The simple version

Start with the plain version. If a few buyers are most of what a company sells, the company depends on them, and that dependence is a real risk to the business. If one buyer is a third of what you sell, your business is partly a bet on that buyer's plans.

Lose them and the hole is not a bad quarter. It is a third of the company.

Securities rules make a company disclose the size of that dependence. They do not make it name the buyer. So the filing hands you the exposure and leaves you guessing about who is on the other side.

The numbers

• Customer concentration, as Broadcom's own annual report states it: one customer accounted for 32%, 28% and 21% of net revenue in fiscal years 2025, 2024 and 2023 (Form 10-K for the fiscal year ended November 2, 2025)
• The report says revenue from that customer sat in the semiconductor solutions segment, and it does not name the customer (same filing)
• Sales to distributors accounted for 48% of net revenue in each of fiscal 2025 and fiscal 2024 (same filing)
• Aggregate sales to the top five end customers, through all channels, were approximately 40% of net revenue in each of fiscal 2025 and fiscal 2024, which the filing states as a belief rather than a measurement (same filing)
• Broadcom has filed no 8-K in October 2026; its most recent is accession 0001730168-26-000076, filed September 2, 2026 (EDGAR, checked October 8, 2026)

Why the rule stops short of the name

The disclosure exists so an investor can size a dependency. Naming the buyer would hand a competitor the commercial relationship, so the requirement covers the magnitude and not the identity.

That is a defensible trade and it has a consequence. You can know that a third of a company's sales rest on one decision-maker, and have no filed document telling you who that is.

Which means the size of the dependence is known and the counterparty is not. Those are different kinds of uncertainty and they do not cancel out.

The actual math on 32 and 40

Two figures in the same report tell two different stories. One buyer at 32% of revenue. The top five buyers together at about 40%.

Subtract and the other four are roughly eight points between them, or about two points each. The skew is the finding: this is not a company with five big customers, it is a company with one, plus four that round to noise beside it.

Treat that subtraction as ours, not the filing's. The 32% comes from the segment note and the 40% from the business section, the 40% is stated as approximate and as a belief, and the two counts use different channel definitions. The direction holds; the precision does not.

What this means

When a company borrows against a program built for one buyer, the lender's real exposure is that buyer's plans, not the borrower's engineering. A reported borrowing of more than $50 billion tied to one customer's chips is a bet on that customer continuing to want them.

The annual report is where you check the shape of that bet. It will tell you how big the dependence already is. It will not tell you whose.

What this is NOT

This article explains what customer concentration risk is and what a filing must and must not disclose, using Broadcom's annual report for the fiscal year ended November 2, 2025 as the example. It is not a recommendation to buy, sell, or hold Broadcom or any security, and it takes no view on the company's prospects or on whether its customer mix is too narrow. The reported borrowing of more than $50 billion is a Wall Street Journal report, not a company statement: Broadcom has filed no 8-K in October 2026, so the amount, the terms and the customer are unconfirmed by any filed document, and this article asserts none of them. It does not identify the 32% customer, because the filing does not. The subtraction producing roughly eight points for the other four customers is our arithmetic across two sections of one filing that use different channel definitions, and the filing does not state it. ClearMoneySchool uses AI tools from Anthropic, and Anthropic competes with OpenAI, which is named in the reporting behind this article. This is not a political endorsement or criticism of anyone.

Sources

• Broadcom Inc., Form 10-K for the fiscal year ended November 2, 2025, accession 0001730168-25-000121, filed December 18, 2025
• U.S. Securities and Exchange Commission, EDGAR filing history for Broadcom Inc., CIK 0001730168, checked October 8, 2026