US Treasuries: market analysis and auction data

The Treasury market moved into a sharply steeper and more volatile regime: the 2-year yield fell 5 basis points while 10-, 20-, and 30-year yields rose 6, 10, and 11 basis points. The central case is continued volatile steep-curve trading, with relative front-end support but persistent long-end sensitivity to financing guidance and upcoming coupon supply. Available evidence does not establish that supply caused the long-end selloff. Current Treasury Regime The regime is sharp curve steepening with a front-end rally, long-end selloff, rising long-duration implied volatility, and a heavy supply backdrop. The 2-year yield ended at 4.28%, down 5 basis points on the week, while the 5-, 10-, 20-, and 30-year yields ended at 4.45%, 4.75%, 5.28%, and 5.27%, respectively, up 2, 6, 10, and 11 basis points. The 2s10s, 5s30s, and 10s30s slopes ended at positive 47, 82, and 52 basis points. Long-duration implied volatility increased materially: VIXTLT rose from 71.94 to 85.9. This indicates greater option-implied uncertainty for long-duration Treasury ETF exposure, but it is not a whole-curve volatility or cash-market liquidity measure. Executive Summary The week produced two distinct curve moves: front-end bull-steepening from the 2-year rally and long-end bear-steepening from the rise in 10-, 20-, and 30-year yields. The evidence supports relative front-end strength and long-end vulnerability, but does not isolate a single cause for either move.…

Central case: The base case is continued volatile, steep-curve trading in which the 2-year sector remains better supported than long duration, while the long end remains sensitive to financing guidance and upcoming coupon supply. The unusually large rise in VIXTLT, positive but modest growth evidence, mixed labor signals, and adequate but heterogeneous auction absorption argue against assuming an orderly directional move [cmsazpkm200xrqf01xqc7u846, cmsazpkm200xsqf01n9821jf1, cmsazpkm200xvqf01wedq4z9i, cmsazpkm100x0qf01pw57mzc4, cmsazpkm100x2qf01h69irlx4, cmsazpkm200x7qf01dgutkw7d]. The August 5 quarterly refunding announcement and subsequent 3-, 10-, and 30-year auctions will test whether the long-end underperformance persists. This case does not assume that supply caused the latest selloff or that the stale term-premium estimates have already risen.

• The prior week’s broad selloff and modest bear-flattening reversed into pronounced steepening. The 2-year yield fell 5 basis points, while the 5-, 10-, 20-, and 30-year yields rose 2, 6, 10, and 11 basis points, respectively. Consequently, 2s10s widened from 36 to 47 basis points, 5s30s from 73 to 82 basis points, and 10s30s from 47 to 52 basis points [cmsazpkm200x8qf011gijz35r, cmsazpkm200x9qf01ogw8ijww, cmsazpkm200xaqf016j8gxcft, cmsazpkm200xbqf01yy5qk83d, cmsazpkm200xcqf01rfs4pitp, cmsazpkm200xdqf01eeq5y9ac, cmsazpkm200xeqf017vbol0d3, cmsazpkm200xfqf017wwv577p]. Real-yield changes were similarly divided: 5- and 10-year real yields declined 5 and 2 basis points, while 20- and 30-year real yields rose 2 and 3 basis points [cmsazpkm200xgqf01i89c2v69, cmsazpkm200xhqf01cjbj376p, cmsazpkm200xiqf01o6hvj188, cmsazpkm200xjqf01701nkhdq]. Five-year, 10-year, and 5y5y inflation compensation each increased 2 basis points, and VIXTLT rose sharply from 71.94 to 85.9 [cmsazpkm200xkqf01dd01td6k, cmsazpkm200xlqf011rbdvcgt, cmsazpkm200xmqf01p9wg50tk, cmsazpkm200xrqf01xqc7u846].
• Auction demand was adequate overall but mixed across maturities and measures. The 13-week bill was the strongest exact-series result: its 3.06 bid-to-cover was at the 100th percentile of seven preceding auctions and 1.82 standard deviations above average; indirect awards were 61.01%, a 2.25 z-score, while dealer awards were 24.83%, a -2.49 z-score [cmsazpkm100x0qf01pw57mzc4]. The 26-week bill also had a 100th-percentile bid-to-cover of 3.12 and a 1.44 z-score, but indirect participation was near average at a -0.09 z-score; dealer awards were below average at a -0.50 z-score and direct awards were 9.65% versus a 6.80% exact-series average [cmsazpkm100x2qf01h69irlx4]. The 4-week bill was ordinary to moderately firm: its 2.80 bid-to-cover ranked at the 63rd percentile of eight comparisons, indirect awards had a 0.63 z-score, and dealer awards a -0.21 z-score [cmsazpkm100x6qf01kef8tj3z]. The 8-week bill normalized after the prior week’s weakness: its 2.74 bid-to-cover ranked at the 38th percentile, while indirect and dealer shares were close to their eight-auction averages at 0.11 and -0.05 z-scores [cmsazpkm200x7qf01dgutkw7d]. Coupon histories were limited to one comparison. The 2-year note’s 2.66 bid-to-cover exceeded 2.64, with higher indirect and lower dealer shares; the 5-year note’s 2.28 was below 2.35, with a somewhat higher dealer share; and the 7-year note’s 2.49 was near 2.50, with indirect awards rising to 63.04% from 49.98% [cmsazpkm100x1qf01cct0f2vj, cmsazpkm100x3qf1
• The curve steepened through two distinct mechanisms: the 2-year rally generated front-end bull-steepening, while the 10-, 20-, and 30-year selloff generated long-end bear-steepening. Weekly changes widened 2s10s by 11 basis points, 5s30s by 9 basis points, and 10s30s by 5 basis points [cmsazpkm200x8qf011gijz35r, cmsazpkm200x9qf01ogw8ijww, cmsazpkm200xaqf016j8gxcft, cmsazpkm200xbqf01yy5qk83d, cmsazpkm200xcqf01rfs4pitp, cmsazpkm200xdqf01eeq5y9ac, cmsazpkm200xeqf017vbol0d3, cmsazpkm200xfqf017wwv577p]. Falling 5- and 10-year real yields are consistent with front and belly support, while rising 20- and 30-year real yields and modestly higher inflation compensation align with greater long-end pressure [cmsazpkm200xgqf01i89c2v69, cmsazpkm200xhqf01cjbj376p, cmsazpkm200xiqf01o6hvj188, cmsazpkm200xjqf01701nkhdq, cmsazpkm200xkqf01dd01td6k, cmsazpkm200xlqf011rbdvcgt, cmsazpkm200xmqf01p9wg50tk]. However, the available term-premium estimates are stale and unchanged, so the residual long-end move cannot be assigned conclusively to term premium, supply, or inflation risk.

Key uncertainty: A broad duration rally led by falling 20- and 30-year real yields, softer inflation compensation, and materially lower VIXTLT would invalidate the expectation of persistent long-end vulnerability.