Rate headlines and interest bills run on different clocks, and I split them in the tracker because I kept mixing up the afternoon move with the bill that hasn't changed. The share price can move the same afternoon, because the discount rate used on future cash flows changed, or because investors changed their mind about the path. The coupon the company pays changes later, and only on the debt that is allowed to change.
I think of three clocks, and the dates are the whole point. Floating-rate debt resets on the contract date, often quarterly. Commercial paper rolls in weeks or months, so the new rate arrives at the next issue. Fixed-rate notes keep the old coupon until maturity, unless a swap has already made them float. Until one of those dates, a higher policy rate is a valuation fact and a refinancing risk, and it isn't yet a higher interest expense.
The announcement itself, the statement, the projections and the press conference, is covered in what a central bank rate decision means for your holdings. This piece is only about the lag between that day and the cash.
Where the lag is short
Disney is a short-lag case. Borrowings were $42.03 billion at September 27, 2025. Swaps had turned $10.56 billion of notes into floating debt, and commercial paper was $2.06 billion, so roughly 30% of borrowings were effectively floating. That slice reprices at the next reset or the next paper issue, and the rest waits for maturity. Experiences is also funding a park buildout, so the rate on new borrowing matters beside the rate on old borrowing. See Disney.
Salesforce added a floating slice with the Informatica deal. At January 31, 2026, debt principal was $14.5 billion, of which $6.0 billion borrowed in November 2025 floats off term SOFR or a base rate plus a margin, and another $4.0 billion is due within 12 months. The new cash sensitivity is the credit agreement, not the subscription model. See Salesforce.
Apple looks quiet if you stop at the fixed-rate notes, which is the easy misread. At September 27, 2025 it had $91.28 billion of term debt and $8.0 billion of commercial paper at a 4.19% weighted average rate, due within nine months. Of the term debt, $12.39 billion comes due in fiscal 2026, so the paper and the near maturity are the clocks that strike soon. A further $12.6 billion of term debt sits in fair value hedges, which means the swap footnote can change the word fixed. See Apple.
Oracle is the case where the next issue matters more than the old coupon. At May 31, 2026, future principal payments were $130.1 billion, with $7.2 billion due in fiscal 2027, and capital spending was $55.7 billion against $32.0 billion of operating cash flow. The gap was funded largely with new borrowings. Higher rates hit Oracle's next bond, not the ones it already sold. With that much borrowing ahead, the next bond is what matters. See Oracle.
Where the company's own interest bill is the small number
Home Depot shows the trap of reading only the debt footnote, and this is where most people get the story backwards. At February 1, 2026, floating-rate principal after swaps was $5.4 billion, and the filing says a one point rise in rates would add about $54 million of annual interest expense. The same filing says high rates and home prices kept housing turnover historically low, which cuts projects tied to buying and selling a home. The $54 million is real, but it's small. What actually hurts Home Depot is people not moving house. See Home Depot.
GE Aerospace has $20.49 billion of borrowings, with senior notes at a 4.00% average rate. The rate story in the filing sits with airline finances, because services are the core of the business and shop visits follow flying hours. A higher funding cost for an airline can defer an overhaul, so that cash leaves GE's revenue line, not its interest line.
Banks, and the sign of the cash
For a deposit-funded bank the cash effect is net interest income, and it can move against the equity-market habit of treating "higher rates" as one direction.
- JPMorgan Chase earned $95.4 billion of net interest income in 2025, slightly more than half of net revenue. The earnings-at-risk table shows that income rising about $2.1 billion under an immediate 100 basis point parallel increase, and falling about $2.4 billion under a 100 basis point decrease, so a cut costs more, in that table, than a rise adds. See JPMorgan.
- Bank of America estimates that an immediate 100 basis point parallel rise would add about $0.7 billion to banking book net interest income, while a 100 basis point drop would cut it by about $2.0 billion. The asymmetry is larger. See Bank of America.
- Wells Fargo's table shows net interest income over the next 12 months rising $1.9 billion under an instantaneous 100 basis point rise and falling $2.3 billion under a 100 basis point drop. See Wells Fargo.
These tables are the banks' own scenarios, not a forecast, and they assume a parallel move rather than the curve twist a real meeting often produces. They're still the page I open. A line that says higher rates help banks is the sector prior, while the table says the bank has measured a cut as the more expensive side.
Goldman Sachs is the break in the group. Net interest income was about 23% of 2025 net revenue, and underwriting, advisory and trading swing more. Most long-term borrowings are fixed but swapped to floating, so funding does reprice, and it still isn't the main earnings line. See Goldman Sachs.
Cash that earns interest, not debt that pays it
A lower policy rate can cut income at companies that hold large amounts of cash and short bills, which is the opposite of the relief story people tell about rate cuts. Berkshire Hathaway's insurance and other businesses held $369.0 billion of cash, cash equivalents and US Treasury bills at the end of 2025, and the filing says insurance investment income fell in 2025 on lower rates. The equity story that lower rates are a relief doesn't describe that pile. See Berkshire.
Costco is a smaller version of the same pile. Interest expense was $154 million in fiscal 2025, and interest income and other was $589 million. Long-term debt was $5.8 billion, almost all fixed, with $75 million current, so the interest income line is where a cut shows up, modestly, as less income.
What to pull from the debt note
The annual report's debt footnote is enough for this check, and you don't need the whole filing. The five sections that carry risk more broadly are in how to read a company filing for risk. For the rate clock, I look for the maturity table first, meaning how much principal is due in the next twelve months and how much in the year after that, then commercial paper and other short-term borrowings, which is the slice that reprices by being rolled over. I also want the fixed versus floating split after swaps, because a note labelled fixed can be floating once the hedge is included. Amgen, for example, had interest rate swaps with a notional amount of $6.7 billion on a $54.60 billion stack. Next is the rate on new borrowing the company has already said it needs, and Abbott's plan to borrow about $20 billion for Exact Sciences is a future coupon, not a current one. Last is the sentence about customers: if the filing talks about mortgage demand, airline financing, or equipment loans, that is a revenue clock, and it won't appear in the maturity table.
A hedge buys time. It doesn't make the risk go away. A swap that runs eighteen months means the cash effect waits until it rolls off, while the share price doesn't wait, so watching only interest expense makes the rate move look like it did nothing until the hedge expires.
Write two dates next to a holding. One is the next reset or maturity that changes interest expense, and the other is the business cycle the filing names, such as housing turnover or a customer's refinancing. When a decision lands, ask which date just got closer. The stock reacts to both at once, so you have to pull them apart yourself.
Educational reading of how policy rates reach interest expense and customer demand. Not investment advice. Figures come from the filings cited on each stock page and can be updated in the next annual report.