US Treasury Weekly Assessment — Week Ending July 24, 2026

Current Treasury Regime

The current regime is a real-yield-led selloff with positive bear-flattening, elevated near-term supply pressure, and rising long-duration option-implied volatility. The assessment confidence is 0.82. The characterization is descriptive: 2-, 5-, 10-, and 30-year nominal yields all rose, while 5-, 10-, 20-, and 30-year real yields also increased. The largest nominal increases were in the 2- and 5-year sectors, and the largest real-yield increase was at five years.

The curve remained positively sloped but flattened modestly during the selloff. The 2s10s, 5s30s, and 10s30s slopes closed at 36, 73, and 47 basis points, respectively. The supplied weekly yield changes imply narrowing of roughly 1, 5, and 4 basis points in those slopes.

Long-duration volatility rose materially: VIXTLT increased from 66.07 to 71.94. This measures option-implied uncertainty for the long-duration Treasury ETF rather than cash-Treasury liquidity or the entire curve.

Executive Summary

Treasuries sold off across the curve. The 2-year and 5-year yields rose 15 basis points to 4.33% and 4.43%; the 10-year rose 14 basis points to 4.69%; and the 30-year rose 10 basis points to 5.16%. Rising real yields across supplied maturities make higher real discount rates the clearest common feature of the move.

The available labor signal was firmer but narrow. Initial claims fell to 187,000 from 209,000, the four-week average declined to 207,500, and insured unemployment was nearly unchanged at 1.796 million. This reduces downside growth pressure on the supplied evidence, but does not establish a broad change in growth conditions.

The immediate financing calendar is substantial. Known July 27–29 offerings total $479 billion, including $213 billion in 2-, 5-, and 7-year securities. Treasury's most recent borrowing estimate projected $671 billion in privately held net marketable borrowing for July through September. Recent demand was mixed, with a strong 13-week bill result but acute weakness in the 8-week bill and a softer descriptive result in the 20-year reopening.

The central case remains volatile, supply-sensitive trading with relatively greater vulnerability in the front and belly, preserving a positive curve slope while retaining modest bear-flattening pressure. This is not a conclusion that the supplied Fed-path or term-premium measures shifted: both were unchanged in the available, July 17-dated observations.

Previous Thesis Review

No previous report was supplied. Consequently, no prior expectations can be scored, and no prior thesis can be confirmed, rejected, or carried forward.

What Changed During the Week

Nominal yields rose broadly: 2-year and 5-year yields each increased 15 basis points, the 10-year increased 14 basis points, and the 30-year increased 10 basis points. The relative pattern produced a modest bear-flattening rather than a long-end-led bear-steepening.

Real yields also rose materially: five-year real yields increased 16 basis points, followed by 12 basis points at 10 years, 9 basis points at 20 years, and 8 basis points at 30 years. Inflation compensation was mixed rather than uniformly higher: the five-year breakeven declined 3 basis points, while the 10-year breakeven and 5y5y forward measure rose 2 and 7 basis points.

The weekly claims release showed lower initial claims and a lower four-week average. Separately, VIXTLT rose by 5.87 points, increasing the price of long-duration rate uncertainty and hedging.

Auction Demand and Dealer Absorption

Auction demand was heterogeneous. The 13-week bill was the strongest exact-series result: its 3.00 bid-to-cover ratio was at the 100th percentile of six preceding comparable auctions, indirect awards were 50.92% versus a 47.57% average, and dealer awards were below average at 34.10% versus 37.95%.

The 26-week and 4-week bill auctions were closer to recent norms. The 26-week bid-to-cover was 2.94, at the 67th percentile of six comparisons; its dealer award share was below average. The 4-week bid-to-cover was 2.79, at the 57th percentile of seven comparisons; its indirect share was above average and dealer share below average.

The 8-week bill was the clear weak point. Its 2.31 bid-to-cover ranked at the zero percentile of seven comparable auctions and was 2.92 standard deviations below average. Indirect awards were 36.32%, a -3.89 z-score, while dealers absorbed 50.87%, a +3.60 z-score. This establishes localized dealer absorption pressure, not broad dealer stress across all supplied auctions.

The 20-year reopening was descriptively softer than its single comparison: bid-to-cover was 2.64 versus 2.75, indirect and direct shares were lower, and dealer awards increased to 14.53% from 8.44%. The one-observation history prevents statistical classification. The new 10-year TIPS issue had no exact-series history, so its 2.30 bid-to-cover and 8.82% dealer award share cannot be ranked.

Indirect awards should not be treated as a direct measure of foreign demand, and direct awards do not isolate domestic demand. No ownership, custody, or fund-flow evidence was supplied.

Yield Curve and Market Structure

The week was a positive bear-flattening. The 2-year and 5-year yields each rose 15 basis points, compared with 14 basis points at 10 years and 10 basis points at 30 years. The resulting 2s10s, 5s30s, and 10s30s slopes remained positive at 36, 73, and 47 basis points.

Higher real yields were concentrated in the front and belly of the supplied real-yield series. The five-year real yield rose 16 basis points and the 10-year real yield rose 12 basis points, compared with 8 basis points at 30 years. Mixed inflation compensation limits any conclusion that the move was a uniform inflation-compensation shock.

The supplied Kim-Wright fitted one-year forward rate and 2-, 5-, and 10-year forward term-premium estimates were unchanged. Those observations were dated July 17, preceding the later cash-yield moves, and are model estimates subject to specification and revision risk. They therefore do not explain the subsequent selloff.

Positioning risk is elevated and two-sided. Asset-manager positioning was at the 100th percentile in 10-year futures but near the bottom of its range in 2- and 5-year futures. Leveraged-fund positioning was at the 100th percentile in 2-year futures and the 98.1st percentile in 5-year futures, while 10-year leveraged positioning was less extreme. These measures do not identify cash positions, options exposure, or trader intent.

Macro, Fed, Issuance, and Fiscal Conditions

The available labor evidence points to a firmer near-term backdrop but is insufficient for a broad growth assessment. Initial claims fell by 22,000 to 187,000, the four-week average declined by 7,250 to 207,500, and insured unemployment was nearly unchanged. No GDP, consumption, business-activity, broader employment, or realized inflation release was supplied.

The supplied model-implied Fed-path measure was unchanged: the Kim-Wright fitted instantaneous forward rate one year ahead remained at 4.1852%. Firmer claims and higher 2-year yields may be consistent with increased policy sensitivity, but they do not establish a change in the expected Federal Reserve path.

Treasury projected $671 billion of privately held net marketable borrowing for July through September. Its May refunding guidance anticipated stable nominal coupon and FRN auction sizes for at least several quarters, while financing seasonal and unexpected needs through bills and cash-management bills. The same guidance anticipated incremental July increases in bill offering sizes.

Known July 27–29 offerings total $479 billion. They include $92 billion of 13-week bills, $79 billion of 26-week bills, $69 billion of 2-year securities, $70 billion of 5-year notes, $95 billion of 6-week bills, $44 billion of 7-year notes, and an additional $30 billion 2-year security. The 2-, 5-, and 7-year offerings total $213 billion. Offering amounts were unavailable for some later bill auctions in the forward window.

Geopolitical and Cross-Market Evidence

Geopolitical evidence is unavailable. No qualifying high-trust geopolitical Treasury-transmission item was supplied, and no direct safe-haven, cross-asset risk, Treasury fund-flow, or foreign-custody evidence was provided.

Higher yields across the curve mean that safe-haven buying did not dominate the observed weekly price action. They do not establish the absence, size, or direction of underlying safe-haven demand.

Current Base Case

The base case is a volatile, supply-sensitive market in which the front and intermediate sectors remain relatively weaker than the long end. The expected result is continued modest bear-flattening pressure while the curve remains positively sloped.

This case is conditioned on mixed rather than uniformly strong auction absorption, elevated real yields, and long-duration volatility that remains high. Known July 27–29 supply is concentrated materially in bills and short-to-intermediate notes. The prior week's auction evidence was mixed, with the weak 8-week result demonstrating that demand can be uneven.

The case does not require a further rise in the supplied expected-policy or model term-premium measures. Those measures were unchanged in the available observations and were dated July 17. A single firmer claims release is also insufficient to establish a persistent shift in the policy path.

Bullish-Duration Scenario

A duration rebound would require evidence that real-yield pressure is reversing alongside softer labor conditions and broadly stronger auction absorption. Relevant confirmation would include stronger exact-series bid-to-cover and indirect participation across several auctions, consistently below-average dealer awards, and a decline in VIXTLT.

A policy-sensitive rally led by the 2- and 5-year sectors would favor bull-steepening in 2s10s and 5s30s, while allowing outright long-duration yields to decline. This is a lower-confidence alternative scenario, assessed at 0.22.

Continued firm labor evidence, persistently high real yields, or broadening dealer absorption would invalidate this bullish-duration scenario.

Bearish-Duration Scenario

A broader duration selloff would require weak demand across several upcoming auctions, materially higher dealer absorption beyond the localized 8-week and 20-year results, continued increases in real yields, and persistent firm labor evidence. For the selloff to become long-end-led, long-horizon inflation compensation or the supplied model term-premium estimates would need to rise.

Without a long-end extension in inflation or term-premium compensation, pressure concentrated in front and intermediate supply would preserve the current bear-flattening pattern. With such an extension, the curve could shift toward bear-steepening. This is a lower-confidence alternative scenario, assessed at 0.23.

Strong auction absorption, lower dealer awards, and stable or declining long-horizon inflation and term-premium measures would invalidate the broad bearish-duration scenario.

Thesis Risks and Invalidation

The base case would be weakened by broad strengthening in upcoming auctions: high exact-series bid-to-cover and indirect shares alongside consistently below-average dealer awards would challenge the assumption that supply is a material near-term restraint.

A reversal lower in real yields accompanied by weaker labor evidence would undermine the front/intermediate-led bearish component. A sustained long-end-led selloff accompanied by rising model term-premium estimates or broader long-horizon inflation compensation would invalidate the bear-flattening framework in favor of bear-steepening.

A material decline in VIXTLT together with stable auction absorption would weaken the expectation of elevated and uneven price action. Conversely, current evidence also resists a uniformly bearish reading: five-year breakevens declined, the 13-week bill auction was strong, and supplied forward-rate and term-premium measures were unchanged.

Next-Week Catalysts

July 27: auctions of $92 billion in 13-week bills, $79 billion in 26-week bills, $69 billion in 2-year securities, and $70 billion in 5-year notes.

July 28: auctions of $95 billion in 6-week bills and $44 billion in 7-year notes. July 29: an auction of a $30 billion 2-year security and a scheduled 17-week bill auction for which the offering amount was unavailable.

July 30: scheduled 4- and 8-week bill auctions, with offering amounts unavailable in the supplied forward records. The 8-week result is especially relevant following the prior week's weak exact-series outcome. The July 30 unemployment-claims release will test whether the decline in initial claims persists.

Evidence and Source Notes

Evidence consists of supplied official Treasury daily par-yield, quarterly-refunding, auction-result, and upcoming-auction records; supplied DOL unemployment-claims data; and supplied approved FRED and Cboe series. Auction comparisons use the exact-series samples stated in each record; the 20-year reopening and 10-year TIPS records have insufficient respective histories for statistical ranking.

Breakevens are inflation-compensation measures that include inflation-risk and liquidity premia rather than pure inflation forecasts. Kim-Wright forward-rate and term-premium observations are model-derived, dated July 17, and subject to model specification and revision risk. VIXTLT measures options on TLT, not whole-curve cash-market liquidity.

No direct evidence was supplied for cash-Treasury bid-ask spreads, market depth, price impact, repo conditions, dealer inventory, foreign custody, investor fund flows, or geopolitical transmission.

Methodology and Disclosure

This report uses only the supplied assessment, thesis, and evidence records, with a data cutoff of 2026-07-24T00:00:00.000Z. It does not incorporate unsupplied information. Yield-curve descriptions use supplied weekly changes and closing spreads; auction interpretations retain the supplied exact-series comparison limits.

The report was prepared with AI assistance and subject to manual review. Source data may be revised, including model-derived series, and conclusions are consequently conditional on the supplied records and their stated timing and limitations.

This material is informational only and is not investment advice, a recommendation, or a solicitation to buy or sell securities.