What moved
The 30-year Treasury yield reached 5.24 per cent on Thursday, up from 4.9 per cent at the end of June and within reach of a 19-year high set days earlier. The Treasury’s unscheduled plan to at least double buybacks of 10 to 30 year securities, from $2bn to $4bn an issue from 9 September, bought a rally that lasted a day. National debt has passed $40tn, deficits run near 6 per cent of GDP, and the dollar fell on both days. Thirty-year gilts moved in sympathy, to 5.81 per cent.
Why it matters for pricing
The US long end is the reference rate beneath a large share of the world’s intercompany debt, including in groups with no US operations at all. When it moves more than 30 basis points in under two months and an intervention cannot hold it, that is not an American problem. It is the base leg of your USD pricing.
What matters is where the move sits. The selling is concentrated at the long end while the front end stays anchored — the term premium is back after a decade in which it barely existed. Through that decade a ten-year intercompany loan priced barely above a three-year, and the habit of taking a short benchmark, adding a flat margin and applying it across tenors survived because the error was small. It is not small now. Long-dated USD intercompany debt priced that way is understated, and the understatement grows with tenor.
The obvious response is to shorten. It is what the Treasury itself is expected to do — fund the buybacks by issuing more at the front end — and it is being criticised for it, because swapping a known long-term cost for rollover risk is not a saving. The same criticism reaches you, with an extra edge: tenor is a delineation question. Where the funding need is plainly long-term and the paperwork says 364 days rolling, the delineated transaction is the one the facts support, not the one the documents describe. Shortening intercompany tenor to reduce a rate is the most legible tax-motivated structuring decision available right now.
Our call. USD long-end rates stay elevated and the term premium widens further absent genuine fiscal consolidation, which nobody expects this year. Price to the matching point on the curve, and expect the spread between three-year and ten-year intercompany pricing to keep opening.
What this means for your business
- Price to the tenor. A ten-year USD intercompany loan benchmarked off a short reference plus a flat margin is now visibly understated. Match the benchmark to the term.
- Before shortening any intercompany facility, establish what the funding need actually is. Tenor that does not match the commercial purpose is the delineation argument you lose.
- The spillover is not confined to USD. Long-end sovereign stress correlates — 30-year gilts moved with Treasuries on the same days. Re-run long-dated GBP and EUR pricing too.
- US groups: model 163(j) capacity at current rates before the next funding decision, not at the year-end close when the options have gone.
Sources
Financial Times reporting on the US Treasury buyback and long-end bond market, 20–21 August 2026, with US Treasury department debt data. Section 163(j): Internal Revenue Code as amended by the One Big Beautiful Bill Act (Grant Thornton analysis, July 2025). Yields as at 21 August 2026.