Alphabet could crush estimates on Cloud strength, Anthropic windfall: BofA
The EPS gap is largely driven by an estimated $80 billion boost to second-quarter operating income from the revaluation of Alphabet's Anthropic stake, after Anthropic's valuation rose from $380 billion in the first quarter to $965 billion in the second. Cloud growth estimates were raised to 70%, supported by demand indicators and a backlog suggesting at least $230 billion in revenue over the next eight quarters. For full-year 2026, Bank of America raised its net revenue estimate by 1% to $427 billion and its EPS estimate by 36% to $19.70, now projecting 16% full-year search growth and 72% Cloud growth. For 2027, the bank raised net revenue estimates by 3% to $537 billion and EPS by 1% to $14.70, with Cloud revenue from the second quarter of 2026 through the first quarter of 2028 now projected at $290 billion, above the current backlog. Given accelerating AI demand, higher component pricing for items like memory, and Alphabet's recent capital raise, Bank of America believes the company could raise its 2026 capex range by roughly 5% to $190 billion to $200 billion.
Wall Street Breakfast Podcast: TSMC Chips In $100B More
Taiwan Semiconductor Manufacturing (TSM) plans to invest an additional $100B to expand its U.S. chipmaking capacity. This expands the company’s total investment plan to $265B. The additional investment will fund the construction of four new chip fabrication plants to produce logic chips with 2-nanometer process technology. Uber Technologies (UBER) launched a public takeover offer for Germany's Delivery Hero (DLVHF), valuing the food delivery company at about $14.8B in equity value. The acquisition is conditional on securing acceptances representing at least 50% plus one share of Delivery Hero's (DLVHF) outstanding stock. Uber (UBER) currently holds an economic interest of about 37% in Delivery Hero (DLVHF), including equity derivatives, while Prosus, which owns about 17% of the company, has agreed to tender its shares into the offer.
In the AI Data Center Buildout, Amazon.com Inc (AMZN) Has a Cost Advantage
Amazon Web Services introduced Loom for AWS, an open‑source platform to help enterprises build and deploy AI agents with strong security controls and governance frameworks. The brokerage expects Amazon to spend $159 billion on AI infrastructure buildout in 2026, unchanged from the previous projection. For 2027, Amazon is now projected to spend $230 billion, up from the prior estimate of $196 billion. All together, Amazon is projected to spend $389 billion on AI capacity expansion between 2026 and 2027. According to BofA Securities, it costs anywhere between $25 billion and $45 billion to build 1 GW of data center capacity. The brokerage believes Amazon has the lowest data center building costs among the top hyperscalers. It estimates Amazon's costs at $25 billion per GW, compared to $37 billion per GW for Google and $45 billion per GW for Meta Platforms.
Meta's data center plans could help solve its AI spending problems
Meta spent a whopping $72.2 billion on capital expenditures, with the bulk of that going toward its AI build-out. That prodigious outlay is going toward facilities like Meta's upcoming Canadian data center announced last week, as well as toward expanding its massive Hyperion data center in Louisiana, which will support 5 gigawatts of capacity. In 2025, Meta spent a whopping $72.2 billion on capital expenditures, with the bulk of that going toward its AI build-out. And it plans to spend even more this year: between $125 billion and $145 billion. If Meta does end up leasing out its data center space, the next question becomes, for how long? SpaceX's deal with Anthropic technically runs through May 2029, but either party can call it off with 90 days' notice.
Visa backs Open USD with new stablecoin platform as Circle faces fresh competition
Visa introduced the Visa Stablecoin Platform on Thursday, giving financial institutions tools to issue, store and transfer Open USD stablecoins. The platform combines wallet infrastructure, stablecoin issuance and Visa's payment network to simplify blockchain-based payments and settlement. Circle shares fell about 5% as competition in the stablecoin market intensified following Visa's support for Open Standard's Open USD. Visa introduced a new platform aimed at making it easier for banks, fintech companies and crypto firms to build products using stablecoins, expanding its push into blockchain-based payments as competition in the sector intensifies. The company announced on Thursday that it was launching the Visa Stablecoin Platform (VSP), an enterprise service that allows institutions to issue, store, transfer and redeem stablecoins through a single Visa-managed system. The platform launched with support for Open USD (OpenUSD), a recently introduced stablecoin from Open Standard, and includes tools for minting and redeeming the token along with wallet infrastructure for managing onchain assets. CEX trading volumes rose for the first time in five months in June, with spot climbing 15.3% to $1.11T and RWA perpetual volumes surging to a record $311B.
Nvidia-backed Fireworks hits $17.5 billion valuation as companies pursue cheaper AI models
The cost of the latest artificial intelligence models is increasingly breeding anxiety among finance executives, who have started directing employees to consider open-source alternatives. That's boosting cloud startup Fireworks, which competes with Amazon and Google to host models that developers can weave into applications. The Nvidia-backed company said Thursday that it has exceeded $1 billion in annualized revenue, five times what it had last year, and it has now raised a $1.5 billion round at a $17.5 billion valuation. Fireworks is much smaller than Anthropic and OpenAI, which investors have valued above $800 billion each this year, nor is it close to the top names in technology, whose market capitalizations are counted in the trillions. But the startup's revenue milestone suggests that companies aren't completely satisfied with the models coming out of the top labs. Shares of easy-to-use cloud infrastructure vendor DigitalOcean are up 149% so far this year as growth has accelerated. CoreWeave, which rents out Nvidia graphics processing units, or GPUs, raised $1.5 billion in an initial public offering last year and is now worth $42 billion. Fireworks gives developers an easy way to adopt models from Chinese companies such as DeepSeek, MiniMax and Z.ai. Open-weight models OpenAI released last year are also available. The idea is for clients to bring their own data that frontier labs don't have and refine models until they deliver state-of-the-art performance for specific tasks, Qiao said. While Anthropic and OpenAI serve up "generalized intelligence," Fireworks can unlock "specialized intelligence," she said. A former Meta director, Qiao and six of her co-founders started Fireworks in 2022. The company employs around 200 people. Qiao expects the head count to reach 600 by the end of 2026. Fireworks now handles 40 trillion AI tokens per day, Qiao said. Google disclosed in May that its AI models were processing about 19 billion tokens per minute for developers, implying more than 27 trillion per day. OpenAI announced in March that its developer tools were working through 15 billion tokens per minute, which would suggest about 22 trillion per day. Each token equates to about three-quarters of a single word. As of last year, about half of Fireworks' revenue came from AI coding startup Cursor, which has become less dependent on OpenAI and Anthropic and built a custom model named Composer. "We are much more diversified right now," Qiao said. In June, Elon Musk's SpaceX agreed to acquire Cursor in a $60 billion stock deal, with the transaction set to close this quarter.
SK Hynix and SanDisk Sink 7%, Micron Falls 5% as China’s CXMT Readies an $8.6B Memory IPO
Micron just reported Q3 FY2026 revenue of $41.5 billion, up 346% year over year (YoY), with GAAP gross margin of 85%, and guided Q4 FY2026 revenue to $50 billion at the midpoint. CXMT’s $8.6 Billion IPO Rattles the Memory Trade
Dallas Fed's Lorie Logan calls for higher interest rates
Dallas Federal Reserve President Lorie Logan called Thursday for modestly higher interest rates, arguing that inflation remains too far above the Fed's 2% target and that current policy is not doing enough to bring it back down. "I currently believe modestly higher interest rates would better balance the outlook and risks for the FOMC's dual mandate goals," Logan said in prepared remarks for a Houston speech. "Every month of above-target inflation has compounded the strain on Americans' budgets." Logan, a voting member of the Federal Open Market Committee this year, said her best judgment is that inflation is heading toward the "mid 2's" — not all the way back to 2%. She pointed to several measures to support that view: core PCE inflation stands at 3.4% and has risen since December, the New York Fed's multivariate core trend model also puts the persistent component of inflation at 3.4%, and market-based non-housing core services inflation has made no progress on a 12-month basis since mid-2024. Logan acknowledged that June's Consumer Price Index data showed a monthly decline, driven in part by falling energy prices and softening housing costs. But she said one positive month was not sufficient. "One month of relief is not enough. It is time to finish the job of restoring price stability," she said. She described the labor market as solid, with the unemployment rate averaging 4.3% in the first half of the year and employers adding an average of 92,000 jobs per month. That stability, she argued, removes a key obstacle to tightening policy. "If inflation is not heading all the way to 2 percent on its own, then at least some policy restriction is needed to help get it there," Logan said. "Better modest restriction now than severe restriction later."
TSMC Adds $100 Billion to U.S. Expansion
Taiwan Semiconductor Manufacturing (TSM, Financials), the world's largest contract chipmaker and a key supplier of advanced processors to companies such as Nvidia and Apple, said it will invest another $100 billion in the United States after strong AI demand helped drive record second-quarter profit. The new commitment adds to the $165 billion already planned for Arizona. The company said the expansion could include four more facilities for advanced chip production and packaging, although the timing will depend on market demand. TSMC also raised its 2026 capital spending forecast to between $60 billion and $64 billion, up from an earlier range of $52 billion to $56 billion. Management now expects full-year revenue in U.S. dollar terms to grow slightly more than 40%, compared with its previous forecast of more than 30%.
Andy Jassy Says Amazon's Chip Business Already Has $225 Billion in Commitments
Amazon's custom chip unit -- Graviton processors, Trainium artificial intelligence (AI) accelerators, and Nitro networking chips, all deployed inside Amazon Web Services (AWS) -- has an annual revenue run rate above $20 billion, growing at triple-digit percentage rates year over year. And customers have lined up. Jassy said in the company's first-quarter earnings call that it now holds more than $225 billion in revenue commitments for Trainium. Numbers like those suggest Amazon is building something bigger than an internal cost-saving project.
Alibaba and Baidu shares jump in Hong Kong on Apple AI partnership
Shares of Chinese tech giants Alibaba and Baidu rose Thursday on their partnership with Apple for deploying their AI tools. Hong-Kong listed shares of Alibaba rose 5% after the company confirmed that its Qwen AI model would be integrated into Apple services in China. U.S.-listed shares of Alibaba had closed slightly higher overnight after an Alibaba spokesperson told CNBC that "Qwen will be integrated into Apple Intelligence experiences within iOS, iPadOS, macOS, and vision OS for users in China." Baidu's Hong Kong-listed shares gained 4% as the company confirmed that it was working with Apple on Apple Intelligence features for iPhones in China. This comes amid reports in late June that its artificial intelligence chip unit Kunlunxin is targeting an initial public offering in the city, which could value its affiliate at $50 billion.
JPMorgan Just Reported $21.2 Billion in Q2 Net Income -- Up 41% -- and CEO Jamie Dimon Said the Economy Is "Close to as Good as It Gets."
JPMorgan Chase generated a record net income of $21.2 billion, up 41% year over year. Revenue also set a record, coming in at $57.3 billion, up 28% year over year and significantly above estimates of $51.1 billion. The bank's credit quality also improved. Net charge-offs, which are bad loans unlikely to be repaid, fell by $44 billion year over year. JPMorgan Chase lowered its net charge-off rate in Card Services to 3.2%, down from its previous guidance of 3.4%. The bank also raised its net interest income guidance for fiscal 2026 from $103 billion to $105.5 billion. Net interest income rose a robust 10% to $25.6 billion, but the real alpha came from noninterest or fee revenue, which surged 45% to $32.4 billion. Within the trading business, equity market trading revenue skyrocketed 86% to $6 billion, fueled by a major market rally in April and May.