The Fed Got Good News. The Ten-Year Didn't Care.

Housekeeping first. I owed you the scorecard for the August path on October 1. It went out a day late, and that's on me: Six Frozen Windows, Graded. This letter is about the week.

On Monday, CME FedWatch put the odds of another Fed hike on October 28 near 70%. Trump had rejected Iran's Hormuz proposal over the weekend (AP). On Monday oil jumped, yields rose, and gold dropped more than 3% in a day (Yahoo Finance).

After Friday's jobs report, the odds were 17%, down from 36% a week earlier (CNBC).

Three reports helped.

Tuesday: job openings fell to 7.08 million in August (BLS). A Reuters poll had expected about 7.2 million.

Wednesday: core PCE, the inflation number the Fed watches most, came in at 3.0% for the year to August (BEA). Forecasters expected 3.3%.

Friday: the economy added 29,000 jobs in September (BLS). The Dow Jones poll expected 84,000. Unemployment rose to 4.2%, and July was revised from a small gain to a loss of 10,000.

The inflation number needs a footnote, and it's a big one.

On the same day, the Bureau of Economic Analysis changed how it measures prices for three things: fees for managing investments, software, and legal services. It applied the new methods back to 2021 (BEA).

Take investment fees. The old price index reflected asset values and fee rates, so a rising stock market could show up as rising prices. BEA now estimates how much service was provided from employment and hours worked, then works out the price from revenue.

Here's what that did. July's core PCE was first reported at 3.3%. In the revised series, July and August both round to 3.0%. So the 0.3-point drop in the headlines isn't August cooling. It's the new ruler, applied to both months.

Before the release, RBC estimated the changes would take about 0.2 points off core inflation (RBC). Afterward, Capital Economics' Stephen Brown put it at 0.3 (Yahoo Finance).

I don't think the new method is a trick. The old way of pricing investment fees was odd. But a 3.0% that comes from re-measuring isn't the same as a 3.0% that comes from prices slowing down. On Tuesday, before the release, Fed Governor Michael Barr said he still didn't see a clear trend toward a timely return to 2% (Federal Reserve).

The same day also brought second-quarter growth of 2.2%, revised up from 1.5% (BEA). So on paper, in one week, growth got better, inflation got lower and jobs got weaker. For a Fed that hiked in September, that's about as friendly as a week gets.

The ten-year yield.

The two-year, which follows the Fed, did what you'd expect: 4.92% on Monday, 4.78% by Thursday (FRED). The ten-year was 5.17% last Friday and reached 5.29% on Wednesday, its highest daily reading since May 2002 (FRED). On Friday it dipped when the jobs number landed, then finished the day higher, near 5.28% (CNBC).

Credit didn't relax either. The extra yield investors ask for holding junk bonds went from 2.93% last Friday to 3.24% on Thursday (FRED).

Gold had another bad week. GLD lost 3.4%.

Stocks split. The S&P 500 ended the week down about 0.3%, even after Friday's rally. Semiconductors had a good week: SOXX gained 2.8%.

My read: the two-year followed the lower hike odds. The ten-year didn't. Oil and the Treasury's financing needs may have mattered more to long yields than the weak data did. Brent settled above $100 on Thursday, on reports of a third US carrier heading for the Middle East and China's fuel-export curbs (Reuters), then dipped under $100 on Friday morning.

Last week I said the boring rate story fit better than my interesting one. This week it fit again. Real yields rose, with the ten-year TIPS yield going from 2.83% to 2.93% by Wednesday (FRED), and gold fell with it.

The entry window. The September 17-18 test finished Intermediate, and I'm still holding the call. SPY ended September at 762.63 and this week at 769.64, fifty-five cents below the September 4 level of 770.19. It got above that level last week. Now it's back between that and the 754.79 shock line.

Bitcoin. Last week I set two lines: above $86,000 with inflows still coming would say demand is getting through, below $83,000 as the flows fade would say it isn't. No completed daily reading through Thursday crossed either line. It dipped under $83,000 on Monday and pushed above $86,000 on Friday, and came back both times. The ETF inflow streak reached nine sessions before a $149 million outflow on Wednesday (Farside). Friday's reading and ETF flows aren't final yet. No verdict.

Gold. It broke my $4,300 line last week and kept going. Weaker again. I hold GLD, and I'm not going to pretend this is fine.

Late September. The frozen path needed yields and credit to stop getting worse. They got worse. The grade is in the scorecard.

The next frozen window runs from October 3 to November 13. In my September 16 letter I gave it a 4/5 attention rating, with peak dates on October 2-4, October 24-30 and November 13. The Fed meets October 27-28, inside the second one. The rating says where attention clusters. It doesn't say which way prices go.

What would change my read. A ten-year back under 5.17% with hike odds still low would weaken it. A close above 5.29% with junk spreads still widening would strengthen my read that the pressure goes beyond the Fed.

Disclosure: I hold long positions in ether (since 2025) and in GLD. No bitcoin position.

Market data: ETF and index closes from Yahoo Finance chart data. Yields and the high-yield spread are FRED daily series (DGS10, DGS2, DFII10, BAMLH0A0HYM2), which run through Thursday. Bitcoin figures are CoinGecko UTC-day readings, not exchange closes.

Public forecast record and personal judgment, not personalized investment advice or an instruction to transact. Past samples and historical observations do not guarantee future performance.

This Free website mirror preserves the Perchery article with original content ID perchery-2026-10-02-fed-good-news-ten-year-didnt-care, dated 2026-10-02. The first-person interpretation remains Perchery's. This import does not convert the source into a Quant-native empirical audit or create or resolve a research ledger record.

Substack: https://perchery.substack.com/p/the-fed-got-good-news-the-ten-year

X: https://x.com/readperchery/status/2106200107238715767

Bluesky: https://bsky.app/profile/perchery.bsky.social/post/3mwwrly3frs2f

Original Substack publication: 2026-10-03T01:47:42.454Z. Information cutoff: 2026-10-02 US close (FRED series through 2026-10-01). Other market-data cutoffs and frozen-record wording remain as written. Inline link markup is adapted to plain text; original source URLs follow in order of appearance.

https://perchery.substack.com/p/six-frozen-windows-graded

https://apnews.com/article/c529a0a729600230559354b80c005bad

https://finance.yahoo.com/markets/commodities/articles/gold-falls-below-4-200-095210430.html

https://www.cnbc.com/2026/10/02/fed-rate-hike-odds-decline-after-september-jobs-report.html

https://www.bls.gov/news.release/jolts.nr0.htm

https://www.bea.gov/news/2026/personal-income-and-outlays-august-2026

https://www.bls.gov/news.release/empsit.nr0.htm

https://apps.bea.gov/scb/issues/2026/09-september/0926-nea-annual-update.htm

https://www.rbc.com/en/economics/us-analysis/us-featured-analysis/pce-deflator-update-a-retroactive-revision-not-a-disinflationary-signal/

https://finance.yahoo.com/economy/policy/article/pce-inflation-eases-to-34--cooling-the-case-for-another-fed-rate-hike-125513195.html

https://www.federalreserve.gov/newsevents/speech/barr20260929a.htm

https://www.bea.gov/news/2026/gdp-third-estimate-industries-corporate-profits-state-gdp-and-state-personal-income-2nd

https://fred.stlouisfed.org/series/DGS2

https://fred.stlouisfed.org/series/DGS10

https://www.cnbc.com/2026/10/02/treasury-yields-bonds-nonfarm-payrolls.html

https://fred.stlouisfed.org/series/BAMLH0A0HYM2

https://www.brecorder.com/news/amp/40442265

https://fred.stlouisfed.org/series/DFII10

https://farside.co.uk/btc/

Editorial research and personal judgment, not personalized investment advice or an instruction to transact.