The one-page fundamentals report for every US stock

Impact 4 · 22 Oct, 3:14 PM

Moderna’s stock tumbled after leaked late-stage melanoma vaccine data suggested a 42% reduction in risk of cancer recurrence or death when combined with Merck’s Keytruda, though earlier trial data showed 49%. The leak sparked a sell-off that wiped more than $23 billion in market value and sent Moderna toward its biggest one-day drop on record. The stock had surged earlier in August after a blast of positive trial news; the breach has traders questioning the full data, which will be released at a major cancer conference in Madrid. Merck, while affected, is not as dependent on the vaccine as Moderna. If the price breaks below key levels, attention shifts to potential downside targets such as $174 and $116, with full data due Saturday."

Impact 4 · 22 Oct, 1:01 PM

Meta Platforms’ shares slid about 5.7% after pre-market gains faded, as regulators disclosed roughly $825 billion in future AI-related capex commitments—up from about $700 billion—covering data centers, energy infrastructure, and computing capacity. The obligations include $355.7 billion in uncommenced leases and $454.5 billion in non-cancelable contracts, with $121.4 billion due in 2027. Third-quarter capex reached $34.91 billion, driving free cash flow negative at $2.12 billion and pushing 2026 capex guidance higher to $135–$145 billion. Revenue rose 26% to $64.60 billion, with Q4 guidance beating consensus. GAAP EPS came in at $3.83 (miss due to a $9.89 billion youth-litigation charge; ex-items, $7.40 would beat). Meta also raised full-year expense guidance to $175–$179 billion and unveiled a new non-voting Class C stock to finance long-term infrastructure. Muse engagement metrics and Watermelon hype fueled optimism, but the liability load weighed on sentiment.

Impact 4 · 22 Oct, 12:53 PM

Five major U.S. banks—Morgan Stanley, JPMorgan Chase, Bank of America, Wells Fargo, and Citi—are restricting or blocking third-party AI agents from logging into customer accounts after detecting automated activity. They differentiate between merely reading data and acting on it, with read-only access via approved open-banking channels already in place in some cases, while credential-based access is being tightened. Citi explicitly forbids third-party AI agents from logging in or transacting, Chase warns customers not to share login details, and Morgan Stanley has begun blocking agents as a precaution. The industry is moving toward scoped, narrowly defined permissions that let an agent view specific information without control over transactions, though standardized frameworks for such access are not yet established. Governments have signaled broader scrutiny of who can access financial data.”

Impact 4 · 22 Oct, 12:28 PM

IBM says large enterprises are tightening AI spend scrutiny, seeking clear ROI and avoiding out-of-control costs. Enterprises are increasingly favoring open-weight models, which are cheaper to run than frontier models, using modest infrastructure (two to three racks) to operate them either on-premises or on private clouds. This shift could curb hyperscalers’ dominant AI infrastructure growth and reshape purchasing choices. Open AI and Anthropic are rolling out cheaper, mid-range options like GPT-6 variants and Claude Haiku 5.5 to meet demand. The trend comes as AI token costs fall and companies weigh computing power against revenue gains, profits, and cost reductions. OpenAI and Anthropic also eye IPO trajectories amid this pricing-and-value compression in enterprise AI.