📈 Analysis Of Rising Bond Yields
🏦 Weakening Jobs And Softening Inflation Reduce Rate Hike Odds
💼 Key Takeaways From This Week’s Earnings
📊 Q2 Earnings Tracker
💾 Micron Earnings Show That Memory Shortage Has No End Date
QUOTE OF THE WEEK:
“Never before in history has the S&P 500 generated 13.6% profit margins. I think that's a return on AI, and again it's because so many of these companies are using AI in their businesses to be more productive and, as a result, they are more profitable.” - Portfolio Manager of the Bullseye American Ingenuity Fund
KEY US ECONOMIC EVENTS NEXT WEEK:

MARKET CLOSE:

WEEKLY MARKET WRAP:
Good Afternoon. Mixed week. Nasdaq led at +0.45% and closed near all-time highs at 27,190; the S&P was modestly red at −0.27%, and the Dow lagged at −1.26%. Monday through Thursday, all three indexes drifted lower on continued bond market pressure and government shutdown risk going into Oct. 1, with the Dow hitting a weekly low of 50,547 Thursday morning. Friday brought a strong tech-led rally, with Nasdaq up 1.19% on the day. YTD, the Nasdaq is up 17.00%, the S&P 12.84%, and the Dow 6.48%. More on rising yields in the curated insights and analysis section below.
Growth is strong, but hiring is cracking. GDP came in hot at 2.2%, well above the 1.5% forecast. But the labor market softened sharply — only 29K jobs added versus 89K expected, unemployment ticked up to 4.2%, and wage growth cooled to 0.1% m/m. Core PCE came in at 0.2%, below forecast, giving the Fed some cover. The combination of weakening jobs and softening inflation gives the Fed room to pause after its September hike. CME FedWatch now prices an 80% probability of a hold at the October 28 FOMC, with just 20% odds of another hike — down sharply from ~71% hike odds before Friday's jobs print.For the week:

CNN's Fear & Greed Index now stands at 31 (Fear) out of 100, down 6 points from last week. Details here
The top five trending stocks on Reddit are SPY, QQQ, Micron, SGOV, and DTE Energy. Read More
Liquidity:
Banking Reserves + ON RRP: Banking reserves remain at approximately $2.8 trillion. ON RRP balance remains immaterial.
Standing Repo Operations: The New York Fed’s standing repo operation (primarily reflecting SRF take-up) is $0.
Here is a summary of this week’s key economic releases:

Target Rate Probabilities for October 28th FOMC Meeting:

CURATED INSIGHTS & ANALYSIS:
A 5% 10-year is historically normal. Today's Debt Load Is Not:
The 10-year Treasury yield averaged 4.99% in September, its highest monthly average since 2007. The number sounds alarming. History suggests the level is closer to normal than the last 15 years made it feel. What has changed is why yields are rising, and the size of the debt they apply to.
The level is not unusual. From 1962 to 2007, the 10-year averaged 7.0%. From 1998 to 2007, it averaged 4.9%. Today's yield sits close to that pre-crisis norm.
The low-rate era was the exception. From 2009 to 2021, the 10-year averaged 2.3%. That period had long stretches of near-zero policy rates, large-scale bond buying, and low inflation.
Inflation compensation is normal. Real yields are not. A bond's yield has two parts: compensation for expected inflation, and the real return investors earn on top of it. The inflation part is 2.35% today, almost exactly where it was in 2003–2007. The real return part is 2.64%, up from 2.06% in 2003–2007 and 1.75% a year ago. Over the past year, nearly all of the 10-year rise came from higher real yields, while breakeven inflation was essentially unchanged.

Long yields rose even as short rates fell. From September 2024 to September 2026, the effective fed funds rate fell about 1.4 points, even after the Fed's 0.25-point increase on September 16. Over the same period, the 10-year rose about 1.3 points. The long end reflects more than today's policy rate. It likely includes expectations for future short rates and a higher term premium, the extra pay investors want to lock up money for ten years.

The term premium has risen modestly. Estimates differ by model. The NY Fed's ACM model read 0.89% at the end of September, up 0.28 points from a year earlier; on September averages it rose only about 0.1 point. The Fed's Kim-Wright model averaged about 0.94% in September, up roughly 0.5 points. Part of the rise in real yields therefore likely reflects higher expected policy rates, not term premium alone.
The cost is different this time. Federal debt held by the public is about 100% of GDP, against roughly 35% in 2007. CBO projects net interest above $1 trillion in fiscal 2026, about 3.3% of GDP.
Is the term premium too low?
It depends on which past you compare it with. The ACM term premium is close to its 2004–2007 average of about 0.9%. It sits well below its 1992–2007 average of 1.64%, which reflects the 1990s. Inflation risk was real then, and bonds often fell with stocks.
Some 1990s conditions are back:
Stocks and bonds have moved together recently. The 1-year correlation is +0.33, against an average of −0.37 in 2016–2019.
Deficits are near 6% of GDP at low unemployment.
Foreign investors hold 29.7% of federal debt held by the public, down from about 45% in the 2010s.
Under its September directive, the Fed reinvests into bills, which offers limited support for long bonds.
Others are not:
Breakeven inflation is stable near 2.35%, and the Fed has had a formal 2% target since 2012.
The foreign share is close to its 1992–2007 average of 31%.
The Fed's total Treasury holdings rose about $355 billion, comparing September averages.
Long Treasuries often provide a refuge in growth-driven sell-offs.

With rate expectations held steady, a term premium at its 2004–2007 level leaves the 10-year near 5.0%. At its 1992–2007 level, the 10-year would sit near 5.7%. These are simplified sensitivity cases, not forecasts.
In short, a 5% 10-year is historically familiar in nominal terms. Over the past year, nearly all of the increase came from higher market-implied real yields, while breakeven inflation was essentially unchanged. With federal debt near 100% of GDP, the same rate now carries a much higher fiscal cost than before 2008. Whether yields climb further depends on how many 1990s conditions return. The early signs, larger deficits and weaker stock-bond diversification, suggest the balance of risk is tilted upward.
Sources: Federal Reserve H.15 and H.4.1, Federal Reserve Board (Kim-Wright), NY Fed ACM, BLS CPI-U, CBO. Yields are September 2026 monthly averages; the ACM term premium is the end-of-September reading; foreign share as of January 2026. Simplified sensitivities hold the expected rate path constant.
Q2 earnings tracker:
Q2 is effectively closed: four more Russell 3000 companies reported this week, leaving all three universes at 99% complete, with no change to any headline figure. The median S&P 500 constituent grew GAAP net income 14.2% year over year, with 69% of reporters improving, while the Russell 3000's median of 17.3% again sat above it on the weakest participation of the three at 64%.
Key takeaways from this week’s earnings:
The memory shortage no longer has an end date
Micron said more than 75% of its next-year shipments are already committed and that it has no line of sight to when supply and demand balance. It expects 2028 to be tighter than 2027. Most of its added construction spending is for clean rooms that open in late 2028 or later. The company can no longer say when the shortage will end.
AI agents are boosting demand for ordinary servers
Until now, most AI spending went to specialized AI chips. Micron said AI agents, software that carries out tasks on its own, run mostly on ordinary servers instead. That is pushing server shipments up in the high teens, and every server needs memory. Micron said memory is now the main bottleneck in data centers. Jabil expects its data-center revenue to grow about 52% next year.
AI is adding work for consultants, not yet removing it
Accenture grew 7% last quarter, booked $22.2 billion of new work, and started advanced AI projects with more than 400 clients this year. Its headcount rose 5% while revenue per employee also increased. But next year’s organic growth guide is only about 1–4%, so AI work has not yet lifted its overall growth.
Shoppers are choosing trips over things
Carnival is half booked for next year at record prices, and its chief executive called vacations “sacrosanct.” Nike guided full-year revenue down high single digits and said the consumer is cautious. McCormick said shoppers are using what is already in their pantry and repurposing leftovers as gas prices stay high. Spending is holding up for experiences while everyday goods soften.
Food cost inflation is still rising
McCormick raised its cost inflation outlook to 6–7% from mid-single digits, citing input, freight and other costs. A week earlier, General Mills said its costs would rise about 6% by its fourth quarter. Carnival offset higher fuel prices with $150 million in efficiency gains. Fuel and freight costs are now showing up in food companies’ outlooks.
FRONT PAGES:
OpenAI Unveils Dots and GPT-6.1 Sol: OpenAI used its Monday DevDay keynote to launch Dots, a set of always-on autonomous agents, and GPT-6.1 Sol, a lower-cost reasoning model with per-task compute budgeting. The company also rolled out Codex Cloud for background coding tasks, an Agents API, and an app-like plugin store that moves ChatGPT closer to Microsoft 365's territory. OpenAI is separately reported to be in talks to raise $30 billion at a $1.4 trillion valuation.
Hedge Funds Rebuild Yen Shorts: Hedge funds boosted bearish bets against the yen for a third straight week, rebuilding short positions that were cut after August's US-Japan joint intervention. CFTC data showed speculators adding net shorts as USD/JPY weakness extended and the US-Japan 2-year yield spread continued to widen. The move revives the carry-unwind risk that briefly roiled markets in August and keeps intervention watch back on the table.
Tesla Delivers Record Q3 Vehicles: Tesla reported 486,532 Q3 deliveries, beating the 461,974 consensus and marking its best quarter on record. Energy storage deployments hit 13.7 GWh, its second-strongest quarter in that segment. The beat eases pressure after two softer quarters, though some demand was likely pulled forward ahead of the US EV tax credit expiration at year-end. Tesla also secured $30 billion in new credit lines to fund scaling Cybercab and Optimus.
Lyft Pays $272.5M in Driver Case: Lyft agreed to pay $272.5 million to settle a long-running California lawsuit over driver misclassification, the largest payout of its kind in the state. The settlement resolves claims dating to the pre-Proposition 22 period and covers back wages, expense reimbursement, and penalties. Uber faces similar exposure. The ruling validates California's enforcement posture but leaves the broader gig-classification question unresolved at the federal level.
Apple Pay Launches in India: Apple Pay went live in India on Monday through a partnership with Axis Bank, ending years of regulatory stalemate and giving the country's roughly 25 million iPhone users tap-to-pay functionality for the first time. The launch stops short of UPI integration but signals Apple's willingness to work within India's RBI-mandated local processing rules, opening the door for additional bank partners.
EARNINGS UPDATE:

Micron. EPS $33.42 versus $31.77; revenue $54.2B versus $51.3B, both beats. More than 75% of next year’s shipments are already committed, and it raised high-bandwidth memory prices significantly for 2027. Management said it has no line of sight to when supply catches up with demand. Shares rose about 3%. Demand is booked; supply is the question.
Jabil. Core EPS $4.40 versus $4.07; revenue $10.6B, up 29% and a beat. AI infrastructure revenue grew 56%, and the company expects to add more than $8.5B of revenue next year. Its first-quarter earnings guide of $3.80–4.20 sits below this quarter’s $4.40. Shares fell about 10%.
Accenture. EPS $3.29 versus $3.18; revenue $18.7B, up 7% in local currency and a beat, with $22.2B of new bookings. Next year’s growth guide is 3–6%, with 2–2.5 points coming from acquisitions. Shares jumped about 16%, then gave back 6% the next day.
Nike. EPS $0.48 versus $0.43; revenue $11.2B, down 4% and a slight miss. Greater China fell 26%, and Jordan fell in the mid-teens, while performance products grew in the high single digits. It guided full-year revenue down high single digits and announced a $2.5B cost program. Shares fell about 4%.
Carnival. EPS $1.43 versus $1.35; revenue $8.4B and a beat, with yields up about 2.5%. Next year is already half booked at record occupancy and pricing, and customer deposits hit a record $7.6B. Shares jumped about 13%.
McCormick. EPS $0.86 versus $0.76; revenue $2.02B and a beat, helped by its Mexico acquisition; organic growth was 2%. It raised its cost inflation outlook to 6–7% from mid-single digits. Shares fell about 5%.
EARNINGS PREVIEW:
Date | Symbol | Name | Time |
6-Oct | RPM | RPM International | Before Open |
6-Oct | STZ | Constellation Brands | After Close |
8-Oct | PEP | PepsiCo Inc | Before Open |
9-Oct | DAL | Delta Air Lines | Before Open |
VIDEO’s OF THE WEEK:
Blu Dot surpasses 2,000% ROAS with self-serve CTV ads
Home furniture brand Blu Dot blew up on CTV with help from Roku Ads Manager. Here’s how:
After a test campaign reached 211,000 households and achieved 1,010% ROAS, the brand went all in to promote its annual sales event. It removed age and income constraints to expand reach and shifted budget to custom audiences and retargeting, where intent was strongest.
The results speak for themselves. As Blu Dot increased their investment by 10x, ROAS jumped to 2,308% and more page-view conversions surpassed 50,000.
“For CTV campaigns, Roku has been a top performer,” said Claire Folkestad, Paid Media Strategist, Blu Dot. “Comping to our other platforms, we have seen really strong ROAS… and highly efficient CPMs, lower than any other CTV partner we've worked with.”
Using Roku Ads Manager, the campaign moved from a pilot to a permanent performance engine for the brand.
Please Share This Newsletter With Your Friends.
Also, check my blog here.
This newsletter's content is for informational and educational purposes only and should not be considered trading or investment recommendations. All the opinions in this newsletter are personal and do not belong to any organization.




