply unknown from the record.

The labor math runs the same direction and it is counterintuitive. Cherokee is expected to create up to 1,000 jobs when fully operational, which computes to one job per 2,200 square feet, against roughly one per 830 square feet at Mesa. The bigger, more automated plant is the less labor-dense one per foot. Hadrian says Opus is engineered so that technicians with limited prior manufacturing experience reach full productivity within 30 days. That is a company claim and not a PSV finding, and no independent record tests it. But it is the claim that makes a rural county with no legacy machinist pool a viable site at all, and if it holds even loosely it widens the site-selection map to exactly the markets where obsolete industrial trades cheapest.

If Hadrian is a template rather than an exception, the question for an owner or an acquisitions desk is which buildings in your footprint fit that template before someone else prices them. That is a screening problem, and it is the kind PSV builds. The inputs are all records: county assessor and parcel files for industrial buildings above a size threshold and below a year-built threshold, utility interconnection queues and substation capacity filings for available service, rail-served flags from Class I and shortline network data, permit and listing records for clear height, crane capacity and floor loading, state and local incentive awards for what a jurisdiction has actually paid before, and county labor data for the trades inside a commute shed. The output is a ranked shortlist with a one-page evidence card per site, and every attribute on that card links to the record it came from with the date it was pulled.

The reviewer is the acquisitions lead, not the model. The approval gate is explicit and it has two locks. No site advances to a letter of intent until a person has confirmed the power position with the utility directly, because interconnection queue data goes stale faster than anything else on the card, and until a person has verified the ownership chain and current control, because the whole point of a stranded plant is that title, liens and control are usually messier than the assessor file suggests. Everything the system produces is a candidate and a citation. Nothing it produces is a decision. The gain is that a two-person desk can look seriously at every qualifying building across eleven states instead of the forty a broker sends over, which is the same pattern a screening layer produces everywhere: more looked at, not more approved.