Democrats Took the Senate, and SpaceX's $10 Billion Federal Contract Book Just Became a Re-Competition Risk — the Market Priced It at 7%

Democrats clinched the Senate on the night of Nov. 3, 2026, and won the House, ending a Republican era in Washington that had been unusually good for one family's federal revenue. The market's opening verdict on SpaceX came fast: about 1.8% lower in postmarket trading, down to $136.42 overnight, briefly4.15% lower at $135.45, and then 6.9% lower in premarket trading on Nov. 4, with Tesla down about 5% on the same print. In other words, the market priced a new Congress at roughly 7% of SpaceX's market value and moved on.

Seven percent looks decisive until you price what Democrats actually control. They cannot cancel a contract, and the president retains the agencies that sign them. They can subpoena award files, rewrite the rules that decide who may bid, bar the money that funds political campaigns by contractors, and shape the appropriations lines that renew the work. Every one of those channelsacts on a procurement cycle rather than on a trading session, which is why the tape move is the smallest part of the story.

Alexandra Merz of L&F Investor Services put the limit plainly when she noted that Congress cannot simply veto a private merger. What a Democratic Congress can do is hold hearings, seek documents and investigate the agencies that award contracts. The distinction matters: the merger's compensation structure is a private-ordering question settled by two special committees, the SEC and Tesla shareholders, while the contracts that underpin the valuation are a public-ordering question that Congress can reopen.

The sharpest weapon in the new majority's kit is not a hearing about Musk. It is a hearing about award justification. On Oct. 19, 2026, four Senate Foreign Affairs Committee Democrats — Jeanne Shaheen, Richard Blumenthal, Sheldon Whitehouse and Chris Coons — wrote to SpaceX vice president of legal Christopher Cardaci, demanding answers on the restrictions that hit Ukrainian Starshield access between Oct. 10 and 15, 2026, including devices moving above 100 mph, and asking directly whether the awards were single-source or competed. Blumenthal then opened a formal inquiry, and the Pentagon inspector general declined to take it up, calling it a contractual matter — which leaves the senators' inquiry as the surviving channel.

That thread is not isolated. On Oct. 23, 2026, a bipartisan, bicameral group of House members urged the Pentagon to restore and expand Starshield access for Ukraine. And on Nov. 2, 2026, the day the senators' response deadline arrived, a new bill landed that would make the money link between contractors and candidates a legal problem rather than a scandal. Every one of these is an inquiry into how a contract was awarded, not into whether it should exist.

On Feb. 2, 2026 SpaceX closed its acquisition of xAI, and the consolidated numbers show what that did. The Super Intelligence segment went from $2,561 million in the second quarter to $7,560 million in the third, total revenue reached $14,575 million, and contracted backlog stood at $62,959 million with 67% scheduled to convert within a year. At the Goldman Sachs Communacopia conference in September, chief financial officer Bret Johnson put annualized compute income at $41.1 billion and signed a further hosting deal worth $13.3 billion a year beginning Dec. 1, on top of existing commitments of $1.25 billion a month for Anthropic, $920 million a month for Google Cloud and $150 million a month for Reflection AI.

The politically exposed line inside that income statement is smaller than the headlines suggest and more fragile than it looks. The 10-Q splits revenue into consumer and enterprise-and-government buckets: in the third quarter, enterprise and government revenue was $2,127 million against $3,121 million of consumer revenue, and over the first nine months of 2026 it was $5,042 million, or 18.6% of revenue, and it includes commercial Starlink Mobile service. So the directly federal slice is a minority of a business whose growth engine is AI compute sold to private customers. The exposure is not the contract book; it is the permission structure behind the compute build-out.

Here is the causal chain that turns an election into a contract problem. Concentration bought SpaceX speed in procurement: one owner, one board, a demonstrated ability to deliver at scale, and the political cover to win work without a long competitive fight. That same concentration is now the obvious target of a conflict-of-interest inquiry, and single-source awards are the most vulnerable kind of award to reopen. If the Starshield file is re-examined and a sole-source justification does not survive, the moat narrows at exactly the moment the compute business needs a cooperative federal government. The merger is how Musk monetizes the whole structure, and it needs 50% of all Tesla shareholders.

The timing is not neutral for Tesla holders. Third-quarter operating income fell to $718 million from $2.04 billion a year earlier on flat revenue of $28.05 billion, a 65% decline, which makes the $2 trillion market-cap milestone — and the 20 billion shares issued behind it — the load-bearing element of the pay package. If the political cloud keeps the merger in special committees for another year, the likely outcome is not cancellation but renegotiation: Reuters reported that experts see SpaceX's board able to amend Musk's plan substantially after a transaction, and the change-of-control language in Tesla's plan already allows market-cap-based payouts without an operating milestone.

None of this is a solvency story, and that is precisely why it becomes a multiple story. SpaceX ended the third quarter with $89.9 billion of cash and short-term investments against $38.8 billion of debt, a net cash position of $12.5 billion, and roughly two and a half years of runway at the trailing burn rate of $36.9 billion. A profitable incumbent can absorb a lost contract in a quarter. A company funding a data-center build-out at 1.7 times revenue in capital expenditureabsorbs it in the multiple, and the multiple is where a 7% gap opens.

At 59.5 times sales and 138 times EBITDA, SpaceX is priced on a future that assumes the federal government keeps behaving as it did through 2025 and 2026. Morgan Stanley's published estimates have revenue reaching $175 billion in 2027 and $289 billion in 2028, operating income of $119 billion in 2028 and free cash flow turning positive in 2031; on fiscal 2028 earnings the same note has the shares at 17.6 times against 17.7 times for the S&P 500. A valuation that looks reasonable on 2028 numbers is a bet that the political cycle stays friendly for two more election cycles — precisely the assumption a Senate takeover invalidates.

The asymmetry is the point. The political risk is created in one place — the federal award process — and consumed almost entirely by one company: the vendor whose owner the Senate is now investigating. Suppliers upstream are paid for consumer terminals, turbines and spectrum no matter who wins a re-compete; customers downstream are large enough to absorb a price change. The names with real convexity are the substitutes. Iridium earns $903 million of trailing revenue at a 75.9% gross margin and an 18.1% net margin, with $358 million of free cash flow and a 1.34% dividend yield, at 13.5 times enterprise value to EBITDA. Rocket Lab has $769 million of trailing revenue at a 37.3% gross margin but a $223 million operating loss and capital spending at 19.3% of revenue. One is a cash-generating incumbent waiting for a procurement decision; the other is a funded challenger with no cushion.

Put the pieces together and the thesis is not that Democrats will cancel SpaceX's contracts; the president still controls the agencies that sign them. The thesis is that the ownership structure which made those contracts exclusive is now the most expensive thing the company owns. Concentration bought speed in procurement; the same concentration is now the fastest route to a conflict-of-interest probe, a sole-source challenge and a re-compete, and the compute engine that replaced launch as the profit center was assembled on the assumption that the federal government would keep behaving as it did in 2025 and 2026. With the Senate flipped, that assumption now has a committee attached to it, and the 7% move is too small a discount for a franchise whose exclusivity is the asset.