WSJ : OpenAI Closes Silicon Valley’s Largest-Ever Funding Round

OpenAI Closes Silicon Valley’s Largest-Ever Funding Round
The $122 billion round includes Amazon, Nvidia, SoftBank, wealthy investors and a money manager that plans to add the startup to its exchange-traded funds

  • OpenAI completed the largest funding round in Silicon Valley history, raising $122 billion ahead of an expected blockbuster IPO.
  • The company further diversified its shareholder base, raising more than $3 billion from wealthy investors and is set to join ARK Invest ETFs.
  • OpenAI is shifting its strategy to build a “superapp” for developers and business users, expecting roughly half its revenue to come from enterprises by year-end.

OpenAI completed the largest funding round in Silicon Valley history, raising $122 billion ahead of a blockbuster IPO expected by the end of the year.

The deal came with an additional perk: greater access to individual investors. As part of the financing, OpenAI raised more than $3 billion from wealthy investors through banks and said it would be included in several exchange-traded funds managed by ARK Invest, the investment firm led by technology bull Cathie Wood.

The deal shows how the ChatGPT-maker is diversifying its shareholder base ahead of its planned public listing, specifically among individual investors keen to gain exposure to some of the hottest names in the artificial-intelligence boom. OpenAI shares are also a small part of mutual funds and ETFs run by T. Rowe Price and Fidelity.

Amazon, Nvidia, and SoftBank committed $110 billion to the funding round, while additional cash came from Silicon Valley and Wall Street investment firms.

The new funding puts OpenAI on strong financial footing, giving it a crucial cash cushion to continue spending aggressively on AI chips for its research and products. It is also a signal of the company’s enduring appeal among top investors, even as rivals begin to close the gap in the AI race.

OpenAI is in the middle of a major strategy shift to concentrate its resources around building a new “superapp” tailored toward developers and business users. The company recently ditched its much-hyped Sora video app, and is looking at other areas to cut so it can focus its efforts on building so-called productivity tools, particularly coding assistants.

The company said it expects roughly half of its revenue to come from enterprises by the end of the year.

“AI is driving productivity gains, accelerating scientific discovery, and expanding what people and organizations can build,” OpenAI said. “This funding gives us the resources to continue to lead at the scale this moment demands.”

WSJ : Ares Management Raises Nearly $10 Billion for Opportunistic Private-Credit

Ares Management Raises Nearly $10 Billion for Opportunistic Private-Credit Strategy
The firm targets credit and equity deals that provide liquidity to stressed and healthy companies

  • Ares Management has raised over $9.8 billion for its opportunistic credit strategy, closing its third fund for such deals along with related transaction vehicles.
  • The fund secured $8.3 billion in equity commitments, with Calpers pledging $1.5 billion and other institutional investors participating.
  • Ares uses its opportunistic credit strategy for high-yield debt and equity investments in stressed public and private businesses.

Ares Management has raised over $9.8 billion for its opportunistic credit strategy, closing its third fund for such deals along with related transaction vehicles even as private credit as an asset class has come under increased scrutiny.

The fund
The firm collected $8.3 billion in equity commitments for Ares Special Opportunities Fund III, with the remaining capital coming from commitments to the related vehicles. The total doesn’t include anticipated leverage.

Ares has already committed over $1.8 billion from the new fund, according to the firm. The firm began raising the fund in May 2024 and held a first close for it in December of that year. It initially targeted $7.1 billion for the vehicle, according to Ares.

In the broader market, questions in recent months over the risks of backing some types of software businesses have rocked private-credit investment vehicles geared to individual investors. But the new Ares fund drew its capital largely from traditional institutional investors.

Participants in the fund include the California Public Employees’ Retirement System, which pledged $1.5 billion, according to the WSJ Pro Private Equity LP Commitments database. The Virginia Retirement System and the Teachers’ Retirement System of Louisiana also made commitments of $275 million and $150 million, respectively.

The firm’s predecessor fund for the strategy collected $7.1 billion by the time it closed in 2022, according to a news release.

The strategy
Ares pursues high-yield debt and equity investments in publicly traded as well as private businesses through its opportunistic credit strategy, including backing stressed or distressed companies during market dislocations.

The firm finds deal opportunities in companies stressed by temporary pressures, such as elevated borrowing costs or near-term refinancing needs. While distressed companies face more acute balance sheet or liquidity issues, Ares views them as still offering pathways to recovery with infusions of new capital, restructuring or operational improvements.

Ares’s opportunistic strategy is largely industry agnostic. Through the strategy, the firm backs midsize companies to support expansion, provide refinancing and return capital to investors in the businesses.

Target companies typically have earnings before interest, taxes, depreciation and amortization that range from $50 million to $250 million.

Ares prefers to back companies that operate in sectors undergoing transformational changes or businesses with balance sheets that have become overly complex, according to the firm’s website.

The firm writes checks averaging from $150 million to $500 million through the strategy, and can scale them up with participation of other Ares capital pools.

Ares plans to use the new fund to provide debt, equity and hybrid financing in deals that can fill gaps between private-equity control investments and more traditional corporate lending. The firm also uses the fund to purchase stressed publicly traded corporate debt.

The context
Opportunistic credit has found recent fundraising success across asset managers, with notable raises by blue-chip firms in the private-equity industry.

Blackstone raised $5.9 billion for an opportunistic credit fund that closed in January. In 2024, Carlyle Group closed its third opportunistic credit fund with more than $7 billion.

The Ares opportunistic credit strategy, co-led by Aaron Rosen and Craig Snyder, became part of the firm’s extensive credit operations in February 2024. It had been housed in the firm’s private-equity group. Ares manages over $405 billion in credit assets with a team that has more than 560 investment professionals.

Since inception, the strategy has deployed more than $17 billion.

>>> PrismML 1-bit Bonsai — Investment Implications for Hyperscalers & Infra

PrismML 1-bit Bonsai — Investment Implications for Hyperscalers & Infra

The Core Tension
This is a genuine efficiency breakthrough, not marketing noise — Caltech math, Khosla conviction, Cerberus (ex-Google Silicon, ex-Nvidia SoC) on the cap table. The question for equity positioning is the classic Jevons Paradox vs. Substitution debate, but with sharper edges than usual.

NVIDIA (NVDA) — Net Ambiguous, Short-Term Headwind Narrative
Bear case the market will price:
  • 1-bit models run on CPUs, NPUs, mobile silicon — bypassing the GPU stack entirely for inference
  • 8x speed improvement + 75-80% energy reduction at the model level = you need far fewer H100s for equivalent inference throughput
  • The "future hardware designed for 1-bit" comment is the real threat: if the compute primitive shifts from FP16/BF16 matrix multiply to integer add/subtract, NVIDIA's architectural moat (Tensor Cores, CUDA ecosystem optimized for multiply-accumulate) gets structurally challenged
  • Edge/on-device inference cannibalizes the data center inference build-out thesis that's driving NVDA's 2025-2026 earnings narrative
Bull case / Jevons offset:
  • Cheaper inference → more applications → more aggregate model calls → training demand doesn't compress (1-bit is an inference optimization, training still needs FP precision)
  • Hyperscaler capex is as much about sovereign/competitive signaling as utilitarian ROI — unlikely to stop building
  • PrismML's model is 8B parameters — the frontier training race (GPT-5 class, 1T+ params) is completely unaffected
  • NVIDIA still wins if 1-bit hardware never materializes at scale (ARM, Intel, Qualcomm would need to build dedicated silicon)
Verdict: Short-term negative sentiment catalyst, especially if the open-source traction is strong. Medium-term the Jevons offset likely dominates, but the inference-as-a-moat narrative for NVDA weakens. Watch how management frames inference vs. training revenue split next earnings.

Hyperscalers (MSFT/Azure, AMZN/AWS, GOOGL/GCP)
Structurally mixed but leaning negative on unit economics:
  • If enterprise clients can run comparable models locally or on cheaper edge silicon, cloud inference revenue per query compresses
  • The 2TB → 150GB compression means storage and egress costs drop dramatically — deflationary for cloud ARPU on AI workloads
  • However, hyperscalers own the training infrastructure — this doesn't move
  • Google is interesting: Bonsai 8B was trained on Google v4 TPUs — Google captures training revenue even as PrismML disrupts inference economics
Potential positive: Hyperscalers could license or integrate 1-bit models to dramatically lower their own serving costs and expand margin on AI APIs — net positive if they're on the right side of the stack.

Winners in the 1-bit World
Name Angle
Qualcomm (QCOM) NPUs in Snapdragon designed for on-device inference — direct beneficiary of models that run on phones
ARM Holdings Edge silicon architecture wins if 1-bit deployment explodes
Apple Neural Engine + privacy story — on-device LLMs become genuinely capable
SMCI / Dell Edge server buildout for industrial/robotics applications
Cerebras / Groq Specialized inference silicon could be redesigned for 1-bit primitives

WSJ : Caltech Researchers Claim Radical Compression of High-Fidelity AI Models

Caltech Researchers Claim Radical Compression of High-Fidelity AI Models
PrismML says its 1-bit large language model achieves radical compression without sacrificing performance, lowering energy consumption

  • PrismML open-sourced a 1-bit AI model that radically compresses its size without compromising performance.
  • The 1-bit technology enables AI to run locally on consumer devices and industrial-edge devices, boosting data center efficiency.
  • PrismML’s Bonsai 8B model reduces memory to 1 gigabyte, boosts processing speeds by eight times, and raised $16.25 million.

A team of researchers led by California Institute of Technology computer scientist and mathematician Babak Hassibi says it has created a large language model that radically compresses its size without compromising performance.

The company, PrismML, came out of stealth Tuesday and open-sourced its 1-bit technology model, enabling others to use it.

PrismML has developed an extreme form of compression that allows AI to run locally on phones, laptops and other devices, and enables data center build-outs that can do more with fewer resources and avoid ballooning energy costs, according to Hassibi.

“We spent years developing the mathematical theory required to compress a neural network without losing its reasoning capabilities,” said Hassibi, chief executive of the venture. “We are creating a new paradigm for AI: one that adapts to diverse hardware environments and delivers maximum intelligence per unit of compute and energy,” he said.

The other Caltech-affiliated co-founders include Sahin Lale, Omead Pooladzandi, and Reza Sadri, who is also vice president, strategy.

The intellectual property is owned by Caltech, and PrismML is the sole exclusive licensee, Hassibi said.

The company raised $16.25 million in a SAFE and seed round with investors Khosla Ventures, Cerberus Capital and Caltech. A SAFE, or Simple Agreement for Future Equity, occurs when an investor gives a startup money in exchange for the right to receive equity later.

AI’s future won’t be defined by who can build the largest data centers, but by who can deliver the most intelligence per unit of energy and cost, according to investor Vinod Khosla. “So this is not a minor iteration. This is a major technical breakthrough,” Khosla said. “It’s a mathematical breakthrough, not just another tiny model.”

PrismML answers a need, Khosla said, for fast and small but high-performing models that serve a range of applications from voice conversation to some aspects of finance.

The models PrismML developed are designed to operate on consumer devices such as smartphones and laptops, as well as industrial-edge devices. The idea is to enable applications in robotics, wearables, and personal computing that were previously impractical, the company said.

“You can fit a much better model on a phone. That’s a big deal. Of course on your phone or a mobile device, energy consumption is a very, very big deal,” Khosla said.

The same efficiency gains that enable local deployment also allow data centers to operate more effectively, PrismML said.

While the broader tech industry fiercely debates whether the future of AI lies in transformers, diffusion models, or newer concepts, PrismML’s mathematical framework can be applied to any of them, according to Hassibi.

How It Works
One way to describe an AI model is in terms of bits, which refers to the amount of code needed to render a numerical value in ones and zeros, the language of computing. Most AI models are written with a 16-bit level of precision, although some approaches employ 4-bits or fewer. PrismML has achieved a mathematical breakthrough that achieves a 1-bit model without compromising the reasoning, coding, and general knowledge capabilities of full-precision models, according to Hassibi. The mathematics are proprietary, but Hassibi said the effect was much like compressing a digital photograph without losing visual fidelity.

When it comes to running AI models, delays, known as latency, and energy consumption are tied to moving data in and out of memory. By reducing the units of data, or model weights, to a single bit represented by +1 or -1, PrismML’s flagship 1-bit Bonsai 8B model can boost processing speeds by as much as eight times compared with a 16-bit model, Hassibi said. It can also achieve reductions in energy consumption of up to 75% to 80% on current hardware platforms, Hassibi said. If future hardware is designed specifically for one-bit models, it will eliminate the need for complex mathematical multiplications altogether, he said. The hardware would only need to perform simple addition and subtraction, which would drive energy consumption and latency down even further, according to Hassibi.

Amir Salek, senior managing director at Cerberus Capital Management, said he was convinced PrismML achieved a major mathematical breakthrough with the potential to improve the economics of AI.

By employing a 1-bit architecture, a two-terabyte model instantly becomes 150 gigabytes, according to PrismML. “Your bandwidth requirement significantly drops, your memory storage size significantly drops and the energy that you consume to move the data…is reduced in a big way,” Salek said. He was previously founder and Head of Silicon for the Google Technical Infrastructure and Google Cloud businesses. Before that, he was founder and head of Nvidia’s System-on-a-Chip Design organization.

Developers, researchers, and other users can download PrismML’s open source 1-bit Bonsai 8B model free.

Bonsai 8B is an 8-billion parameter large language model, trained using Google v4 TPUs.

The model achieves high-fidelity reasoning and language understanding comparable to 16-bit models, but with a memory footprint of 1 gigabyte vs 16 gigabytes, according to PrismML. High-fidelity reasoning is the capability to successfully perform complex reasoning.

The company said it is also releasing 1-bit Bonsai 4 billion parameter and 1.7 billion parameter models, with 0.5 gigabytes and 0.24 gigabytes memory footprints, and even higher intelligence density.

>>> US Gapping downGapping down

Gapping down
In reaction to earnings/guidance:
  • PHR -25.4%, TE -16.3%, JFIN -2.7%, CEG -2.4% (guidance), ANNA -2.3%, HDL -2%, ALTI -1.7%
Other news:
  • PEPG -42.1% (topline results from lowest dose (5 mg/kg) MAD cohort in the ongoing Phase 2 FREEDOM2 Study)
  • SVC -17.3% (stock offering)
  • IONR -15.2% (secures court ruling upholding permit for Rhyolite Ridge project)
  • SATL -10.7% (entered into $50 mln Sales Agreement)
  • TRMK -7.3% (names new CFO and COO)
  • PALI -2.8% (topline data from phase 1b clinical study of PALI-2108)
  • JBSS -2.2% (declares special cash dividend of $1.50/share)
  • MTB -2% (authorizes new $5 bln share repurchase program)
  • BIIB -1.5% (Apellis Pharmaceuticals to be acquired by Biogen (BIIB) for $41.00/share in cash at closing)

>>> US Gapping up

Gapping up
In reaction to earnings/guidance:
  • SOGP +26.7%, MNSO +8.8%, SPCE +7.8%, FDS +5.6%, TOYO +3.6%, MKC +3.6%, PRGS +2.2%, BITF +1.6%, CHA +1.1%
Other news:
  • APLS +140.4% (Apellis Pharmaceuticals to be acquired by Biogen (BIIB) for $41.00/share in cash at closing)
  • CNTA +48.6% (Centessa Pharmaceuticals to be acquired by Eli Lilly (LLY) for $38/share in cash plus CVRs worth up to $9/share) SOGP +25.7% (announces special cash dividend)
  • SRRK +12.6% (resubmitted the BLA for apitegromab for the treatment of children and adults with SMA to the FDA)
  • AGIO +9.9% (moves mitapivat toward potential accelerated approval in sickle cell disease)
  • FDS +5.6% (strategic partnership with Finster AI to power new AI banking platform)
  • NBIS +3.6% (to build 310 MW AI data center in Finland, expands capacity pipeline)
  • USAU +3.1% (delivers robust feasibility study for CK Gold Project highlighting attractive economics and detailing relative low development risk)
  • RKLB +2.4% (receives regulatory approval to acquire Mynaric)
  • GLPG +2.4% (Galapagos NV and Gilead (GILD) enter collaboration to advance T cell engager program)
  • TNXP +2.2% (reports presentation of Phase 1 data and outlines planned adaptive Phase 2 field study of TNX-4800 for the prevention of Lyme Disease, at the World Vaccine Congress Washington 2026)
  • RCAT +2.1% (discloses Share Purchase Agreement details with respect to the acquisition of all the shares in the capital of Quaze Technologies)
  • OCS +2% (receives EMA PRIME designation for Privosegtor in optic neuritis)
  • CGAU +1.9% (COO to depart)
  • SNY +1.9% (receives EU approval for Rezurock in chronic graft-versus-host disease)
  • AXP +1.5% (named official payments partner of the NFL)
  • ARRY +1.2% (opening of the new headquarters of APA Solar)
  • RTX +1% (awarded a $3.8 bln modification to Navy contract; also awarded a $125 mln U.S. Transportation Command contract)
  • LLY +1% (Centessa Pharmaceuticals to be acquired by Eli Lilly (LLY) for $38/share in cash plus CVRs worth up to $9/share)

>>> US Research Calls I

Research Calls I
  • Upgrades:
    • 10x Genomics (TXG) upgraded to Outperform from Market Perform at William Blair
    • Beiersdorf (BDRFF) upgraded to Neutral from Sell at UBS
    • Croda (COIHY) upgraded to Hold from Reduce at Kepler Cheuvreux
    • Diageo (DEO) upgraded to Buy from Hold at Deutsche Bank
    • Enagas (ENGGF) upgraded to Equal Weight from Underweight at Morgan Stanley
    • Highwoods (HIW) upgraded to Equal Weight from Underweight at Morgan Stanley, tgt $23
    • Perimeter Solutions (PRM) upgraded to Buy from Neutral at UBS, tgt $30
    • Shake Shack (SHAK) upgraded to Neutral from Underperform at BofA Securities, tgt $101
    • Universal Music Group (UMGNF) upgraded to Buy from Neutral at Guggenheim
    • Western Digital (WDC) upgraded to Outperform from Market Perform at Bernstein, tgt $340
  • Downgrades:
    • BASF (BASFY) downgraded to Hold from Buy at Kepler Cheuvreux
    • Colgate-Palmolive (CL) downgraded to Hold from Buy at TD Cowen, tgt $85
    • Lincoln Electric (LECO) downgraded to Hold from Buy at Jefferies, tgt $280
    • Phreesia (PHR) downgraded to Neutral from Overweight at JPMorgan, tgt $16
    • Phreesia (PHR) downgraded to Neutral from Buy at Citigroup, tgt $10
    • Phreesia (PHR) downgraded to Hold from Buy at Truist, tgt $11
    • Phreesia (PHR) downgraded to Neutral from Outperform at Robert W. Baird, tgt $10
    • Rockwell Automation (ROK) downgraded to Hold from Buy at Jefferies, tgt $380
    • Vertiv (VRT) downgraded to Hold from Buy at Jefferies, tgt $260
  • Others:
    • American Tower (AMT) initiated with a Buy at Truist, tgt $205
    • Amphenol (APH) assumed with Buy from Hold at Jefferies, tgt $145
    • Arista Networks (ANET) initiated with a Buy at Truist, tgt $161
    • Block (XYZ) initiated with a Buy at Loop Capital, tgt $75
    • Bridger Aerospace (BAER) initiated with a Buy at Stifel, tgt $3.35
    • Cantor Equity Partners II (CEPT) initiated with a Buy at Benchmark, tgt $16
    • Cisco (CSCO) initiated with a Buy at Truist, tgt $94
    • Corning (GLW) initiated with a Hold at Truist, tgt $125
    • Crown Castle (CCI) initiated with a Hold at Truist, tgt $90
    • Dell Technologies (DELL) initiated with a Hold at Truist, tgt $170
    • Digital Realty (DLR) assumed with a Buy at Truist, tgt $207
    • DigitalBridge (DBRG) assumed with a Hold at Truist, tgt $16
    • DocuSign (DOCU) reinstated with an Underperform at BofA Securities, tgt $52
    • Emerson (EMR) assumed with Buy from Hold at Jefferies, tgt $175
    • Envoy Medical (COCH) initiated with a Buy at H.C. Wainwright, tgt $2.50
    • Equity LifeStyle (ELS) initiated with an Outperform at Mizuho, tgt $72
    • Fiserv (FISV) initiated with a Hold at Loop Capital, tgt $62
    • GE Aerospace (GE) initiated with a Neutral at Daiwa, tgt $301
    • Global-e Online (GLBE) reinstated with a Buy at BofA Securities, tgt $43
    • HeartBeam (BEAT) initiated with a Buy at B. Riley, tgt $4
    • HPE (HPE) initiated with a Buy at Truist, tgt $31
    • HubSpot (HUBS) reinstated with a Buy at BofA Securities, tgt $300
    • Ingram Micro (INGM) initiated with a Hold at Truist, tgt $25
    • Jack Henry (JKHY) initiated with a Buy at Loop Capital, tgt $197
    • Keysight Technologies (KEYS) initiated with a Hold at Truist, tgt $310
    • Mastercard (MA) initiated with a Buy at Loop Capital, tgt $631
    • MiniMed (MMED) initiated with a Buy at Goldman, tgt $24
    • MiniMed (MMED) initiated with an Outperform at William Blair
    • MiniMed (MMED) initiated with an Overweight at Wells Fargo, tgt $26
    • MiniMed (MMED) initiated with an Outperform at Evercore ISI, tgt $20
    • MiniMed (MMED) initiated with a Buy at Citigroup, tgt $23
    • MiniMed (MMED) initiated with an Overweight at Morgan Stanley, tgt $19
    • MiniMed (MMED) initiated with an Outperform at Mizuho, tgt $21
    • MiniMed (MMED) initiated with a Neutral at Piper Sandler, tgt $16
    • MiniMed (MMED) initiated with a Buy at BofA Securities, tgt $27
    • Motorola Solutions (MSI) initiated with a Buy at Truist, tgt $540
    • NexMetals Mining (NEXM) initiated with an Outperform at Raymond James
    • NiSource (NI) initiated with an Overweight at KeyBanc, tgt $52
    • Paccar (PCAR) reinstated with a Neutral at BNP Paribas, tgt $126
    • Passage Bio (PASG) initiated with an Outperform at Oppenheimer, tgt $30
    • PayPal (PYPL) initiated with a Hold at Loop Capital, tgt $46
    • Regeneron (REGN) assumed with an Overweight at Piper Sandler, tgt $875
    • SBA Communications (SBAC) initiated with a Hold at Truist, tgt $193
    • Sun Communities (SUI) initiated with an Outperform at Mizuho, tgt $143
    • Terex (TEX) resumed with an Overweight at Barclays, tgt $65
    • Toast (TOST) initiated with a Hold at Loop Capital, tgt $26
    • Veralto (VLTO) assumed with Buy from Hold at Jefferies, tgt $110
    • Versant (VSNT) initiated with a Buy at Seaport Research, tgt $45
    • Visa (V) initiated with a Buy at Loop Capital, tgt $387

>>> Berkshire Hathaway: Warren Buffett CNBC interview

Berkshire Hathaway: Warren Buffett CNBC interview
Mr. Buffett said:
  • Berkshire Hathaway is the largest bidder for T-bills.
  • Stocks do not look cheaper to him right now.
  • The recent decline is "nothing" compared to recent large declines in the market.
  • If there is another big decline in the market, he will deploy cash.
  • Berkshire bought $17 bln in Treasury bills this week.
  • He sold some shares of Apple (AAPL) too soon, but it is still the largest investment.
  • The Fed is going to print money.
  • He would care about inflation and the stability of the banks if he were at the Fed.
  • Troubles with parts of the banking system can spread to other parts of the banking system.

>>> PHARMA/BIOTECH M&A — QUICK UPDATE

═══════════════════════════════════════════════════════════════════
LAURENT CHEKROUN | 31 MAR 2026
PHARMA/BIOTECH M&A — QUICK UPDATE
═══════════════════════════════════════════════════════════════════

Markets selling off today. Two more deals just dropped anyway.

TWO NEW TODAY
─────────────
LLY / CENTESSA | Up to $7.8bn
$38/sh cash + CVR $9 | Premium 40.5% to 30d VWAP
Asset: Cleminorexton (OX2R agonist) — Ph2a narcolepsy/IH
Read: LLY building CNS franchise alongside GLP-1
Orexin = direct intervention on sleep-wake master switch

BIIB / APELLIS | $5.6bn | $41/sh all-cash
Assets: Syfovre (GA, $500m+ rev) + Empaveli (PNH)
Read: Viehbacher buying a second act — MS declining,
Leqembi slow. Revenue + complement pipeline.

MARCH 2026 SCORECARD
─────────────────────
8 deals / ~$31bn+ / 25 days / no macro catalyst

Servier → Day One $2.5bn Rare Onco
Novartis → Synnovation $3bn Breast Ca.
Merck → Terns $6.7bn CML
Novartis → Excellergy $2bn Allergy
Gilead → Ouro $2.2bn Autoimmune
Otsuka → Transcend $1.2bn CNS/PTSD
Lilly → Centessa ★ $7.8bn CNS/Sleep
Biogen → Apellis ★ $5.6bn Rare/Kidney

RUMOUR BOARD (active this week)
─────────────────────────────────
NBTX JNJ studying takeover (La Lettre 25 Mar) — denied
RVMD MRK at $28-32bn (Jan) — talks stalled, deal expected
ABVX LLY at €15bn — LLY now on Centessa, watch AZN/BMY/SNY

⚡ IVA FP — UNUSUAL VOLUME TODAY
──────────────────────────────────
Inventiva (IVA FP) -10.42% intraday today (31 Mar)
Open 4.80 → Low 3.85 (09:42) → Recovery to ~4.40 → Fade to 4.30
Heavy swings on numbers; should not move the stock like that
Classic pattern for Biotech
IVA is a NASH Ph3 play (lanifibranor, NATiV3 readout 2026) — already
on the structural M&A candidate list. Unusual activity without news
is worth watching on a day when 2 deals just printed.

THE POINT
──────────
The bid follows patent calendars, not the tape.
$200-300bn in drug LOE by 2030. $1.3tn dry powder. Clock is fixed.
Every sell-off = lower VWAP reference for next bid. Buyers win.

Buy quality biotech on weakness. The floor is structural.

─────────────────────────────────────────────────────────────────
Laurent Chekroun | Freienbach | Confidential / Inst. only
═══════════════════════════════════════════════════════════════════