CASH & CHAOS

MARKETS, MONEY AND THE MADNESS IN BETWEEN.

THE WEEKEND BRIEFING

SpaceX rattles telecom. Earnings are next.

Good morning. T-Mobile fell 13.27% on Friday. Crown Castle jumped 15.60%. SpaceX’s spectrum deal sent investors selling wireless carriers and buying the towers a new network might need. The same competitive threat landed very differently on opposite sides of the industry.

That split was more revealing than SPY’s 0.60% gain. Elsewhere, Teva won approval for a new monthly schizophrenia treatment, Netflix reportedly prepared a 5% staff cut, and Lumentum’s booked capacity put AI infrastructure demand back in focus. Next week brings bank earnings, TSMC and inflation data—a test of whether stronger demand is reaching profits, and what investors will pay for them.

THE BIG STORY · TELECOM

SpaceX’s spectrum deal sends carriers sliding—and towers soaring

Illustrative hardware, not the proposed mobile service. Ka23 13 / Wikimedia Commons, 2025. · CC BY 4.0

T-Mobile fell 13.27% to $148.58 Friday. AT&T dropped 10.82% to $22.18; Verizon lost 10.14% to $41.65. Those are October 9 regular-session closes and changes against Thursday. Investors are repricing the possibility that Starlink becomes a broader wireless competitor.

Grain Management announced Thursday that SpaceX agreed to buy its nationwide 800 MHz spectrum portfolio. Spectrum is the licensed radio capacity networks use to carry signals. The agreement needs FCC approval; Grain did not disclose a price.

Low-band frequencies are useful for broad coverage. Pairing them with a satellite network opens more ways to reach customers, although owning licenses is only one part of providing a reliable mobile service. Equipment, network capacity, distribution and customer support still have to arrive.

The earnings concern for incumbents is straightforward. A credible entrant can force lower prices or richer plans before it wins a large subscriber base. That squeezes revenue per customer. If carriers then spend more to defend service quality, the pressure reaches cash flow from both directions.

T-Mobile’s June-quarter revenue was $22.79 billion, with net income of $3.24 billion. Those numbers establish the size of the existing business; they do not measure how much this deal puts at risk. Friday’s price move reflects changed expectations for future returns.

The other side of the trade was striking: Crown Castle gained 15.60%, American Tower 9.30%, and SBA Communications 7.34%. A new network could need terrestrial infrastructure. Investors therefore marked down the companies selling service and marked up businesses that might rent space to whoever builds it.

For dividend investors, the next carrier reports deserve more attention to customer additions, churn, promotional costs and cash available after capital spending. For tower investors, announced contracts and deployments will matter more than another ambitious coverage map. Friday priced a competitive threat; it did not establish its eventual market share.

Read Grain’s agreement and closing conditions → · See T-Mobile’s June-quarter accounts →

HEALTHCARE · TEVA

Teva’s new monthly schizophrenia treatment adds a growth opportunity

Illustrative research photography: Teva.

Teva rose 4.71% to $41.10 Friday after announcing FDA approval of WELTRUZA, its once-monthly injectable formulation of olanzapine for adults with schizophrenia. The company expects U.S. availability in the coming weeks.

This is a product story with a practical commercial angle. Teva says the formulation requires no loading dose, oral supplementation or post-injection monitoring. Reducing the work required around treatment could help adoption, but it is a regulatory milestone rather than proof of future sales.

The financial context is Teva’s shift toward innovative medicines. June-quarter revenue was $4.142 billion, down 1% in dollars. Its established UZEDY injectable generated $77 million in U.S. quarterly revenue, up 43%. That gives the company an existing commercial foothold in long-acting schizophrenia treatments.

A launch can build on relationships with clinicians and payers already familiar with that franchise. That may make each additional sales dollar less expensive to pursue than starting from scratch. The relevant question is whether the new medicine expands the franchise rather than merely moving prescriptions between its products.

There is no verified WELTRUZA revenue forecast in this briefing. Approval must still turn into reimbursement, patient starts and repeat prescriptions. Launch spending also comes before the full sales benefit, so a positive regulatory headline need not immediately lift earnings.

The medicine retains material safety risks, including weight and metabolic effects, and a boxed warning concerning elderly patients with dementia-related psychosis. Those constraints belong in any assessment of adoption; convenience does not erase prescribing considerations.

Teva’s next results call is scheduled for November 3 at 8 a.m. ET. Look for launch timing, access arrangements and whether management changes its broader outlook. UZEDY’s experience provides an operating baseline, but investors will need product-specific evidence before assigning the new treatment the same trajectory.

Read the approval announcement and safety information → · See the existing franchise’s quarterly sales →

STREAMING · NFLX

Netflix reportedly plans a 5% staff cut ahead of earnings

Archive photo, 2018: Coolcaesar / Wikimedia Commons. · CC BY-SA 4.0

Netflix closed Friday at $70.30, down 1.77%. Puck reports that the company is preparing to cut roughly 5% of its workforce, with an announcement possible as early as next week. Treat the plan as reported, rather than a confirmed company restructuring.

The timing gives investors something concrete to ask about at the October 20 results release. A staff reduction can help expenses, but it cannot answer whether subscriber economics, advertising and programming returns are improving.

Netflix’s June-quarter filing shows $12.56 billion of revenue and $3.40 billion of net income. Revenue grew about 13% year over year, supported by memberships, pricing and advertising. This is a profitable business examining its cost structure, rather than a filing that establishes a sudden collapse in sales.

The mechanism matters. Spending less on personnel can lift operating profit if revenue holds up. But the share of employees affected is not the percentage by which total costs fall. Compensation differs across roles, and content production, licensing and technology expenses remain substantial.

Savings must be weighed against severance and disruption. We do not have a verified savings estimate or restructuring charge. Applying the headcount percentage to Netflix’s entire expense base would produce a number with little meaning.

The filing also explains why comparisons of headline net income need care: the first quarter included a $2.8 billion termination fee related to the ended Warner Bros. Discovery transaction. That one-off payment does not establish recurring earning power.

Netflix will post results on October 20 at approximately 4:01 p.m. ET, followed by its management interview at 4:45 p.m. ET. Revenue growth, operating profitability, advertising progress and any quantified restructuring costs should do the talking. A leaner organization helps shareholders when it supports those results; fewer desks alone are not a growth strategy.

Read the reporting on the proposed cuts → · Check Netflix’s filed quarterly accounts → · See the confirmed earnings schedule →

MAG 7 + THREE

Friday’s big-tech scoreboard

October 9 regular close · USD · Changes versus October 8. Mag 7 plus AMD, Broadcom and Micron; our editorial watchlist.

AI HARDWARE · LITE

Lumentum’s booked capacity makes optics an earnings story

Company cleanroom photo: Lumentum.

Lumentum gained 5.22% Friday after CEO Michael Hurlston told Bloomberg its optical-component capacity is booked through early 2029. The company makes components that help transmit data around increasingly demanding AI systems.

Processors need fast connections to work effectively together. That makes lasers and optical links part of the computing bottleneck—and creates an opportunity for suppliers whose names get less attention than the chips they connect.

The latest reported quarter already showed the effect: revenue of $1.006 billion, more than double a year earlier, and a 27.8% GAAP operating margin. Management previously guided the following quarter to $1.225 billion–$1.275 billion of revenue. That outlook remains a forecast.

A full schedule offers visibility, but it also limits immediate sales unless usable capacity expands. Factory spending must become qualified, shippable product before demand becomes revenue. Bookings far into the future are not the same as cash already collected.

Watch production growth and operating margin together. More output with preserved economics would support the investment case; costly expansion or shifting customer schedules would weaken it. The next quarterly update can show whether the company is converting this shortage into profitable deliveries.

Read Bloomberg’s CEO interview reporting → · Compare demand claims with reported margins →

AI BUSINESS

OpenAI’s revenue definitions leave suppliers with a cash-flow question

Archive photo, 2019: Steve Jennings / TechCrunch. · CC BY 2.0

Bloomberg reports that OpenAI expects $70 billion or more in annualized revenue by year-end. Reporting this week put its recent pace near $50 billion. The current figure and the year-end expectation describe different dates.

Annualized revenue projects a recent sales pace over twelve months. It is not a full year of revenue already earned. Comparisons also need consistent accounting: Bloomberg’s Friday report says Anthropic counts gross cloud-partner sales, while OpenAI recognizes its own share.

This distinction reaches public stocks because chipmakers and cloud providers are committing resources to AI demand. The revenue recognized by one company does not automatically become cash available to fund all the infrastructure around it.

Friday’s response was mixed: Oracle gained 4.21%, Microsoft 2.38%, while Nvidia fell 0.52%. Those closing moves describe the session; they do not prove that one revenue headline caused each change.

The useful question is how much customers actually pay, how much the AI provider retains, and when suppliers collect. A consistent revenue definition is the starting point. Margins, cash consumption and contractual payment obligations will determine whether a bigger sales pace supports the spending attached to it.

Read the reported year-end outlook → · See the cloud-partner accounting comparison →

NEXT WEEK · EVENTS & EARNINGS

Banks, inflation and TSMC take the stage

October 12–16 · All times Eastern. The next batch of results puts bank earnings and AI infrastructure economics side by side. These are scheduled events, not reported results; the list is a selection of the week’s major catalysts.

What to watch: For banks, compare net interest income with deposit costs and credit-loss provisions. Higher loan yields help only if funding costs and bad debts do not absorb the benefit. Goldman and Morgan Stanley add a read on trading, deal activity and client assets.

TSMC previously guided third-quarter revenue to $44.6 billion–$45.8 billion and gross margin to 65%–67%. Watch the fourth-quarter outlook and capacity spending: strong AI demand must translate into profitable wafer production. Those ranges are company guidance, not actual results or consensus estimates.

CPI and producer prices can change the rate outlook as those earnings arrive. The combination matters: stronger company results alongside renewed inflation pressure can still leave stock valuations facing higher financing costs.

MACRO

Households feel the pressure from prices and borrowing

Michigan’s preliminary October sentiment index fell to 46.3 from 48.1 in Friday’s release. One-year inflation expectations rose to 4.7% from 4.6%; longer-run expectations reached 3.5%. The survey points to weak buying conditions for durable goods. For consumer stocks, that raises the importance of pricing, product mix and financing costs. Sentiment measures how households feel, rather than their actual spending. Final October results arrive October 23 at 10 a.m. ET.

See the survey and inflation tables →

PREDICTION MARKETS

October’s Fed hold remains the favorite

Polymarket’s October contracts priced no change at 83.5% and a quarter-point increase at 15.5%, retrieved October 10 at 06:00 UTC / 2 a.m. ET. These are separately traded contract prices, not official forecasts.

The contracts resolve using the change in the upper bound of the target federal funds rate after the October 27–28 meeting. A hold this month would not settle the path beyond it; companies still face the financing costs already in place.

Check contract prices and resolution rules →

UPGRADES & DOWNGRADES

Two firms, two American Express targets

October 9 analyst actions: BMO initiated American Express at Outperform with a $380 target. RBC trimmed its target to $400 from $415, keeping Outperform. One is new coverage; the other is a target reduction, not a rating downgrade.

Both ratings remain positive despite different valuations. Targets are analyst opinions, rather than company guidance or promised returns.

BMO initiation · RBC target change

CRYPTO OVERVIEW

Crypto keeps trading through the weekend

Bitcoin: $82,751. Ether: $2,496. Coinbase spot snapshots retrieved October 10 at 06:00 UTC / 2 a.m. ET; rounded to dollars. They are weekend prices, not Friday stock closes.

Coinbase shares finished Friday at $179.39, up 4.30%. Exchange earnings depend on trading volume and fee mix as well as coin prices. The shares pause for the weekend while the underlying assets keep moving, leaving Monday’s stock opening exposed to developments since Friday.

Follow bitcoin on Coinbase →

UNTIL NEXT TIME

Keep your weekend. We’ll keep the watchlist.

Forwarded this email? Get Cash & Chaos in your inbox.

Stock figures: October 9 regular-session closes; daily changes versus October 8. Crypto and prediction markets: October 10, 06:00 UTC. For information, not personalized investment advice.