Biotech & AI Fuel Big Stock Moves

Biotech names jump on trial data and revenue surprises

Several healthcare stocks delivered outsized gains after company‑specific catalysts. ImmunityBio’s shares rose 137% week‑on‑week to a record high, supported by a 700% jump in preliminary 2025 revenues from its treatment Anktiva, which generated $113 million on strong demand. The company also reported promising results for its CD19 CAR‑NK therapy combined with rituximab in a rare blood cancer, with disease control in all four enrolled patients and durable complete remissions in two cases without further therapy after the initial dosing cycles.

Erasca advanced 42.3% over the week to a three‑year high as four investment firms issued bullish ratings. Clear Street, HC Wainwright, Piper Sandler and Stifel all assigned “buy” or “overweight” recommendations and lifted price targets, in part reflecting positive preliminary results from Erasca’s pan‑RAS molecular glue degrader ERAS‑0015, with topline data expected in the first half of 2026. The upgraded targets, which now cluster between $5 and $11, underscored growing optimism around the company’s clinical pipeline.

Johnson & Johnson highlighted additional data from its neuropsychiatry portfolio, reporting that CAPLYTA (lumateperone), when added to an antidepressant, produced significantly higher remission rates in adults with major depressive disorder over six weeks compared with placebo plus antidepressant, with benefits maintained through six months in an extension study. The data were among 11 abstracts presented at a major neuropsychopharmacology meeting and fit into J&J’s broader strategy of supporting growth with new medicines across immunology, neuroscience, oncology and other areas.

AI infrastructure build‑out drives sharp rallies in data center and hardware stocks

Companies providing compute and infrastructure for artificial intelligence also saw strong share price reactions. Bitdeer Technologies climbed 37.6% over the week after announcing adoption of NVIDIA GB200 NVL72 systems in Malaysia, which it said would support demanding AI workloads and advance its global AI infrastructure strategy. The firm is converting former cryptocurrency mining sites in Washington and Tennessee into GPU‑optimized data centers and building new facilities in Clarington and Tydal, with combined planned accelerated computing capacity in the hundreds of megawatts.

CoreWeave gained 26.3% to about $101, a two‑month high, as investors responded to Taiwan Semiconductor Manufacturing Company’s plan to raise 2026 capital expenditure to $52–56 billion to meet AI chip demand. TSMC’s fourth‑quarter 2025 net income rose 35% year‑on‑year and net sales increased 20.5%, reinforcing expectations of sustained demand for high‑performance and AI chips that underpin CoreWeave’s cloud services. Separately, Galaxy Digital advanced 37.57% after securing approval from Texas’s grid operator to lift approved capacity at its Helios data center to 1.6 gigawatts, with initial committed power to major client CoreWeave expected in early 2026.

Ultra Clean Holdings, a supplier of subsystems and ultra‑high purity manufacturing services to semiconductor fabs, rose 27.89% to a 52‑week high. The move followed TSMC’s capex announcement and Ultra Clean’s investor presentations emphasizing its role in AI‑related semiconductor equipment, which investors interpreted as positioning it to benefit from increased spending on advanced fabs.

Mixed valuation signals for large‑cap chip and retail leaders

Several long‑established large‑caps extended multi‑year gains while drawing scrutiny on valuation. Applied Materials’ shares traded around $327, leaving the stock up 72% over 12 months and more than 200% over five years. A discounted cash‑flow model cited in recent analysis estimated intrinsic value at about $139.94 per share, implying the stock was roughly 134% above that estimate. On price‑earnings metrics, Applied Materials’ P/E of about 37 times sits below semiconductor peers and the industry average, but above one modelled “fair” multiple of about 32 times, leading that framework to characterize the shares as expensive.

Costco Wholesale also continued a strong run, with the stock last closing at $963.61 and delivering five‑year returns of 181.4%. A cash‑flow‑based fair value estimate of $790.27 per share implied the stock was about 22% above that level, and the retailer scored 0 out of 6 on one set of valuation checks. Its current P/E ratio of 51.52 times earnings stands well above consumer retail and peer averages and above a modelled fair P/E of 36.14 times, leading that analysis to flag the shares as overvalued even as long‑term narratives and some community estimates point to higher potential fair values.

Chipmakers, software platforms deepen AI‑focused investment plans

Micron Technology outlined an aggressive capacity expansion centered on AI memory demand. The company has broken ground on a planned $100 billion megafab complex in New York that could eventually house up to four fabs, and it signed a letter of intent to acquire Powerchip’s P5 fab in Taiwan for $1.8 billion. Meaningful DRAM wafer output from that site is not expected until the second half of 2027, and commentary emphasized that both projects are intended to align Micron’s future high‑bandwidth memory capacity with AI and data center needs while acknowledging the risk from heavy capital intensity in a cyclical market.

In software, GitLab announced general availability of its GitLab Duo Agent Platform, an AI‑driven suite that coordinates automation and assistance across planning, code review, security and CI/CD workflows. The launch introduces GitLab Credits, a usage‑based model for Premium and Ultimate customers that operationalizes the company’s hybrid seat‑plus‑usage pricing. The platform is positioned as an effort to address what GitLab calls the “AI paradox” in software delivery by applying AI beyond code generation. Analysts and community forecasts cited alongside the launch project substantial revenue and earnings growth for GitLab through 2028, while also highlighting execution and competitive risks.