Broadcom Wasn't Blocked for Being Foreign. It Was Blocked for Being Cheap.

On March 12, 2018, a president of the United States reached into a corporate fight that had no signed merger agreement, no agreed price, and no completed deal — and killed it anyway. The order cited 'credible evidence' that Broadcom 'might take action that threatens to impair the national security of the United States,' and for good measure it barred all fifteen of Broadcom's nominees from standing for Qualcomm's board.¹ No one had to wait for a transaction to close. The government stopped the deal at the proxy stage, before the wedding, while the couple were still arguing about the ring.

The official story is that a foreign company was blocked from buying an American one for security reasons. Almost every word of that framing points the wrong way. Broadcom was Singaporean, not Chinese — and was literally in the middle of moving to the United States. The thing the government feared wasn't who Broadcom was. It was what Broadcom intended to do with the budget once it owned the company.

The threat wasn't a spy. It was a spreadsheet.

Read the letter CFIUS filed publicly and the security argument is not about backdoors or stolen blueprints. It is about accounting. The committee flagged three things: Broadcom's 'private equity-style' approach — code for an acquirer that buys engineering-heavy companies and trims the R&D to lift margins — would cut the very research that keeps Qualcomm ahead in 5G; the $106 billion in debt behind the bid, the largest corporate acquisition loan on record at the time, would force short-term profit pressure to service it; and the deal could disrupt Qualcomm's supply to Defense Department programs.⁴ Strip away the legal varnish and the worry is plain. Slash Qualcomm's labs to pay down the loan, and the company that helps write the global 5G standard goes quiet. The vacuum has a name. Its name is Huawei.

That qualifier — 'this particular transaction only' — is the tell.⁸ CFIUS was at pains to say it had nothing against Broadcom the company or Broadcom the employer. The objection was the structure: a high-leverage, R&D-skeptical buyout of a strategic asset. The committee was not policing a nationality. It was policing a business model, and declaring, for the first time, that the level of a company's research spending could itself be a matter of national security.

The redomiciliation that came too late on purpose

Here is the move almost everyone misreads. The story goes that Broadcom could have saved the deal by finishing its relocation to America — become domestic, escape the foreign-acquirer review. CEO Hock Tan had even stood beside the president at the White House on November 2, 2017 to announce that very redomiciliation, days before the hostile Qualcomm bid surfaced publicly on November 6.⁷ It was choreographed to look like a homecoming. But the order was signed while the move was nearly complete, and it applied to Broadcom regardless of where its headquarters sat. CFIUS did not arrive too slowly. It arrived deliberately early, to stop the relocation from quietly erasing its own jurisdiction. The government wasn't outrun by paperwork. It pre-empted the paperwork.

And it moved through an unusual door. Normally CFIUS reviews a deal the parties bring to it. Here, Qualcomm — the target, fighting for its life — filed a unilateral CFIUS notice on January 29, 2018, asking the committee to examine Broadcom's proxy campaign to seat a majority of friendly directors.⁶ The defending company weaponized the security review against its acquirer. On March 4, CFIUS forced Qualcomm to delay its shareholder meeting by thirty days so it could investigate — and eight days later the presidential order made the delay moot.⁵ It was the first time in CFIUS history a transaction was blocked before any merger agreement was even signed.⁵

Wasn't this just protectionism in a security costume?

The fair objection is that this was industrial policy dressed up as defense — a government picking a domestic favorite, calling it security, and inventing a threat where there was only a cheap-money buyout. There is truth in that, and it should make anyone uneasy. CFIUS itself conceded the danger was indirect: Broadcom was an ally's company, not an adversary's, and 'private-equity-style R&D cuts' is a thesis about future strategy, not evidence of present harm.⁴ But the honest counter is that the security and the economics had genuinely fused. By 2018, 5G was not just a product cycle — it was the standard that would govern a decade of global networks, and whoever wrote it would hold a structural advantage. If you accept that 5G leadership is a national interest, then anything that predictably hollows out the leader's research is a national-security event, even when no spy is involved. The price was the proof: Broadcom's bid drifted from an equity value near $103 billion in November to a final, blocked figure of $117 billion²³ — a deal whose own arithmetic depended on the leverage that CFIUS said would force the cuts. The objection wasn't that Broadcom was foreign. It was that Broadcom had to economize on exactly the thing the United States had decided it could not afford to lose.

Broadcom dropped the bid on March 14, said it was 'disappointed' and would comply, and walked.⁸ It had done everything a foreign suitor is supposed to do to become acceptable — moved toward American soil, stood in the Oval Office, raised its offer. None of it mattered, because none of it answered the actual question. The government was never afraid of where Broadcom was incorporated. It was afraid of what Broadcom would have to cut to pay for the company. The deal didn't fail a loyalty test. It failed an arithmetic test — and the lesson is that in the industries a nation has decided it cannot lose, the most dangerous thing an acquirer can bring is a plan to spend less.