Meta (META)
Shares rose almost 4% after the report. EPS came in at $3.83, a beat. The estimate range was wide, from about $3.45 up to $6.50, because the legal expenses for the third quarter were hard to weigh.
Revenue for Q3 hit $64.6 billion, above the $63.3 billion expected. Ad revenue was $63.06 billion, beating the $61.85 billion expected. Operating income was $12.9 billion against a $20.85 billion expectation. Family of apps revenue was $64.16 billion, above the $62.75 billion estimate, but its operating income of $17.12 billion missed the $23.3 billion estimate. Reality Labs revenue missed at $443 million versus $517.7 million expected, but its loss narrowed to $4.13 billion, better than the $5.13 billion estimated loss.
Guidance: Q4 revenue of $73 to $76 billion versus a $73.57 billion estimate, so the midpoint is higher. 2026 capital spending is set at $135 to $145 billion, dead on expectations.
The one-off item distorting the numbers was about $10 billion in legal expenses tied to a settlement with 52 attorneys general. Some had expected the bottom line to carry noise from this multi-state settlement, so part of it was already priced in. The settlement seems mostly behind the company now.
Meta will declare a one-time stock dividend of one share of Class C common stock for each outstanding Class A and Class B share. This looks similar to Google's structure with voting rights on Class A and separate C shares, likely an effort to raise capital. Of the new issuance, 20% will stay with the owners and 80% will reach the market over time. This does not dilute current shareholders in value or voting rights, and mechanically the A and C shares should not disconnect much in price. It funds the AI buildout without adding debt to the balance sheet, which is smart in an environment critical of debt raises. This is a key reason for the strong bid.
Meta will not provide a fiscal year 2027 revenue outlook. Guidance usually comes on the conference call, but it does not look like that revenue guidance will be issued, and the market is shrugging that off.
Meta's AI app saw 10 million downloads in 40 days. Advertising expectations around it are good. A big market question is whether this report would be a clearing event for the AI trade, since Meta is one of the big spenders and many want proof of monetization. One view from an analyst: on the call, investors want to hear Zuckerberg take the same approach Elon Musk did with SpaceX and the Grok bot, calling the AI product the fastest growing product he has ever seen. Such comments would be very positive for the stock.
Separately, Zuckerberg said he may leave California if Prop 40 passes. Prop 40 is a billionaire tax imposing at least a 5% tax on certain taxpayers with assets over $1 billion. It has not passed, and there is a lot of opposition to it.
Meta closed the session up about 3%.
Lam Research (LRCX)
Shares fell about 5.5% to 6%, a sharp drop despite a decent report. The first quarter EPS was $2.20, beating estimates by a penny. The initial number that crossed the wire was wrong; the correct figure is $2.20.
Revenue rose 54% year-over-year to $8.22 billion, topping the $8.13 billion estimate. Systems revenue was $5.73 billion, up 62% year-over-year, above the $5.6 billion estimate. Customer support revenue was $2.49 billion, up 40% year-over-year, but missed the $2.6 billion estimate. Adjusted gross margins expanded to 52.4% from 50.6% a year ago, slightly above the above-52% estimate. Adjusted operating margins were 40.2%. Net income was $2.75 billion.
Guidance: revenue of $8.2 to $9 billion versus an $8.6 billion estimate, right at the midpoint. Adjusted gross margin of 51.5% to 53.5% versus 51.9% expected. Adjusted operating margins of 39.5% to 41.5% versus 39.7% expected, roughly in line.
China sales are around 35% of total revenue and have been improving over recent quarters. Korea sales are around 20% of total revenue. Nothing in the headline numbers explains the pullback. It may be commentary on the call about supply chain issues and possible price increases.
Lam closed down 5.8%.
Tesla (TSLA)
Reported a little ahead of its scheduled time. Q3 EPS was 48 cents, beating the 44 cent estimate but down from 50 cents a year ago. Revenue was $28.05 billion, just under the $28.13 billion estimate and down 0.2% year-over-year.
Automotive revenue of $20.56 billion beat the $19.83 billion estimate. Energy generation and storage revenue was $2.88 billion, a big miss against the $3.91 billion estimate, more than a billion dollar shortfall. Services and other revenue was $4.61 billion, beating the $4.3 billion consensus. Gross margins were 19.6% versus 17.8% expected.
Free cash flow was positive at $1.53 billion, against an estimate of negative $4.99 billion, a $6.5 billion swing from expectations. This turned positive after being negative the prior quarter; Q1 was $1.44 billion. The surprise negative free cash flow last quarter had been a major concern, and this result nearly returned the company to Q1 levels.
Active full self-driving subscriptions reached 1.74 million, beating the estimate of just over 1.5 million. FSD attach rate exceeds 75% in North America. The quarter ended with the largest passenger vehicle backlog in four years. Tesla has begun production of Optimus in Fremont and the Tesla Semi in Nevada. Austin fab equipment commissioning starts in 2027. Tesla plans to reach 400 megawatts of AI training compute by year end.
Margin improvement stood out, especially since US subsidies are gone, showing normalization without government stimulus. FSD adoption beat handily and is a services-type revenue that could grow substantially; what seemed a fantasy a couple years ago is now real. The energy storage miss was not surprising, since analysts can forecast that business two to three quarters out given its scale. Tesla delivered more vehicles than it produced, drawing down inventories, which is a tailwind. Overall a strong report.
Key questions for the call: color on where capex is going, Optimus production lines, the cyber cab ramp, the terafab factory, FSD adoption given European developments, next target cities for robotaxis and cyber cabs, and an update on Optimus humanoids, production, deliveries, and deployments. On a possible Tesla/SpaceX merger, Musk hinted last night he would support it. There are questions about how closely the two stocks could trade in tandem, shareholder approval, and other parties involved, including China.
Tesla moved up from near 2% to close about 1.3% higher.
IBM (IBM)
Shares jumped about 6.5% to 7%. EPS was $2.95, beating the $2.86 estimate and up 11.32% year-over-year. Revenue rose almost 5% to $17.1 billion, above the $16.83 billion expected.
Software revenue grew more than 8% to $7.8 billion, a beat. Infrastructure rose 4.2% to $3.71 billion, above the $3.57 billion estimate. Consulting was $5.34 billion, slightly below the $5.39 billion estimate. Free cash flow was a standout at $3.03 billion, well above the $2.63 billion estimate; investors wanted earnings growth turning into cash.
About two-thirds of mainframe and associated software deals that slipped in Q2 closed in Q3. CFO Jim Kavanaugh said the company is optimistic about most of the rest. IBM reaffirmed full-year constant currency revenue growth of 4% to 5%, in line with consensus. Beyond the deferred deals, 75% of net new deals are still ahead in coming quarters, so they backfilled order flow too. The key for IBM is sticking to its timeline for onboarding customers and closing deals. Software revenue growth of about 7% to 8%, up from a 6% to 8% range, is a low bar for an industry that usually pushes for 15% comps, but this is a traditional business the market understands. The deferred sales showing up rather than slipping further avoided what would have been a big red flag.
IBM closed up about 7%.
Texas Instruments (TXN)
Shares were fractionally higher, near the unchanged line, up about 1.84% by the close. EPS was $2.62, beating estimates of $2.41 to $2.44, up 77% year-over-year. Revenue was $6.06 billion, up 28% year-over-year, above the $5.93 billion estimate.
Analog revenue was $4.91 billion, up 32% year-over-year, a beat. Embedded processing revenue was $858 million, up 21%, a beat. Operating profit was $2.87 billion versus $2.61 billion expected. Free cash flow was $2.96 billion, well above the $2.25 billion estimate. Capex was $463 million, below the $650.6 million the street expected.
Q4 guidance: revenue of $5.8 to $6.3 billion and EPS of $2.45 to $2.79, both beating consensus ranges. The CEO said revenue grew 11% sequentially and 28% year-over-year, with broad growth in industrial, automotive, and data center, and that trailing 12-month operating cash flow of $9.9 billion showed the strength of the business model.
The stock has traded well over the past six months and is now consolidating. It is an analog business, the nuts and bolts of getting compute running rather than a new AI frontier. The analog business run may be short-lived, like memory. It is unclear how much premium the market will bid up for it over the next couple weeks as other big movers like AMD report.
