A capex headline is not bond supply.资本开支（capex）的头条，不等于债券供给。

Next week packs a Fed decision and the biggest AI spenders into about 48 hours, and the preview coverage is already telling you where to look: all eyes on capex guidance. For the question most people are really asking, is AI borrowing what pushed long-term interest rates up, that is the wrong gauge.

Here is the calendar. The Federal Reserve's rate-setting committee meets Oct 27-28, with the decision on the 28th. Alphabet has confirmed it reports third-quarter results after the close on Oct 28. Earnings calendars put Microsoft and Meta the same afternoon and Amazon on the 29th. So the Fed's statement and four mega-cap capex updates arrive within a day of each other.

A long Treasury yield has two parts: the path of short-term rates the Fed is expected to set, and a term premium, the extra yield investors demand to hold a long bond. Capex guidance, how much a company plans to spend on chips and data centers, speaks to the equity story: whether the AI bet is paying off. It does not, by itself, add bonds to the market. Only new long-dated debt issuance competes with Treasuries for capital and moves the term premium. Spending funded from a company's own cash flow does not.

So two different numbers answer two different questions. If long yields jump on Oct 28, check what moved them: a hawkish or dovish Fed statement, the path, or fresh issuance plans, the premium. A capex beat with no new bonds is an equity event wearing a macro costume.

The strongest counter: PIMCO's Marc Seidner and Pramol Dhawan wrote in August that AI-linked issuance, much of it longer-dated, has grown large enough to compete with government bonds for capital, even if it is not the dominant driver of yields. If capex guidance comes bundled with billions in new bond sales, the two can move together, and over months the supply channel could matter more.

Capex tells you about the AI bet; only issuance tells you who it is crowding out of the bond market.

If you catch yourself reacting to the Oct 28 capex headlines as the reason rates moved, run one test first: separate the Fed statement from any issuance news before you credit the spending number.

We'd revise if the 10-year term premium rises more than about 5 basis points over Oct 27-31 on a capex-driven day with no Fed surprise and no new large long-dated issuance.

Research, not investment advice.