Thu-Fri: claims at 197,000, a middling 30-year auction, and a plan with nothing to grade.
Written Tuesday October 13, covering Thursday October 8 and Friday October 9. Thursday's plan was committed before the 8:30 AM data and is graded here: initial claims printed 197,000 (DOL) and the $22 billion 30-year reopening drew a 2.54 bid-to-cover and 72.3 percent indirects (Treasury Fiscal Data), both inside the plan's no-call bands, so nothing triggered. The 10-year fell 4.6 basis points to 5.231 percent anyway, almost all in real yields. Friday, reported not graded: Michigan sentiment fell to 46.3 with year-ahead inflation expectations at 4.7 percent, and gold rose about $58.
Written Tuesday October 13, covering Thursday October 8 and Friday October 9. Thursday's plan, graded below, was committed before Thursday's 8:30 AM ET data; no plan was registered for Friday, so Friday is reported, not graded. Thursday's plan had two scheduled tests, and both came in between the lines it drew. Initial jobless claims printed 197,000, inside the 191,000 to 214,000 band the plan defined as "neither." The $22 billion 30-year reopening drew a 2.54 bid-to-cover and a 72.3 percent indirect share, neither weak nor strong by the plan's definitions. Nothing triggered, so nothing is graded as a hit or a miss. The market moved anyway: the 10-year fell 4.6 basis points to 5.231 percent and the 30-year 5.5 basis points to 5.606 percent, the dollar slipped and gold rose. That is the direction the plan's "high claims" row would have called, on a claims number that pointed slightly the other way.
This continues Wednesday's note on the hawkish minutes and the strong 10-year auction, which ended with two conditional statements about Thursday. Both are scored below.
Thursday: what the data said
- Initial jobless claims (8:30 AM ET), week ending October 3: 197,000, down 2,000 from the prior week, which was revised up from 197,000 to 199,000. The four-week average fell 2,500 to 198,000. Insured unemployment (continuing claims) for the week ending September 26 rose 17,000 to 1,716,000; the insured unemployment rate held at 1.1 percent (source: Department of Labor, Unemployment Insurance Weekly Claims, October 8). The plan's consensus reference was 200,000 (Investing.com). For what a weekly claims number can and cannot carry, see our guide to weekly jobless claims.
- Wholesale trade (10:00 AM ET), August. Inventories rose 0.5 percent to $964.2 billion, below the 0.7 percent advance estimate; sales rose 1.8 percent to $817.5 billion; the inventories-to-sales ratio was 1.18, against 1.28 a year earlier (source: Census, Monthly Wholesale Trade, August 2026). No call was made on it.
- EIA natural gas storage (10:30 AM ET), week ending October 2: a net injection of 85 Bcf to 3,500 Bcf, 68 Bcf above the five-year average and 130 Bcf below a year earlier (source: EIA Weekly Natural Gas Storage Report). Trading Economics put the preliminary estimate at 79 Bcf. No call.
- Fed speakers. Governor Waller, speaking at 4:30 AM ET in Istanbul, said more rate increases are likely needed to bring inflation back to 2 percent but that they need not come at consecutive meetings (sources: investingLive, Daily Sabah). St. Louis Fed President Musalem, at a Bloomberg event in New York, said rates "ought to be going up further in an appropriate period of time in the next six to nine months" and declined to say what should happen at the October 27-28 meeting (Reuters, as carried by WMBD). Both are hawkish in direction and both left the October meeting open. The plan made no call on either.
Thursday: the 30-year reopening
Treasury sold $22 billion of the 30-year bond (CUSIP 912810UW6, a reopening) at a high yield of 5.618 percent, 31 basis points above September's 5.308 percent. Bid-to-cover was 2.54, down from 2.61. Indirect bidders took 72.3 percent of the competitive amount accepted ($15.87 billion of $21.94 billion), down from 79.5 percent in September; direct bidders 20.9 percent; primary dealers 6.8 percent, up from 2.2 percent (source: Treasury Fiscal Data, auctions query). The when-issued yield at 1:00 PM was 5.617 percent, so the auction tailed by 0.1 basis point (source: investingLive; Treasury does not publish the WI level). By the standards in our guide to auction demand, that is an unremarkable sale: priced on the screen, softer than September's on every share measure, but not a failure.
The 30-year had already done its moving. It fell from about 5.656 percent at 11:50 AM to about 5.616 at 12:50 PM, then moved about 0.2 basis points in the half hour after the auction (Yahoo ^TYX 30-minute bars, which sit on :20 and :50).
Thursday: the tape
| Market | Wed Oct 7 reference | Thu Oct 8 close | Change |
|---|---|---|---|
| 10Y yield | 5.277% | 5.231% | -4.6 bp |
| 30Y yield | 5.661% | 5.606% | -5.5 bp |
| 5Y yield | 5.021% | 4.991% | -3.0 bp |
| DXY | 102.25 | 102.14 | -0.11 |
| Gold (GC=F, 4 PM ET) | ~4,137 | ~4,158 | ~+21 |
| USD/JPY (4 PM ET) | 158.05 | 157.86 | -0.19 |
| EUR/USD (4 PM ET) | 1.1201 | 1.1215 | +0.0014 |
| Brent (BZ=F, 4 PM ET) | ~100.95 | ~103.92 | ~+2.97 |
| S&P 500 | 7,801.77 | 7,765.36 | -0.47% |
Sources: Yahoo Finance query1 daily closes for ^TNX, ^TYX, ^FVX, DX-Y.NYB and ^GSPC; Yahoo hourly 4:00 PM ET bar for GC=F, USDJPY=X, EURUSD=X and BZ=F, the convention set in the plan. The DXY reference is the plan's 102.25; Yahoo now shows Wednesday's daily close as 102.24, which changes nothing below.
Treasury's official curve tells the same story with one extra piece: the 10-year fell from 5.28 to 5.22 percent, the 10-year TIPS yield from 2.92 to 2.87 and the breakeven from 2.36 to 2.35 (sources: FRED DGS10, DFII10, T10YIE). Almost all of the drop was in real yields, the same component that carried the rise earlier in the week.
Brent is the outlier. It settled at $104.28 on Thursday, up from $100.20, about 4 percent (Yahoo BZ=F daily), and it was already up about 3.3 percent before 8:30 AM ET when the plan was written. We did not find a verified account of the cause and do not offer one. An oil rally of that size on a day when yields fell and breakevens barely moved is worth recording: the bond market did not treat it as an inflation input on the day.
Grading Thursday's plan
The plan, committed Thursday at about 4:50 AM ET, had two legs and a conflict rule, graded at the 4:00 PM ET close against Wednesday's reference levels. Leg A classified initial claims: low at 190,000 or below (10-year up at least 3 basis points, DXY up, gold down, USD/JPY up), high at 215,000 or above (the mirror image), no call in between. Leg B classified the 30-year reopening: weak if bid-to-cover was below 2.40 or indirects below 65 percent (30-year up at least 3 basis points), strong if bid-to-cover was at least 2.60 and indirects at least 77 percent (30-year at or below 5.661 percent), no call in between.
Classification: claims at 197,000 are neither. The auction, at 2.54 and 72.3 percent, is neither: above both weak bars, below both strong bars.
| Line in the map | Expected | Actual at close | Grade |
|---|---|---|---|
| Leg A low (claims ≤ 190k): 10Y, DXY, gold, USD/JPY | up 3 bp, up, down, up | claims 197,000 | Not triggered |
| Leg A high (claims ≥ 215k): 10Y, DXY, gold, USD/JPY | down 3 bp, down, up, down | claims 197,000 | Not triggered |
| Leg A neither (191k to 214k) | no call | met | No call |
| Leg B weak (cover < 2.40 or indirects < 65%): 30Y | up 3 bp or more | 2.54, 72.3% | Not triggered |
| Leg B strong (cover ≥ 2.60 and indirects ≥ 77%): 30Y | at or below 5.661% | 2.54, 72.3% | Not triggered |
| Leg B neither | no call | met | No call |
| Conflict rule | void rates if legs oppose | neither leg triggered | Not triggered |
| Brent noise note | record further oil moves before 1 PM | Brent ~104.25 at writing, ~103.84 at 12:30 PM bar open, ~103.92 at 4 PM | Recorded, not graded |
The tally is zero hits, zero misses: nothing triggered. Two things to say about that, plainly.
- A no-call day is not a success. The bands were set wide on purpose, so that only a clear surprise would count. That keeps the plan from scoring noise, and it also means a session like Thursday, with a 4.6 basis point move in the 10-year, produces no test at all. Had the high-claims row triggered, all four lines would have been hits at the close; had the low-claims row triggered, all four would have missed. The actual print, 197,000, was a shade firmer than consensus, which is the low-claims direction. The market went the other way, and we have no claim-based explanation for why.
- The 30-year strong row would have hit. The 30-year closed at 5.606 percent, below the 5.661 reference. But the auction was not strong by the plan's definition, and most of the 30-year's fall came in the hour before the auction, not after it. We do not credit it.
Event-window record (not graded). Claims window, 8:30 to 9:00 AM ET: DXY from 102.26 to 102.38, gold from about 4,155 to about 4,143, USD/JPY from 158.16 to 158.22 (Yahoo 30-minute bars), and the 10-year from 5.324 to 5.320 percent between the 8:20 and 8:50 bars. The first half hour after claims moved the dollar up and gold down, the low-claims direction, and the close reversed both. Auction window, 1:00 to 1:30 PM ET: the 30-year moved from 5.616 to 5.614 percent and the 10-year from 5.252 to 5.237 percent (12:50 and 1:20 bars), while DXY fell from 102.25 to 102.06 and USD/JPY from 158.01 to 157.57. That 1:00 PM dollar drop was the largest half-hour dollar move of the day. We could not tie it to a verified headline. For why these windows are reported, see our guide to event windows.
The thesis carry-forward. The asymmetry hypothesis (a hawkish input moves the 10-year up by more than a dovish input moves it down) gets no observation from Thursday: neither input was classified. For the record, the 10-year fell 4.6 basis points on the day and about 0.4 basis points in the claims window.
Wednesday's two conditional statements
- "If the 30-year reopening draws an indirect share of at least 77 percent and a bid-to-cover of at least 2.60 ... then Wednesday's 10-year result was part of a run of strong long-end demand ... Wrong if either figure falls short." Both fell short (72.3 percent, 2.54). Graded wrong. On this evidence, Wednesday's 10-year result was not the start of a run.
- "If Treasury's official October 7 curve shows the 10-year breakeven unchanged at about 2.36 again, then a fourth straight session has kept inflation compensation flat while nominal yields moved ... Wrong if the breakeven moved by 3 basis points or more." FRED's T10YIE for October 7 is 2.36, unchanged from October 2, 5 and 6. Graded right. It slipped to 2.35 on Thursday.
Friday: reported, not graded
No plan was registered for Friday, so nothing below is scored.
- University of Michigan consumer sentiment, October preliminary: 46.3, down from 48.1 in September. Current conditions fell to 44.7 from 50.9; expectations rose to 47.3 from 46.3. Year-ahead inflation expectations rose to 4.7 percent from 4.6 percent and long-run expectations to 3.5 percent from 3.4 percent (source: University of Michigan Surveys of Consumers). investingLive put consensus at 47.8. The final October reading is due Friday October 23 at 10:00 AM ET. For how this survey compares with other expectations measures, see our guide to inflation-expectations surveys.
- USDA WASDE (cross-asset context only). USDA raised its US corn yield estimate 2.7 bushels per acre to 181.2, against a Dow Jones survey average of 177.7, and put 2026-27 corn ending stocks at 1.849 billion bushels. December corn briefly traded limit-down and closed 20 1/2 cents lower at $4.79 3/4 (source: DTN; the 1.677 billion pre-report ending-stocks expectation is per Newsquawk). A large supply surprise in one commodity; we see no sign it reached rates or the dollar.
| Market | Thu Oct 8 | Fri Oct 9 | Change |
|---|---|---|---|
| 10Y yield | 5.231% | 5.244% | +1.3 bp |
| 30Y yield | 5.606% | 5.600% | -0.6 bp |
| 5Y yield | 4.991% | 5.021% | +3.0 bp |
| DXY | 102.14 | 102.21 | +0.07 |
| Gold (GC=F, 4 PM ET) | ~4,158 | ~4,216 | ~+58 |
| USD/JPY (4 PM ET) | 157.86 | 158.25 | +0.39 |
| Brent (BZ=F settle) | 104.28 | 104.72 | +0.44 |
| S&P 500 | 7,765.36 | 7,811.54 | +0.59% |
Sources: Yahoo Finance query1, same conventions as above; Brent here is the daily settle. On the official curve, the 10-year rose from 5.22 to 5.24 percent, the 10-year TIPS yield from 2.87 to 2.91 and the breakeven fell from 2.35 to 2.33 (FRED DGS10, DFII10, T10YIE). That is an odd pairing with a Michigan survey that showed higher inflation expectations: the bond market's own inflation compensation went down on the day consumers said theirs went up. Gold's $58 rise is the largest move in the table, and we have not found a verified cause for it.
What we do not know
- Why yields fell on Thursday. Claims were firm, two Fed officials leaned hawkish, and the 30-year auction was soft. None of those points down. Bloomberg's Thursday headline said lofty yields lured demand to the 30-year sale; that is a description of buying, not an explanation of it.
- What moved Brent 4 percent on Thursday. Not verified.
- Who stepped back from the 30-year. The indirect share fell 7 percentage points from September and dealers took three times the share. "Indirect" covers many kinds of buyer, so we cannot say whether it was foreign demand.
What would change the read
The next scheduled test is September CPI, Wednesday October 14 at 8:30 AM ET (source: BLS release schedule). A conditional map for it was committed to the repository on Tuesday evening, before the data. Monday and Tuesday are covered in the next note in this thread. One conditional statement from this piece:
- If the 10-year breakeven (FRED T10YIE) stays at or below 2.33 through Thursday October 15 while Brent holds above $100, then the bond market is still treating the oil rally as a growth or risk-premium story rather than an inflation one. Wrong if the breakeven rises 3 basis points or more over that span.
What to read alongside
- Bond market holidays: what changes when stocks trade and Treasuries don't (the reference piece published with this catch-up)
- Weekly jobless claims
- Treasury auction demand
- Event windows
Nothing on this site is investment advice or a recommendation to trade. Most retail traders lose money. Reaction maps published here are falsifiable hypotheses, graded in public, not signals.