Which forces are lifting U.S. borrowing costs, and how much control does the Federal Reserve have over them in the near term?

My position is that the Fed retains substantial near-term influence: a credible downward surprise to the expected policy-rate path should reduce borrowing costs. The latest maturity pattern supports policy repricing over an exclusively long-term-premium explanation, although the stronger claim of expectations dominance remains provisional. This is a conditional transmission argument, not a recommendation to cut regardless of inflation. (federalreserve.gov)

First, distinguish today’s episode from 2023. AP reports that the ten-year Treasury reached 5% this week for the first time in three years. More revealingly, from September 8–14, 2026, official constant-maturity yields rose 22 basis points at one year, 26 at two years, 23 at five years, 17 at ten years, and only nine at thirty years. The yield curve—the relationship between yields and maturities—therefore flattened between two and ten years. My inference is that this front-loaded increase fits higher expected policy rates better than an isolated increase in compensation for holding long bonds. Ten-year inflation breakeven, the nominal-minus-inflation-protected yield difference, remained 2.37%; it measures inflation compensation, not pure expected inflation. (apnews.com)

Second, model uncertainty need not make decomposition arbitrary. A Federal Reserve comparison of ACM, Christensen–Rudebusch, and Kim–Wright models proposes an approximate rule: the ten-year expectations change equals half the two-year yield change, while the premium change also incorporates curve steepening. Applying that historical rule to this window gives roughly 13 basis points of expectations and four of term premium—compensation for interest-rate risk beyond expected short rates. This is an illustrative calculation, not an updated model estimate or proof of cross-model robustness. The strongest objection matters: for July–October 2023, all four models examined assigned more of the increase to premiums than expectations. That episode cannot honestly establish my stronger claim. (federalreserve.gov)

Third, identified policy surprises demonstrate practical transmission. Gilchrist, López-Salido, and Zakrajšek’s Federal Reserve research estimates that conventional easing lowering the two-year yield ten basis points lowered the ten-year yield about four, and real three- and ten-year investment-grade corporate yields about seven and five, respectively. Its two-factor specification attributes conventional-policy effects primarily to path surprises rather than the immediate rate decision. These historical estimates are evidence of influence, not guaranteed contemporary coefficients. (federalreserve.gov)

For households, Freddie Mac’s September 10 application-based averages were 6.76% for thirty-year mortgages and 6.09% for fifteen-year mortgages. Lower benchmarks can lower new quotes, but credit spreads—borrower yields above comparable benchmark yields—and mortgage-specific charges can absorb some benefit; investment-grade results do not guarantee equal relief for weaker borrowers. (freddiemac.com)

The stakes are meaningful financing relief, not mechanical one-for-one pass-through. I have not established that recent auction weakness or issuance explains comparatively little: that requires matched auction-event and supply-surprise evidence. Current curve movements and historical transmission support substantial Fed control; they do not yet settle every component of the yield increase.

B’s strongest point is that my historical curve heuristic does not independently identify September 2026 expectations and term premiums. That objection has genuine merit: regression approximations cannot replace contemporaneous model estimates. Nevertheless, uncertainty about decomposition does not establish weak policy transmission, and additional evidence strengthens my interpretation of the latest increase. (federalreserve.gov)

First, the maturity pattern is broader than a two-versus-ten-year comparison. From September 8–14, yields increased 18 basis points at six months, 22 at one year, 26 at two years, 29 at three years, 23 at five years, 17 at ten years, and nine at thirty years, while the effective federal funds rate stayed at 3.63%. The yield curve—the relationship between maturity and yield—therefore repriced most strongly around one to five years. My inference is that anticipated policy tightening better explains this pattern than an isolated long-duration premium shock. Ten-year inflation breakeven, nominal yield minus inflation-protected yield, remained 2.37%; that constrains an inflation-compensation explanation without excluding real term-premium changes. (federalreserve.gov)

Second, we can test sensitivity rather than rely on the average heuristic. Applying each model’s published historical regression coefficients to the 26-basis-point two-year increase and nine-basis-point curve flattening produces approximate expectations increases of 16.5 basis points for ACM, 14.8 for Christensen–Rudebusch, 13.0 for Kim–Wright, and 11.3 for the survey regression. Each exceeds half the observed ten-year increase. These are my calculations using historical coefficients—not updated model decompositions, confidence intervals, or independent measurements. They establish robustness across these regression approximations only, but show that the directional conclusion does not require selecting one favorable model. (federalreserve.gov)

Third, issuance evidence weakens the specific hypothesis of a fresh coupon-supply surge. Treasury’s August 5, 2026 refunding statement maintained nominal coupon auction sizes; September’s planned ten- and thirty-year auctions were $39 billion and $22 billion, identical to June and July. Predictable issuance can still affect yields, and this announcement does not measure auction-demand surprises. I therefore cannot yet assign auction weakness a small numerical contribution. (home.treasury.gov)

Finally, borrower transmission has evidence beyond benchmark yields. Gilchrist, López-Salido, and Zakrajšek estimate that conventional easing lowering the two-year yield ten basis points reduces nominal thirty-year mortgage rates approximately six basis points, after incomplete secondary-to-primary-market transmission. Their three-year corporate coefficients imply nominal yield declines around nine basis points for both A- and BBB-rated firms. Credit spreads—borrower yields above comparable benchmarks—need not erase relief. These are historical estimates, not current quotes or guarantees for distressed borrowers. (federalreserve.gov)

My conclusion remains conditional: credible additional easing can meaningfully lower new borrowing costs. Policy-expectations dominance is supported for this narrow window, not conclusively established for the entire multiyear rise; a cut that damages inflation credibility falls outside my claim.

B’s strongest argument is that an overnight cut does not guarantee cheaper long-term financing: term premiums, mortgage intermediation costs, and borrower-risk spreads can offset it. Moreover, historical yield-curve regressions do not independently identify today’s decomposition, so conditional policy influence is not proof of reliable relief across borrowers. These are valid limitations. (federalreserve.gov)

My position remains stronger on the narrower causal question: what happens when the Fed credibly lowers the expected policy path? B’s 2024 counterexample does not identify that experiment. Fannie Mae reports that mortgage rates rose from 6.09% on September 19, 2024, to 6.84% on November 21, while markets revised their expected end-2026 policy rate from below 3% toward 3.5–4%. That supports distinguishing the announced cut from expectations of subsequent policy; issuance concerns also contributed, so this is not exclusive attribution. (fanniemae.com)

The September 8–14, 2026 maturity pattern likewise favors policy repricing: one-, two-, three-, five-, ten-, and thirty-year Treasury yields increased 22, 26, 29, 23, 17, and nine basis points, respectively. The yield curve—yields across maturities—flattened beyond two years. Ten-year inflation breakeven, the nominal-minus-inflation-protected yield difference, remained 2.37%. My inference is that these changes fit higher expected policy rates better than an isolated long-duration compensation shock, although unchanged breakevens cannot exclude real-premium increases. (federalreserve.gov)

Applying the Fed study’s four historical regression mappings gives approximately 11.3–16.5 basis points of expectations increases against the observed seventeen-basis-point ten-year rise. This is sensitivity analysis, not contemporaneous cross-model identification. Indeed, the same study finds term-premium dominance in July–October 2023: compensation beyond expected short rates can dominate, but that historical result does not establish September 2026 dominance. (federalreserve.gov)

Treasury’s August 5 statement maintained coupon auction sizes, including September ten- and thirty-year offerings of $39 billion and $22 billion. That weakens a fresh issuance-size explanation, but does not quantify demand weakness or anticipated future supply; my claim that auctions explain comparatively little remains unproven. (home.treasury.gov)

Policy-surprise research supplies the stronger transmission evidence: conventional easing lowering two-year yields ten basis points reduced estimated thirty-year mortgage rates about six basis points and three-year A/BBB corporate yields approximately nine. Credit spreads—the margin above comparable benchmarks—therefore did not erase historical investment-grade relief. These estimates are not current quotes or guarantees for distressed borrowers, and my argument is empirical, not a recommendation to cut irrespective of inflation. (federalreserve.gov)

I would reverse my decomposition judgment if contemporaneous ACM, Christensen–Rudebusch, and Kim–Wright estimates consistently assigned most of this increase to premiums, corroborated by policy-futures and survey expectations. I would reverse my transmission judgment if repeated, clean easing surprises durably lowered expected policy rates while matched mortgage and business quotes barely declined, with measured premium or spread increases explaining the offset.