US sells USD 25bln of 30-year bonds; Tail 0.4bps

• Tail: 0.4bps (prev. 0.1bps, six-auction avg. -0.1bps)
• High Yield: 5.441% (prev. 5.618%, six-auction avg. 5.211%); WI 5.437%
• B/C: 2.36x (prev. 2.54x, six-auction avg. 2.44x)
• Dealer: 11.39% (prev. 6.79%, six-auction avg. 9.49%)
• Direct: 20.5% (prev. 20.89%, six-auction avg. 20.02%)
• Indirect: 68.11% (prev. 72.32%, six-auction avg. 70.49%)

For a USD 25bln 30-year sale, a 0.4bps tail against the WI is a mildly soft result rather than a failed one, but the demand mix is weaker than the stop-out alone: cover at 2.36x is below both the prior auction and the recent average, indirects at 68.11% are lighter than usual, and dealers at 11.39% are left with more paper than in the prior sale and above the recent norm. Long-bond auctions historically carry a wider tail distribution than bills and front-end coupons, so small positive tails are common when the market is concession-led; the useful distinction is between a lower clearing level, with the high yield down from the previous auction, and the quality of sponsorship, which here looks thinner. In comparable episodes, a modest tail plus heavier dealer retention tends to matter less for the first move than for the next session, when intermediation, futures basis and hedge flows decide whether the cheapening is absorbed or pressed into the 10s/30s leg and swap spreads. A repeat pattern across the next coupon supply, or signs that foreign-style demand proxies remain soft while dealer takedown rises, is the tell that concession is being built rather than cleared. The immediate reference points are the post-auction behaviour of the long end versus the belly, indirect and dealer shares in the following long-dated sale, and any change in coupon and bill mix signalled around refunding rather than one auction in isolation.