Will the US Take a Stake in TSMC Before 2027? Odds Fell to 14%
TSMC lost 14 percentage points in three days despite holding the only active JV structure that could satisfy resolution criteria.
TSMC's Odds Were Cut in Half in Three Days, and That Might Be a Massive Mistake
TSMC is the only company in the "Which companies will the US take a stake in before 2027?" prediction market that has a White House photo-op, a $165 billion U.S. investment pledge, and an active joint-venture structure with Intel that could directly satisfy the resolution criteria. And yet, in three days, the market slashed its implied probability from 28% to 14%, a 14-percentage-point collapse with no identifiable negative catalyst.
The drop is especially stark across platforms. Kalshi prices TSMC at 9%. Polymarket prices it at 18%. That 9-point spread between the two largest prediction platforms suggests deep disagreement about what the question even means. When a spread that wide persists, it usually signals definitional confusion rather than consensus repricing. In this case, the confusion may be fundamental: traders appear to be reading "TSMC investing in the US" and concluding that is not the same as "the US taking a stake in TSMC." They are right about the distinction. They may be wrong about where the deal flow actually points.
What 'The US Taking a Stake in TSMC' Actually Means, and Why Traders May Be Getting It Backwards
The market's resolution criterion is specific: the U.S. government must take an equity stake in a company before December 31, 2026. A subsidy is not a stake. A grant is not a stake. A tax credit is not a stake. The question demands ownership, and most of TSMC's headline-grabbing activity in the U.S. involves the company spending its own capital to build fabs in Arizona. That is TSMC investing in America, not America investing in TSMC.
This distinction is real, and it explains the surface logic of the selloff. But it misses the one deal structure already in motion that flips the direction of capital entirely. In April 2025, TSMC and Intel discussed a preliminary agreement for a joint venture in which TSMC would take a 20% stake in Intel's U.S. fabs. A JV of that scale, involving two companies already receiving CHIPS Act funding, creates a natural entry point for U.S. government equity participation. The U.S. government took equity stakes in General Motors during the 2009 bailout and in major airlines after September 11, 2001. In both cases, the mechanism was a restructuring that created a new entity or converted government support into ownership. The Intel-TSMC JV is precisely that kind of vehicle.
If the U.S. government converts any portion of its CHIPS Act support for Intel's foundry operations into equity in a joint entity that includes TSMC, the market resolves yes. That is not a hypothetical pathway. It is the logical structure of the deal already being discussed.
Inside the Deal Machine: How TSMC Became the Most Concrete Candidate in This Market
No other candidate in this market has TSMC's combination of political proximity, deal momentum, and structural fit. Start with the political layer. In March 2025, TSMC CEO C.C. Wei stood alongside President Trump at the White House to announce the original $100 billion commitment, which has since grown to $165 billion. That photo-op was not a courtesy call. It was a signal that TSMC sits at the center of U.S. industrial policy for semiconductors, a bipartisan priority that survived the transition from the Biden administration's CHIPS Act to the Trump administration's tariff-driven reshoring agenda.
Then consider the operational facts. TSMC's Arizona fab is already employing more than 3,000 people on 1,100 acres. The company plans three additional fabs, two advanced packaging facilities, and a major R&D center. TSMC holds 71% of the global foundry market, with Samsung a distant second at under 7%. Its client roster includes Apple at 25% of revenue, NVIDIA at 11%, Qualcomm at 8%, AMD at 7%, Broadcom at 7%, and Intel at 6%, according to 2025 revenue breakdowns.
The Intel JV talks are the critical piece. A joint venture in which TSMC operates Intel's U.S. fabs would create exactly the kind of hybrid entity where government equity participation becomes not just possible but strategically logical. Washington has every incentive to ensure that a TSMC-Intel foundry JV remains anchored to U.S. interests. An equity stake is the most direct mechanism to guarantee that.
The Bear Case: Why TSMC's Odds Might Deserve to Be This Low
The strongest argument against TSMC resolving this market is simple: the U.S. government has shown no public interest in taking an equity stake in a foreign-headquartered semiconductor company. TSMC is a Taiwanese corporation. Even with massive U.S. operations, the political optics of the U.S. government owning a piece of a Taiwanese firm are complicated, particularly given the cross-strait tensions that make TSMC's geopolitical position so sensitive in the first place.
There is a second, more structural objection. The CHIPS Act disbursements to date have taken the form of grants, loans, and tax credits, not equity. The Commerce Department's agreements with Intel, Samsung, and TSMC for CHIPS funding did not include equity conversion clauses in their publicly disclosed terms. For the market to resolve yes, the government would need to deviate from its established disbursement model, either by renegotiating existing agreements or by structuring the Intel-TSMC JV with a novel equity component.
Third, the Intel JV talks themselves are preliminary. "Discussed a preliminary agreement" is corporate-speak for early-stage conversations that may never close. Intel's own financial instability, with its foundry division posting billions in operating losses through 2025, adds execution risk. If Intel's foundry restructuring collapses or pivots in a different direction, the JV vehicle disappears, and with it the most plausible pathway to a U.S. stake in a TSMC-affiliated entity.
These are real risks. The market is not irrational for pricing them in. But a 14-point drop in three days, with no new negative information, suggests something other than careful risk assessment. It looks like a correction driven by traders who saw headlines about TSMC spending money in America and concluded the question had been answered in the wrong direction.
Resolution Path: Nine Months, One Question
The market resolves on December 31, 2026. Nine months remain. TSMC's stock trades at $338.79, reflecting steady confidence in the company's $122.4 billion annual revenue base and 30%-plus growth projections for 2026. The company is not in distress. It is not seeking a bailout. The bear case is correct that this makes a government equity stake less intuitive than in the GM or airline precedents.
But the Intel JV changes the calculus entirely. If that deal advances to a definitive agreement in the next nine months, the question of U.S. government equity participation in the resulting entity becomes a live policy discussion, not a hypothetical. At 14% implied probability, with a 9-point cross-platform spread between Kalshi and Polymarket, the market is pricing TSMC as a long-shot candidate. The evidence says it is the front-runner in a field where no other candidate has a comparable deal structure, political relationship, or operational footprint. Traders who sold from 28% to 14% may have answered a question that was never asked.