A daily market update from FS Insight — what you need to know ahead of opening bell

“Embrace iteration as the road to improvement, but don’t let that lull you into rolling out poorly-thought-out crap.” — Kate O’Neill

Over the weekend

In an unprecedented attack, Iran launched a swarm of some 350 drones and missiles at Israel on Saturday; Israel repelled the vast majority with help from U.S. and Arab partners (WSJ)

China develops new light-based chiplet that could power artificial general intelligence — where AI is smarter than humans (LS)

Brokerages shift Fed rate-cut expectations to September after consumer price data (RT)

U.S. Steel shareholders approve sale to Japan’s Nippon Steel; the $14.1 billion deal still faces regulatory reviews, opposition from United Steelworkers union, members of Congress (WSJ)

OpenAI is now 4% of Ark Venture Fund, just behind Anthropic (BBG)

Two brothers, a big biotech bet, and an $8 billion payout (WSJ)

UAE state-owned ADNOC recently eyed $110B oil giant BP as takeover target (RT)

Clearlake Capital Group is exploring a renewed takeover offer for $4.1B cloud software provider BlackBaud (BBG)

CVC looks to raise at least $1.3 billion in Amsterdam IPO (WSJ)

Blackstone, CVC among potential bidders for European festivals organizer Superstruct Entertainment, valued at as much as $1.9B (RT)

Brookfield Asset Management is in advanced talks to buy a majority stake in private credit manager Castlelake for over $1.5B (FT)

CGN New Energy Holdings’ parent China General Nuclear Power considers taking the $1.3B HK-listed power producer private (BBG)

French lender BPCE will acquire Société Générale’s professional equipment financing business for $1.2B (RT)

European self-storage company Shurgard agreed to acquire UK-listed rival Lok’nStore in a $474M all-cash deal (RT)

Investment firm General Atlantic will take a minority stake in marketing software provider HighLevel (WSJ)

Cheap mortgages deter workers from relocating for $250,000 jobs (BBG)

Franklin Templeton turns toward private investing as mutual-fund business shrinks (WSJ)

Zero-emission truck developer Windrose Technology raised a $110M Series B led by HSBC New Economy Fund (PRN)

Factorial, a Catalan business and HR management platform for SMBs, raised $80M in funding from General Catalyst (FN)

Arcadia, a global utility data and community solar platform, raised a $50M growth round from Macquarie Asset Management and more (PRN)

Anrok, a global sales-tax platform for software companies, raised a $30M Series B led by Khosla Ventures (BW)

Kenyan solar irrigation startup SunCulture raised a $27M Series B from InfraCo Africa, Acumen Fund, and more (FN)

Abre, a K-12 data platform, raised a $24M Series A led by PeakSpan Capital (PRN)

Mimo, a fintech automating admin behind B2B payments, raised a $19.6M round led by Northzone (EU)

Veteran AXA IM equity income manager George Luckraft to retire after 22 years (IW)

Safety science firm UL Solutions raised $946M in an upsized IPO (BBG)

PACS Group shares rose ~10% in its trading debut, giving the nursing home operator a $4B valuation after its $450M IPO (RT)

Italian bank UniCredit received ECB approval for a $3.3B share buyback (RT)

Norwegian hydrogen technology company HydrogenPro raised $8.3M in equity from ANDRITZ AG via a private placement (PRN)

CarMax kicked-off a $1.3B ABS sale just hours after reporting a Q4 earnings miss (BBG)

Dish Network, a satellite-TV provider with over $20B in debt, received at least one proposal from private credit firms offering over $1B in financing (BBG)

El Salvador sold $1B in bonds linked to an IMF deal or credit upgrade (BBG)

First news

• Japanese chipmaking in collaboration with the U.S. is heating up, while China blows cold air on the use of Western chips, and certain voices support the LNG export pause
• Northrop Grumman et al are even now making AI-driven drone versions of submarines and fighter planes
• Two startups are here to 1. quantify the value AI models provide to the business sphere, and 2. provide mathematical proof of how the models evolve, the better to furnish a more efficient development process for new models.

Chart of the Day

MARKET LEVELS

MORNING INSIGHT

Good morning!

This week is an important fragility test for markets as equities sit at 50dma.

While de-risking is a knee-jerk response, this dip will likely be bought.

In contrast to 2022 Ukraine-war eve, the Fed is dovish.

Click HERE for more.

TECHNICAL

Simply stated, we don’t make much of the recent selloff and feel that U.S. Equities should be close to bottoming, which should happen sometime next week. The reasons are as follows:

-Technology has held up admirably vs. the SPX despite its underperformance in recent months. Charts of MSFT, META, GOOGL, NVDA look quite bullish, and recent evidence of AAPL 1.18%  having broken out looks constructive for Technology.

-Financials and Healthcare are both nearing support after recent underperformance. These two groups account for 25% of SPX.

-Insufficient strength in Defensive groups like Staples and REITS has been seen to argue that the market should begin a larger selloff.

-Our former cycle composite that showed 4/20 as being important very well could help markets bottom on/near that date, before turning higher.

-TRIN readings for Friday registered a 1.86, showing outsized downside volume. This normally has been a great predictor of possible market turns when it gets to 2 or higher, and should be watched early next week.

Click HERE for more.

CRYPTO

In our latest video, we discuss the monetary and fiscal levers affecting markets, potential impacts of spot ETFs in Hong Kong, and the rising tensions in the Middle East.

Click HERE for more.

FIRST NEWS

Eastern Promises. At a recent summit, the U.S. and Japan intensified cooperative efforts intended to fend off authoritarian states. The two nations will integrate air defense systems and collaborate more closely on other defense initiatives. All this while former researchers at Google are looking to help Japan catch up in the AI race. Their pitch, using smaller, cheaper models will allow more companies there to tap into the technology. The venture has won supercomputing grants and blue-chip partners seeking to catch up in artificial intelligence.

While we’re on the topic of chips, an old Japanese ceramics maker, Maruwa, is now projected to be worth $2.75 billion as its heat-dissipating materials are increasingly used to cool AI chips.

In building circuit boards, Maruwa uses materials made up of ceramic composites containing metals that make them effective at absorbing heat. Keeping semiconductors at an optimal temperature is vital to their performance, and data centers spend much of their electricity bills on cooling. Shares in Maruwa have risen in line with demand for data center capacity.

Analysts from Goldman Sachs say Maruwa has 60% of the global market share in high-temperature ceramics, and its 200-plus-year expertise in manufacturing the materials provides a competitive advantage.

In the meantime, the Chinese government is continuing to put pressure on its telecom companies to stop using foreign chips, such as those made by Intel and AMD. China Mobile and others have been ordered to phase them out by 2027. It’s the latest tit-for-tat move in the tech Cold War between the U.S. and China.

Following Prime Minister Fumio Kishida’s visit to Washington last week, the U.S. and Japan will also collaborate on developing advanced offshore wind technology. The U.S. pause on LNG export terminal permitting has been a source of worry to Japan, a major LNG importer.

On the other side of the issue, Ray Mabus, the top U.S. Navy official under President Barack Obama, led a group of retired admirals and generals in a letter to Congressional leaders supporting the pause. Increasing U.S. LNG exports, which are “now increasingly being [sent] directly to the country’s rivals, like China, allowing that country to gain leverage and build power globally while weakening our nation’s ability to respond in a crisis” is not a good idea says the letter.

At the same time, The Biden administration has approved a large crude oil export terminal in the Gulf of Mexico, to be operational by 2027, serving customers for U.S. oil in Europe and Asia.Semafor, NZZ, Bloomberg

Eternal Brother, Strong to Save. Northrop Grumman has unveiled an unmanned submarine capable of long-range, silent, undersea missions. Taking its main design idea from its namesake, The Manta Ray seeks to be low-energy, able to hibernate anchored to the sea bed for long periods before “delivering payload capability from the sea.” The U.S. is increasingly invested in human-free weapons. For example, The Air Force Secretary is set to fly in a modified F-16 piloted by AI, a high-water mark for AI’s capabilities and a high-profile way to test them. The Pentagon plans to retrofit older aircraft (e.g. the F-16) as unmanned drones, to act as “loyal wingmen” to a single piloted jet. Each drone like this would cost a quarter to a third of a new F-35. Northrop Grumman

But Does It Provide Value? Vals.ai is a startup that wants to assess the actual efficacy of AI models across diverse domains such as accounting, law, and finance. Seems like a simple matter, yet currently, the tech industry lacks a standardized, impartial evaluation system to gauge the performance and functionalities of different AI models and services. This absence complicates decision-making for companies looking for the most suitable AI systems for their specific requirements.

Vals.ai aims to solve this issue by establishing a neutral, third-party review mechanism for appraising large language models. Working with researchers and industry specialists, Vals.ai has devised testing frameworks that leverage both academic and industry-specific datasets.

The co-founders, Rayan Krishnan and Langston Nashold, noticed that even prominent AI models like GPT-4 and Gemini Pro encounter difficulties with, say, tax-related inquiries. To them, that signaled a need for more rigorous, domain-specific testing to understand the practical capabilities and limitations of these AI systems.

Given the lack of industry consensus on the optimal approach to evaluating AI models, Vals.ai’s push to furnish nuanced, impartial performance metrics addresses a crucial gap in the swiftly advancing AI landscape. Bloomberg

Chemistry vs. Alchemy. On a related note, Khosla Ventures-backed Symbolica AI, co-founded by former Tesla Autopilot engineer George Morgan, is building a new AI-assisted coding tool that it says uses a new method of machine learning that works completely differently from the cutting edge foundation models made by OpenAI, Google, and other major AI companies.

With this new approach, Morgan says Symbolica’s models won’t require the same massive hunger for computation companies are currently spending tens of billions of dollars to secure for the most advanced AI models.

Those making the investments are essentially speculating that, with enough data and enough compute resources, later versions of these models will grow more intelligent, yet without mathematical proof showing how these things work, the process is more like alchemy, says Morgan. “That’s exactly what AI models are today: you mix a bunch of random stuff together, you test it, you see if it does the thing or not. If it doesn’t, you try something else. What Symbolic is doing is bringing this into the era of chemistry.”

There is a big market for what Symbolica is trying to build. While consumers may want to chat with AIs about any topic, businesses want the opposite, with language interfaces that are narrowly focused and entirely reliable. In other words, in many enterprise contexts, hallucination is simply unacceptable. Semafor