US Treasury Weekly Assessment — Week Ending 11 September 2026

Current Treasury Regime

The current regime is a front-end-led bearish repricing with bear flattening, firm inflation pressure and strong coupon-auction absorption. At the 11 September close, the two-year yield was 4.63%, the five-year 4.78%, the ten-year 4.96%, the twenty-year 5.38% and the thirty-year 5.35%. Weekly yield increases were largest at two and five years, while the long end sold off less severely in relative terms.

The curve remained positively sloped but flattened over the week. Closing 2s10s, 5s30s and 10s30s spreads were +33 bp, +57 bp and +39 bp, respectively. The supplied weekly maturity changes imply roughly 8 bp of 2s10s flattening, 13 bp of 5s30s flattening and 7 bp of 10s30s flattening.

Executive Summary

August inflation releases made the inflation backdrop more concrete: headline and core PPI each rose 0.4% month over month, with year-over-year readings of 5.4% and 5.1%; headline CPI rose 0.4% month over month and core CPI 0.3%. Five-year inflation compensation increased 3 bp and ten-year inflation compensation increased 1 bp, although the five-year/five-year-forward measure declined 1 bp. These measures include risk and liquidity premia and should not be treated as pure inflation expectations.

The nominal selloff was accompanied by higher real yields. Five-, ten-, twenty- and thirty-year real yields rose 12 bp, 12 bp, 10 bp and 9 bp, respectively, in observations through 10 September. Nominal yields and inflation-compensation observations are dated 11 September, whereas real-yield observations are dated 10 September; consequently, this is not an exact same-date nominal-real-inflation decomposition.

Primary-market coupon demand was a counterweight rather than a source of visible stress. The three-year auction had 88th-percentile coverage, the near-ten-year reopening ranked at the 100th percentile and the near-thirty-year reopening at the 92nd percentile versus their respective preceding 24-auction histories. Dealer awards at the two longer reopenings were exceptionally low. This supports relative long-end resilience, not an outright bullish duration conclusion, because long yields and real yields still rose.

Previous Thesis Review

The prior framework was confirmed in several respects: the curve remained positively sloped; two- to five-year maturities remained vulnerable; the long end outperformed the belly in relative terms; and coupon supply proved an important event risk. However, the coupon-auction test was resolved materially more positively than a cautious absorption framework would have implied.

The prior thesis no longer remains valid as the central framework. The selloff became substantially larger and front-end-led, while real yields rose across five through thirty years. The expectation that relatively contained long real yields could support duration did not materialize, and the prior possibility of stable-to-modestly-steeper 2s10s was contradicted by approximately 8 bp of flattening.

Key questions remain unresolved. The only supplied policy-path measure is a model-fitted one-year forward rate dated 4 September, unchanged at 4.4119%; it is neither current-week evidence nor a meeting-specific futures or OIS probability. Likewise, available Kim-Wright term-premium estimates predate the report-week selloff, and geopolitical evidence does not provide measured Treasury flows or event-window attribution.

What Changed This Week

The market move shifted from a belly-led selloff to a larger front-end-led bearish repricing. Two-year and five-year yields rose 26 bp and 24 bp, compared with increases of 18 bp at ten years, 13 bp at twenty years and 11 bp at thirty years. That distribution generated bear flattening while preserving a positive curve slope.

Inflation evidence strengthened. PPI and CPI each registered firm August month-over-month readings, and five- and ten-year breakevens moved modestly higher. Initial claims fell to 206,000, the four-week average declined to 206,000, and insured unemployment was nearly unchanged. This narrow labour evidence does not establish broad growth strength, but it does not show material labour deterioration.

Strong coupon-auction outcomes replaced unresolved supply absorption as a central risk. The near-ten-year and near-thirty-year reopenings combined very high historical coverage rankings with high indirect awards and unusually low dealer awards. Bills were less uniform, with the 13-week reopening notably weaker than other supplied bill results.

Auction Demand

Coupon-auction absorption was strong. The $58 billion three-year sale recorded a 2.72 bid-to-cover ratio, at the 88th percentile of the preceding 24 exact-series auctions. Direct awards were 26.77%, above the 18.60% average; the supplied comparison also shows above-average indirect participation and below-average dealer awards.

The $39 billion near-ten-year reopening was the strongest relative result: its 2.71 bid-to-cover ratio was at the 100th percentile, indirect awards were 78.98% against a 63.30% average, and dealer awards were 4.30% against a 9.59% average. The $22 billion near-thirty-year reopening recorded 2.61 coverage at the 92nd percentile, 79.33% indirect awards against a 58.28% average, and only 2.20% dealer awards against a 10.98% average.

The evidence supports strong primary-market absorption and light coupon dealer burden, but allocation categories require care. Indirect awards are not a pure foreign-demand measure, while direct awards do not establish domestic end-investor demand or bidder nationality. Auction allocations also do not identify secondary-market flows, dealer inventory after distribution, or broader cash-market liquidity.

Bill demand was mixed. The 13-week reopening had 21st-percentile coverage, below-average indirect participation and modestly above-average dealer participation. The 26-week, four-week and eight-week offerings had above-average indirect participation and below-average dealer awards, while their coverage rankings were around the middle of available history.

Yield Curve and Market Structure

The week’s market structure was broad bearish repricing with relative long-end outperformance. The front end and belly rose more than the long end, consistent with a policy-sensitive flattening pattern, though the supplied evidence does not uniquely identify the catalyst. Inflation releases and stable claims are consistent with front-end pressure, but current meeting-specific policy pricing is unavailable.

Real discount-rate pressure participated across the curve: five-, ten-, twenty- and thirty-year real yields rose through 10 September. Ten-year inflation compensation rose only 1 bp through 11 September, whereas five-year compensation rose 3 bp. Because these observations are not aligned to the same day, no precise decomposition of nominal changes into real-rate and inflation-compensation components is supportable.

Term-premium direction during the report week is unknown. Kim-Wright estimates dated 4 September were positive in level—0.5898% at two years, 0.8263% at five years and 1.6006% at ten years—but were unchanged versus the preceding available observation and predated the subsequent selloff. Treasury-volatility conditions are also unknown because the only supplied VIXTLT reading is dated 30 July and marked stale.

Macro, Policy Expectations, Issuance and Fiscal Backdrop

The available macro evidence is inflation-heavy. August PPI was +0.4% month over month in headline and core terms, while August CPI was +0.4% headline and +0.3% core. Claims data were stable to marginally firmer, with seasonally adjusted initial claims down 1,000 to 206,000. Broader activity, payroll and unemployment-rate data were not supplied, so overall growth pressure remains insufficiently measured.

The Federal Reserve-path conclusion should remain qualified. The two-year yield rose 26 bp, but the supplied model-fitted one-year forward rate was unchanged at 4.4119% as of 4 September. That dated model estimate is not a Fed funds futures probability and cannot establish the precise within-week policy-path repricing.

Treasury projects $739 billion of privately held net marketable borrowing for July–September and $628 billion for October–December. Current refunding guidance anticipates maintaining nominal coupon and FRN auction sizes for at least several quarters, with September reductions in shorter-dated bill auction sizes expected before increases across the bill curve in October. Near-term announced supply includes $92 billion of 13-week bills, $79 billion of 26-week bills, $75 billion of six-week bills, a $13 billion near-twenty-year bond and a $19 billion near-ten-year TIPS reopening. Amounts for certain later scheduled bill offerings were not supplied.

Geopolitical and Cross-Market Considerations

The supplied geopolitical records indicate material oil-transport and energy-supply disruption risk around the Strait of Hormuz and related tanker attacks. These developments provide a potential inflation and policy-rate transmission channel, particularly alongside firm inflation releases.

They do not establish realized Treasury flows, safe-haven demand, or causal attribution for the week’s yield changes. Nominal yields rose across the supplied two-, ten- and thirty-year maturities, but this price action cannot distinguish weak safe-haven demand from stronger offsetting inflation, real-rate or policy pressure. No Treasury ETF, mutual-fund, custody-flow or event-window market-response data were supplied.

Current Base Case

The base case is for a positively sloped curve with continued relative vulnerability in the front end and belly, rather than an assumption that the report week’s selloff repeats at the same magnitude. Firm inflation releases, modestly higher five- and ten-year inflation compensation, stable claims and the concentration of weekly yield increases at two and five years support that conditional bias.

The long end may continue to outperform in relative terms if coupon demand remains firm. The near-ten-year and near-thirty-year reopenings showed very strong coverage and unusually low dealer absorption. This is a relative-value observation, not a call for outright duration strength: real yields rose across all supplied long maturities and the fiscal financing requirement remains large.

Confidence is moderated by mixed bill demand, large two-sided futures positioning, stale term-premium and volatility measures, and absent current meeting-specific policy pricing. Leveraged funds remained heavily net short across supplied sectors, while asset managers remained net long but reduced five- and ten-year exposure; these data identify potential amplification risk rather than trader intent.

Bullish-Duration Scenario

A bullish-duration outcome would require evidence that current inflation pressure is easing, labour conditions are deteriorating materially rather than remaining stable, and a valid current policy-path measure is shifting lower. Lower five- and ten-year inflation compensation together with declining real yields would strengthen that case.

Strong coupon demand and sizeable leveraged-fund net shorts could amplify a rally, but neither is sufficient to establish one. Futures positioning does not reveal cash-market ownership, leverage, trade motivation or stop levels. Depending on where the rally concentrates, a front-end-led rally could steepen 2s10s, while strong long-duration demand could flatten longer curve segments.

Bearish-Duration Scenario

A bearish-duration extension would be supported by persistently firm inflation data or higher inflation compensation, continued increases in five- through thirty-year real yields, and updated policy-path data showing further restrictive repricing. Weakening coupon auctions—lower coverage and end-investor participation alongside higher dealer awards—would add a direct primary-market stress signal.

A policy-led continuation would tend to preserve flattening pressure in 2s10s. If higher real yields and weaker duration absorption instead shift the selloff toward the long end, 5s30s and 10s30s could steepen. The present auction evidence contradicts that second pathway for now: long-duration reopenings were strongly absorbed.

Thesis Risks and Invalidation

The base case would be challenged by a sustained front-end rally accompanied by materially softer labour evidence, weaker inflation releases or a lower valid current policy-path measure. It would also be challenged if subsequent coupon auctions weakened materially, particularly through lower coverage and higher dealer absorption.

The relative-long-end-resilience component would be invalidated by a renewed long-end-led selloff alongside rising long real yields and widening 5s30s or 10s30s. Conversely, a broad decline in real yields and five- and ten-year inflation compensation would weaken the current bearish-duration impulse.

Important evidentiary limitations are themselves risks to conviction: current policy-path pricing, report-week term-premium estimates, broad cash-market liquidity, safe-haven flows and current Treasury volatility were not measured in the supplied record. Geopolitical evidence establishes potential energy risk, not Treasury-flow causality.

Next-Week Catalysts

The 14 September auctions of $92 billion in 13-week bills and $79 billion in 26-week bills will test whether the weak 13-week reopening was idiosyncratic or persistent and whether stronger 26-week indirect participation is sustained.

On 15 September, the $75 billion six-week bill and $13 billion near-twenty-year bond auctions will test short-dated and long-duration absorption. The planned $19 billion near-ten-year TIPS reopening on 17 September will provide a direct demand test for inflation-linked duration after the rise in real yields.

The 17 September claims release will test whether the supplied stable labour signal persists. Offering sizes for the scheduled 17-week bill and four- and eight-week bill reopenings were not supplied, so the full near-term bill financing requirement remains incomplete.

Evidence and Source Notes

This assessment uses supplied official Treasury daily par-yield-curve, auction, upcoming-auction and quarterly-refunding records; supplied BLS CPI and PPI releases; supplied Department of Labor claims data; supplied FRED-series observations for real yields, inflation compensation, fitted forward rates and Kim-Wright term premium; supplied CFTC Treasury-futures positioning records; and supplied geopolitical records marked valid with warnings.

The evidence cutoff is 11 September 2026 at 21:15:54 UTC. Market observations have different dates: nominal yields and breakevens are through 11 September; real yields are through 10 September; fitted forward-rate and term-premium observations are through 4 September; positioning is through 8 September; and VIXTLT is dated 30 July and stale. These timing differences limit comparisons and prevent exact decomposition or current-state claims where the observations are older.

Auction percentiles, z-scores and bidder shares describe outcomes relative to supplied exact-series auction histories. They do not identify bidder nationality, ultimate beneficial ownership, secondary-market flows, dealer inventory, or broader market liquidity.

Methodology and Disclosure

This is an evidence-led, informational market assessment generated with AI assistance and subject to manual review. It is not personal financial advice, a recommendation, an offer, or a solicitation to buy or sell any security or derivative.

The assessment privileges measured outcomes over causal attribution. Yield, curve, inflation, real-yield, auction, issuance, fiscal and positioning statements are limited to the supplied records. Scenario language is conditional and identifies evidence that would support or invalidate each framework rather than forecasting with certainty.

The report may be revised if supplied source records are corrected, updated or superseded. Missing measures are treated as unknown rather than as evidence of absence; this is particularly relevant to safe-haven flows, cash-market liquidity, current volatility, meeting-specific policy pricing and report-week term premium.