The U.S. government’s bet on Intel is paying off, at least on paper. Washington now holds roughly a 10% stake in the chip maker through its CHIPS Act investment, and that position has gained nearly $30 billion in value, according to an analysis published by The Wall Street Journal.
The government bought into Intel when the company’s shares were depressed and its foundry business was burning cash, a purchase that was politically controversial at the time. The stake, built up over the past year through a series of share purchases, is now worth well above its cost, the Journal found. For a government industrial policy that has spent most of its existence defending itself against critics, the paper gain is a visible validation.
The numbers tell part of the story. Washington’s roughly 10% ownership makes the federal government one of Intel’s largest shareholders, and the nearly $30 billion paper gain ranks among the biggest returns on any government investment in a private company in recent memory. The original purchase, announced in the spring of last year, was itself unprecedented: the government had never taken a direct equity position in a major chip maker before, and the mechanics of the deal, including how the stake would be managed and sold, were worked out as the program went along.
Barron’s, which has tracked the position, has noted that the investment also carries a geopolitical logic. Intel’s advanced manufacturing coming online in the United States reduces the concentration of leading-edge chip production in Taiwan, the central goal of the CHIPS Act, which set aside $52.7 billion to revive domestic semiconductor manufacturing. Each percentage point of advanced production moved onshore is, in Washington’s framing, a point of risk reduced.
The strategic arithmetic is straightforward. Taiwan produces most of the world’s most advanced chips, and U.S. policymakers have spent years worrying about what happens if a conflict interrupts that supply. Intel is the only American company capable of manufacturing leading-edge chips at scale, and its fabs in Arizona and Ohio are the main vehicles for the CHIPS Act’s ambition. The company’s 18A process, its bet to close the gap with TSMC, is central to that plan, and Intel has said it expects the node to carry its foundry business into its next phase.
The harder question is the one the Journal’s analysis leaves open: whether Intel can match TSMC’s yields on advanced nodes by 2027. Intel’s foundry unit continues to post heavy operating losses as it ramps production and courts external customers, and yields on leading-edge processes are the industry’s hardest technical problem. TSMC has spent decades building its advantage, and its customers have shown little patience for second-best manufacturing. Analysts said the government’s equity stake changes the incentive structure: Washington now has a direct financial interest in Intel’s manufacturing execution, not just a policy interest.
Skeptics point out that paper gains can reverse. The government’s stake is subject to the same swings that move any semiconductor stock, and Intel’s shares have been volatile as it works through a turnaround that has included thousands of job cuts. A government as shareholder also creates awkward dynamics, from questions about how the Treasury votes its shares to the optics of Washington profiting from a company that has shed workers. The real test, analysts said, is commercial: whether Intel wins outside customers for 18A and whether those customers can match the yields and pricing TSMC offers.
The politics of the CHIPS Act have shifted since it passed in 2022. Some lawmakers have called for clawing back unspent funds, and the program’s administration has been reorganized under the current White House. A visible gain on the Intel stake gives supporters of the program ammunition in those fights, though defenders caution that one profitable investment does not prove the whole policy worked.
The comparison that matters is not with past government bailouts but with the private market. TSMC has spent more than $100 billion building leading-edge capacity in Taiwan, Arizona and Japan, and its market value reflects the strength of that franchise. Intel’s challenge is not just to match a process but to match a supply chain, a customer base and a reputation built over four decades. Analysts said the government’s stake gives Intel credibility with customers who worried the foundry effort might be abandoned, which could prove more valuable than the paper gain itself.
Intel’s own executives have said the foundry business is on track, and the company has signed agreements with a small number of external customers for 18A, though it has disclosed few details about volume or pricing. The coming quarters will show whether those relationships deepen into the kind of orders that make a foundry profitable.
For now, the arithmetic favors Washington: nearly $30 billion richer and a little less dependent on Taiwan. Whether that holds depends on a yield curve in Arizona and Ohio that no government spreadsheet can control.
