US Treasury Weekly Assessment — Real-Yield-Led Bear Steepening and Auction Absorption Strain

Current Treasury Regime

The market regime is a real-yield-led bear steepening with uneven but material primary-market absorption strain and elevated financing pressure. The 2-year yield rose 5 bp to 4.81%, while the 5-, 10-, 20-, and 30-year yields rose 12 bp, 16 bp, 16 bp, and 15 bp, respectively. The resulting 2s10s slope ended at +36 bp and 5s30s at +51 bp; 10s30s remained positive at +32 bp but narrowed slightly on the week.

Real yields increased by 15 bp at 5 years, 17 bp at 10 years, 15 bp at 20 years, and 12 bp at 30 years through September 24. Five- and 10-year breakevens rose only 3 bp and 1 bp through September 25, while the 5y5y forward inflation-compensation measure declined 1 bp. These different observation dates mean the nominal move cannot be treated as an exact same-date decomposition, but the evidence supports describing the week as predominantly real-yield-led.

Executive Summary

The central development was a reversal from the prior front-end-led flattening framework to broad duration underperformance, concentrated from the belly through 20 years. The smaller 2-year increase relative to the 10-year increase widened 2s10s by about 11 bp. Steepening was not uniform beyond 10 years: 5s30s widened by about 3 bp, while 10s30s narrowed by about 1 bp.

Auction results reinforced an absorption concern rather than demonstrating a uniform demand failure. The 5-year note had 0th-percentile exact-series coverage, indirect awards 2.08 standard deviations below average, and dealer awards 1.69 standard deviations above average. The 4-week bill likewise had weak coverage, very low indirect awards, and elevated dealer awards. The 13-week bill and 2-year note were relatively more resilient, though neither showed unambiguously strong indirect participation.

The available policy-path and term-premium measures are dated September 18 and were unchanged at that observation. They therefore do not explain the subsequent selloff. Current meeting-specific policy pricing, measured domestic cash demand, foreign purchases outside auctions, safe-haven flows, and secondary-market liquidity measures were not supplied.

Previous Thesis Review

The prior thesis is not retained as the central framework. Its positive-curve, fiscal-supply, auction-risk, and positioning-amplification elements remained relevant. However, its defining flattening bias, expectation of relative ultra-long resilience, and expectation that front-end and belly maturities would remain more vulnerable than the long end were contradicted by the week's price action.

The 10- and 20-year sectors led the nominal selloff, each rising 16 bp, while the 30-year yield rose 15 bp. Real yields also rose across every supplied maturity. At the same time, auction weakness broadened through the 5- and 7-year coupon sectors and was visible in several bill and reopening results.

The unresolved elements remain the current expected-policy-rate versus term-premium split, the persistence of dealer burden after settlement, and whether geopolitical developments produced either inflation effects or Treasury safe-haven flows. The supplied evidence establishes possible transmission channels, not realized price or flow effects.

What Changed This Week

Nominal yields rose across all supplied maturities, but the move was substantially larger after the front end. The 2-year yield rose 5 bp; the 5-year rose 12 bp; the 10- and 20-year yields each rose 16 bp; and the 30-year yield rose 15 bp. The curve therefore steepened materially through 10 years.

Primary-market outcomes became more concerning in the belly. The 5-year auction was the clearest weak result, while the 7-year auction also showed low coverage and below-average indirect participation. In bills, the 4-week auction was notably weak, whereas the 13-week result was comparatively stable.

The real-yield evidence shifted materially: 5- through 30-year real yields all increased sharply. Inflation compensation remained relatively contained and mixed, with only modest increases in spot 5- and 10-year breakevens and a small decline in the 5y5y forward measure.

Auction Demand and Dealer Absorption

Auction demand weakened across several maturities when evaluated against each security's preceding 24 exact-series auctions. The 5-year note recorded 0th-percentile bid-to-cover, a -2.08 indirect-award z-score, and a +1.69 dealer-award z-score. The 7-year note had 8th-percentile coverage, a -1.42 indirect z-score, and a +0.43 dealer z-score. These results indicate weak auction absorption, although they do not reveal subsequent dealer distribution or secondary-market liquidity.

The 4-week bill was the weakest supplied bill outcome, with 13th-percentile coverage, a -2.33 indirect z-score, and a +2.22 dealer z-score. The 26-week bill also had 8th-percentile coverage and below-average indirect awards. By contrast, the 13-week bill showed median coverage and allocations near its exact-series averages. The 2-year note had 54th-percentile coverage and near-average dealer awards, though indirect awards remained below average.

Indirect awards are not a clean foreign-demand measure, and direct awards do not establish aggregate domestic end-investor ownership. The auction evidence is therefore best read as primary-market absorption and dealer-intermediation evidence, not as a complete account of domestic or foreign Treasury demand.

Yield Curve and Market Structure

The curve remained positively sloped and bear steepened through 10 years. At the September 25 close, 2s10s was +36 bp, 5s30s was +51 bp, and 10s30s was +32 bp. Weekly yield changes imply roughly 11 bp of 2s10s steepening and 3 bp of 5s30s steepening, offset by about 1 bp of 10s30s flattening.

The supplied real-yield and inflation-compensation records support a real-rate-led interpretation but do not provide an exact nominal-yield decomposition. Real yields were observed through September 24; breakevens were observed through September 25; nominal Treasury par yields were also observed on September 25. Differences in instruments, maturities, and observation dates are material limitations.

Current secondary-market liquidity cannot be classified. Weak auction outcomes are evidence of primary-market absorption strain, not direct evidence of impaired cash-market liquidity. No bid-ask, market-depth, price-impact, repo, fails, or inventory-turnover measure was supplied. The only options-based volatility observation is stale, dated July 30.

Macro, Federal Reserve, Issuance and Fiscal Setting

The supplied labour evidence remains resilient but narrow. Seasonally adjusted initial claims fell to 197,000, the four-week average declined to 202,250, and insured unemployment edged up by 2,000. This does not show a clear near-term labour deterioration, but one weekly claims release cannot establish the broader growth path.

The Kim-Wright fitted one-year instantaneous forward rate was unchanged at 4.8182% as of September 18. It is a model-derived measure, not meeting-specific futures or OIS pricing, and predates most of the report-week selloff. It should not be interpreted as a current policy forecast or proof of the cause of the weekly yield move.

Treasury's latest supplied borrowing estimate projected $739 billion of privately held net marketable borrowing for July–September and $628 billion for October–December. Treasury anticipated maintaining nominal coupon and FRN auction sizes for at least several quarters, while indicating that bill sizes could rise across the bill curve in October. Announced September 28–29 bill offerings total at least $316 billion across 13-, 26-, 6-, and 52-week maturities.

Geopolitical and Cross-Market Considerations

Geopolitical and trade-policy evidence is present, including sanctions legislation affecting major buyers of Russian oil and a separate trade-related development. These developments provide potential inflation, growth, and safe-haven transmission channels, but the supplied evidence does not establish an actual energy-price effect, Treasury flow, or causal contribution to yields.

No measured safe-haven Treasury flow was supplied. The 10- and 30-year yields rose 16 bp and 15 bp during the week, inconsistent with an observable duration rally, but this does not rule out offsetting safe-haven purchases. Safe-haven demand should therefore remain classified as unknown rather than absent.

Current Base Case

The base case is for a positively sloped curve with continued vulnerability in the belly and long end and a modest bear-steepening bias through 10 years. Supporting conditions are broad real-yield increases, weak 5- and 7-year auction absorption, elevated dealer awards in several auctions, and a heavy near-term bill calendar followed by scheduled coupon auctions.

Conviction is moderate rather than high. Inflation compensation is contained, the 13-week bill and 2-year note auctions were relatively better, current policy-path and term-premium updates do not cover the selloff, and large leveraged-fund short positions create reversal and squeeze risk. Supply pressure and weak auctions are risk factors consistent with higher required compensation; the supplied evidence does not prove that they caused the yield move.

Bullish-Duration Scenario

A duration rally would become more credible if subsequent labour data weakened materially, a current policy-path measure shifted lower, real yields declined broadly, and coupon-auction absorption improved through stronger coverage and indirect participation with lower dealer awards. Large leveraged-fund short positions could amplify such a reversal.

A policy-led rally would likely concentrate more at the front end and flatten 2s10s. A rally led by improved duration demand and lower long real yields could instead produce more pronounced bull flattening in 5s30s and 10s30s. These are conditional market-structure outcomes, not forecasts.

Bearish-Duration Scenario

The selloff could extend if resilient labour evidence persists, updated policy measures show stable or higher expected short rates, real yields continue rising across intermediate and long maturities, and the coming 3-, 10-, and 30-year coupon auctions show weak coverage and indirect participation with higher dealer absorption.

Under this outcome, 2s10s could steepen further if duration absorption and real-yield pressure remain concentrated beyond the front end. If an updated policy measure instead points to a more restrictive short-rate path, the move could become more parallel or bear-flattening. The current evidence does not determine which channel would dominate.

Thesis Risks and Invalidation

The base case would be materially weakened by a broad nominal- and real-yield rally accompanied by weaker labour evidence, lower current policy pricing, and stronger coupon auctions with improved indirect participation and contained dealer awards.

A sustained front-end-led selloff that reverses the recent widening in 2s10s would invalidate the expected relative 2-year resilience and bear-steepening bias. A material rise in inflation compensation, rather than real yields, would also invalidate the characterization of the current move as predominantly real-yield-led.

Further long-end underperformance combined with a current and sharply higher term-premium estimate would require reassessing the modest base-case framing. Conversely, measured safe-haven flows or evidence of impaired secondary-market liquidity would require reassessment because neither is currently observed in the supplied evidence.

Next-Week Catalysts

The September 28 auctions of $95 billion in 13-week bills and $82 billion in 26-week bills will test whether the prior relative resilience of the 13-week sector and weakness of the 26-week sector persist. On September 29, $85 billion of 6-week bills and $54 billion of 52-week bills are scheduled, bringing announced supply for those two days to at least $316 billion.

The September 30 17-week bill auction and October 1 reopenings of 4- and 8-week bills will provide further evidence on bill absorption, though offering amounts were not supplied. The October 1 unemployment-claims release will test whether the limited resilience in claims data continues.

The scheduled October 6–8 auctions of a 3-year note, a 9-year 10-month note reopening, and a 29-year 10-month bond reopening are the next major coupon-duration tests. An updated meeting-specific policy measure or refreshed Kim-Wright policy-path and term-premium readings would also help distinguish expected-short-rate pressure from duration compensation.

Evidence and Source Notes

This assessment uses supplied official Treasury daily yield-curve, auction, issuance, and quarterly-refunding records; supplied FRED-approved real-yield, inflation-compensation, and Kim-Wright measures; the supplied Department of Labor claims release; supplied CFTC Treasury-futures positioning records; and supplied geopolitical reporting. The evidence cutoff is September 25, 2026 at 21:16:14 UTC.

Auction comparisons are against the supplied 24 preceding exact-series auction histories. Indirect awards are an imperfect proxy for foreign demand and do not identify foreign official, foreign private, or domestic ownership. Dealer awards measure auction allocation rather than post-settlement inventory, financing conditions, or secondary-market liquidity.

The supplied Kim-Wright measures are dated September 18 and the real-yield measures are dated September 24; they should not be used to assign a precise decomposition to the September 25 nominal yield close. The VIXTLT observation is dated July 30 and marked stale, so no current volatility regime is inferred.

Methodology and Disclosure

This is an informational, evidence-led market assessment, not personal financial advice, a recommendation, or a solicitation to transact. It uses only the supplied evidence and separates observed market data from conditional scenario analysis. Causal claims are not made where the evidence shows correlation, timing, or a potential transmission channel only.

AI assistance was used to organize and draft this report. The assessment is subject to manual review and may be revised if supplied source data are corrected, updated, or supplemented. Model-derived measures, including Kim-Wright estimates, are revision-prone and should not be treated as direct policy forecasts.

The report distinguishes unknown conditions from absent evidence. In particular, geopolitical developments were supplied, but measured safe-haven flows were not; auction strain was measured, but secondary-market liquidity was not. Nominal yields, real yields, and inflation compensation may have different observation dates and therefore are not assumed to be an exact decomposition.