Weekly US Treasury Market Assessment — Week Ending September 4, 2026

Current Treasury Regime

The current regime is an belly selloff associated with higher inflation compensation with mixed curve changes and persistent financing pressure. At the September 4 close, the five-year yield had risen 6 basis points from the prior week, exceeding the 3-basis-point increase in two-year yields, the 5-basis-point increase in ten-year yields, and the 2-basis-point increase in thirty-year yields.

The curve remained positively sloped: 2s10s closed at 41 basis points, 5s30s at 70 basis points, and 10s30s at 46 basis points. The weekly move therefore combined roughly 2 basis points of 2s10s steepening with approximately 4 basis points of 5s30s flattening and 3 basis points of 10s30s flattening.

Executive Summary

Inflation compensation rose across the supplied horizons, led by a 7-basis-point increase in the five-year breakeven to 2.37%. The ten-year breakeven increased 4 basis points to 2.35%, while 5y5y inflation compensation rose 1 basis point to 2.33%. These measures include inflation-risk and liquidity premia and should not be treated as pure inflation forecasts.

Labor data were firm overall: August payrolls increased by 162,000 after 21,000 previously, unemployment held at 4.1%, and monthly average hourly earnings increased 0.3% after 0.2%. Claims moved modestly higher, tempering but not reversing the resilience signal.

The available five-year real-yield observation declined 3 basis points through September 3, while ten-, twenty- and thirty-year real-yield observations were unchanged. Because real-yield observations are dated September 3 and nominal and breakeven observations are dated September 4, the evidence does not support an exact same-date nominal-real-inflation decomposition.

Review of the Previous Thesis

The prior framework was directionally useful in identifying a positive curve slope, continued two- to five-year vulnerability, elevated financing needs, and heterogeneous primary-market demand. The five-year sector did underperform, and bill results ranged from strong at 52 weeks to weak at 26 weeks.

It does not remain valid as the central framework. Inflation compensation reversed higher, nominal duration sold off rather than rallied, and the curve response became mixed: 2s10s steepened, whereas 5s30s and 10s30s flattened. The revised framework emphasizes relative long-end resilience rather than outright long-duration support.

Unresolved questions remain material. The latest supplied one-year-ahead fitted forward rate and Kim-Wright term-premium estimates are dated August 28, before the September 4 labor release and final-week pricing. They therefore cannot establish how expected policy rates or term premium contributed to the subsequent selloff.

What Changed This Week

The selloff broadened beyond the front end but remained concentrated in the belly. Five-year nominal yields rose more than other supplied major maturities, while the thirty-year sector sold off least. This relative configuration produced flattening from five to thirty years despite modest 2s10s steepening.

Five- and ten-year inflation compensation increased as real yields remained contained in the latest available September 3 observations. This supports describing part of the nominal selloff as associated with higher inflation compensation, but not assigning a precise decomposition across mismatched dates.

Firm payrolls, unchanged unemployment and faster monthly wage growth coincided with the repricing. Modestly higher initial claims, the four-week average and insured unemployment supplied a limited counter-signal rather than a comprehensive contrary growth assessment.

Auction Demand

Bill-auction demand was heterogeneous. The 52-week bill was the strongest supplied result, with 88th-percentile bid-to-cover, indirect awards 1.09 standard deviations above its comparison average, and dealer awards 1.46 standard deviations below average. The 8-week bill was also firm, with 71st-percentile coverage, strong indirect participation and below-average dealer awards.

The 26-week bill was the clear weak point: its bid-to-cover ranked at the 8th percentile of the prior 24 exact-series auctions, indirect awards were 2.47 standard deviations below average, and dealer awards were 2.20 standard deviations above average. The 13-week result was near the middle of its history, while the 4-week result was moderately firm.

These outcomes are aggregate bill-auction measures. Indirect awards are not equivalent to foreign purchases, and the supplied sample contained no coupon-auction result. Accordingly, the data cannot establish domestic or foreign ownership flows, nor investors' willingness to absorb coupon duration.

Yield Curve and Market Structure

The curve's level remained positive across the supplied spreads, but the change in slope was mixed. The five-year sector was the relative weak point, while the long bond outperformed the belly on a relative basis despite a higher nominal yield.

Available Kim-Wright estimates were positive and elevated in level—0.561% at two years, 0.81% at five years and 1.6015% at ten years—but unchanged through August 28. They predate the late-week market move and cannot be used to attribute that move to term premium.

Positioning presents two-sided amplification risk. Leveraged funds added to already substantial net shorts in two- and five-year futures, while ten-year asset-manager net length stood at the 98.1st percentile of its 52-week range. Conversely, leveraged funds reduced ten-year shorts and dealers reduced net shorts in two- and five-year futures but remained net short. Futures data do not reveal cash ownership, trade intent, or basis hedging.

Macro, Federal Reserve, Issuance and Fiscal Backdrop

The supplied labor releases point to resilience rather than a broad deterioration, but do not provide GDP, consumption, business-activity, or inflation-release coverage. The unchanged August 28 fitted one-year forward rate of 4.332% does not confirm a post-payroll change in the expected Federal Reserve path; no current futures, OIS, policy decision, or Federal Reserve communication was supplied.

Treasury estimates $739 billion of privately held net marketable borrowing for July through September, $68 billion above its May estimate. Treasury also expects nominal coupon and floating-rate-note auction sizes to remain unchanged for at least several quarters, while anticipating shorter-dated bill-size reductions during September.

Near-term announced coupon supply includes $58 billion of three-year notes, $39 billion of near-ten-year notes and $22 billion of near-thirty-year bonds. These are direct upcoming absorption tests, but announced supply alone does not predetermine auction outcomes or secondary-market yield direction.

Geopolitical and Cross-Market Considerations

Supplied geopolitical records describe reported developments concerning a US-Venezuela oil arrangement and identify a potential energy-inflation transmission channel. The records do not provide oil-price movements, Treasury flow data, event-window asset returns, or evidence of causation.

The presence of geopolitical evidence should not be equated with measured safe-haven demand. Higher ten- and thirty-year nominal yields also do not prove that safe-haven flows were absent, because other factors may have offset or dominated any such demand.

Current Base Case

The base case is for the curve to remain positively sloped, with the five-year sector comparatively vulnerable rather than an assumption of a uniform duration selloff. Firm labor readings and elevated five-year inflation compensation constrain the front end and belly, while the latest available long real yields remained unchanged and the thirty-year sector outperformed the belly as its yield rose less.

This relative long-end resilience faces an immediate supply test. The scheduled three-, near-ten- and near-thirty-year auctions will provide information on duration absorption that this week's bill-only auction sample cannot supply. Mixed bill demand and two-sided futures positioning argue against extrapolating either a clean rally or uninterrupted selloff.

Confidence is moderate because this framework depends on future auction performance and lacks current post-payroll policy-path pricing, current term-premium estimates, and same-date nominal-real data.

Bullish-Duration Scenario

A duration rally would require clearer labor deterioration, lower current policy-path pricing, a reversal lower in inflation compensation, and strong coupon-auction absorption with low dealer awards. The modest rise in claims, contained real yields, and sizable leveraged-fund shorts in two- and five-year futures identify possible supporting conditions, not confirmation that this scenario is underway.

If a rally were led by the front end and belly, 2s10s and 5s30s would tend to steepen. A long-end-led rally following especially strong long-duration auctions could instead flatten 5s30s and 10s30s.

Continued labor resilience, further increases in inflation compensation, or weak coupon auctions with high dealer absorption would invalidate this scenario.

Bearish-Duration Scenario

The selloff could broaden into long duration if inflation compensation remains firm, long real yields rise, and the near-ten- and near-thirty-year auctions show weak coverage or end-investor participation alongside elevated dealer awards. Current evidence establishes the relevant supply calendar and a higher nominal-yield week, but not weak coupon-auction demand, because those auctions occur after the assessment cutoff.

Extreme ten-year asset-manager length could add liquidation risk in such a scenario, although positioning alone is not directional evidence. A long-end-led selloff would generally steepen 5s30s and 10s30s; simultaneous policy repricing could keep 2s10s stable or flatter.

Falling inflation compensation and real yields, softer labor readings, or strong long-duration auction outcomes with low dealer absorption would invalidate this scenario.

Thesis Risks and Invalidation

The central expectation of persistent belly pressure would be challenged by a broad front-end rally accompanied by materially softer labor evidence or a lower current policy-path measure. It would also be challenged by a same-direction decline in inflation compensation and strong coupon-auction outcomes.

The interpretation that the latest weakness was associated principally with inflation compensation would be undermined by a sustained rise in five- through thirty-year real yields. Conversely, the relative-long-end-resilience view would be undermined by weak ten- and thirty-year auctions combined with higher long real yields.

Direct measures of cash-market liquidity are unavailable. Planned Treasury buybacks are a policy tool, not evidence of current market depth, bid-ask spreads, price impact, financing conditions, or settlement stress. The only supplied Treasury-volatility indicator is dated July 30 and flagged stale.

Next-Week Catalysts

The September 8 three-year auction is a $58 billion coupon-supply test. The September 9 near-ten-year auction is a $39 billion duration-absorption test, followed by a $22 billion near-thirty-year auction on September 10. Their coverage, bidder allocations and dealer awards will be more informative for coupon-duration demand than this week's bill results.

The September 10 unemployment-claims release will test whether the latest modest increases in initial claims, the four-week average and insured unemployment persist. Updated post-release policy-path and real-yield measures would also materially improve assessment of the belly selloff's drivers, but no such future observations are assumed here.

Evidence and Source Notes

The report cutoff is September 4, 2026 at 21:15:48 UTC. Official Treasury daily par yields and curve spreads are observed at the September 4 close. Inflation-compensation observations are dated September 4, while supplied real-yield observations are dated September 3 and fitted forward-rate and term-premium observations are dated August 28.

Auction comparisons use the 24 preceding exact-series auctions. The available weekly auction sample comprises bills only, so it is not evidence of coupon-duration demand. Indirect awards should not be interpreted as a complete measure of foreign demand, and dealer awards do not independently measure dealer balance-sheet capacity.

Geopolitical material is treated solely as evidence of a potential transmission channel. It contains neither measured Treasury flows nor sufficient event-window evidence to establish causation.

Methodology and Disclosure

This assessment synthesizes only the supplied evidence available by the stated data cutoff. It distinguishes observed market moves from proposed drivers, identifies missing measurements, and does not infer causation from timing alone. Where observation dates differ, nominal yields, real yields and inflation compensation are not represented as an exact decomposition.

AI assistance was used to draft and review this report against retained evidence through an administrator-authorized workflow. This is not independent human analyst review. Source data, calculations, interpretations and wording may be revised as additional or corrected evidence becomes available.

This document is informational only. It is not personal financial advice, an offer, a recommendation, or a representation of independent human analyst authorship.