Two companies reported almost the same cloud growth this week, and the market treated them like opposites. Microsoft rose on Azure, Alphabet suffered its worst day since its 2004 listing, and the difference came down to one number: free cash flow. That's why we built Winvesta Crisps, to cut through the noise and tell you what actually moved markets and why. 60,000+ investors from across India are already in. What about you?

Subscribe now!

🔔 Don't miss out!

Add winvestacrisps@substack.com to your email list so our updates never land in spam.

Big Tech's earnings week turned into a single test: show that the AI spending is paying for itself, or pay the price. Alphabet failed it and lost roughly $500 billion in market value in one session, Microsoft passed it and rose, and Amazon was marked down as much as 10% after hours on Thursday for burning cash. All of this landed on top of a Fed that held rates but saw two officials vote for a hike, and an inflation print that kept moving the wrong way.

📊 Market recap

Monday and Tuesday were the calm before the earnings flood. The S&P 500 rose 0.70% on Monday to 7,676.77 and added another 0.60% on Tuesday to 7,722.60, while the Nasdaq Composite gained 0.82% and then 1.18% to close Tuesday at 26,477.81, per Tapeboard's daily recaps. The VIX eased from the low-18s to around 17.5 over the two sessions. The one big casualty was Boeing, which fell close to 11% on Tuesday to around $168 after its Q3 results showed bigger-than-expected losses in commercial jets and a cut to its delivery target, per XTB and 24/7 Wall St, one of the stock's worst days on record.

Wednesday belonged to the Fed. The Federal Open Market Committee held its target range at 3.75% to 4.00% in a 10-2 vote, with Governor Lisa Cook and Cleveland Fed President Beth Hammack both voting for a quarter-point hike, per CNBC and CNN. Chair Kevin Warsh said politics played no role ahead of next week's midterm elections. Stocks shrugged and edged higher: the S&P 500 added 0.28% to 7,743.98, the Nasdaq gained 0.27% and the Dow rose 0.19% to the 52,400s, per Cappnotes. The ten-year Treasury yield, by contrast, climbed about 5 basis points to just above 5.0%, and CME FedWatch pricing for a December hike jumped to the low-to-mid 80s in percentage terms, per Yahoo Finance. Vertiv was the session's standout large cap, up about 16.5% after third-quarter sales rose 61% and its orders more than doubled.

Then the earnings landed after Wednesday's bell. Microsoft reported revenue of $92.3 billion, up 19%, with Azure growing 48%, and rose in late trading. Alphabet reported Google Cloud revenue more than doubling, but Search growth slowed to 12% from 17% and free cash flow turned negative under the weight of AI spending, per StockTitan and Yahoo Finance. Shares fell around 10% after hours.

Thursday was the reckoning. Alphabet closed down about 12.6%, its worst post-earnings reaction since listing in 2004, erasing roughly $500 billion in market value, per Yahoo Finance. Microsoft finished up roughly 2 to 3%, and Meta gained 2 to 3% midday. A September PCE print showing headline inflation at 3.6% year on year, up from 3.4%, and core at 3.1%, up from 3.0%, added pressure early, even though both matched forecasts. Q3 GDP came in at a strong 3.4% annualised pace. The Nasdaq fell about 1.4%, ending a three-day winning run, while the S&P 500 slipped about 0.3% to the low 7,700s and the Dow lost about 0.25% to the low 52,300s. Caterpillar fell as much as 12% intraday on a soft Q4 sales outlook, ServiceNow dropped about 25% on bookings worries, and Fortinet lost around 21% after cutting its service revenue forecast.

Read more

Five years ago, analysing Alphabet meant tracking search ad prices and how much cash Google would hand back through buybacks. Not anymore. This quarter Google Cloud doubled its revenue, Alphabet started selling its own AI chips to rivals such as Anthropic, and the company that used to buy back its stock is now issuing it to pay for data centres. That's why we built Winvesta Crisps - to decode what's actually driving the companies you own, in plain language, before the consensus catches up. 60,000+ investors from all over India are already in. What about you?

Subscribe now!

🔔 Don't miss out!

Add winvestacrisps@substack.com to your email list so our updates never land in spam.

Most investors still think of Alphabet as the world's biggest advertising machine, a company that prints money from Google Search and YouTube and returns the surplus to shareholders. That framing is accurate and dangerously incomplete. On 28 October Alphabet reported revenue up 25% to 127.6 billion dollars, a Google Cloud business that grew 103% in a single year, and negative free cash flow of 9.1 billion dollars, because it spent a record 51.4 billion dollars on capital expenditure in three months. The shares fell roughly 10% in after-hours trading. The market was not reacting to a weak business. It was reacting to a company changing what kind of business it is, and to the bill for that change.

🏭 What Alphabet actually is today

Alphabet still runs the Google most people use every day: Search, YouTube, Android, Chrome, Maps, Gmail and the Play Store, funded almost entirely by advertising and, increasingly, by subscriptions. That Google Services segment produced 96.3 billion dollars of revenue in the third quarter of 2026 and 39.2 billion dollars of operating income, per Alphabet's Q3 2026 earnings release. It remains one of the most profitable businesses ever built.

Sitting beside it is a second company that now looks very different from the one investors knew even two years ago. Google Cloud no longer only rents computing capacity and sells Workspace seats. It sells access to Google's Gemini models, it runs AI workloads for other model builders, and it now sells complete systems built on Google's own Tensor Processing Units, or TPUs, the custom AI chips Google designed for its own use a decade ago. On the earnings call, Google Cloud chief executive Thomas Kurian named Anthropic, Apple, Citadel Securities and Jump Trading among the customers buying TPU capacity.

Around both sits a capital programme that has turned Alphabet into one of the largest builders of physical infrastructure on the planet. India is part of that map: in October 2025 Google announced a 15 billion dollar AI hub in Visakhapatnam, its largest investment in the country. The Gemini app, which ties the consumer side to the infrastructure side, has crossed one billion users, per chief executive Sundar Pichai.

🧩 Segment breakdown: Cloud now carries the growth

The segment table makes the shift plain. Google Search and other revenue grew 12% to 63.4 billion dollars, slower than the 17% of the second quarter and below the roughly 15% analysts had expected, per Fortune's coverage of the results. YouTube advertising grew 9% to 11.2 billion dollars. Google Network, the ads Google places on other publishers' sites, was almost flat. Subscriptions, platforms and devices, which includes YouTube Premium, Google One and Pixel, grew 11% to 14.3 billion dollars. Google Cloud, by contrast, more than doubled, from 15.2 billion dollars to 30.7 billion dollars.

Cloud is now roughly 24% of Alphabet's revenue, and it is no longer a margin drag. Its operating income rose to 11.1 billion dollars, a 36.2% operating margin against 23.7% a year earlier, per the earnings release. Google Services still earns about 40% operating margins on a much larger base, so it remains the profit engine. Cloud is where nearly all of the growth now comes from: of the 25.2 billion dollar increase in Alphabet's total revenue this quarter, Cloud supplied about 15.6 billion dollars, by our calculation from the reported segment figures.

⚡ The core strategic bet: selling the picks and shovels of AI

The bet Alphabet is making is that the most valuable thing it owns is no longer the search box. It is the stack underneath it: custom chips, data centres, networking, frontier models and a cloud business that can sell all of it to anyone. Google Cloud ended the quarter with a contracted backlog of 561 billion dollars, of which management expects to recognise just over 55% as revenue over the next 24 months, per chief financial officer Anat Ashkenazi on the earnings call. For a business that booked 30.7 billion dollars of revenue this quarter, that backlog is an extraordinary amount of work already signed.

Read more

Most investors read a rate decision as a verdict on the economy: strong means hike, weak means cut. This week that rule breaks. Inflation is running at 3.7% because of petrol, hiring has nearly stalled, and the Fed has to pick a side tonight. That's why we built Winvesta Crisps, to break down what's actually moving markets, in plain language, before the consensus catches up. 60,000+ investors from all over India are already in. What about you?

Subscribe now!

🔔 Don't miss out! Add winvestacrisps@substack.com to your email list so our updates never land in spam.

On 16 September, the Federal Reserve raised its policy rate by a quarter point to a range of 3.75% to 4.00%, the first hike since 2023, under Chair Kevin Warsh. Six weeks later, the committee sits down again with a problem that textbooks file under "awkward". Prices are rising faster, and jobs are being created more slowly. The decision lands at 2:00 pm in Washington today, which is 11:30 pm in India, so most of you will wake up to the answer. Here is how to read it before it arrives.

🛢️ Two kinds of inflation, and why the difference matters

Inflation is not one thing. Economists split it into two buckets, and the Fed treats them very differently.

Headline inflation counts everything a household buys, including petrol and electricity. Core inflation strips out food and energy, because those prices swing on events the central bank cannot control: a war, a harvest, a pipeline outage. The Fed can make borrowing dearer to cool demand for cars and houses. It cannot make a barrel of oil cheaper.

September's US consumer price report shows the gap clearly. Headline inflation hit 3.7% over twelve months, up from 3.4% in each of the two previous months. Energy was up about 20% on the year and petrol about 34%, per the Bureau of Labor Statistics report as covered by CNBC and Trading Economics. Core inflation was 2.6%, higher than the 2.5% forecast but still far below the headline. Petrol alone accounted for more than half of the monthly rise.

So the first thing to understand is that most of the heat in the headline number is coming from one input. Brent crude has stayed above 100 dollars a barrel for much of the autumn, well above where it traded in the summer. That is the force behind the number.

⚙️ Why the Fed cannot simply look through it

The standard playbook says a central bank should "look through" a one-off energy spike. Prices jump, then level off, and raising rates only adds a recession on top of a price shock. The playbook has a catch. It works when people believe inflation will come back down. If households and firms start to expect higher prices as the norm, they ask for higher wages and raise their own prices, and a temporary shock becomes a lasting one.

That is the Fed's worry now, and the numbers explain it. Inflation on the Fed's own preferred gauge, the PCE index, is still running above the 2% target, and has been for a long time, which is exactly the situation where expectations can slip. The Fed's own projections in September showed 16 of 18 officials expecting at least one more hike by the end of the year.

The other side of the ledger is the labour market. September payrolls rose by just 29,000, well below forecasts, and unemployment ticked up to 4.2%. July was revised to a loss of 10,000 jobs, and August was cut to 133,000 from 162,000, per the jobs report. An oil shock is a tax on every household that drives or ships anything, so it weakens spending at the same moment it lifts prices. Raise rates into that, and you tighten the squeeze.

That combination, prices rising while growth slows, is what economists mean by stagflation. This is a mild version, with core inflation still under 3% and unemployment at 4.2%, but the shape of the problem is the same, and it is why this meeting is hard.

The dynamics covered in this article affect every US stock in your portfolio. Trade from India on the Winvesta app. No US bank account needed!

🚀 Join 60,000+ investors, become a paying subscriber or download the Winvesta app and fund your account to get insights like this for free!

Get the app!

🏆 Who an oil-led squeeze helps and who it hurts

The table above is a framework, not a forecast, and the direction in each row can change if oil reverses.

Energy producers are the obvious beneficiaries while crude stays elevated. Their revenue moves with the barrel, and their costs do not move one for one. Companies with real pricing power, such as dominant software and payments businesses, tend to cope better, because they can pass higher costs on without losing customers.

The pressure falls on businesses where fuel is a large cost and customers are price sensitive. Airlines and logistics firms sit at the top of that list. Consumer discretionary names suffer when petrol takes a bigger bite of household budgets. Highly indebted companies are exposed on a second front, because a Fed that stays tight keeps their borrowing costs high. Long-duration growth stocks, whose value rests on profits far in the future, are the most sensitive to higher yields.

🌏 What this means if you are investing from India

Three channels reach an Indian investor holding US stocks, and oil sits behind all of them.

The first is the dollar. A Fed that hikes or sounds hawkish supports the dollar, and a stronger dollar means your US holdings are worth more in rupees even if the share prices go nowhere. India imports most of its crude, so expensive oil also pressures the rupee from the other side. The rupee has spent much of 2026 under pressure, and a hawkish Fed plus costly oil is the combination that keeps that pressure on. For an unhedged holder of dollar assets, this has been a cushion, not a cost.

The second is yields. The ten-year Treasury yield has been hovering around 5%, after touching a multi-year high above 5.3% in early October. Yields near 5% give investors a real alternative to stocks, which is why valuations on expensive growth names are under pressure whenever the Fed sounds firm. Short-dated yields sitting above the 4.00% top of the Fed's range would tell you the bond market expects more tightening, not less.

The third is your own sector mix. Many Indian portfolios lean heavily on US technology. That is the part of the market most sensitive to yields, and the part with the least direct link to oil. If you hold a broad US index fund, you already own some energy exposure. If you hold only tech, you do not.

🔭 What to watch tonight and over the next few weeks

Tonight, the rate itself is the least interesting part. Market pricing points strongly to a hold, with only a small chance of another hike. What moves markets is the language. Does the statement keep December open? Does Chair Warsh say anything that signals the next step? Chair Warsh has moved away from traditional forward guidance, so his press conference tone is harder to read than usual.

The Fed's next meeting is in December, and it will come with a fresh set of projections. Check the release calendar for the PCE inflation report and the first estimate of third-quarter GDP, which land in the next few days.

Three signals would change the picture. Oil falling back sharply from current levels would weaken the case for another hike. A second weak jobs report would tilt the Fed toward patience. And a ten-year yield that breaks decisively above its early-October high would suggest the bond market thinks the Fed is still behind.

If this changed how you read an inflation number, pass it on.

Share

📊 The bottom line

A rate decision is not a simple read on whether the economy is strong or weak. When inflation comes from an oil shock and hiring is cooling, the Fed is choosing between two imperfect options, and the market trades the language more than the move. Do not try to guess tonight's tone. Check how much of your portfolio is exposed to rising yields, how much to energy prices, and how much to the dollar, because those three variables are what a stagflation-style squeeze actually runs through.

📈 Numbers worth remembering

This article is for educational purposes only and does not constitute investment advice. Winvesta India Technologies Ltd. does not guarantee the accuracy of third-party data cited above. Past performance is not indicative of future returns. Please consult a registered financial adviser before making investment decisions.

Five years ago, buying a Nasdaq-100 fund meant owning a hundred companies and calling it a day. Not anymore. This week, Microsoft, Alphabet, Meta and Amazon all report earnings in the same stretch the Federal Reserve makes its rate call, and between them those four names already carry more weight in your "diversified" fund than your entire mid-cap sleeve. That's why we built Winvesta Crisps, to decode what's actually moving the funds you own, in plain language, before the consensus catches up. 60,000+ investors from all over India are already in. What about you?

Subscribe now!

🔔 Don't miss out! Add winvestacrisps@substack.com to your email list so our updates never land in spam.

Most investors treat "I own a Nasdaq-100 fund" as shorthand for "I own a hundred companies, so I'm fine." That shorthand has been getting less true for years, and this week puts it to an unusually direct test. The Magnificent Seven now make up roughly a third of the S&P 500, a touch above 33% per First Trust's latest index tracking, and the concentration inside a pure tech or Nasdaq-100 fund runs higher still. Four of those seven report earnings this week: Microsoft, Alphabet, Meta and Amazon. Sitting in the same week is the Federal Reserve's policy decision on Wednesday, a meeting with no fresh economic projections attached, so the market's main read on the Fed will come from the statement and the press conference alone.

This is not a routine earnings week. It is the second AI-capex reporting cycle of the year that has arrived with real tension already loaded into it. In April, the same four names collectively guided 2026 capital spending toward a combined figure north of $650 billion. In July, two of them, Meta and Alphabet, followed that up with guidance that unsettled the market rather than reassured it: Meta's free cash flow came in sharply lower even as its capex range moved higher, and Alphabet raised its own capex guide again without giving analysts a clean demand story to hang it on. Both stocks fell on the news. Here's what nobody's telling you: if this week repeats that pattern, the investor who feels diversified because they hold "the whole Nasdaq" will take a concentrated hit, not a diversified one.

🎯 Meet Rhea

Rhea is 32, a product manager at a Bengaluru SaaS company, with a US portfolio worth close to ₹52 lakh built up over three years of steady monthly investing into a Nasdaq-100 index fund. She picked it specifically because "it's not just one stock, it's a hundred of the best companies in the world," and has largely ignored the portfolio since, treating the broad name as its own diversification argument.

She has never opened the fund's actual holdings list. If she had, she would see that a handful of names, not a hundred, decide most of what happens to her money in any given quarter, and that this week, four of those names report earnings on the same calendar as a Fed decision she has not been tracking at all.

Her hidden exposure. Nvidia, Apple, Microsoft, Amazon, Alphabet, Broadcom and Meta combined have, across recent fund disclosures, sat at somewhere above a third of her fund's total weight, even before accounting for how those same names dominate the S&P 500 and, by extension, most "global" and "world" equity funds that also happen to be cap-weighted. Rhea thinks she is diversified across a hundred companies. She is actually running a concentrated bet on roughly seven balance sheets, four of which release their next set of numbers inside the next few days.

📊 What QQQ, VGT and RSP actually give you

QQQ, the Invesco Nasdaq-100 fund Rhea actually owns, is cap-weighted, which means the largest companies in the index get proportionally larger slices of every rupee that goes in. Recent fund disclosures put Nvidia, Apple, Microsoft, Amazon and Alphabet's combined share at somewhere in the mid-to-high-20s percentage-wise on their own, before adding Broadcom and Meta, which is precisely why a single capex-guidance surprise from any of this week's four reporters moves the whole fund, not a sliver of it.

VGT, Vanguard's Information Technology fund, is built the same way but narrower, holding only the tech sector rather than the full Nasdaq-100, which pushes the concentration in the same handful of names higher still rather than lower. It is the more aggressive version of the same bet, not a diversified alternative to it.

RSP, the Invesco S&P 500 Equal Weight fund, is the structural counterpoint. It holds roughly the same 500 companies as the standard S&P 500 index, but rebalances every holding back toward an equal slice regularly, so no single name, however large its market value, can dominate the fund's return. Its largest sector exposures each sit in the mid-teens percentage range, a materially flatter profile than either QQQ or VGT. It is not an anti-tech fund. It is a fund that refuses to let seven earnings calls decide its quarter.

Read more

Five years ago, Cadence Design Systems was a line item fund managers skipped past on the way to the semiconductor names everyone actually talked about. Not anymore. The company reports its third quarter results today with a record $8.1 billion order backlog and revenue growing 24% a year, and the stock is still down for 2026 heading into tonight's print. That gap, between what the business is doing and what the market is paying for it, is exactly what Winvesta Crisps exists to find. That's why we built Winvesta Crisps, to decode what's actually driving the companies you own, in plain language, before the consensus catches up. 60,000+ investors from all over India are already in. What about you?

Subscribe now!

🔔 Don't miss out!

Add winvestacrisps@substack.com to your email list so our updates never land in spam.

Most investors who have even heard of Cadence Design Systems think of it as back office software that chip engineers use and nobody else needs to care about. The reality is sharper than that: no modern chip, not Nvidia's latest GPU, not AMD's newest accelerator, not Apple's next silicon, reaches a fab without first running through Cadence's or Synopsys's design and verification software. Between them, Cadence and Synopsys sit across the vast majority of the global electronic design automation market, and almost every dollar being poured into AI data centres eventually passes through one of their licences. Cadence reports its third quarter results today after the US market closes, coming off a quarter in which revenue grew roughly 24% year on year to $1.58 billion and its order backlog hit a record $8.1 billion, yet the stock was still trading lower for the year heading into tonight's numbers. When the business keeps improving and the share price does not follow, that gap is worth understanding properly.

🔥 Top movers

Cadence shares head into today's print still underwater for 2026, even after management raised full year guidance in July on the back of record bookings, a pattern some analysts have attributed to multiple compression across AI-adjacent software names rather than anything specific going wrong at the company. The print lands in a week stacked with Big Tech earnings, with Microsoft, Alphabet, Meta and Amazon all due to report over the coming days, so broader sentiment on AI capital spending could move Cadence's stock just as much as its own numbers do. Verizon, Nucor and several other large names also report today, keeping Monday's tape busy on both sides of the AI trade.

🧩 Business model: the toll booth before the chip exists

Cadence sells the software that semiconductor companies use to turn an idea for a chip into a design that a factory can actually manufacture. Its core electronic design automation business covers digital and custom integrated circuit design, verification tools under the Verisium brand, and implementation tools under the Cerebrus brand. Alongside that sits an intellectual property licensing business, where Cadence licenses pre-built circuit blocks that chipmakers plug into their own designs rather than build from scratch, and a newer System Design and Analysis business that has grown rapidly since Cadence pushed beyond pure chip design into simulating how chips behave physically once assembled into a system.

The economics are built on multi-year licence agreements and subscriptions rather than one-off sales, which is why Cadence talks about backlog the way a hotel chain talks about advance bookings. At the end of its second quarter this year, that backlog stood at a record $8.1 billion, with roughly $4.2 billion of it expected to convert into recognised revenue over the following twelve months, per the company's own second quarter disclosures. That gives Cadence a degree of revenue visibility that most software companies would envy, because customers are not buying a licence for this quarter's chip. They are buying the tool they will use for the next several years of chip designs, win or lose in the market.

The reason this matters so much right now is structural, not cyclical. Designing an advanced chip today, something with billions of transistors arranged in three dimensional stacks and chiplet architectures rather than a single flat die, cannot be done by hand or with generic software. It requires specialised tools from a small number of vendors who have spent decades building the mathematics and the customer relationships to do it. Cadence and Synopsys are estimated to hold the large majority of that global market between them, with Siemens EDA, formerly known as Mentor Graphics, a distant third, though exact market share figures vary meaningfully depending on which analyst you ask and how the simulation and IP businesses are counted. Whoever wins the AI chip race, whether that is Nvidia, AMD, Broadcom's custom silicon customers or someone not yet on investors' radar, that winner almost certainly used Cadence or Synopsys tools to get there. Cadence gets paid either way.

Read more

The AI trade's financing problem didn't stay contained to one debt-heavy cloud stock for long. This week it reached Meta, a company that beat on revenue and still fell nearly 7%, and Lam Research, which warned on its own supply chain. That's why we built Winvesta Crisps, to decode every shift reshaping the companies you own, in plain language, before the consensus catches up. 60,000+ investors from across India are already in. What about you?

Subscribe now!

🔔 Don't miss out!

Add winvestacrisps@substack.com to your email list so our updates never land in spam.

Big Tech mostly beat this week, and the market sold it anyway. Meta posted 26% revenue growth and still dropped nearly 7% on fresh anxiety over its AI spending, Lam Research tumbled on a supply-chain warning, and CoreWeave took another analyst downgrade less than two weeks after its own debt scare rattled the sector. The financing question that hit AI infrastructure stocks in mid-October has now spread to the hyperscalers writing the cheques.

📊 Market recap

The week opened in a holding pattern. With last week's CoreWeave and Nebius debt scare still fresh and a run of mega-cap earnings due, Monday through Wednesday traded without a clear direction, the kind of quiet that usually means the market is waiting for a specific date rather than running out of reasons to move. That date arrived on Thursday.

Thursday was the week's real session. The Nasdaq Composite fell 2.60% to 26,008.86, its worst single day since June, while the S&P 500 dropped 1.21% to 7,663.43. The Dow Jones barely moved, up 0.02% to 52,873.28, extending its pattern from recent weeks of grinding higher while tech absorbs the selling. Meta Platforms fell nearly 7% despite reporting revenue of $64.6 billion, up 26% year on year, after free cash flow came in negative and the company narrowed its 2026 capex guidance to $135 billion to $145 billion, effectively raising the floor. Lam Research dropped close to 12% on a supply-chain constraint warning that dragged the Philadelphia Semiconductor Index down more than 6% and pulled Nvidia and Micron lower in sympathy, per Fool.com's live coverage. CoreWeave, already down sharply from its mid-October debt scare, took a fresh analyst downgrade the same day. Tesla slipped around 3% on margin concerns following its own Q3 results. Not every report was punished: Pool Corp jumped over 14% on an earnings beat and reaffirmed guidance, Spotify rose roughly 9 to 11% on stronger margins and Q4 revenue guidance, Lockheed Martin gained about 7% on a beat-and-raise quarter, and Procter & Gamble added 5.6% on an EPS and revenue beat. Bloom Energy fell around 7% after hours on its own Q3 print.

The selloff spilled into Friday's Asia session, with Japan's Nikkei down about 1.7% and South Korea's Kospi off 2.4 to 3.3% at the open, per regional market reports, as the same AI-financing worry that hit Wall Street on Thursday worked its way through chip-exposed Asian markets.

😶‍🌫️ Sentiment watch

A precise, verified Fear and Greed reading for this week was not available before this went to print. The gauge was already sitting in "fear" territory as of an early-October snapshot, and Thursday's broad, cross-sector selloff gives no reason to think sentiment has recovered since. The VIX spiked to an intraday high of 22.31 last Friday, its highest level since June, and Thursday's session, the Nasdaq's worst day in four months, extended rather than reversed that elevated-volatility regime.

The ten-year Treasury yield told an unusual story alongside the selloff: rather than rising on Brent crude's jump to a multi-month high, it fell to around 4.95% to 5.0%, one report calling it the sharpest one-day drop since March 2023, as money rotated into bonds. That is a growth-scare signature, not an inflation-scare one, which matters for how this week should be read against the Fed meeting still to come.

The stocks caught in this week's AI-financing scare, from Meta to the chipmakers and cloud names around it, are all tradeable from India on the Winvesta app. No US bank account needed!

🚀 Join 60,000+ investors, become a paying subscriber or download the Winvesta app and fund your account to get insights like this for free!

Read more

Five years ago, Netflix was the company you paid a monthly fee to binge a show without a single advertisement. In the quarter it just reported, Netflix ran the biggest live sports slate in its history, sold enough advertising to matter to Madison Avenue, and signed a ten year, five billion dollar deal to carry WWE's flagship show. That's why we built Winvesta Crisps - to decode what's actually driving the companies you own, in plain language, before the consensus catches up. 60,000+ investors from all over India are already in. What about you?

Subscribe now!

🔔 Don't miss out!

Add winvestacrisps@substack.com to your email list so our updates never land in spam.

Most investors still think of Netflix as a subscription business that sells ad-free television on demand, a company whose stock moves on how many new households sign up each quarter. That framing is accurate and dangerously incomplete. Netflix stopped reporting subscriber numbers altogether in 2025, and in their place it has built an advertising business from close to nothing, bought its way into live sports and wrestling, and started behaving, quarter by quarter, less like a streaming app and more like a television network that happens to be ad supported, on demand and live all at once. The subscription business still funds all of it. It is no longer the only story.

🏭 What Netflix actually is today

Netflix's core business is still the one most people recognise: a paid, ad-free or lower-priced ad-supported subscription to a library of films, series and documentaries, sold in more than 190 countries. The company last disclosed a global membership figure alongside its Q4 2025 results, when paid memberships topped 325 million, per Subscription Insider's coverage of that shareholder letter. Since the start of 2025, Netflix has stopped reporting quarterly subscriber counts entirely, choosing instead to be judged on revenue, margin and engagement.

What has changed underneath that subscription number is the shape of the business funding it. Netflix now runs four reportable geographic regions, United States and Canada (UCAN), Europe, Middle East and Africa (EMEA), Latin America (LATAM) and Asia-Pacific (APAC), each growing at a double digit clip. Layered on top of that core is an advertising tier that barely existed three years ago, a slate of live sports and wrestling rights that did not exist at all before 2024, and a games division still small enough that Netflix does not break out its revenue separately. The company that used to compete only with HBO and Disney+ for prestige television now also competes with NBC, Fox and ESPN for sports rights, and with Google's YouTube for the household remote control itself.

🧩 Segment breakdown: every region is growing, and advertising sits inside all of them

Netflix's third quarter of 2026, the three months to 30 September 2026 and reported on 20 October 2026, showed broad-based growth rather than one standout market. Total revenue reached 12,874.2 million dollars, up 12% year-on-year (11% on a foreign exchange neutral basis), per Netflix's Q3 2026 shareholder letter. UCAN, still the largest region, grew revenue 11% to 5,628.2 million dollars. EMEA grew 11% (9% FX neutral) to 4,106.6 million dollars. LATAM grew the fastest on a reported basis, up 17% (12% FX neutral) to 1,602.6 million dollars, while APAC grew 12% (15% FX neutral, reflecting currency effects) to 1,536.8 million dollars.

Netflix does not break out advertising revenue by region, or disclose it as a separate line in the quarterly letter at all. What the company does say is that advertising revenue sits inside every one of these four regional totals, growing fastest in UCAN where the ad tier launched first, but expanding as Netflix rolls its ad products into new markets. The geographic breakdown looks like a mature, diversified subscription business because it is one. The advertising and live events layer inside it is where the more interesting growth sits, and that is a number Netflix is choosing not to disclose on its own.

⚡ The core strategic bet: turning a subscription app into a TV network

Netflix's advertising business grew from a standing start in late 2022 to more than 1.5 billion dollars of revenue in 2025, more than two and a half times what it generated in 2024. For 2026, Netflix had initially guided to roughly doubling that figure to approximately 3 billion dollars. In its Q3 2026 letter, the company cut that guidance to approximately 2.7 billion dollars, citing a slower rollout of some ad product capabilities, its own words, rather than any weakness in advertiser demand. Netflix says its ad-supported tier now reaches more than 250 million global monthly active viewers, a reach metric built by multiplying ad-tier members by average household size rather than a subscriber count, so it should be read as an audience estimate and not a paid-account figure.

Read more

Most investors treat the Fed's next rate call as a simple question: cut, hold, or hike. This month that question has a complication sitting underneath it. The government shutdown that began on 1 October has now run three weeks, and it has switched off a good part of the data the Fed normally reads before it moves. In seven days the committee has to decide anyway. That's why we built Winvesta Crisps, to break down what's actually moving markets, in plain language, before the consensus catches up. 60,000+ investors from all over India are already in. What about you?

Subscribe now!

🔔 Don't miss out! Add winvestacrisps@substack.com to your email list so our updates never land in spam.

On 1 October, Congress failed to pass a funding bill and the federal government shut down for the second time in under a year. Three weeks on, it is still shut, Washington is no closer to a deal than it was in week one, and the Federal Open Market Committee meets on 28 October to decide what happens to the policy rate that sets the cost of money for every dollar asset an Indian investor holds. The twist is not whether the Fed wants to cut again. This newsletter noted on 6 October that a third cut in 2026 was all but priced in. The twist is that the agencies that normally hand the Fed its evidence, on jobs, on prices, on spending, have mostly stopped producing it, and the committee has to make the call with a meaningfully thinner file than usual.

🏛️ What a shutdown actually switches off

A government shutdown is not a metaphor. When Congress does not pass appropriations bills or a continuing resolution by the start of the new fiscal year, federal agencies lose their legal authority to spend money, full stop. Non-essential staff are furloughed, essential staff keep working without pay until it ends, and anything that depends on a federal agency publishing something on schedule simply stops showing up.

This is not a one-off event. The 2013 shutdown ran 16 days. The 2018 to 2019 shutdown ran 35 days and was, at the time, the longest in US history. Last year's shutdown broke that record outright: it ran 43 days, from 1 October to 12 November 2025, and the Congressional Budget Office put its cost to the economy at 7 to 14 billion dollars, with fourth-quarter annualised GDP growth cut by an estimated 1 to 2 percentage points and the unemployment rate pushed up by roughly 0.4 percentage points. None of that was recovered quickly. A chunk of it, the CBO said, would never be recovered at all.

What makes this year's repeat unusual is not its length so far, three weeks is still short by last year's standard, but its timing. It has landed directly on top of a Fed meeting, for the second time in two cycles, at a moment when the central bank's own credibility is already under the kind of scrutiny this newsletter covered through September's rate hike debate and the Jackson Hole speech that preceded it.

⚙️ How the Fed is supposed to decide, and what happens when half the inputs go dark

The Fed's rate decisions run on a handful of government data series: the monthly jobs report from the Bureau of Labor Statistics, the consumer price index, GDP from the Bureau of Economic Analysis, retail sales from the Census Bureau. During a shutdown, the statistical agencies that produce these numbers stop collecting and publishing them, because that work itself is federal spending that has just lost its authorisation.

The precedent from last year's 43-day shutdown is direct. The BLS delayed its September 2025 jobs report by roughly seven weeks, to 20 November, and said it could not calculate a full October unemployment rate at all because the household survey that feeds it cannot be conducted retroactively once the window for it has closed. The Fed still had a rate decision to make in that window, and it cut anyway, in what reporting at the time described as the first occasion the committee had made a call without the government data it would normally lean on. Chicago Fed president Austan Goolsbee said plainly that the committee would use outside data where the official numbers were missing.

That is exactly the position the FOMC is in again. In place of BLS and Census numbers, the committee leans on private substitutes: ADP's payroll data, which draws on the roughly 25 million workers on its own payroll platform and keeps publishing through a lapse; ISM manufacturing and services surveys; card-spending trackers from Visa and Mastercard; regional Fed surveys; and scheduling data from platforms like Homebase that proxy for small-business hiring. Every one of these is noisier and narrower than the official series it stands in for. None of them covers the full economy the way a BLS household survey does. The Fed is not flying with no instruments. It is flying with a smaller panel of instruments, each individually less reliable, at the exact moment a policy mistake in either direction (cutting into stickier inflation than the committee realises, or holding into a labour market that is weakening faster than the private data shows) is harder to catch and reverse quickly.

The dynamics covered in this article affect every US stock in your portfolio. Trade from India on the Winvesta app. No US bank account needed!

🚀 Join 60,000+ investors, become a paying subscriber or download the Winvesta app and fund your account to get insights like this for free!

Get the app!

🏆 Who the data blackout actually hurts and helps

Treasury markets, oddly, are among the least affected in the short run. Auctions continue on schedule because Treasury debt management is funded separately from the discretionary spending that lapses, so the long end's own deficit-and-issuance story, the one driving the 10-year yield above 5% through October, keeps running on its own logic regardless of whether BLS is open.

Federal contractors and the local economies built around them are the clearest losers. The Washington DC metro area alone carries hundreds of thousands of federal and federally-dependent jobs, and furloughed back pay, guaranteed since 2019 legislation, does not restore spending that households deferred while their paycheques stopped. Airlines and air travel sit in the same bucket: air traffic controllers are classed essential and keep working unpaid, and past shutdowns have produced exactly the kind of staffing-driven delays and flight cuts that eventually force the FAA's hand.

Systematic and data-dependent funds sit in the uncomfortable middle. Strategies built around the monthly jobs print or CPI release lose their anchor entirely when those prints are delayed or thinned out, which tends to show up as choppier, less predictable trading around the dates those releases would otherwise have landed.

Gold and other instruments that benefit from uncertainty rather than from any specific data point are the clearest beneficiaries, for as long as the uncertainty lasts. That is a separate mechanism from the real-yields story this newsletter walked through on 30 September, and the two can point in opposite directions at once, which is part of why gold has traded choppily rather than in a clean trend through this stretch.

🌏 What this means if you are investing from India

An Indian investor holding US equities through the Winvesta app does not get a pass on any of this because the shutdown is a domestic American political story. The FOMC's decision on 28 October lands in every dollar-denominated holding regardless of where the investor sits, and a data blackout does not mean lower uncertainty around that decision. It means the opposite: wider bands around what the Fed might do, and historically, wider bands around how markets react once the decision actually lands.

The currency leg matters just as much. If the Fed cuts again on thinner evidence, a softer dollar is the more likely near-term reaction, which works in favour of unhedged dollar asset holders translating gains back into rupees. If the committee instead holds, judging the private data too unreliable to act on, the dollar's recent strength has more room to persist, which is the same pressure this newsletter detailed on 9 September, when the RBI had already spent an estimated 136 billion dollars defending the rupee this year. A shutdown that drags past 28 October without resolution raises the odds of that second scenario, because a Fed genuinely uncertain about the state of the economy has more reason to wait for clarity than to act on it.

The practical read-through is not to guess which way the Fed goes. It is to expect a noisier few days around 28 October than a normal meeting would produce, and to size any short-term trading around that date accordingly rather than treating a data vacuum as a quiet one.

If this changed how you read the next Fed decision, pass it on.

Share

📊 The bottom line

A government shutdown does not just furlough federal workers. It furloughs the statistics the Fed needs to do its job properly, at precisely the moment the job gets harder to do well. The central bank has shown, in last year's 43-day shutdown, that it will make the call anyway rather than wait indefinitely, leaning on a thinner, noisier set of private substitutes for the official data it would rather have. That is not a reason to assume the Fed gets it wrong on 28 October. It is a reason to expect more uncertainty than usual around whichever call it makes, and to remember that uncertainty, not any specific number, is often the thing markets actually trade on.

📈 Numbers worth remembering

This article is for educational purposes only and does not constitute investment advice. Winvesta India Technologies Ltd. does not guarantee the accuracy of third-party data cited above. Past performance is not indicative of future returns. Please consult a registered financial adviser before making investment decisions.

Hot inflation data, a widening shipping war in the Gulf and a memory-chip stock wobble all hit Wall Street in the same week, and the S&P 500 barely blinked. That is not calm. That is a market choosing which risks to price and which to ignore. That's why we built Winvesta Crisps, to decode every shift reshaping the companies you own, in plain language, before the consensus catches up. 60,000+ investors from across India are already in. What about you?

Subscribe now!

🔔 Don't miss out!

Add winvestacrisps@substack.com to your email list so our updates never land in spam.

Wall Street spent this week pricing three separate shocks, a scorching September inflation print, an escalating tanker war in the Strait of Hormuz, and a China-driven wobble in memory-chip stocks, and still finished within a few points of record territory. Big bank earnings added a fourth complication: Citigroup and Goldman Sachs blew past forecasts while Wells Fargo and JPMorgan stumbled, the clearest split verdict of the earnings season so far.

📊 Market recap

The week's price action read in two acts: a quiet start that gave away nothing, and a Wednesday inflation shock that gave away quite a lot. The S&P 500 spent Monday and Tuesday pinned in the high-7,700s to high-7,800s, within a whisker of the record levels it printed the previous week, before Wednesday's hot CPI reading knocked it back into the high-7,600s to low-7,700s. The Dow and Nasdaq told the same story from different angles: the Dow ground along near 51,700 on thin conviction, while the Nasdaq was dragged by a deepening chip-stock selloff even before the inflation data landed.

Monday was a non-event by design. US markets traded through the Columbus Day holiday on thin volume and the major indices barely moved: the Dow added about 0.38% to roughly 51,850, the Nasdaq slipped about 0.44% to around 27,246, and the S&P 500 closed almost exactly flat at 7,807, per market reports citing Reuters data. The real action was underneath the index level. A scattered, catalyst-light rally in small and mid-caps, FIGS, FICO and CCC Intelligent Solutions were all up double digits on no clear news, per Tapeboard, sat next to a sharp, concentrated selloff in chip and optical-networking names: Cerebras, Semtech, Ciena and Lumentum were all down close to 10% as investors began rotating out of 2026's most crowded hardware trade.

Tuesday brought the bank earnings and the first hard evidence that the AI hardware trade was cracking. Citigroup and Goldman Sachs opened earnings season with results that beat decisively: Citi's Q3 net income jumped roughly 40% year on year to about $5.2 billion on its best third-quarter revenue in a decade, while Goldman's revenue of around $18.7 billion ran nearly $2 billion ahead of consensus. Wells Fargo and JPMorgan told the opposite story. Wells Fargo's net interest income missed estimates for a sixth straight quarter, and JPMorgan posted its first earnings miss in more than a year, dragged down by weak fixed-income trading and $658 million in investment-securities losses, even as its equity trading revenue surged 49%. JPMorgan's shares fell as much as 3% in premarket trading before recovering to close higher once investors absorbed the bank's raised 2026 net interest income guidance. Away from the banks, China's CXMT disclosed plans to push into high-end server memory with a denser chip design, a direct shot at Micron's core business. Micron fell about 5% and SanDisk, which has barely any server-memory exposure, fell about 4% in sympathy. The Philadelphia Semiconductor Index logged its fifth straight losing session, a monthly low.

Wednesday was the week's pivot. September's CPI print landed hot: headline inflation at 3.7% year on year and core at 2.6%, both above where the Fed would like to see them. The ten-year Treasury yield spiked as high as 5.32% intraday on the release before easing back. Equities took the data badly: the S&P 500 fell 1.1% on the day, and semiconductor stocks, already weakened by the CXMT news, dropped a further 4.5%.

Thursday and Friday were about absorbing the shock rather than adding to it. The ten-year yield eased back to around 5.11%, a three-week low, as Fed officials offered a genuinely split read on the inflation print. San Francisco Fed president Mary Daly said rates will likely have to rise again, though not necessarily at the 27 to 28 October meeting. Governor Michelle Bowman said she would rather wait. Chicago Fed president Austin Goolsbee called the CPI print "discouraging." Markets pared back bets on an October hike after the comments, though UBS is now pencilling in a third hike for March 2027. The chip-stock weakness spilled into Asia overnight: South Korea's Kospi fell about 8% and Japan's Nikkei dropped around 5.1%, while Nasdaq 100 futures pointed roughly 1% lower heading into Thursday's US session. Fed chair Kevin Warsh was due to speak late Thursday evening US time, just hours before this went to print in India, so his remarks are a live input the market has not fully absorbed yet.

😶‍🌫️ Sentiment watch

A precise, verified Fear and Greed reading for Friday was not available before this went to print, so treat the mood read as directional rather than exact. The VIX, Wall Street's own volatility gauge, moved from about 15.3 on Tuesday to 16.87 by Thursday, a jump of more than 10% on the week but still well inside what traders call a calm market, nowhere near the readings in the mid-20s and above that usually mark genuine fear.

That is the tell for this week. A hot inflation print, a widening Gulf shipping war and a chip-stock wobble together pushed volatility up by only a few points. Markets are not panicking. They are mildly more alert than they were seven days ago, and arguably should be more alert than that.

The stocks moving markets this week, from the big banks to the chipmakers caught in the China memory fight, are all tradeable from India on the Winvesta app. No US bank account needed!

🚀 Join 60,000+ investors, become a paying subscriber or download the Winvesta app and fund your account to get insights like this for free!

Get the app!

🏆 Best performers

The week's clearest winners were catalyst-light momentum names and one direct beneficiary of the Gulf shipping disruption, not the earnings season's actual headline beats.

DHT Holdings, a crude tanker operator, was the one gainer with an obvious story: oil's climb through the week on Hormuz supply fears lifted freight economics directly. The rest of the list, FIGS, FICO and CCC Intelligent Solutions, moved on volume too thin and news too absent to call a trend. Per Tapeboard's own reporting, none of the week's biggest single-day percentage gainers had a confirmed company-specific catalyst attached.

💔 Worst performers

Read more

Five years ago, Oracle was the vendor your IT department renewed every January without a second thought, the database nobody got fired for buying. In the past year it has raised tens of billions of dollars in debt and equity, signed a compute contract with OpenAI that press reports value at around 300 billion dollars, and turned a 48 year old enterprise software business into one of the most leveraged bets on the AI buildout trading on Wall Street. That's why we built Winvesta Crisps - to decode what's actually driving the companies you own, in plain language, before the consensus catches up. 60,000+ investors from all over India are already in. What about you?

Subscribe now!

🔔 Don't miss out!

Add winvestacrisps@substack.com to your email list so our updates never land in spam.

Most investors still think of Oracle as the company that sells databases to banks, airlines and tax departments, a legacy enterprise vendor that grows in the single digits and gets owned for its dividend, not its upside. That framing is accurate and dangerously incomplete. Oracle has spent the past eighteen months turning its balance sheet into the financing arm of the AI buildout, borrowing tens of billions of dollars to build data centres that it then rents out, with OpenAI alone reportedly on the hook for a five year contract worth roughly 300 billion dollars. The database business still exists, and still throws off real cash. It is no longer the business that moves the stock.

🏭 What Oracle actually is today

Oracle was founded in 1977 by Larry Ellison on the idea of a commercial relational database, and for four decades that database, plus the enterprise applications Oracle bolted on through acquisitions such as PeopleSoft, Siebel and NetSuite, was the whole story. Fortune 500 companies ran payroll, inventory and financial reporting on Oracle software because switching was expensive and the software worked.

That company still exists inside Oracle's results every quarter. What has changed is who runs it and what it spends money on. In September 2025, Safra Catz stepped down after eleven years as sole chief executive to become executive vice chair, and Oracle named two co-chief executives in her place: Clay Magouyrk, who joined from Amazon Web Services in 2014 and had spent more than a decade building Oracle's cloud infrastructure platform, and Mike Sicilia, who runs Oracle's applications and AI products. Larry Ellison kept his titles of executive chairman and chief technology officer but has largely stayed off Oracle's earnings calls through 2026, leaving Magouyrk and Sicilia to field analyst questions.

Promoting an infrastructure engineer to co-chief executive the same year Oracle committed to building hundreds of billions of dollars of AI data centre capacity was not a coincidence. Oracle today reports across four lines: cloud (infrastructure plus applications), software licences, hardware, and services. The cloud line, and specifically the infrastructure piece inside it, is now where nearly all of the growth, the capital spending and the investor attention sit.

🧩 Segment breakdown: the cloud is carrying a company that still sells licences

Oracle's fiscal first quarter of 2027, the three months to 31 August 2026 and reported on 10 September 2026, shows exactly how lopsided the business has become. Total revenue reached 19.3 billion dollars, up 30% year-on-year, per Oracle's Q1 FY27 earnings release. Cloud revenue, combining infrastructure and applications, rose 62% to 11.6 billion dollars. Inside that line, cloud infrastructure, the raw compute and storage business that rents capacity to AI labs and enterprises, jumped 121% to 7.4 billion dollars, while cloud applications such as Fusion ERP and NetSuite grew a comparatively modest 10% to 4.2 billion dollars.

The older parts of the business tell a different story. Software licence revenue, the traditional on-premise business Oracle built its name on, fell 3% to 5.55 billion dollars as customers keep migrating off it. Hardware revenue rose 15% to 774 million dollars, driven by Exadata and AI-adjacent systems rather than a broader server revival, and services revenue rose 5% to 1.41 billion dollars.

Cloud infrastructure alone, a line that barely registered five years ago, is now bigger than Oracle's entire software licence business and growing more than ten times as fast. That single fact is the clearest evidence that Oracle's centre of gravity has shifted.

⚡ The core strategic bet: renting out the AI boom

The number that dominates every conversation about Oracle in 2026 is its remaining performance obligations, or RPO, a measure of contracted revenue that customers have committed to but Oracle has not yet recognised. RPO reached 664 billion dollars at the end of Q1 FY27, up 209 billion dollars year-on-year and 26 billion dollars from the prior quarter, after Oracle booked more than 30 billion dollars of new AI cloud contracts in the quarter alone, per Oracle's own investor materials. Management has said it expects roughly half of that backlog to convert into recognised revenue over the next three years, a multi-year runway most software companies would never dream of disclosing.

Read more

Markets spent the week finding out that a single earnings leak can move more money than an oil war. A Financial Times report that OpenAI's revenue was running some 20 billion dollars short of what had been signalled wiped out a fortnight of record highs in two sessions, even as tankers were getting hit in the Strait of Hormuz and the Fed's own officials were talking about more rate rises. That's why we built Winvesta Crisps, to cut through the noise and tell you what actually moved markets and why. 60,000+ investors from all over India are already in. What about you?

Subscribe now!

🔔 Don't miss out!

Add winvestacrisps@substack.com to your email list so our updates never land in spam.

This was the week the AI trade found out it could lose money two days running. The S&P 500 touched its first close above 7,800 on Monday and a fresh record on Tuesday. By Thursday the Nasdaq had logged its worst session since mid-August, oil was up more than 3 per cent on tanker attacks near the Strait of Hormuz, and the 10-year Treasury yield had touched its highest level since 2002. The Dow, propped up by energy and value names, barely noticed.

📊 Market recap

The index-level picture: the S&P 500 spent the week in the high-7,700s to low-7,800s, the Nasdaq Composite in the high-26,000s to high-27,000s, and the Dow in the low-51,000s. Three different stories, one week.

Monday brought the calm before it. The S&P 500 closed above 7,800 for the first time, helped along by tech gains and yields that had eased off their recent highs, per CNBC.

Tuesday was the high point. The S&P 500 climbed 0.58 per cent to a record close of 7,818.93, and the Nasdaq Composite added 0.45 per cent to close at a record 27,599.79, per CNBC. Nobody on the tape that day would have guessed what was two sessions away.

Wednesday turned. A Financial Times report said OpenAI's annualised revenue was running about 20 billion dollars below the figure that had been doing the rounds a month earlier, down to roughly 50 billion against a widely cited 68 billion, per CNN and CNBC. The Nasdaq fell about 1 per cent as the AI trade took its first real hit of the autumn. The same day, the Fed released the minutes from its 15 to 16 September meeting, where it had lifted rates 25 basis points to 3.75 to 4.00 per cent, its first hike since 2023.

Thursday was the reckoning. The Nasdaq Composite fell 1.25 per cent to 27,193.34, its worst day since mid-August, and the Philadelphia Semiconductor Index dropped 3.39 per cent. The S&P 500 slipped 0.47 per cent to 7,765.36. The Dow, insulated by its heavier energy and consumer weighting, actually rose 0.1 per cent to 51,231.64, per CNBC and HDFC Sky. Brent crude jumped roughly 3.5 to 4 per cent on reports that Iran had stepped up attacks on tankers in and around the Strait of Hormuz, with a Gulf of Mexico hurricane shutting in a further slice of US production for good measure, per CNBC. Oil pared some of that gain after the Trump administration signalled it would not strike Iran before the midterms.

Friday opened with futures pointing higher. S&P 500 futures were up about 0.3 per cent and Nasdaq 100 futures were up roughly 0.5 per cent ahead of the bell, with prediction markets pricing a strong chance of a positive close. Delta Air Lines was due to report its September quarter before the open, with the Street looking for adjusted earnings up close to 15 per cent year on year.

The week's verdict: a genuine AI wobble, not a market-wide panic. The pain stayed concentrated in the names most exposed to the OpenAI story, while energy, staples and a handful of earnings winners carried the broader tape.

😶‍🌫️ Sentiment watch

The CNN Fear and Greed Index closed the week at 38, in "Fear" territory, down from 43 earlier in the week and 40 a week earlier, as cited by Benzinga's daily market wraps. The reading drifted between Neutral and Fear through the week rather than cratering in one move, which tells its own story: this was a sentiment bruise, not a sentiment collapse.

The VIX backs that up. It closed at 15.57 on Thursday, up 3.25 per cent on the day, having traded as high as 16.46 intraday, per HDFC Sky's market data and FRED's official series. That is a contained reading. It sits well below the year's September high of 17.84 and nowhere near the late-July spike above 20. Wall Street's own fear gauge is saying the OpenAI story is a stock-picker's problem, concentrated in a dozen or so AI-adjacent names, not a reason to hedge the whole market.

The stocks moving markets this week are all tradeable from India on the Winvesta app. No US bank account needed!

🚀 Join 60,000+ investors, become a paying subscriber or download the Winvesta app and fund your account to get insights like this for free!

Get the app!

🏆 Best performers

The week's winners clustered in two places: anything that benefits when oil spikes, and the one earnings beat strong enough to swim against a falling tech tape.

Accenture led the pack, up just over 6 per cent after news of a new tie-up with Dell, a rare piece of good news for a stock usually read as a bellwether for enterprise technology spending, per a Wall Street Republic market recap. PepsiCo climbed almost 4 per cent after reporting third-quarter net revenue growth of 5.6 per cent, a clean beat in a week when very little else in consumer-facing earnings was getting rewarded, per the company's own filing. The rest of the list, Marathon Petroleum, Valero and Chevron, is a straight read of the oil shock: refiners and an integrated major all catching the updraft from Brent's jump on the Hormuz tanker attacks, per CNBC's oil coverage.

The common thread: when one trade (AI infrastructure) gets hit this hard, the capital has to land somewhere else. This week it landed on energy and on the handful of companies that could point to a concrete, dated piece of good news.

Read more

Japan has raised interest rates twice this year, taking its policy rate to 1.25 per cent, the fastest pace of tightening the Bank of Japan has run since 1990. That sounds like a domestic story about Japanese savers finally earning something on their deposits. It is not. Cheap yen has spent two decades funding trades everywhere from Wall Street tech stocks to Indian equities, and the last time this specific trade unwound, it erased more than a tenth of the Nikkei in a single session and reached straight into Indian portfolios. That is why we built Winvesta Crisps, to break down what is actually moving markets, in plain language, before the consensus catches up. 60,000+ investors from all over India are already in. What about you?

Subscribe now!

🔔 Get the Wednesday macro breakdown in your inbox every week, before the headlines catch up. Subscribe now!

Most Indian investors who hold US stocks have never placed a single yen trade in their life, and their returns are still exposed to one. The mechanism runs through a currency most of them have never converted, a central bank most of them do not follow, and a trade that is sized, depending on whose estimate you trust, anywhere from a couple of hundred billion dollars to well over two trillion. Nobody actually knows the exact number, which is itself the point. In August 2024, that uncertainty turned into the worst single trading day for Japanese stocks since 1987. Japan is hiking again, faster than it has in thirty five years, and the question worth asking is not whether this happens again, but what it drags down with it this time.

📜 What the yen carry trade actually is

For most of the last two decades, Japan kept interest rates at or near zero while the rest of the world's central banks did not. That gap created one of the most durable trades in global finance: borrow yen for almost nothing, convert it into dollars or another higher-yielding currency, and buy assets that pay more than the yen loan costs. US tech stocks, European bonds, Australian dollars, Indian equities, all of it has at some point been funded, in part, by yen borrowed at close to zero per cent.

The trade works beautifully until two things happen at once: Japanese rates rise, making the yen loan more expensive, and the yen itself strengthens, making the loan more expensive to repay in whatever currency the profits came in. Both happened in early August 2024, when the Bank of Japan raised rates and signalled more hikes to come. The yen surged, funding costs jumped, and leveraged positions across the world got margin calls on the same days. The Nikkei 225 fell 12.4 per cent on 5 August 2024, its worst single day since Black Monday in 1987. The selling did not stay in Tokyo. US and European equities fell hard in sympathy within hours, because the funds unwinding yen positions were selling whatever they held that was liquid, not just Japanese assets.

Estimating the size of the trade today is genuinely difficult, and any number you read should be treated as a range, not a fact. Cross-border yen borrowing, the broadest measure of how much global leverage is funded in yen, has been estimated in the trillions of dollars by some bank research desks. Narrower measures of speculative positioning, which try to isolate the specific bet that the yen stays weak, have put the figure closer to a few hundred billion dollars. The gap between those estimates is the real story. Markets found out how large the trade actually was in August 2024 only after it broke.

⚙️ How an unwind actually spreads

The mechanism is not complicated once you see it, which is part of why it catches people out. A leveraged position funded in yen has two exposures stacked on top of each other: the return on whatever was bought, and the yen's exchange rate against the currency that return is measured in. When the Bank of Japan hikes, it attacks both exposures simultaneously. The cost of servicing the yen loan rises, and the yen tends to strengthen as higher Japanese rates attract their own capital inflows, which means the loan is now more expensive to repay in yen terms too.

For a leveraged fund, that combination forces deleveraging, often through a margin call rather than a considered decision. The fund does not sell only its Japanese holdings. It sells whatever is most liquid and easiest to convert back into yen quickly, which in practice means the most widely held, most heavily traded assets in the world: large cap US technology stocks, liquid emerging market equities, currencies seen as risk-on bets against the dollar. That is the channel through which a Bank of Japan decision reaches a portfolio that has never touched a yen-denominated instrument.

The current cycle looks different from 2024 in one respect worth naming plainly. The Bank of Japan has been signalling its hiking intentions well ahead of each move this time, under Governor Kazuo Ueda, and markets have had weeks rather than hours to reduce yen-funded exposure in response. The federal funds rate sits at 3.50 to 3.75 per cent against Japan's 1.25 per cent, a gap of roughly 225 to 250 basis points, which still leaves the carry trade profitable on paper even after this year's hikes. US Treasury Secretary Scott Bessent has also been pushing Tokyo publicly toward faster tightening, which has added political pressure on top of the economic case. So far, that combination of advance warning and a still wide rate gap has meant a gradual unwind rather than a repeat shock. The gap narrowing further, or a hike landing with less warning than the market expects, is what would change that.

The mechanics of a trade most Indian investors have never heard of are already reaching into every US stock held from India, through exactly this kind of global deleveraging. Trade US markets from India on the Winvesta app, no US bank account needed.

🚀 Join 60,000+ investors, become a paying subscriber or download the Winvesta app and fund your account to get insights like this for free!

Get the app!

⚖️ Who wins and who loses if the unwind accelerates

Leveraged macro funds running yen-funded positions are the clearest losers in a sharp unwind, for the reasons above, and so are the assets they are forced to sell to raise yen quickly. High beta US growth stocks tend to fall harder than the broad market in these episodes, not because anything changed in their underlying business, but because they are among the most liquid things a deleveraging fund can sell in minutes. Emerging market currencies and equities carry the same exposure, since they are common destinations for yen-funded carry positions and get sold alongside everything else when the trade reverses.

Japanese exporters sit in an uncomfortable middle position. A stronger yen makes their overseas earnings worth less when translated back home, which is why Japanese equities, particularly the export-heavy names on the Nikkei, were among the hardest hit in August 2024 even though the shock originated in their own central bank's decision.

The clearer winners are harder to name with the same confidence, which is itself informative. Yen holders and Japanese importers benefit from a stronger currency, as do Japanese households buying anything priced in dollars. US Treasuries tend to catch a safe haven bid during the risk-off days that follow a sharp unwind, even though the hike that triggered the unwind came from Japan, not the Federal Reserve. Beyond that, a carry trade unwind is mostly a story about who gets hurt and by how much, not who profits from it.

🇮🇳 What this means for an Indian investor holding US equities

India is not a bystander to this trade. Market estimates at the time of the August 2024 unwind put India's yen carry trade linked exposure at around 21 billion dollars, close to 2.2 per cent of total foreign institutional investor holdings, a meaningful enough slice that the unwind showed up directly in FII selling data that month.

The 2026 numbers suggest the exposure, or at least the sensitivity, has not gone away. Foreign portfolio investors pulled more than 3.05 lakh crore rupees out of Indian equities across the first nine months of 2026, with selling continuing into October, including a further outflow of roughly 9,232 crore rupees in the opening days of the month alone, according to NSDL linked flow data. The rupee hit a record low near 96.89 to the dollar in May 2026. None of that selling is solely a yen story. US yields, crude oil prices and India's own earnings season have all played a part. But a yen carry unwind adds to exactly the same selling pressure, through the same mechanism described above: funds raising cash fast tend to sell whatever is liquid, and Indian large caps qualify.

For an investor holding US equities through Winvesta, the read-through is about correlation rather than a reason to change allocation. A disorderly yen unwind tends to arrive alongside a broad risk-off day across global equities, US stocks included, and alongside rupee weakness that changes what a given dollar return is worth once converted home. Those two effects can compound in the same direction on the worst days, and recognising the shared cause is more useful than treating a US portfolio dip and a weak rupee as two unrelated events.

👀 What to watch from here

The Bank of Japan's forward signalling is the single most useful thing to track, since Governor Ueda's public comments have been the clearest advance warning markets have had this cycle, a sharp contrast with the limited notice ahead of the August 2024 hike. A hike that lands with less warning than the market has grown used to is the scenario most likely to reproduce 2024 rather than extend the gradual pattern seen so far this year.

USD/JPY itself is worth watching as a level, not just a headline. The pair has been trading in the low to mid 160s, levels that have previously drawn explicit intervention threats from Japan's Ministry of Finance, with 155 to 160 flagged by several trading desks as the zone where intervention risk rises materially. A sudden, sharp yen move in either direction, rather than a gradual drift, is historically the trigger for forced unwinds rather than the level itself.

The Federal Reserve's own rate path matters just as much as Japan's, since the entire trade runs on the gap between the two. A faster pace of Fed cuts than currently priced would narrow the rate differential from the other side and could squeeze carry positions even without any further move from Tokyo. Cross-border yen funding data, published periodically by the Bank for International Settlements, is the slowest moving but most reliable signal of how much leverage is actually built up in this trade at any given time, even though it arrives with a lag.

If this changed how you read a quiet day in Japanese interest rates, pass it on.

Share

💡 The bottom line

The yen carry trade is not a niche derivatives story. It is one of the largest, least visible sources of leverage in global markets, funded by a central bank most Indian investors never think about, and it has already proven, in a single trading day in August 2024, that it can take down US and Indian equities alongside Japanese ones. The Bank of Japan is hiking faster now than at any point since 1990, the rate gap with the Federal Reserve is narrowing, and the trade is unwinding more gradually this time only because Tokyo has chosen to signal its moves well in advance. That pattern can break with one less predictable hike. An Indian investor holding US equities does not need to trade the yen to be exposed to it. The exposure already runs through the portfolio, via exactly the kind of global deleveraging that turned a Japanese rate decision into the Nikkei's worst day since 1987.

📊 Numbers worth remembering

This article is for educational purposes only and does not constitute investment advice. Winvesta India Technologies Ltd. does not guarantee the accuracy of third-party data cited above. Past performance is not indicative of future returns. Please consult a registered financial adviser before making investment decisions.

Five years ago, a Fed rate-cut cycle was simple instruction for anyone holding a bond fund: yields fall, prices rise, sit back and collect the gain. That assumption broke this year. The Federal Reserve has already cut its policy rate twice in 2026, and the ten-year Treasury yield closed above 5.1% in late September, its highest level since 2007. That's why we built Winvesta Crisps, to decode what's actually moving the funds you own, in plain language, before the consensus catches up. 60,000+ investors from all over India are already in. What about you?

Subscribe now!

🔔 Don't miss out! Add winvestacrisps@substack.com to your email list so our updates never land in spam.

Most investors treat "the Fed is cutting" as a single instruction that applies to every bond fund they own. It does not. The Fed's two cuts this year, plus a third all but priced in for its 28 October decision, have pulled short-term yields down roughly on schedule. The ten-year yield has done the opposite: it closed at 5.17% on 25 September, up from 4.96% just two weeks earlier, and the highest since 2007. Widen the lens further and the divergence gets starker. Since September 2024, the ten-year yield has climbed by around 1.3 percentage points even as the Fed has cut short-term rates by roughly 1.5 percentage points over the same stretch, according to RBC Wealth Management's own tracking of the gap.

A government shutdown that began on 1 October, now in its sixth day, has made the picture harder to read, not easier. Roughly 750,000 federal employees are furloughed, the Congressional Budget Office puts the furloughed payroll cost alone at about $400 million a day, and the shutdown has delayed the data releases the Fed would normally lean on before its 28 October meeting. Policymakers are heading into a near-certain cut with a meaningfully incomplete picture of the labour market behind it. None of that stopped the ten-year yield from climbing all September. Here's what nobody's telling you: the fund you bought because "the Fed is cutting" may be the one fund in your portfolio that this cut barely touches.

🎯 Meet Arnav

Arnav is 36, an engineering lead at a Bengaluru fintech, with a US portfolio worth roughly ₹68 lakh built up since 2021. Most of it sits in S&P 500 and Nasdaq-100 index funds, plus a handful of individual megacap names. In March, with the Fed's first cut of the cycle already signalled, he added a 6% position in TLT, the iShares 20+ Year Treasury Bond ETF, reasoning that a Fed cutting cycle was the obvious moment to lock in a long-duration bond trade before prices moved. He also parked a smaller, almost incidental sum in SHY, the short-duration Treasury fund, purely as a place to hold cash he hadn't decided what to do with.

Six months and two rate cuts later, his "obvious" trade is down. TLT has lost ground in a stretch where the Fed has done exactly what he expected it to do. The "boring" SHY position he barely thought about has held its value and quietly paid out a yield north of 3%. His question is specific: has the TLT trade simply not worked yet, with the third cut on 28 October the catalyst that finally turns it around, or did he misread what a Fed rate cut actually does to a 20-year bond?

His hidden exposure. TLT's effective duration runs to roughly 16 to 17 years, which means its price moves far more on long-end yield swings than on the Fed's own overnight rate. Those long-end swings this cycle have been driven by a separate argument, about fiscal deficits, Treasury issuance and the term premium investors demand to hold government debt for decades, not by what the Fed does at its next meeting. Arnav bought a bet on the Fed. He is actually holding a bet on the bond market's patience with Washington's deficit math, and those two bets have been moving in opposite directions all year.

📊 What TLT, SHY and XLU actually give you

TLT is the purest expression of duration risk available in a single ticker: a portfolio of bonds maturing in 20 years or more, which means a given move in long-term yields shows up in its price almost immediately and almost in full. That is precisely why it has struggled this year. It was never built to track the Fed funds rate. It was built to track the long end, and the long end has its own argument underway about who absorbs the growing supply of Treasury debt now that the deficit keeps widening.

SHY sits at the other end of the same asset class. With an effective duration under two years and roughly $11 billion in assets, making it the largest fund of its kind, it is built to track the Fed's own rate path far more closely than the long end's separate dispute with the bond market. Its 30-day yield sits near 3.3%, and because its holdings mature within three years, it has avoided almost all of the price damage TLT has taken this year.

XLU, the Utilities Select Sector SPDR, is the equity-market version of the same tension. Utilities carry a dividend yield of roughly 3%, trade partly on that yield the way a bond does, and get hit when long rates rise even though they sit in the equity sleeve of most portfolios, not the bond sleeve. It is the fund that reveals whether an investor's "bond problem" has quietly become an "entire portfolio" problem.

Read more

Five years ago, a Tesla delivery number was simple: more cars sold meant a better quarter. Not anymore. Today the company's valuation rests on two businesses that barely existed the last time Tesla reported a delivery miss: Cybercab robotaxis now running paid fares in seven US cities, and Optimus humanoid robots just starting their first production run in Fremont. Reading Tesla as a car company in October 2026 means missing most of what is actually moving the stock. That's why we built Winvesta Crisps, to decode what's actually driving the companies you own, in plain language, before the consensus catches up. 60,000+ investors from all over India are already in. What about you?

Subscribe now!

🔔 Add winvestacrisps@substack.com to your email list so our updates never land in spam.

Most investors read Tesla's 486,532 third-quarter deliveries as an unambiguous win. The number beat Wall Street's consensus of 461,100 by roughly 25,000 vehicles, and the stock jumped 4.65% to $370.59 on 2 October on volume nearly 40% above its three-month average. Fewer people noticed that Tesla sold 2.1% fewer vehicles than it did a year earlier, when buyers rushed to beat the expiry of the federal EV tax credit. The headline beat is a shrinking comparison base dressed up as momentum, arriving three weeks before an earnings report that will show whether the robotaxi and robotics businesses Tesla is now priced on can carry the stock once the delivery story fades.

🚗 Three businesses sharing one ticker

Tesla's automotive division is still the business paying the bills, and the Model 3 and Model Y carried that load again this quarter even as deliveries fell versus last year's tax-credit rush. Automotive revenue grew 23.1% year on year to $20.5 billion in the most recently reported quarter (Q2 2026), roughly 73% of total revenue, per the company's own filings. The weaker line sits elsewhere: deliveries of everything outside Model 3/Y, which includes the Cybertruck and Model S/X, fell nearly 48% year on year in Q3, a sign that Tesla's older and newer non-core vehicle lines are both struggling to find buyers at current prices.

Automotive gross margin, excluding regulatory credits, slipped sequentially from 19.2% to 16.3% in Q2, pressured by the absence of one-off benefits from the prior quarter and rising input costs. That is the part of the business a traditional auto analyst would flag as deteriorating. It is also the part the market has stopped paying much attention to.

Energy generation and storage is the quieter growth story. Tesla deployed 13.7 GWh of batteries in Q3, its second-highest quarterly total on record, trailing only the 14.2 GWh deployed in Q4 2025. First-half 2026 deployments of 22.3 GWh were up 53% sequentially. Energy revenue grew by double digits year on year, though analysts differ on the precise figure depending on how they allocate services revenue between segments. Energy margins, however, compressed sharply, from 39.5% to 20.4% in Q2, on warranty true-ups, tariffs and lower average selling prices. Services and other revenue, which includes Supercharging, insurance and parts, rose 50% to $4.6 billion with margins at an all-time high of 14.1%. That is the cleanest growth line in the entire earnings report, and it gets almost none of the attention paid to robotaxis.

Then there is the third Tesla, the one the stock is actually priced for: a bet that autonomous driving and humanoid robotics turn a car company into something closer to an AI infrastructure company. Neither business contributes a meaningful line to revenue yet. Both are the reason analyst price targets on this stock range from $25 to $600, one of the widest spreads of any large-cap name on Wall Street, per multiple sell-side notes.

Read more

Five years ago, analysing AMD meant comparing clock speeds against Intel chips in a personal computer market that barely grew year on year. Not anymore. In 2026, AMD handed one customer a path to ten per cent ownership of the company in exchange for a multi-year chip order, and the market is still working out what that means for every other AI infrastructure stock. That's why we built Winvesta Crisps - to decode what's actually driving the companies you own, in plain language, before the consensus catches up. 60,000+ investors from all over India are already in. What about you?

Subscribe now!

🔔 Don't miss out!

Add winvestacrisps@substack.com to your email list so our updates never land in spam.

Most investors who follow AMD still see a solid, slightly cheaper alternative to Nvidia, a stock that does well when Nvidia cannot build chips fast enough and struggles once supply catches up. That framing is accurate and dangerously incomplete. Through 2026, AMD turned its largest prospective customers into financial partners, wiring their chip orders directly into the AI industry's capital structure. The clearest evidence sits inside a single legal document: a warrant that could eventually hand OpenAI a tenth of AMD, issued not for cash but for a promise to buy AMD's chips at scale.

🏭 What AMD actually is today

AMD builds and sells four distinct product lines, not one. Data Center is EPYC server processors and Instinct AI accelerators sold to cloud providers and AI labs, and it is now the company's main growth engine. Client is Ryzen processors for laptops and desktops, the business most people still associate with the AMD name. Gaming covers semi-custom chips for PlayStation and Xbox consoles alongside Radeon graphics cards, and it is the most cyclical and currently the weakest of the four. Embedded, built around the 2022 acquisition of Xilinx, sells specialised chips into industrial, automotive and aerospace customers who value long product lifecycles over raw speed.

What has changed is the mix. Data Center has gone from a meaningful contributor to the majority of revenue inside two years, and that shift is why AMD now trades, rightly or wrongly, as an AI infrastructure stock first and a PC chipmaker second.

🧩 Segment breakdown

Second-quarter 2026 revenue was concentrated in Data Center to a degree AMD has never shown before, with EPYC and Instinct sales more than doubling year on year while the older consumer-facing businesses grew more slowly or shrank.

Gaming's decline is mostly semi-custom console chip revenue normalising after a strong prior period, per AMD's Q2 2026 earnings release, rather than a sign of weakness in Radeon graphics cards specifically. Embedded's steady growth reflects industrial and automotive demand that moves on multi-year design cycles, largely insulated from the AI capital expenditure cycle driving the Data Center numbers.

⚡ The core strategic bet: six gigawatts and a warrant

In October 2025, AMD and OpenAI announced a partnership for OpenAI to deploy up to six gigawatts of AMD Instinct GPUs over multiple years, starting with one gigawatt of the next-generation MI450 series in the second half of 2026, per AMD's official announcement and SEC filing. Alongside the compute order, AMD issued OpenAI a warrant for up to 160 million AMD shares, roughly ten per cent of the company, at an exercise price of one cent each.

Read more

Five years ago, the rule for owning energy or defense exposure was simple: wait for a headline crisis, buy the spike, sell the relief rally. That rule assumed the crisis would end in weeks. The Strait of Hormuz standoff is now well into its second month with no resolution in sight, and the funds tied to it are behaving like an income position, not a trade. That's why we built Winvesta Crisps, to decode what's actually moving the funds you own, in plain language, before the consensus catches up. 60,000+ investors from all over India are already in. What about you?

Subscribe now!

🔔 Don't miss out! Add winvestacrisps@substack.com to your email list so our updates never land in spam.

Most investors are watching the Strait of Hormuz standoff the way they watch any other headline war: as background noise for a portfolio that lives somewhere else. That assumption stopped being safe weeks ago. Brent crude closed near $107 a barrel on 28 September and West Texas Intermediate near $94, both up sharply from about $57 in January, after the latest round of US-Iran talks over reopening the strait broke down again and war-risk shipping insurance in the Red Sea reportedly tripled. The ten-year Treasury yield has climbed alongside it, closing near 5.01% on 16 September, its highest level since 2023, and pushing higher still through the back half of the month as the standoff dragged on with no resolution.

Two US sector funds have quietly done exactly what a war-and-oil-shock regime is supposed to do to them. The iShares US Aerospace & Defense fund, ticker ITA, is up 16.7% year to date. The Energy Select Sector SPDR, ticker XLE, is up close to 24%. The problem is almost nobody holding a standard US portfolio out of India owns either one.

🎯 Meet Neha

Neha is 34, a lead product manager at a Pune fintech, and has run a US portfolio since early 2023. It is worth roughly ₹56 lakh: a Nasdaq-100 index fund makes up just over half of it, and the rest sits in five megacap names she added one at a time whenever they came up in her company's internal trading Slack channel, Nvidia and Microsoft among them. She has never owned an energy fund, a defense fund, or a barrel of oil in any form. Her reasoning was straightforward: geopolitics is noise, technology is the sector that has actually made her money since 2021, and she does not have the bandwidth to trade headlines.

That reasoning survived three years of a market where the Fed was cutting, oil was cheap, and the ten-year yield sat comfortably under 4.5%. It has not survived this one.

Her question this week is specific. She has watched two funds she has never looked at twice, ITA and XLE, post some of the steadiest double-digit gains in the market this year, while the Nasdaq-100 fund that makes up more than half her portfolio has gone nowhere for six weeks. She wants to know if adding either fund now is chasing a headline that could reverse the day a ceasefire is announced, or catching a structural shift she has been on the wrong side of since January.

Her hidden exposure. Run the actual sector math on that Nasdaq-100 fund and the picture is starker than "some tech." Technology and communication-services names make up close to nine-tenths of the index by weight. Energy and industrials, the two sectors doing the work in XLE and ITA respectively, are close to a rounding error inside it. Neha does not hold a diversified portfolio that happens to be tech-heavy. She holds a single macro bet, dressed up as an index fund, with the two sectors best placed to hedge it at effectively zero weight.

What happened last time this regime showed up. 2022 is the closest precedent: the Fed hiking fast, oil elevated after Russia's invasion of Ukraine, and long-duration growth stocks getting hit hardest by higher discount rates. The Nasdaq Composite fell by roughly a third that year. The S&P energy sector was the only one of eleven S&P sectors to finish positive, up in the region of 60% to 65%. Neha was not investing yet in 2022. She has never lived through the version of this movie where her actual holdings are the ones losing.

📊 What these two funds actually own

Start with what her Nasdaq fund does not give her: real exposure to either the war-spending story or the oil-price story. Not because it is a bad fund, but because a fund that is nine-tenths technology and communication services by design carries no meaningful weight in defense contractors or energy producers, no matter how large it gets.

iShares' US Aerospace & Defense fund, ITA, holds around three dozen names concentrated in prime contractors and their supply chain, Boeing, Lockheed Martin and RTX among the largest. Its return this year is driven less by the day-to-day oil price than by a slower, stickier mechanism: emergency defense appropriations and munitions replenishment orders that, once placed, tend to run for years rather than weeks. Patriot interceptor costs alone reportedly topped $3.2 billion in the first week of the current conflict, and total defense spending tied to it has been estimated at roughly $1 billion a day. None of that reverses the day a ceasefire is signed.

The Energy Select Sector SPDR, XLE, is the more direct trade: large-cap oil and gas producers and refiners whose earnings move close to one-for-one with the crude price. It is up roughly 24% year to date, behind the oil-futures fund USO's steeper gain of around 57%, because XLE's largest holdings hedge part of their own production and USO does not. XLE is also the fund that would give back the most, the fastest, on a genuine de-escalation headline.

Read more

Five years ago, a Nike earnings beat meant the stock went up the next morning. Not anymore. Nike has topped Wall Street's earnings estimate in each of its last four quarters, and the shares fell after two of those releases regardless. That gap between what the numbers say and what the stock does is exactly the kind of mispricing Winvesta Crisps exists to unpack, in plain language, before the rest of the market catches up. 60,000+ investors from all over India are already in. What about you?

Subscribe now!

🔔 Add winvestacrisps@substack.com to your email list so our updates never land in spam.

Most investors watching Nike this week think Thursday's earnings will settle whether Elliott Hill's turnaround is working. It will not, not by itself. Nike beat earnings estimates in each of its past four quarters and the stock fell after two of those prints anyway, sliding to a fresh 52-week low of around $35 in September 2026, down from just under $77 twelve months earlier. The number that will actually decide this week is smaller and less dramatic: whether North America wholesale, up 3 percent last quarter, can keep growing while Greater China revenue is still shrinking by double digits. That is the gap the stock is pricing, not the headline earnings-per-share figure.

👟 What Nike actually sells now

Nike does not run one business anymore. It runs a wholesale business that is recovering and a direct-to-consumer business that is still shrinking, and the mix between the two is the real story behind every headline number.

In the fiscal fourth quarter (ended 31 May 2026), wholesale revenue came in at $6.6 billion, up 4 percent on a reported basis and up 1 percent on a currency-neutral basis, per Nike's own earnings release. NIKE Direct, the stores-and-app business the company spent the past decade building up, brought in $4.1 billion, down 7 percent reported and down 9 percent currency-neutral, with digital sales down 12 percent and owned stores down 7 percent. Converse, the smallest piece, fell 32 percent to $244 million.

That split is not an accident. Under former chief executive John Donahoe, Nike leaned hard into direct-to-consumer, cutting back wholesale accounts and betting the app and Nike-owned stores could carry the brand alone. It did not work as planned: the retreat from wholesale left shelf space open for adidas, On and Hoka to fill, and Nike's own digital channel has now declined for several consecutive quarters. Elliott Hill, who took over as chief executive in October 2024 after three decades at the company, has spent his first two years reversing that call, rebuilding the wholesale relationships Nike walked away from and pushing product back onto the shelves at retailers such as JD Sports, DSG and Foot Locker.

Geographically, the picture is split just as sharply. North America, Nike's largest market, grew 3 percent in the fourth quarter to $4.8 billion. EMEA slipped 1 percent reported (6 percent currency-neutral) to $3.0 billion. APLA, the Asia Pacific and Latin America segment, was roughly flat. Greater China was the outlier in the wrong direction, down 12 percent reported and 17 percent currency-neutral to $1.3 billion, Nike's fourth straight quarter of double-digit declines there. China was once Nike's highest-margin market. It is now the one management cannot yet explain a path back from.

Read more

On 17 September the Federal Reserve's own economists projected interest rates sitting at 4.1 per cent at the end of 2027, exactly where they see this year closing, and a full half a point above the 3.6 per cent path the Fed itself had pencilled in as recently as June. Four days later the Nasdaq closed at a record and the market's own fear gauge sat near a cycle low. That is not how a hawkish surprise used to land. It is the first real test of a Fed chair who has spent four months dismantling the tool that used to make surprises like this move markets. That's why we built Winvesta Crisps: to break down what's actually moving markets, in plain language, before the consensus catches up. 60,000+ investors from all over India are already in. What about you?

Subscribe now!

🔔 Don't miss out!

Add winvestacrisps@substack.com to your email list so our updates never land in spam.

The Federal Reserve raised its benchmark rate by 25 basis points to 3.75 to 4.00 per cent on 16 September, its first increase since 2023, on a unanimous 12 to 0 vote. Nothing about that alone was strange. What was strange sat inside the accompanying projections: the median official now sees rates ending 2027 at the same 4.1 per cent as 2026, not the 3.6 per cent easing path the Fed had shown in June. Bond desks call that a duration shock, a higher-for-longer signal, as opposed to a decision shock, one more hike this year. Five trading days later the S&P 500 gained 1.5 per cent to 7,764.70, the Dow rose 0.7 per cent to 52,048.83, and the Nasdaq closed at a record. Working out why a signal that size stopped moving prices the way it should have is this week's story, and it starts with the Fed chair who has spent this year quietly turning off the microphone.

🔮 The Fed chair who decided not to tell you what happens next

Every Fed chair for two decades has used some version of forward guidance: language in the policy statement, the dot plot, or a press conference line that tells markets roughly where rates are headed before the Fed actually moves them there. Kevin Warsh, confirmed as Fed chair on 22 May 2026, has spent his first year in the job taking that tool apart.

Starting with the June 2026 meeting, Warsh's FOMC statements dropped the forward-looking signal language altogether in favour of what reporters at the time called a just-the-facts read of the decision, no promise attached to what comes next. At the Jackson Hole symposium on 28 August he went further, committing to a quieter Fed and reaffirming the 2 per cent inflation target while arguing that forward guidance boxes a central bank into decisions it later has to walk back. He has since launched five internal task forces to review how the Fed communicates, how it manages its balance sheet, which data it relies on, and how it frames its inflation goal.

The balance sheet is the other half of the shift. Warsh has said publicly that shrinking the Fed's roughly 6.7 trillion dollar balance sheet is a priority tool, one he considers more useful than jawboning about the future path of rates. In practice that reduction has stalled: after falling to about 6.54 trillion dollars in late 2025, the balance sheet has been growing again since the start of this year. The rhetoric points one way, the actual lever has not moved much yet, and that gap matters for how much weight markets should put on anything the Fed says about its intentions right now.

⚙️ What breaks when a central bank stops promising

The rate the Fed actually sets, the overnight funds rate, is one number. What moves your portfolio is the entire yield curve and the multiple the market pays for future earnings, and both of those are forward-looking. A stock priced on cash flows five years out is really priced on where the market expects rates to sit over those five years, not on today's funds rate alone. Forward guidance let a central bank move that whole curve, and every asset priced off it, without lifting the funds rate at all. A single sentence promising low rates for an extended period did the work a dozen rate moves would otherwise have taken.

Take that promise away and the dot plot becomes something else entirely. The quarterly Summary of Economic Projections still exists, an anonymous median of what individual officials expect the rate to be, but Warsh has explicitly downplayed it as a commitment. A dot plot with no promise attached to it is a survey, not a policy. Markets can still read it, but they cannot trade it with the same confidence they once traded an explicit guidance sentence, because nobody at the Fed has said the path will actually hold.

That is the mechanism behind this week's disconnect. The September dot plot was hawkish on paper: 16 of 18 officials pencilled in at least one further hike this year, and the 2027 median stopped easing altogether. A shift that size, delivered as explicit guidance five years ago, would have repriced every long-duration asset within the hour. Instead the CBOE Volatility Index sat at 14.81 into decision week, close to a cycle low, and the market absorbed the hike and the hawkish path within four trading sessions without a drawdown. Part of that is the guidance vacuum: investors are discounting the SEP's signalling value more heavily than they used to discount an explicit statement line. Part of it is separate and simpler, and a U.S. Bank Asset Management strategist put it plainly this week: solid corporate earnings and the productivity gains investors expect from AI deployment are, for now, outweighing the drag from higher rates in how the market prices equities.

History offers a caution against reading too much into four calm days either way. Across six tightening cycles since 1994, the S&P 500 has fallen an average of roughly 4 per cent in the six weeks after the first hike of a cycle before recovering those losses over the following five to six weeks. Twelve months out the average gain has been 6.7 per cent, with a median closer to 10.7 per cent, and returns were positive in every cycle bar one. The point is not that this time is safe because the market shrugged on day one. It is that markets have always needed a few months to actually price a new hiking cycle, and this cycle removed the one tool that used to tell them, in advance, roughly how far that cycle would run.

The dynamics covered in this article affect every US stock in your portfolio. Trade from India on the Winvesta app. No US bank account needed!

🚀 Join 60,000+ investors, become a paying subscriber or download the Winvesta app and fund your account to get insights like this for free!

Get the app!

🏆 Who higher for longer helps, and who it doesn't

A rate path that stays near 4 per cent through 2027 rather than easing back toward 3.6 per cent does not affect every corner of a portfolio the same way. The table below sketches the split, and it is an editorial read of the mechanics, not a forecast.

The uncomfortable middle bucket is the mega-cap AI names carrying the market. Their earnings are currently growing fast enough to absorb a higher discount rate, which is the U.S. Bank strategist's argument for why stocks shrugged this week. That is a statement about current earnings momentum, not a permanent exemption from what higher-for-longer normally does to long-duration valuations.

🇮🇳 What this means for an Indian investor holding US equities

Start with the currency, because a hiking Fed and a hawkish rate path change the maths on every dollar of US equity exposure an Indian investor holds. The rupee traded at 95.71 to the dollar on 23 September, essentially flat over the past month but down 7.81 per cent over the past twelve. A Fed that holds rates higher for longer widens the gap between US and Indian yields, which is generally rupee-negative pressure, working against the currency tailwind that boosted rupee returns on US portfolios earlier in the cycle.

The Reserve Bank of India has been actively managing the other side of that. Reserves stood at a record 785.7 billion dollars as of 11 September, lifted in part by roughly 133 billion dollars raised through a diaspora deposit scheme launched earlier this year, alongside state-run bank dollar sales and RBI spot and swap intervention. That defence has a cost and a limit, and it is happening at the same time foreign investors have been net sellers of Indian equities, pulling out 1.81 billion dollars in September and 25.87 billion dollars so far in 2026. None of that is really about the Fed's forward guidance decision specifically. It is the backdrop an Indian investor is running US equity exposure against while that decision plays out.

The second point is about what to do with the information itself. Indian investors have historically leaned on Fed communication, dot plots, guidance language, press conference wording, to time entries into US equities around rate decisions. A Fed that has deliberately made that signal less reliable is not a reason to stop watching the data. It is a reason to weight the Fed's own words less and the balance sheet and the incoming inflation and jobs prints more, since those are the levers Warsh has said he actually intends to use.

🔭 What to watch

Five things, each checkable without needing to trade around a Fed speech.

The five internal task forces Warsh launched on communications, the balance sheet, data reliance and the inflation framework. Their conclusions, whenever they land, will say more about how the Fed intends to talk to markets going forward than any single statement between now and then.

Whether the balance sheet actually starts shrinking again. It fell to roughly 6.54 trillion dollars in late 2025 and has been growing since, despite Warsh's stated preference for the balance sheet over rate guidance as his primary tool. A renewed contraction would be the first sign that rhetoric and action are lining up.

The December Summary of Economic Projections. If the 2027 median holds at 4.1 per cent or rises further, the duration shock this article describes is confirmed as a trend rather than a one-off. A move back toward easing would say the September shift was noise.

Hyperscaler capital spending guidance in the late October earnings round. The U.S. Bank argument for why equities shrugged this week rests on AI-driven earnings growth outrunning the rate headwind. That argument stands or falls on the next round of capex and revenue guidance from the companies actually spending the money.

The VIX itself. It sat at 14.81 heading into a hawkish decision. A sustained move higher from here, without a fresh shock to explain it, would suggest the market is starting to price the guidance vacuum rather than ignore it.

If this changed how you read the next Fed statement, pass it on.

Share

🏁 The bottom line

A Fed chair who does not want to promise anything about the future path of rates has made every dot plot, every statement and every press conference line worth less than it used to be as a trading signal. That is a deliberate choice by Kevin Warsh, not a market accident, and this week was the first time markets were handed a genuinely hawkish shift under that new regime and largely shrugged it off within days.

The shrug does not mean the shift was harmless. It means the cost of a rate path that stays near 4 per cent for two more years than the Fed itself expected in June has not been priced into equities yet, either because investors do not fully believe the dots, or because AI-driven earnings are currently strong enough to absorb it, or both. Either way, an Indian investor holding US equities is now running that exposure with one fewer tool for reading what the Fed intends to do next, at exactly the moment the rate path got both higher and longer.

The practical response is not to trade the next FOMC statement for a signal it was designed not to give. It is to watch the balance sheet, the incoming data and the earnings that are supposedly justifying today's valuations, because those are the inputs actually driving where rates and equity prices go from here.

📊 By the numbers

• 3.75 to 4.00 per cent: the Fed funds target range after the 16 September hike, a unanimous 12 to 0 vote and the first increase since 2023

• 4.1 per cent: the median SEP projection for both 2026 and 2027 year-end, versus 3.8 per cent and 3.6 per cent respectively projected in June, the duration shock behind this week's move

• 16 of 18: Fed officials who pencilled in at least one further hike before the end of 2026

• 14.81: the VIX close on 18 September heading into decision week, near a cycle low despite the hawkish shift

• 7,764.70, 52,048.83: where the S&P 500 and Dow closed on 21 September, with the Nasdaq closing at a record the same session

• 4.97 per cent: the 10-year Treasury yield on 22 September

• 95.71: the rupee to the dollar on 23 September, down 7.81 per cent over twelve months but broadly flat over the past month

• 785.7 billion dollars: RBI foreign exchange reserves as of 11 September, a record, supported by roughly 133 billion dollars raised through a diaspora deposit scheme

• 25.87 billion dollars: net foreign investor selling of Indian equities so far in 2026, including 1.81 billion dollars in September alone

Disclaimer: All content provided by Winvesta India Technologies Ltd. is for informational and educational purposes only and is not meant to represent trade or investment recommendations. Remember, your capital is at risk. Terms & Conditions apply. Sector impact figures in this article are illustrative editorial models, not reported or verified data.

Five years ago, the rule for owning a bank fund was simple: wait for the Fed to cut, then buy financials, and let a steeper curve and cheaper funding do the rest. That rule broke on 16 September, when the Federal Reserve raised its benchmark rate for the first time since 2023, reversing weeks of dovish pricing, and the financial sector barely blinked. That's why we built Winvesta Crisps, to decode what's actually moving the funds you own, in plain language, before the consensus catches up. 60,000+ investors from all over India are already in. What about you?

Subscribe now!

🔔 Don't miss out! Add winvestacrisps@substack.com to your email list so our updates never land in spam.

Most investors treat a Fed rate hike as bad news for every part of the market, and good news only for cash. That was close enough to true in 2022. It was not true on 16 September 2026, when the Federal Reserve's rate-setting committee voted unanimously to raise the federal funds rate by a quarter point, to a range of 3.75% to 4.00%, its first increase in three years. Fed funds futures had priced roughly one-in-three odds of that outcome as recently as mid-August, when a weak jobs report pushed traders toward pricing cuts, not hikes. The hike happened anyway. Sixteen of eighteen Fed officials now see room for at least one more increase this year, and rate-futures pricing has moved to around 55% odds of a follow-up in October.

The reaction told you who had actually read the mechanism correctly. The dollar index jumped from about 99.6 to above 100.5 within two trading sessions. The ten-year Treasury yield closed near 5.01%, its highest since October 2023. Real estate funds gave back roughly half their year's gains in six weeks. And financials, the sector every 2023-era playbook says should suffer when the Fed tightens, held up. The regional bank fund is up close to 14% for the year. The broad financial sector fund has beaten the S&P 500 over the past three months even though it still trails on a full-year basis.

That divergence is the story this week. It is also, unusually, a divergence most Indian investors are positioned to miss entirely, because almost nobody holding a US portfolio out of India owns a share of either fund.

🎯 Meet Aditi

Aditi is 33, a product manager at a Gurugram fintech, and has run a US portfolio since 2021. It is worth about ₹52 lakh: roughly two-thirds in a broad S&P 500 index fund, the rest split across three megacap technology names she picked because she uses their products every day. She has never owned a financial-sector fund, a bank stock, or anything resembling one. Her reasoning was straightforward: banks felt like a domestic Indian story, she already gets that exposure through HDFC and ICICI in her Nifty investments at home, and the interesting growth in the US has been in technology, not lenders.

That reasoning was sound for most of the past three years. It stopped being sound the week the Fed did the opposite of what the market had priced.

Her question this week is specific. She has watched the regional bank fund climb close to 14% this year, quietly, without a single headline she noticed, while she was reading about AI chip demand instead. Financials are not new money chasing a hike that just happened; the sector has been working since the summer. She wants to know if she is already late, whether the case still holds now that the hike has actually landed, and which of the two financial funds she keeps seeing mentioned, XLF and KRE, actually fits what just happened.

📊 What a rate-hike winner actually looks like

Start with what her S&P 500 fund already gives her, because it is less than it sounds. The index-level financial weight sits in the mid-teens, spread across dozens of names from insurers to card networks to money-centre banks, diluted enough that a genuine sector move barely registers in her total return. Aditi has broad exposure to "financials" the same way she has broad exposure to "energy": present, and too small to matter on its own.

The two dedicated funds behave differently from each other, and that difference is the whole trade.

The Financial Select Sector SPDR, ticker XLF, holds roughly seventy names weighted toward the largest banks, insurers, payment networks and asset managers, with money-centre banks near the top. It is the orderly-policy version of the trade: diversified funding sources and fee income smooth out the pure interest-rate sensitivity, and a fund that also owns Visa and Berkshire Hathaway's financial arm behaves less like a bet on the yield curve and more like a bet on the economy holding together while rates stay higher.

The SPDR S&P Regional Banking ETF, ticker KRE, is the concentrated version: several dozen regional and community lenders, equal-weighted rather than cap-weighted, so a mid-sized bank in Ohio moves the fund about as much as a larger one in California. KRE's business model is borrowing short from local depositors and lending long against local commercial property, which is precisely the balance sheet that a steeper curve helps and a funding-cost spike hurts, often within the same news cycle.

That is why KRE has actually outperformed XLF this year on a simple return basis, near 14% year to date against XLF's high single digits, while carrying a rougher ride to get there. It is the higher-beta version of the same regime, not a safer alternative to it.

Read more

Three months ago Micron delivered a record quarter and told Wall Street the next one would be bigger. It was right. On the call it went further and told customers there is no more high bandwidth memory left to sell them until 2028, because Micron, Samsung and SK Hynix have all already booked their 2027 capacity. The stock is still trading roughly 16% below the high it set on the day of that first record quarter. That gap between what the company can sell and what the market is paying for it is the whole story. That's why we built Winvesta Crisps, to decode what's actually driving the companies you own, in plain language, before the consensus catches up. 60,000+ investors from all over India are already in. What about you?

Subscribe now!

🔔 Don't miss out!

Add winvestacrisps@substack.com to your email list so our updates never land in spam.

Most investors still file Micron under boring, cyclical hardware: a memory chip maker that makes money when DRAM prices go up and loses it the moment they come back down. That description stopped being accurate months ago. In the quarter reported on 24 June 2026, Micron posted revenue of $41.46 billion, up 346% year-over-year, and told analysts it has already sold every gigabyte of high bandwidth memory it can build through the end of 2027. Samsung and SK Hynix, the two rivals that still outsell Micron in plain DRAM, have reportedly done the same. None of that has stopped the stock from trading well below the high it set on results day itself.

🧩 What Micron actually sells now

Micron makes two products, DRAM and NAND flash, and licenses neither name from anyone: it is one of only three companies on earth that can manufacture leading-edge memory at scale, alongside Samsung and SK Hynix. DRAM is the working memory every computer needs to run anything. NAND is the flash storage that holds data when the power is off. For twenty years the business was a commodity: prices moved in vicious cycles, margins swung from spectacular to negative, and investors mostly avoided the stock between cycles.

What changed is a specific product inside DRAM called high bandwidth memory, or HBM, which sits directly on top of an AI accelerator and feeds it data at a speed ordinary DRAM cannot match. Every AI training or inference chip that matters, Nvidia's, AMD's, Google's own silicon, needs it, and there is no substitute. Micron says HBM4 shipments have already passed $1 billion cumulatively and that its HBM4 12-high volume ramp is running at roughly twice the speed of the prior generation, HBM3E. Chief executive Sanjay Mehrotra has told analysts the company has completed agreements on price and volume for its entire calendar 2026 HBM supply and has already booked meaningful 2027 capacity too, which is an unusual thing for any chip company to say about a product that has not been built yet.

Micron reports results across four business lines rather than a simple DRAM and NAND split, and the pattern across all four in the quarter ended 28 May 2026 tells the same story: AI demand is lifting every part of the business, not just the parts touching HBM directly.

Cloud Memory, the line that carries most of the HBM revenue sold directly to data centre operators, was the largest at $13.77 billion. Core Data Center and Mobile & Client were tied at $11.52 billion each, and Automotive & Embedded, the steadiest and least AI-exposed line, still grew to $4.63 billion. Add them up and the total is $41.44 billion against a reported $41.46 billion, close enough that the small gap is rounding rather than an omission. Roughly three-quarters of that revenue is DRAM and the remainder NAND, per industry trackers, though Micron does not disclose that exact split in its own release.

🚀 The supercycle that will not end on schedule

Every memory cycle in the industry's history has ended the same way: prices rise, all three producers add capacity to chase the margin, supply eventually overshoots demand, and prices collapse. Investors who have watched Micron for a decade have that pattern burned in, which is a large part of why the stock trades the way it does. The argument for why this cycle runs longer is specific and, for once, comes with numbers attached rather than just a story.

First, the demand side is structural rather than seasonal. An AI training server does not use marginally more memory than a laptop, it uses many multiples more, because every additional accelerator needs its own bank of HBM sitting inches away from the compute. A single Nvidia rack-scale system in the NVL72 family carries more than 20 terabytes of HBM; AMD's competing Helios rack carries roughly 31 terabytes, per company disclosures. Multiply that by the pace at which hyperscalers are building data centres and the memory bill scales in a way ordinary PC or phone demand never did.

Second, the supply side cannot respond quickly even if producers wanted it to. Micron's own next new fab, part of a roughly $200 billion multi-year US investment plan spanning Idaho, New York and Virginia, does not begin DRAM production until around mid-2027. Samsung and SK Hynix are in a similar position: new capacity from either does not ramp before the second half of 2027, according to industry reports. Deloitte has estimated combined capital spending across Micron, Samsung and SK Hynix could rise nearly 340% between 2024 and 2027, and even that spending will not show up as usable output for another year or more. Reports citing supply chain research firm Digitimes suggest customers across the industry are currently receiving only around 60% to 70% of the DRAM and HBM volumes they have requested for 2027.

Micron itself has put a number on the opportunity: it forecasts the total HBM market growing at roughly 40% a year, from about $35 billion in 2025 to around $100 billion by 2028. Whether that number proves exactly right matters less than the direction. A company selling a product it cannot make fast enough, with the next wave of competing capacity still more than a year from mattering, is not the same company that used to swing from feast to famine every eighteen months.

Read more