Every modern economy runs on chips. Understanding why requires going inside the machines that make them, understanding who controls those machines, and realizing what that control is worth.

Phones need them, cars need them, factories need them, hospitals need them, banks need them, satellites need them, gaming consoles need them, cloud platforms need them, and every serious AI model depends on them.

That is why semiconductors have moved from being a technical supply-chain topic to becoming one of the most important industries in the world. Chips now sit inside the biggest conversations in markets: artificial intelligence, national security, inflation, energy demand, industrial policy, data-centre capex, and the future of computing.

The industry has always been cyclical. Demand rises, companies build capacity, inventories reset, prices fall, and the market starts again. That cycle still matters. What has changed is the size of the demand driver now sitting on top of the old cycle.

AI has turned compute into a scarce resource. It has also turned memory bandwidth, advanced packaging, power efficiency, and leading-edge manufacturing into strategic assets.

This creates the main debate for investors.

Semiconductors have some of the strongest profit pools in public markets, but the value is not shared equally across the chain. The companies that control scarce technology, design software, AI architecture, leading-edge manufacturing, advanced memory, and equipment chokepoints capture the best economics. The weaker players still move with the cycle and can look cheap right before earnings fall.

Our view is clear: semiconductors remain one of the most important sectors for the next decade, but investors need to map the value chain before buying the story. ASML is not Micron. Micron is not NVIDIA. NVIDIA is not TSMC. TSMC is not Amkor. They belong to the same industry, but they do not have the same economics.

The short version

Semiconductors solve one large problem: the world keeps demanding more computation, more storage, more connectivity, and more control over electronic systems. That demand now comes from AI training, AI inference, cloud computing, smartphones, PCs, cars, factories, industrial machines, defence systems, medical devices, and connected products.

The business model changes depending on where the company sits in the chain. Some companies license IP. Some sell design software. Some design chips. Some manufacture chips. Some sell the machines used to manufacture chips. Some package and test the final product. Some distribute chips into fragmented customer markets.

The strongest profit pools are usually found in 5 areas: EDA and IP, fabless AI design, semiconductor equipment, leading-edge foundry, and high-bandwidth memory. These are the places where technical difficulty, customer dependency, scale, and switching costs create strong pricing power.

The industry also has real risks. Fabs are expensive. A leading-edge fab can cost tens of billions of dollars. EUV tools can cost hundreds of millions. Manufacturing cycles are long. Inventory can build quickly. Memory prices can rise fast and fall fast. Export controls can change demand overnight. Taiwan remains the biggest geographic risk in the global technology stack.

The next 5 years will probably be decided by 6 questions:

• Can hyperscalers keep spending on AI infrastructure?

• Can memory suppliers turn HBM demand into more stable long-term contracts?

• Can TSMC keep leading at advanced nodes while managing Taiwan risk?

• Can ASML continue to control the EUV and High-NA transition?

• Can China’s mature-node expansion pressure pricing in older chips?

• Can advanced packaging become a bigger profit pool as chiplets grow?

That is the whole industry debate.

What semiconductors actually do

A semiconductor controls electrical signals. That sounds basic until you realise that modern life is built on the ability to control signals at enormous scale and speed.

Some chips process information. CPUs, GPUs, NPUs, AI accelerators, and custom ASICs fall into this group. They decide how fast a system can calculate, train, infer, render, simulate, or respond.

Some chips store information. DRAM, NAND, and high-bandwidth memory sit here. The AI world is learning this lesson very quickly: compute without memory bandwidth is limited.

Some chips move information. Networking chips, modems, RF chips, optical components, and connectivity chips allow devices and data centres to communicate.

Some chips manage power. Power semiconductors help convert, distribute, and control electricity. They matter in electric vehicles, factories, power grids, renewable energy, and data centres.

Some chips sense the physical world. Image sensors, accelerometers, temperature sensors, radar chips, and industrial sensors turn physical signals into digital information.

That is why the semiconductor market is much broader than AI GPUs. AI is the current centre of attention, but the industry also includes automotive chips, industrial chips, analog chips, memory, sensors, communications chips, microcontrollers, and power devices.

The useful investor question is simple: which chips are scarce, differentiated, and hard to replace?

That is usually where the money goes.

How the industry is structured

The semiconductor industry has 8 main layers.

Before walking through each one, here is the full value chain in one view:

1. Materials and chemicals

This is the base of the industry. Chip manufacturing needs silicon wafers, specialty gases, photoresists, wet chemicals, deposition materials, rare gases, ultra-pure water, and cleanroom infrastructure.

Key players include Shin-Etsu, SUMCO, Siltronic, GlobalWafers, Linde, Air Liquide, Air Products, JSR, Tokyo Ohka Kogyo, and other specialist suppliers.

This layer rarely gets the same attention as NVIDIA or TSMC, but it matters. A fab cannot run without the right materials at the right purity. Even small contamination can ruin wafers that have already absorbed weeks of process time and huge production cost.

Materials also create geopolitical exposure. Neon, helium, bromine, tungsten, rare earths, and specialty chemicals can become supply-chain issues very quickly when trade, war, export rules, or logistics routes change.

2. EDA software and IP

Before a chip can be manufactured, it must be designed.

That is where Electronic Design Automation, or EDA, comes in.

The main players here are Synopsys, Cadence, and Siemens EDA. Arm also sits in this upstream layer through processor architecture and IP licensing.

This is one of the best business models in the industry. EDA tools are deeply embedded inside engineering workflows. If a chip company is spending hundreds of millions on a new design, it will not risk failure by using weak tools. The cost of the software is small compared with the cost of a failed chip.

That gives EDA companies strong pricing power, high margins, recurring revenue, and exposure to the whole industry. They do not need to guess which AI chip wins. If the world keeps designing more complex chips, demand for design software and IP keeps rising.

3. Fabless design

Fabless companies design chips but outsource manufacturing.

This is where NVIDIA, AMD, Qualcomm, Broadcom, Marvell, MediaTek, Apple’s internal silicon teams, and many AI chip startups sit.

The model is attractive because fabless companies avoid the huge capex burden of owning fabs. They spend on architecture, software, engineers, product design, and customer relationships. If the product wins, returns on capital can be very high.

NVIDIA is the clearest example today. The company’s advantage is not only the GPU — it is the full AI computing stack: GPUs, networking, systems, CUDA, software libraries, developer adoption, and deep customer relationships with hyperscalers.

The weakness of the fabless model is dependence. A great chip design still needs TSMC, Samsung, advanced packaging, HBM supply, substrates, and testing capacity. The chip can be designed perfectly and still be constrained by manufacturing or packaging supply.

4. Semiconductor equipment

This is the machinery layer.

Fabs need tools for lithography, etching, deposition, inspection, metrology, cleaning, ion implantation, process control, and testing.

Key players include ASML, Applied Materials, Lam Research, Tokyo Electron, KLA, Advantest, and Teradyne.

This is one of the strongest parts of the value chain because the equipment is hard to build and hard to replace. A fab can change some suppliers over time, but it cannot avoid using highly specialised tools.

ASML has the most obvious chokepoint. It is the only supplier of EUV lithography systems, which are needed for leading-edge chips. No EUV means no leading-edge production at scale.

KLA also matters because fabs need inspection and metrology to control yield. In a high fixed-cost industry, yield is money. A small yield improvement can have a large impact on gross margin because the wafer cost is already incurred.

5. Foundries and IDMs

Foundries manufacture chips for other companies.

TSMC is the leader. Samsung Foundry is the other major advanced-node competitor. GlobalFoundries, UMC, and SMIC focus more on mature and specialty nodes.

IDMs design and manufacture their own chips. Intel, Samsung Electronics, Texas Instruments, Analog Devices, STMicroelectronics, Infineon, NXP, and Micron all have elements of this model, depending on the product category.

Manufacturing is where the industry becomes physically difficult. Fabs require extreme precision, cleanrooms, power, water, process knowledge, expensive tools, long production cycles, and large engineering teams.

The economics are powerful when utilisation is high. Fixed costs get spread across more wafers, and margins can expand quickly. The same model becomes painful when demand falls. Fixed costs remain even when volumes drop.

That is why foundries and IDMs must be analysed through utilisation, capex, yield, node leadership, and customer commitments.

6. Memory

Memory stores data.

The main categories are DRAM and NAND. The main public players are Samsung, SK Hynix, and Micron.

Historically, memory has been more cyclical than logic. Supply gets tight, pricing rises, profits expand, capacity gets added, supply catches up, pricing falls, and margins compress.

AI has changed the memory debate because high-bandwidth memory, or HBM, is now essential for AI accelerators. GPUs need HBM to feed data fast enough. Without memory bandwidth, expensive compute sits underused.

This has made memory more strategic. Micron, SK Hynix, and Samsung are now trying to secure longer-term customer commitments, pricing floors, and take-or-pay structures. That could reduce some cyclicality, but investors should be careful. Memory has improved, but it has not become software.

The key question is whether HBM creates a more disciplined memory industry, or whether the old cycle returns once supply catches demand.

7. Advanced packaging and testing

After wafers are processed, chips must be cut, packaged, connected, tested, and qualified.

This is where OSAT companies such as ASE Technology, Amkor, and JCET operate. TSMC, Samsung, Intel, and memory companies also invest heavily in advanced packaging internally.

This layer used to be viewed as lower-margin back-end work. That view is changing because AI systems now depend on advanced packaging.

Chiplets, HBM, 2.5D packaging, 3D stacking, interposers, and CoWoS-like technologies allow multiple dies to work together inside one package. As transistor scaling becomes harder and more expensive, packaging becomes a bigger part of performance.

The industry is moving from “make the transistor smaller” to “connect more specialised silicon more efficiently.”

That makes advanced packaging one of the most important areas to watch.

8. Distributors and end customers

Distributors such as Arrow Electronics and Avnet help move chips into fragmented customer markets. Their margins are usually lower than the margins earned by IP owners, equipment makers, or leading chip designers.

End customers include Apple, Samsung, Tesla, Microsoft, Amazon, Google, Meta, Oracle, automotive OEMs, industrial companies, defence contractors, telecom equipment makers, and thousands of smaller electronics producers.

The demand signal starts here. If hyperscalers keep spending, AI chips, HBM, networking, storage, power chips, and advanced packaging stay tight. If automotive demand slows, mature-node chips feel it.

Where the money is made

The semiconductor industry has several profit pools, but they are not equal.

The strongest economics sit where customers have few alternatives.

The highest and most consistent returns often come from EDA/IP and selected equipment companies. These companies sell into the whole industry without taking full fab risk.

Fabless leaders can also produce exceptional returns when they own the demand layer. NVIDIA is the current example.

Foundries are strong when they lead at advanced nodes, but they must keep reinvesting heavily.

Memory can produce huge profits in tight markets, but the valuation must reflect the cycle.

Why the economics are attractive

The semiconductor industry has strong economics because the products are essential, the barriers are high, and the best companies control scarce capabilities.

Technical difficulty. Modern chips are hard to design and hard to manufacture. This protects incumbents because new entrants cannot simply raise money and compete at the leading edge.

Scale. Large companies spread R&D, manufacturing learning, customer support, process development, compliance, and supplier relationships across huge revenue bases.

Switching costs. Chip designers do not easily switch EDA tools. Fabless companies do not easily move leading-edge production away from TSMC. Customers do not easily redesign products around a new chip supplier. Fabs do not easily replace key equipment vendors.

Operating leverage. Once fixed costs are covered, incremental revenue can flow strongly into margins. This works beautifully in an upcycle. The same leverage cuts both ways. When demand falls, utilisation drops, inventory rises, and fixed costs stay.

That is the danger of buying a cyclical stock on peak earnings.

Why the economics are dangerous

The industry is capital-intensive.

Fabs cost billions. Tools cost hundreds of millions. Process development takes years. Product cycles can be long. Inventory can become a problem before management teams admit it.

The cash conversion cycle is also long for many semiconductor manufacturers. A wafer can take months to move through production, packaging, testing, delivery, invoicing, and collection.

Fabless companies usually have better capital efficiency because they do not own fabs. Equipment makers can sometimes benefit from customer deposits and long backlogs. Foundries and IDMs carry much heavier capital burdens.

This is why investors must separate the business model before comparing valuation multiples.

A 25x earnings multiple for an EDA company is not the same as a 25x multiple for a memory company.

A low P/E in memory can be a warning.

A high P/E in EDA can still be justified if revenue is recurring, margins are high, and customer lock-in is strong.

AI changes the industry

AI is the largest current demand driver in semiconductors. The market often talks about AI as if it is only a GPU story. That is too narrow.

AI affects the full chain.

It affects NVIDIA through accelerators and systems. It affects TSMC through leading-edge manufacturing demand. It affects ASML through advanced lithography. It affects Micron, SK Hynix, and Samsung through HBM and DRAM. It affects Broadcom and Marvell through networking and custom silicon. It affects Synopsys and Cadence through chip complexity. It affects ASE, Amkor, TSMC, Intel, and Samsung through advanced packaging. It affects power semiconductors because AI data centres need more power management. It affects cooling, energy, and grid infrastructure because AI data centres consume far more electricity than traditional compute facilities.

The real question is whether AI capex produces enough return for hyperscalers.

If Microsoft, Amazon, Google, Meta, Oracle, and other large buyers keep increasing AI infrastructure spending, the semiconductor chain stays supported. If AI returns disappoint, spending can slow quickly — and that would hit the sector because semiconductors sit upstream of the AI buildout.

Memory has become the market’s signal

Micron has become more important because memory is now one of the cleanest signals for AI infrastructure demand.

HBM is no longer a side product. It is a key input for AI accelerators.

This has changed investor perception of memory companies. In the past, memory was often treated as a commodity cycle. Now, large customers are signing longer-term agreements because they need secure supply.

A take-or-pay agreement changes the quality of demand. Customer deposits, pricing floors, and longer commitments can reduce some of the old boom-bust behaviour.

But discipline is still needed.

Memory supply can increase. Competitors can add capacity. China’s CXMT can become more relevant in standard DRAM. HBM technology leadership can shift. AI customers can renegotiate if the cycle weakens.

The memory story is better than before. It is still cyclical. That is the balance investors need to keep.

Chiplets and advanced packaging are becoming strategic

Moore’s Law is becoming harder and more expensive.

The old playbook was simple: shrink transistors, improve performance, reduce cost per function, repeat.

That still matters, but the economics are more difficult at the leading edge.

Chiplets offer another path. Instead of building one large monolithic chip, companies split the system into smaller dies and connect them inside an advanced package. The performance-critical parts can use the newest node. Less demanding functions can use older, cheaper nodes.

This can improve yield, reduce cost, shorten design cycles, and allow companies to combine different technologies more efficiently.

AMD has shown the value of chiplets. Intel is investing heavily in packaging. TSMC has become central through CoWoS and other advanced packaging technologies. Memory suppliers also benefit because HBM must be integrated close to AI accelerators.

Investors should watch advanced packaging capacity the same way they watch wafer capacity. In AI, both can become bottlenecks.

Geopolitics now shapes the sector

Semiconductors used to be organised around efficiency: design in the US, equipment from the Netherlands, Japan, and the US, manufacturing in Taiwan and South Korea, assembled across Asia.

That model lowered cost and created dependency.

Now governments see chips as strategic assets.

The US has the CHIPS and Science Act — roughly $280 billion, with $52.7 billion specifically for domestic semiconductor manufacturing. Europe has the EU Chips Act, mobilising €43 billion. China is pushing self-sufficiency with over $150 billion in public funding and a target of 80% self-sufficiency by 2030. Japan is rebuilding its role in materials, equipment, and manufacturing. South Korea remains central to memory.

Taiwan remains central to advanced foundry capacity — and producing over 90% of the world’s most advanced chips through TSMC alone. This concentration is the single largest systemic risk in global technology. Any disruption to Taiwan creates cascading failures across every industry that depends on modern electronics.

This creates a more regional and more expensive industry. More fabs will be built outside Taiwan. More supply chains will be duplicated. More governments will subsidise capacity. More export rules will shape where companies can sell.

Geopolitics is no longer a footnote. It is part of the financial model.

The main public players

ASML controls one of the gates into leading-edge manufacturing. Its EUV monopoly is its core asset. The bull case is advanced-node demand, High-NA EUV adoption, and service revenue. The risk is export controls, China exposure, and customer capex timing.

TSMC is the world’s leading pure-play foundry. Its strength is manufacturing execution, customer trust, scale, and yield. The bull case is AI, custom silicon, and pricing power at advanced nodes. The risk is Taiwan concentration.

NVIDIA is the leading AI accelerator company. Its strength is GPUs, networking, systems, CUDA, software libraries, and customer relationships. The bull case is continued hyperscaler capex and inference growth. The risk is valuation, customer concentration, custom silicon competition, and export controls.

Synopsys, Cadence, and Arm sit upstream. They benefit from chip complexity. AI chips, automotive chips, chiplets, and custom silicon all require more design work. The bull case is that complexity keeps rising. The risk is valuation and future AI-assisted design disruption.

Applied Materials, Lam Research, Tokyo Electron, and KLA sell tools into fabs. The bull case is global fab expansion, advanced nodes, and AI demand. The risk is capex cyclicality and China restrictions.

Micron, SK Hynix, and Samsung are the memory leaders. AI has made HBM strategic. The bull case is tight HBM supply, disciplined DRAM production, and long-term customer contracts. The risk is supply catching up, pricing falling, and China capacity growing. Memory is better than before. It is still memory.

Intel is the turnaround case — scale, customer relationships, manufacturing assets, and strategic value to the US. The bull case is government support and process recovery. The risk is executing on too many fronts simultaneously.

AMD, Broadcom, Marvell, and Qualcomm sit across high-performance computing, networking, custom silicon, mobile, and connectivity. The bull case is custom silicon and diversified AI infrastructure demand. The risk is competition and customer concentration.

ASE and Amkor sit in packaging and test. Advanced packaging is improving the strategic value of the back end. The bull case is chiplets, HBM integration, and AI accelerator complexity. The risk is that standard packaging remains competitive and margin-limited.

How to analyse a semiconductor company

Investors should avoid analysing all chip companies with one metric.

Revenue growth alone is not enough. P/E alone is not enough.

The 5-year scenarios

Base case (60% probability): AI demand remains strong but uneven. Hyperscaler capex continues, but investors become more selective. Memory remains healthier than past cycles because of HBM and longer-term contracts, but pricing still moves in waves. Equipment demand remains supported because governments and companies keep building fabs across regions. The winners are companies with pricing power, advanced technology, strong balance sheets, and exposure to AI infrastructure without relying on perfect conditions.

Bull case (20% probability): AI spending continues to exceed expectations. Training grows, inference grows, enterprise AI adoption becomes measurable, sovereign AI expands, custom silicon grows alongside GPUs, HBM remains tight, and advanced packaging becomes a major bottleneck. This would support fabless AI leaders, HBM suppliers, leading-edge foundries, EDA companies, advanced packaging players, and equipment leaders. The danger is overpaying for the obvious winners.

Bear case (20% probability): AI spending slows before revenue catches up. Hyperscalers reduce capex growth, memory supply catches demand, foundry utilisation weakens, equipment orders get pushed out, and China pressures mature nodes. This would punish companies priced for perfect growth. The weakest stocks would be companies with high cyclicality, low pricing power, heavy capex, or stretched valuations.

What BE Invested Labs is watching

For the semiconductor sector, we are tracking 12 signals:

1. Hyperscaler capex from Microsoft, Amazon, Google, Meta, Oracle, and other large AI buyers

2. NVIDIA data-centre growth and forward commentary

3. HBM pricing, capacity, and supplier share across SK Hynix, Micron, and Samsung

4. Micron’s customer commitments and whether memory contracts reduce cyclicality

5. ASML bookings, backlog, China exposure, and High-NA EUV adoption

6. TSMC utilisation, advanced-node pricing, and CoWoS capacity

7. Semiconductor equipment orders from Applied Materials, Lam Research, Tokyo Electron, and KLA

8. DRAM and NAND pricing outside HBM

9. Export-control updates from the US, Netherlands, Japan, China, and the EU

10. Advanced packaging capacity and lead times

11. Inventory trends across automotive, industrial, consumer electronics, and data-centre customers

12. Power, water, labour, and permitting constraints for new fabs

The most important signal remains AI capex. If AI infrastructure spending holds, the sector has support. If AI infrastructure spending slows, the market will quickly test which semiconductor companies have real durability.

Our view

Semiconductors are still one of the most important sectors in the market.

The customer problem is real. The world needs more compute, more memory, more networking, more power efficiency, and more manufacturing capacity.

The industry also remains cyclical, capital-intensive, politically exposed, and valuation-sensitive.

Our view is constructive but selective.

The best companies control scarce parts of the chain. EDA controls design workflows. ASML controls EUV. TSMC controls advanced manufacturing scale. NVIDIA controls the AI accelerator stack. Memory leaders control HBM supply, though the cycle still matters. Advanced packaging is becoming more important as chiplets grow.

For investors, the cleanest approach is to separate business quality from price. A great semiconductor company can still be a bad buy at the wrong valuation. A cyclical semiconductor company can look cheap right before earnings fall.

The winners will show 4 things clearly:

1. Customers need what they sell.

2. Competitors cannot easily copy it.

3. Margins survive normal cycle pressure.

4. Free cash flow grows over time.

Map the value chain. Find the chokepoints. Follow the profit pools. Then ask whether the price gives you enough margin of safety.

A note on this collaboration: this breakdown was co-written with Gidion from Behind The Chip, a Substack publication dedicated to exploring the people, physics, and principles behind semiconductor innovation. His knowledge of the industry's technical depth made this article better. If you want to understand what's happening inside the machines that power this supply chain — not just the economics — his newsletter is the right place to go. Subscribe at behindthechip.substack.com.

Try the BE Invested Labs platform: beinvestedlabs.com

Subscribe now

Subscribe to BE Invested Labs for more industry breakdowns, company research, bull and bear cases, valuation notes, and AI audio briefings.

— Buyce & Emmanuel, BE Invested Labs

References and data notes

Sources used for this breakdown include BE Invested Labs semiconductor industry reports covering business models, core economics, value-chain mapping, unit economics, regulation, TAM, future growth scenarios, geopolitics, and macro outlook. We also reviewed recent public market reporting on semiconductor sales, AI capex, HBM demand, memory supply commitments, Micron, SK Hynix, ASML, TSMC, and SEMI commentary.

Disclaimers & Disclosures

AI Transparency BE Invested Labs uses AI to aggregate financial data and format reports. The AI structures the information. Users determine their own investment thesis, risk management, and market outlook. We audit system outputs to reduce AI errors. We use these exact reports to manage our own capital.

Legal Disclaimer This platform provides educational and informational content. BE Invested Labs is not a registered investment adviser and does not provide financial, investment, tax, or legal advice. Nothing in this article is a recommendation to buy, sell, or hold any security. Investors should do their own research and consult a qualified professional before making investment decisions.

Every investor loves company with a moat until one is actually tested with supply chain constrains. It's easy to call a company defensible when the cycle is calm and competitors are quiet. The real test is what happens when demand goes vertical, customers have nowhere else to go, and the company holding the chokepoint has to decide how hard to press its advantage.

That test is happening right now, in real time, at ASML and TSMC.

Every hyperscaler, earnings call promises more compute. Every chip launch promises more performance. However, the two companies standing at the literal physical chokepoint of this entire industry are: the only lithography maker capable of printing leading-edge chips, and the only foundry that can run those machines at hyperscaler scale are telling anyone who will listen that they cannot make enough, fast enough, for years to come.

That’s not a company under siege. That’s a moat doing exactly what a moat is supposed to do.

When Christophe Fouquet, ASML's CEO, says "demand for chips is outpacing supply," and when TSMC's C.C. Wei tells shareholders the chip shortage will last "for years," neither is apologizing. They're describing a market structure that has handed two companies extraordinary pricing leverage over the most important capital cycle of this decade and, notably, both are choosing restraint over maximum extraction, which is itself a sign of how secure that position really is.

This piece breaks down both businesses, what their latest numbers say about the strength of their respective moats, why their CEOs are drawing such a sharp line around the word "bottleneck" even as they hold all the leverage, and what it means for how you should be thinking about positioning in this sector.

Chapter 1: Two moats standing at the choke point

ASML — the company that makes the machines that make the chips

We provided a deep business dive in our article below:

Simply put it ASML has monoploy on extreme ultraviolet (EUV) lithography systems fire a laser at tin droplets 50,000 times per second to generate plasma that emits 13.5-nanometer wavelength light. EUV is used to etch nano features producing leading-edge AI chips at TSMC, Samsung, and Intel.

ASML reported its full 2025 results, 2025 total net sales came in at €32.7 billion with a gross margin of 52.8% and net income of €9.7 billion. The fact that ASML raised the bottom of its own range by €2.8 billion (31.4%) after just one quarter tells you the order book is moving faster than even ASML expected. Q4 2025, quarterly net bookings came in at €13.2 billion, of which €7.4 billion was EUV that backlog is what's giving ASML the confidence to keep raising guidance even as quarterly shipments.

The capacity commitment of ASML according to CFO Roger Dassen states that the company should be able to ship 60 of its bestselling low-NA EUV tools in 2026, 25% more than in 2025, and will have capacity to ship 80 in 2027. This provides a positive of outlook for investors in upcoming quarters.

TSMC: the foundry through which the entire AI stack runs.

ASML makes the lithography machines, TSMC is the company that uses these machines to fabricate the chips that are essential for major AI accelerator desinger such as Nvidia, AMD, Apple Google, Amazon, Broadcom. None of whom own their own leading-edge fabs, hence fabless.

TSMC reported first-quarter 2026 revenue of $35.9 billion, up 6.4% sequentially and slightly ahead of guidance, with gross margin expanding 3.9 percentage points to 66.2%, exceeding the high end of prior guidance by 120 basis points, while operating margin rose to 58.1%. The mix shift is the real story here: HPC hit 61% of revenue in Q1 2026, up from 53% in Q4 2025 and 51% in Q3 2025, while smartphones collapsed to 26% of the mix. The HPC (AI/data center) slice growing by roughly 5 points a quarter while the smartphone slice shrinks at almost the same rate. This shows that TSMC leads the market; it's TSMC's existing capacity being systematically reallocated toward the highest-margin i .e. HPC, which is the highest-demand customer base TSMC has ever had.

Guidance from TSMC management shows an increase in outlook of production. Management raised the full-year 2026 revenue guide from "mid-20% year to year" to "above 30% year to year", and capex guidance for the full year sits at $52-56 billion, targeting the high-end for capacity expansion in Fab in Arizona, Japan and Germany. For context on how fast that capital intensity is accelerating: that $52-56 billion is up roughly 30% from the $40.9 billion invested in 2025, setting a new historical record.

TSMC chokehold on the market: TSMC’s (CoWoS) Chip-on-Wafer-on-Substrate for advanced packaging which is the process that physically bonds high-bandwidth memory (HBM) on GPU die. This has become arguably challenging than etching the leading-edge logic wafers themselves. TSMC’s capacity for CoWoS is sold out until 2027 for new customers. This single handedly reveals that even if TSMC could magically produce more raw silicon, it can't ship a finished AI chip without packaging capacity it has already fully allocated for the next two years.

Chapter 2: The clearest proof of the moat from both CEOs.

The following are details that should make any investor sit up and pay attention. Both companies are sold out years in advance. Both could, in theory, raise prices aggressively and let customers fight over scraps like how makers of memory like Micron increased prices. Instead, both CEOs are spending unusual amounts of public energy insisting they are not the constraint, and pointing calmly, specifically at other parts of the chain instead. That's not modesty. Only a company with no real competitive threat to its own position has the luxury of redirecting blame elsewhere rather than defending its turf.

ASML CEO Fouquet framed it as “the biggest bottleneck seems to be in chip manufacturing”. Fouquet’s message is clear: the AI boom is pushing the semiconductor industry into a prolonged supply-constrained phase. At a tech event in Antwerp, Fouquet state that "Demand on AI is coming so strongly that we will be in a supply-limited market for quite a while,"

ASML is still under pressure, however, because its ability to support customers depends on scaling its own supply chain, capacity, and execution speed. In practical terms, ASML is saying: fabs may be the immediate constraint, but we cannot allow lithography equipment availability to become the next limiting factor. This framing also explains Fouquet’s comments on projects such as Musk’s proposed “Terafab”: massive new AI-driven fab capacity is an opportunity for ASML, he stated, “New projects are an opportunity as long as you're not supply limited”

That's a CEO explicitly naming the risk that his own company could become the reason a customer's mega-project stalls and committing, publicly, to not let that happen. It's a confident or positive posture: only the company holding the only machine of its kind can credibly promise a customer like Musk that it won't be the reason the project slips.

TSMC’s Wei framed it as “we are doing our best to ensure TSMC does not become a bottleneck“.

This reveals that TSMC is under extreme AI-driven demand pressure, but it is working to ensure it does not become the next structural bottleneck in the semiconductor supply chain. The important nuance is pricing discipline: unlike memory makers (such as (Samsung, SK Hynix, Micron) that allowed HBM scarcity to trigger sharp, commodity-like price spikes, TSMC is signaling controlled and predictable price increases. Wei pushed back against the kind of abrupt pricing moves seen in memory chips, signaling instead measured, predictable increases that fund TSMC's capex without damaging long-term customer relationships.

What this restraint actually reveals about the moat

Read together, both above statements are doing the same strategic job: each company is secure enough in its position that it can spend its leverage on reputation and customer trust instead of squeezing every available dollar out of the shortage. That’s a luxury weaker competitors don’t have. A company genuinely worried about a substitute emerging would be racing to extract maximum value before the window closes. ASML and TSMC are doing the opposite playing a long game, because they know there’s no real second option for their customers regardless of how patient they are.

The AI stack is not constrained by one simple bottleneck like wafer capacity. The real pressure points are spreading across HBM, advanced packaging, shortage of electricity, copper, and critical gases, helium rationing in Taiwan and South Korea, and geopolitical supply chains. This is why ASML and TSMC’s “we are not the bottleneck” framing matters: they are positioning themselves apart from the more fragile, cyclical, and substitution-prone parts of the AI supply chain. Their ability to avoid panic pricing or public capacity blame signals the strength of their moats.

Chapter 3: How does a moat look like from above

Zoom out of the ASML/TSMC story, the market position of these company does not conform to the classic moat of price hiking. The managment of these companies understand the effect of repricing the across entire chip industry.

Scale of the shift: total global semiconductor revenues are forecast to reach $1.29 trillion in 2026, up 52.8% year over year from $842.8 billion in 2025, with DRAM revenues alone projected to nearly triple to $418.6 billion, driven by demand for high-bandwidth memory and DDR from hyperscalers.

The memory side is its own story, and it’s brutal for anyone buying chips downstream. Suppliers are reporting record gross margins of 60-70% for HBM much higher than standard DRAM as SK Hynix, Micron, and Samsung have preallocated their entire 2026 capacity, effectively engineering scarcity and escaping the “commodity trap” that plagued memory for decades. HBM demand grew 130% year-on-year in 2025 and is projected to grow a further 70% in 2026, while DRAM capital expenditure cycles run 18-36 months from investment decision to production output meaning capacity simply cannot respond to demand spikes within a single year.

What this mean: notice the memory scarity is a function of the memory manufacturer oligopoly choosing to under-build capacity, which is reversible the moment incentives change. ASML and TSMC's scarcity is a function of physics, accumulated expertise, and capital intensity that no amount of incentive can replicate quickly. One is a cyclical squeeze. The other is a moat. Investors should price them differently. ASML CEO is confident stating during a TechCrunch interview “No one is coming focus“. This indicates thethat no competitor is coming for the ASML’s position of innovation.

Chapter 4: Key insights for investors, how to monitor whether the moat holds

For investors, the key question is whether ASML and TSMC’s moats remain durable as AI capex increases or accelerates. The strongest signals are not headline quarterly beats, but upward guidance revisions, capacity execution, and margin quality. ASML’s higher 2026 outlook and TSMC’s raised revenue growth guide suggest forward demand is still strengthening. Investors should monitor ASML’s EUV shipment ramp and TSMC’s capex, advanced-node capacity, and CoWoS expansion as early indicators of whether supply can keep pace.

The two companies also differ in how AI demand flows through margins. TSMC’s margin expansion is a cleaner signal of HPC mix strength, while ASML’s profitability is lumpier because it depends on which lithography systems ship in a given quarter. Both companies repeated “we won’t be the bottleneck” messaging is itself a strong moat signal: unlike HBM, packaging, power shortage, or raw materials, ASML and TSMC are not relying on scarcity pricing alone. Their advantage is structural, not merely cyclical. Despite geopolitical export control.

The main risks to watch are geopolitical and execution-related. U.S. export controls could pressure ASML’s China sales, while TSMC’s geographic diversification is constrained by permitting, labor shortages, and the difficulty of moving leading-edge capacity abroad. Still, both companies now have unusually long visibility into demand, with backlogs and capacity commitments stretching into 2030 and beyond. That makes this AI cycle different from a normal semiconductor upswing: the demand signal is stronger, longer, and more capacity-demand.

The Conclusion

ASML and TSMC sit at the pinnacle of the entire AI value chain. One makes machines that prints leading chip and the other is the only foundry that has capacity to run those machines at the scale hyperscalers are demanding. Both have backlogs sold out for years. Both have pricing power to charge more than they are but are exercising patience and modesty. Both CEO’s insist that they are not the bottlenecks. Their technology moat is increasingly wide with secured positions on the value chain.

That’s the real signature of moat power in 2026 is not aggressive pricing hikes from these companies, but the ability to choose restraint and still come out ahead with more demand from their customers. For investors, the takeaway isn’t just where the next scarcity premium shows up but rather it’s that ASML and TSMC have already passed the hardest test a moat can face. Demand went vertical, customers had nowhere else to go, and neither company needed to overreach to come out stronger. We strongly believe that these companies have stronger moat in unprecedented times.

Sources

• ASML Holding NV, Q1 2026 Financial Results press release and investor presentation, April 15, 2026 (SEC Form 6-K)

• ASML Holding NV, Q4 2025 / Full-Year 2025 Financial Results press release, January 28, 2026

• Reuters, “ASML lifts 2026 forecast as surging AI chip demand boosts new orders,” April 15-16, 2026

• Reuters, “Exclusive-ASML CEO sees tight supply in booming chip market as AI demand soars,” May 20, 2026

• Bloomberg, “ASML CEO Warns Musk’s Terafab Project May Face Supply Constraints,” June 17, 2026

• TechCrunch, “ASML CEO Christophe Fouquet on his company’s monopoly: no one is coming for us,” May 5, 2026

• TSMC Q1 2026 earnings call transcript and shareholder meeting remarks, April–June 2026 (via Investing.com, Bloomberg, Reuters)

• Bloomberg, “TSMC CEO Says Company’s Chip Supply Won’t Meet AI-Fueled Demand for Years,” June 4, 2026

• Yahoo Finance / Reuters, “TSMC CEO says AI chip demand will outpace supply for years,” June 2026

• IDC, “Semiconductor Market Forecast 2026: The AI Supercycle Arrives,” April 2026

• Deloitte Insights, “2026 Semiconductor Industry Outlook,” February 2026

• Manufacturing Dive / Omdia SemiDynamics 2026 Q1 Report, “The great data center delay,” April 2026

• PatSnap, “HBM technology landscape 2026: market and AI demand,” March 2026

• Center for a New American Security (CNAS), “American AI Companies Can’t Get Enough Chips,” May 2026

  ---

Disclaimers & Disclosures

AI Transparency: B.E Invested Labs uses AI to aggregate financial data and format reports. The AI structures the information. Users determine their own investment thesis, risk management, and market outlook. We audit the system outputs to reduce AI errors. We use these exact reports to manage our own capital.

Legal Disclaimer: This platform provides educational and informational content. B.E Invested Labs is not a registered investment advisor and does not provide financial, investment, tax, or legal advice.

Customers speak to sales. They open support tickets. They click emails. They visit websites. They attend demos. They complain. They renew. They leave.

Somewhere inside all of that activity is the truth about the customer relationship.

That is the job CRM software gets paid to organise.

Customer Relationship Management, or CRM, sounds boring until you realise how much money sits behind it. The best CRM platforms sit close to revenue, retention, support cost, sales productivity, marketing spend, forecasting, and customer data. That is why the category has produced Salesforce, HubSpot, Microsoft Dynamics, ServiceNow, Adobe Experience Cloud, Oracle CX, SAP Customer Experience, Zoho, Zendesk, Freshworks, and a long list of specialised tools.

The industry is moving again because AI is now sitting inside the workflow.

A good CRM used to tell the sales team what happened.

The next version of CRM will be judged by what it can do without waiting for a human to click 12 buttons.

That creates the main debate for investors.

CRM and cloud software have some of the best economics in public markets: recurring revenue, high gross margins, low capital intensity, strong switching costs, and negative working capital. The risk is that AI changes how customers pay for software, how many seats they need, and which vendor controls the customer record.

Our BE Invested Labs view is clear: CRM is still one of the strongest enterprise software profit pools, but the winners will change. The best companies will own customer data, workflow depth, integrations, trust, security, and measurable AI adoption. The weaker companies will become expensive dashboards sitting on top of messy data.

The short version

CRM software solves 5 problems:

1. Customer data is scattered across emails, spreadsheets, calls, support tools, websites, and sales notes.

2. Sales teams waste time updating pipelines and chasing weak leads.

3. Marketing teams send campaigns without enough customer context.

4. Support teams answer customers without seeing the full history.

5. Management struggles to forecast revenue, churn, customer behaviour, and team performance.

The business model is mostly subscription SaaS.

Customers pay monthly or annually. Pricing is usually based on users, product tier, features, data usage, AI usage, automation volume, or enterprise contracts. Vendors then make more money through add-ons, integrations, data products, AI features, professional services, and partner marketplaces.

The economics are attractive because software is cheap to deliver once built.

A top CRM company can generate gross margins in the 70% to 85% range. The main costs sit in engineering, sales, marketing, customer success, cloud hosting, compliance, and support. As revenue scales, the best operators can turn more of each extra dollar into operating profit and free cash flow.

The industry has 2 different market structures at the same time.

Enterprise CRM is controlled by a small group of large platforms: Salesforce, Microsoft, Oracle, SAP, Adobe, ServiceNow, and a few others.

Mid-market and SMB CRM is more fragmented. HubSpot, Zoho, Freshworks, Monday.com, Pipedrive, Zendesk, and specialised vertical players can still grow because smaller companies want simpler tools, lower prices, and faster setup.

The next 5 years will probably be decided by 4 questions:

Can CRM vendors turn AI agents into paid usage?

Will customers pay more for automation if seat growth slows?

Can incumbents defend their data and workflow advantage against AI-native startups?

Will regulation and data sovereignty make scale more important?

That is the whole industry debate.

What CRM actually does

CRM is the system companies use to manage customer relationships across sales, marketing, service, commerce, and analytics.

For a small business, that may mean tracking leads and sending follow-up emails.

For a large enterprise, CRM becomes a large operating layer across sales reps, service agents, account managers, marketing teams, commerce teams, finance, data teams, and senior management.

A bank may use CRM to manage customer onboarding, branch activity, complaints, loan offers, and compliance records.

A software company may use CRM to track leads, pipeline stages, product usage, renewals, support tickets, customer health scores, and expansion opportunities.

A retailer may use CRM to segment customers, run loyalty campaigns, personalise offers, and handle post-purchase support.

A hospital group may use CRM to manage patient communications, appointment follow-ups, insurance interactions, and service feedback.

The value is simple: better customer memory.

The more a business remembers about its customers, the better it can sell, support, retain, and forecast.

That memory has financial value.

Sales leaders care because CRM affects pipeline accuracy and sales productivity.

Marketing leaders care because CRM affects segmentation and campaign return.

Support leaders care because CRM affects resolution time and customer satisfaction.

Finance leaders care because CRM affects revenue visibility, renewal risk, and forecasting.

Investors should care because CRM sits close to the customer budget.

A company can delay a new office. It can delay new furniture. It is harder to delay the system that tracks revenue, customer history, support tickets, and renewals.

How the industry is structured

The CRM and cloud software market has 4 layers.

1. Cloud infrastructure

This is the base layer: AWS, Microsoft Azure, Google Cloud, Oracle Cloud, and other infrastructure providers.

CRM vendors use cloud infrastructure for compute, storage, databases, networking, security, AI workloads, and global availability. This layer is capital intensive. It requires data centres, chips, energy, cooling, fibre networks, and long-term infrastructure spending.

The infrastructure providers have scale advantages because they buy hardware, power, and networking capacity at levels smaller companies cannot match.

For CRM vendors, cloud infrastructure is both an enabler and a cost line.

The cost line matters more now because AI workloads can be expensive. If AI agents become part of everyday CRM usage, inference cost, latency, model routing, and data retrieval will become important margin drivers.

2. Data and integration

This layer includes data warehouses, integration tools, API platforms, ETL tools, identity tools, middleware, and security systems.

Examples include Snowflake, Databricks, MuleSoft, Informatica, Workato, Okta, Cloudflare, and many others.

This layer matters because CRM data rarely lives in one clean place.

Sales data may sit in Salesforce. Product usage may sit in Snowflake. Billing may sit in Stripe or NetSuite. Support may sit in Zendesk. Marketing may sit in Adobe, HubSpot, or Braze. Finance may sit in SAP or Oracle.

The CRM platform becomes more useful when these systems speak to each other.

This is why Salesforce bought MuleSoft in 2018, Tableau in 2019, Slack in 2021, and later moved again into data management with Informatica. The direction is clear: Salesforce wants to own more of the customer data layer, not only the sales database.

3. CRM and application software

This is the visible layer.

Salesforce sells Sales Cloud, Service Cloud, Marketing Cloud, Commerce Cloud, Data Cloud, Slack, Tableau, MuleSoft, and Agentforce.

Microsoft sells Dynamics 365 and connects it with Microsoft 365, Teams, Azure, Power Platform, Copilot, LinkedIn, and its enterprise identity stack.

HubSpot sells marketing, sales, service, content, and operations software, with a strong position in SMB and mid-market.

ServiceNow is stronger in workflow, IT service management, enterprise service delivery, and customer service operations.

Adobe is stronger in marketing, analytics, content, personalisation, and experience management.

Oracle and SAP are stronger where CRM connects deeply into ERP, finance, supply chain, procurement, and large enterprise systems.

This layer earns the best margins when the software becomes part of daily work.

4. Implementation and services

Large CRM projects need consultants, system integrators, migration teams, developers, admin teams, and training.

This is where Accenture, Deloitte, Capgemini, Cognizant, IBM Consulting, Infosys, Wipro, EPAM, and specialised Salesforce and Microsoft partners make money.

The services layer usually has lower margins than software, but it matters for adoption.

A CRM tool that is poorly implemented becomes an expensive database with low usage.

A well-implemented CRM tool can change sales behaviour, support quality, renewal discipline, and management visibility.

Where the money is made

The best profit pool is core SaaS software.

The reason is straightforward. Once the platform has been built, the cost of adding another customer is low compared with the subscription revenue that customer pays.

A mature CRM vendor can sell the same product to thousands of companies, keep improving the code, and charge recurring fees every year.

The strongest revenue streams are:

The best CRM companies turn one product into several products.

A customer may start with sales automation, then add support, marketing, commerce, data, analytics, Slack, AI agents, and industry-specific workflows.

That expansion is what investors want to see.

A good CRM business does not rely only on new customers. It grows inside existing accounts.

That is why net revenue retention matters.

If a company keeps customers and expands annual spend inside those customers, the sales engine becomes more efficient. The company does not need to replace lost revenue every year before growing.

Why the economics are so attractive

CRM and cloud software usually have 5 economic advantages.

High gross margins

Many SaaS companies can generate gross margins between 70% and 85%.

That does not mean they are automatically profitable.

Sales and marketing, R&D, stock-based compensation, cloud hosting, data costs, and customer success can consume a lot of revenue.

The gross margin still gives the business room to produce strong profit when growth costs become more efficient.

Recurring revenue

Annual or monthly subscriptions give better revenue visibility than one-time licence sales.

This makes the business easier to forecast.

It also makes the company more valuable in public markets, especially when churn is low and renewal rates are high.

Negative working capital

Many SaaS customers pay in advance.

The vendor receives cash before recognising all the revenue. That creates deferred revenue on the balance sheet and can improve operating cash flow.

This is one of the quiet advantages of SaaS.

The company can use customer prepayments to fund product development, support, sales, and infrastructure.

Low capital intensity

A CRM company does not need to build factories for every new customer.

Most capital goes into software development, people, cloud infrastructure, internal systems, offices, and acquisitions.

Salesforce is a good example. Its annual revenue is measured in tens of billions, while capital expenditure remains small relative to revenue.

The main investment is human talent and product development.

Strong switching costs

CRM sits inside sales processes, support processes, customer data, compliance records, workflows, dashboards, integrations, and management reporting.

Changing CRM is painful.

Data has to move. Teams need training. Integrations need rewriting. Reports break. Sales operations slow down. Customer records get messy.

This gives incumbents pricing power.

It also gives buyers a reason to stay with a tool they complain about.

That is why CRM can be unpopular with users and still financially powerful.

The 7 powers view

Here is how BE Invested Labs would read the CRM industry through the 7 Powers framework

The strongest powers are switching costs, scale, brand, and partner networks.

Switching costs protect incumbents.

Scale funds R&D and compliance.

Brand helps win enterprise deals.

Partner networks make the product more useful than the software alone.

The weakness is user frustration.

Many CRM systems are powerful but heavy. Sales teams complain about data entry. Support teams complain about fields and processes. Managers complain about data quality. This creates room for simpler tools, AI-native interfaces, and vertical-specific software.

A company can be protected and vulnerable at the same time.

That is the tension inside CRM.

The main public players

Salesforce

Salesforce remains the reference company for cloud CRM.

It created the modern SaaS CRM category and still has one of the strongest platforms in enterprise software. Salesforce has sales, service, marketing, commerce, analytics, integration, collaboration, data, and AI products.

The current debate is growth.

Salesforce is already huge. It reported $41.5B in annual revenue for fiscal 2026 and guided fiscal 2027 revenue around $45.8B to $46.2B. In Q1 FY2027, Salesforce reported around $11.1B of revenue, up about 13% year over year, with Agentforce ARR reportedly around $1.2B.

The company also announced a $50B share repurchase programme and used a large amount of cash on buybacks early in fiscal 2027.

The bull case: Salesforce has the customer base, data layer, workflow depth, and balance sheet to turn AI into another paid product cycle.

The bear case: organic growth has slowed, AI-native tools may pressure seat-based pricing, and large acquisitions may be needed to keep the platform competitive.

Microsoft

Microsoft is dangerous because Dynamics does not need to win alone.

Dynamics sits next to Office, Teams, Outlook, LinkedIn, Azure, Power Platform, Copilot, GitHub, security, and enterprise identity. Microsoft can sell CRM inside a larger enterprise relationship.

For many customers, Microsoft’s advantage is convenience.

The company already has the CIO relationship, the user base, the productivity tools, and the cloud infrastructure.

The risk for Salesforce is that Microsoft can make CRM feel like part of the wider enterprise software contract.

HubSpot

HubSpot is strongest in SMB and mid-market.

Its product is easier to start with than many enterprise CRM platforms. The company has built a strong content, marketing, sales, and service suite for companies that want growth tools without a large implementation project.

HubSpot’s challenge is moving upmarket while keeping the product simple.

That balance is hard.

If HubSpot becomes too complex, it loses part of what made it attractive. If it stays too simple, it leaves larger enterprise budgets to Salesforce and Microsoft.

ServiceNow

ServiceNow is not a pure CRM company, but it matters.

It owns workflow inside IT, operations, employee service, and customer service use cases. Its strength is enterprise workflow automation.

ServiceNow can compete where customer service, support processes, internal approvals, and operational workflows overlap.

Its advantage is workflow discipline.

Its challenge is that CRM buyers may still prefer platforms built around sales, marketing, and customer data.

Adobe

Adobe matters because customer experience includes content, campaigns, analytics, personalisation, commerce, and marketing workflows.

Adobe is stronger in marketing and creative workflows than traditional sales CRM.

Its advantage is content and digital experience.

Its risk is complexity and competition from Salesforce, HubSpot, Microsoft, and newer marketing automation tools.

Oracle and SAP

Oracle and SAP matter because CRM often connects to ERP.

Large companies do not buy CRM in isolation. Finance, billing, procurement, supply chain, HR, product, and customer records all need to work together.

Oracle and SAP have an advantage when CRM decisions are tied to the broader enterprise stack.

Their challenge is user experience, speed, and competing against cloud-native and AI-native tools.

Zoho, Freshworks, Zendesk, Monday.com, Pipedrive, and vertical CRM

The rest of the market is more fragmented.

These companies can win by being cheaper, faster to implement, easier to use, or more focused on a specific customer segment.

Vertical CRM is especially interesting.

A CRM built for doctors, lawyers, real estate agents, wealth managers, local services, car dealerships, or schools can beat a general platform if it understands the exact workflow better.

The question is whether those vertical tools can become large enough public-market winners.

AI changes the industry

AI has brought the CRM debate back to life.

For years, the CRM story was mostly cloud migration, subscription growth, multi-product expansion, and operating margin improvement.

Now the debate is more direct.

If AI agents can update records, summarise calls, qualify leads, draft emails, answer support tickets, book meetings, analyse customer health, trigger workflows, and forecast churn, then CRM vendors can sell more automation.

They can also face pressure on seat-based pricing.

A sales team may need fewer admin seats if AI agents do more of the work. A support team may need fewer human agents for simple tickets. A marketing team may automate more campaign work.

This does not automatically kill CRM.

It changes what customers are willing to pay for.

The old CRM pricing model was often based on users.

The AI CRM model may shift more toward usage, outcomes, automation volume, AI actions, data volume, or workflow execution.

That could help the winners.

It could also hurt vendors that sell expensive seats with low daily usage.

The most important AI truth in CRM is simple: AI needs clean data.

A company with poor customer records, broken integrations, duplicate accounts, bad permissions, weak governance, and unclear processes will struggle to get useful AI output.

This is why the data layer matters.

AI makes CRM more useful when the data is clean. It makes bad CRM implementations more obvious when the data is poor.

The AI risk is execution

There is a lot of excitement around AI agents.

The hard part is reliability.

CRM tasks are messy. A customer may have 4 open tickets, 2 subsidiaries, a renewal due in 60 days, a discount exception, a compliance flag, and an angry email from last week.

An AI agent handling that customer needs context, permissions, rules, memory, and auditability.

This is where enterprise software gets difficult.

A 2024 CRMArena benchmark tested AI agents on realistic CRM tasks across service agent, analyst, and manager personas. The benchmark found that leading LLM agents succeeded in less than 40% of tasks with ReAct prompting and less than 55% even with function-calling.

That should calm some of the hype.

AI agents are improving, but enterprise CRM requires accuracy, rule-following, permissions, and trust.

A customer service AI that gives the wrong refund, exposes private data, ignores a policy, or changes the wrong account can create real cost.

That is why large CRM vendors may still have an advantage.

They already own enterprise permissions, customer records, audit logs, security certifications, workflow rules, and CIO trust.

AI-native startups can move faster. Incumbents can sell trust.

The winner may need both.

The bull case for CRM and cloud software

The bull case is strong.

Companies still need systems to manage customers. Cloud adoption is still growing globally. AI can increase the value of customer data. Regulation can help large vendors because compliance is expensive. Enterprise buyers keep consolidating vendors to reduce software sprawl.

The best CRM companies can still grow through:

1. More products per customer.

2. AI usage fees.

3. Higher enterprise contract sizes.

4. Industry-specific products.

5. International adoption.

6. Partner marketplaces.

7. Data and analytics add-ons.

8. Customer service automation.

9. Workflow automation across sales, marketing, support, and operations.

The economic model is still attractive.

High gross margins, recurring revenue, low capital intensity, and strong cash conversion make CRM a strong business model when growth is disciplined.

Investors should like companies that can combine moderate revenue growth with margin expansion and buybacks.

That is why Salesforce remains interesting even with slower growth.

A 10% revenue grower with high margins, strong free cash flow, and disciplined capital returns can still create shareholder value if bought at the right price.

The bear case

The bear case is also serious.

CRM has maturity risk.

Salesforce is no longer a small hypergrowth company. Many large enterprises already have CRM systems. Growth comes from selling more products, raising prices, adding AI, winning replacements, and expanding internationally.

That is harder than selling cloud CRM for the first time.

There is also seat compression risk.

If AI reduces the number of people needed to handle sales admin, support tickets, reporting, and follow-ups, some vendors may lose seat growth.

There is pricing risk.

Customers may resist paying large AI premiums until the return is measurable.

There is competition risk.

Microsoft can bundle. HubSpot can simplify. ServiceNow can win workflow. Adobe can win marketing. Oracle and SAP can win where CRM connects to ERP. AI-native startups can attack narrow pain points with faster products.

There is implementation risk.

Many CRM projects fail because people do not use the system properly. Bad data, poor training, weak internal process, and over-customisation can destroy the value of the software.

There is regulation risk.

CRM systems hold personal data. Privacy rules, data localisation, AI governance, cybersecurity rules, and industry-specific regulations all increase compliance cost.

Large vendors can absorb that better than small vendors, but the cost is still real.

How to analyse a CRM company

Investors should avoid judging CRM companies only by revenue growth.

The better checklist is:

For Salesforce specifically, we would watch 5 items:

1. Organic subscription growth excluding acquisition contribution.

2. Agentforce ARR and actual production usage.

3. cRPO growth.

4. Free cash flow after buybacks and debt changes.

5. Customer willingness to pay for AI usage.

For HubSpot, we would watch upmarket progress and net retention.

For Microsoft, we would watch Dynamics adoption inside larger enterprise bundles.

For ServiceNow, we would watch workflow expansion into customer service and front-office processes.

For Adobe, we would watch whether marketing and experience products can defend budgets against AI-native tools.

For Oracle and SAP, we would watch CRM attachment to ERP and data platforms.

What makes a winner

A winning CRM and cloud software company needs more than a clean interface.

It needs durable customer value.

The best operators usually have 6 traits:

1. Strong retention

Retention proves the software matters.

If customers stay through budget pressure, pricing changes, and management turnover, the product is doing a real job.

2. Expansion revenue

The best CRM companies grow inside existing accounts.

One team starts using the product. Then another team joins. Then more modules get added. Then AI, analytics, data, workflow, or service products are attached.

3. Deep integrations

CRM becomes harder to replace when it connects to billing, support, marketing, ERP, data warehouses, identity systems, and communication tools.

Integration depth creates stickiness.

4. High data quality

AI makes this more important.

A CRM platform with clean account records, permissions, activity history, support data, product usage, and billing context has a better chance of producing useful AI output.

5. Efficient sales motion

Selling enterprise software can be expensive.

The best companies manage customer acquisition cost, partner channels, product-led adoption, and sales productivity carefully.

6. Trust

CRM vendors handle sensitive customer data.

Trust is not a soft factor. It affects procurement, regulation, renewal, and enterprise adoption.

Security breaches, AI mistakes, poor uptime, or weak governance can damage the thesis quickly.

The 5-year scenarios

Base case: platform consolidation and measured AI adoption

In the base case, CRM remains a core enterprise software category.

Large platforms keep winning major enterprise budgets. AI adoption grows, but customers demand proof before paying materially more. Seat growth slows in some areas, while usage-based AI revenue offsets part of the pressure.

Salesforce, Microsoft, ServiceNow, HubSpot, Adobe, Oracle, and SAP remain important.

The best returns come from companies that combine moderate growth, margin discipline, strong free cash flow, and credible AI monetisation.

Bull case: AI increases the value of customer data

In the bull case, CRM vendors successfully turn AI agents into paid production workloads.

Support agents handle more tickets. Sales agents reduce admin work. Marketing agents create and test campaigns. Customer success agents flag churn risk earlier. Managers get better forecasts.

Customers pay because the value is measurable.

This would favour vendors with clean data platforms, large installed bases, strong integrations, and trusted enterprise relationships.

Salesforce, Microsoft, and ServiceNow could benefit. HubSpot could also benefit if AI makes smaller teams more productive without heavy implementation.

Bear case: AI-native tools pressure the old model

In the bear case, buyers become more sceptical of traditional SaaS pricing.

AI-native tools handle narrow tasks faster and cheaper. Seat growth slows. Enterprise buyers push back against add-ons. Data remains messy. AI demos look good, but production results disappoint.

In that world, software vendors with weak product usage, poor customer satisfaction, and low net retention get punished.

The market would stop rewarding “AI attached to SaaS” and start asking for proof.

What BE Invested Labs is watching

The CRM debate is now about proof. Not press releases. Not demos. Not naming every product “AI”.

We are watching:

1. Agentforce, Copilot, HubSpot AI, ServiceNow AI, and Adobe AI production usage.

2. AI revenue separated from rebranded software revenue.

3. Net retention trends across CRM vendors.

4. cRPO and backlog quality.

5. Sales and marketing efficiency.

6. Free cash flow margin.

7. Cloud infrastructure cost as AI usage grows.

8. Customer support automation results.

9. Seat count pressure from AI.

10. Regulatory action around customer data and AI decision-making.

The most important question is whether AI increases what customers pay for CRM or reduces the need for traditional SaaS seats.

Both can happen at the same time.

That is why this industry is so interesting.

Our view

CRM and cloud-based software are still among the best business models in the market.

The economics are strong. The customer problem is real. The products sit close to revenue and retention. The best vendors have scale, switching costs, recurring revenue, pricing power, and large partner networks.

The industry is also more contested than it was 5 years ago.

AI brings new pricing models, new competitors, and new expectations from customers. It may make the best CRM platforms more useful. It may expose weaker platforms that were mostly selling seats, dashboards, and admin burden.

Our view is balanced. We like the structure of the industry. We are careful with valuation.

For investors, the cleanest approach is to separate business quality from price.

Salesforce can be a strong business and still be a difficult stock if expectations are wrong. HubSpot can be a great product and still be expensive. Microsoft can be the strongest strategic competitor even if Dynamics is only one part of the story. ServiceNow can benefit from workflow automation without being a pure CRM name.

The winners will show 4 things clearly:

1. Customers stay.

2. Customers spend more.

3. AI usage becomes paid.

4. Free cash flow keeps improving.

That is the CRM thesis.

The customer record is still worth owning. The next question is who turns it into the most profitable workflow.

— Buyce & Emmanuel
BE Invested Labs

References and data notes

Sources used for this breakdown include BE Invested Labs CRM industry reports on business models, core economics, value chain mapping, unit economics, regulation, TAM, future growth scenarios, and macro outlook. We also reviewed recent Salesforce earnings coverage, CRM market commentary, public SaaS benchmarks, and AI-agent research including CRMArena.

Disclaimers & Disclosures

AI Transparency
BE Invested Labs uses AI to aggregate financial data and format reports. The AI structures the information. Users determine their own investment thesis, risk management, and market outlook. We audit the system outputs to reduce AI errors. We use these exact reports to manage our own capital.

Legal Disclaimer
This platform provides educational and informational content. BE Invested Labs is not a registered investment advisor and does not provide financial, investment, tax, or legal advice.

Brookfield is easy to respect and hard to understand.

At first glance, it looks like a holding company. Then you realize it owns pieces of an asset manager, an insurance business, infrastructure platforms, renewable power assets, real estate, private equity businesses, credit strategies, and listed partnerships.

After that, the picture becomes clearer. Brookfield raises long-term capital, buys real assets, improves them, sells mature assets, and reinvests the proceeds.

That is the Brookfield model.

The BE Invested Labs research platform rates Brookfield Corporation (BN) as Neutral, with a probability-weighted fair value estimate of $48 per share. The bull case is $65. The bear case is $30. At the report price of $45.03, the implied upside to fair value was around 6.6%.

That matters because Brookfield is not a weak business. It is one of the most respected real-asset investment firms in the world.

The rating is Neutral because quality and complexity are both high. The business is impressive. The valuation gives less margin of safety than we would want for a company with this much debt, internal capital movement, non-IFRS reporting, and exposure to credit markets.

The useful question is simple:

Can Bruce Flatt and the Brookfield team keep investing larger amounts of capital at strong returns without the structure becoming too complex for public investors to properly value?

Our answer: probably yes over the long term, but the current setup still requires trust, patience, and valuation discipline.

The short version

Brookfield Corporation is a global investment firm focused on real assets and alternative investments.

It makes money from 5 main sources:

Management fees from Brookfield Asset Management.Performance fees and carried interest when funds perform well.Investment returns from capital Brookfield invests alongside clients.Insurance spread from Brookfield Wealth Solutions.Distributions and asset sales from operating businesses. The structure looks complicated because it is complicated. Brookfield Corporation owns roughly 73% of Brookfield Asset Management, owns and controls Brookfield Wealth Solutions, owns large stakes in Brookfield Infrastructure, Brookfield Renewable, Brookfield Business Partners, and owns Brookfield Property Group.

The center of the thesis is Brookfield Asset Management.

BAM raises money from pension funds, sovereign wealth funds, insurance companies, endowments, wealthy individuals, and private wealth channels. BAM invests that money into infrastructure, renewable power and transition, real estate, private equity, credit, and other private-market strategies. BAM charges fees on the capital and earns carried interest if fund returns pass the required hurdles.

Brookfield Corporation then benefits in 3 ways: it owns most of BAM, invests its own capital into BAM strategies, and uses BAM’s platform to invest insurance and operating business capital.

This is the strength of Brookfield. It is also the risk.

If BAM keeps producing strong returns, the whole Brookfield system can keep compounding. If BAM stumbles, the weakness spreads across fees, insurance returns, carried interest, reputation, and BN’s own investment returns.

The story: from Brazil to global alternatives

Brookfield’s history starts long before most modern asset managers existed.

The company traces its roots to 1899, when São Paulo Tramway, Light and Power Company was incorporated. The original business helped build electricity and tram infrastructure in Brazil. Over time, the company became Brazilian Traction, then Brazilian Light and Power, then Brascan.

The name Brascan came from Brazil and Canada.

The company slowly moved from operating utilities in Brazil to owning and managing real assets across different markets. In 1997, Brascan merged with The Edper Group to form EdperBrascan. In 2000, the company changed its name back to Brascan Corporation. In 2005, it became Brookfield Asset Management.

The Bruce Flatt era began in 2002, when he became CEO.

That period changed the scale of the company. Brookfield moved deeper into infrastructure, renewable power, real estate, credit, distressed assets, and private equity. It built listed vehicles for retail and institutional investors. It bought and sold assets across cycles. It became one of the largest alternative asset managers in the world.

Then came the 2022 split.

Brookfield Asset Management Inc. changed its name to Brookfield Corporation, ticker BN. The asset management business was separated into a new publicly listed Brookfield Asset Management, ticker BAM. BN kept a large ownership stake in BAM and continued to own its operating businesses and wealth solutions platform.

This is why BN and BAM should not be treated as the same company.

BAM is the asset manager. BN is the parent investment company that owns BAM shares, operating assets, insurance exposure, and invested capital across the Brookfield group.

That distinction matters for valuation.

BAM is cleaner. BN is more complex. BAM earns fees. BN gets fees indirectly, investment returns, insurance earnings, operating business distributions, carried interest, and asset sale gains.

BN has more ways to win. It also gives investors more moving parts to understand.

How Brookfield is structured

Brookfield Corporation has 3 main parts.

Asset Management

This is Brookfield Asset Management.

BAM manages more than $1 trillion of assets across infrastructure, renewable power and transition, private equity, real estate, and credit. It earns management fees on fee-bearing capital and can earn carried interest when funds deliver strong returns.

This is the highest-quality part of Brookfield’s structure because fee-related earnings are recurring, high-margin, and linked to capital that is often locked up for long periods.

For BN, the asset management business matters in 4 ways:

1. BN owns around 73% of BAM.

2. BN receives value as BAM’s fee-related earnings grow.

3. BN can earn carried interest through BAM’s fund performance.

4. BN invests its own capital into and alongside BAM-managed funds.

At year-end 2025, Brookfield valued its stake in the Asset Management business at $61.5B. BAM generated $3.0B of fee-related earnings, up 22% year over year. Brookfield also had $11.6B of accumulated unrealized carried interest before direct costs, with $9.4B attributable to the Corporation.

That is one of the big valuation arguments.

If BAM keeps growing fee-bearing capital and turns unrealized carry into realized cash over time, BN’s value should rise.

Wealth Solutions

This is Brookfield’s insurance business.

The simple version: Brookfield takes in long-duration insurance and retirement liabilities, invests the capital, and tries to earn more on the assets than it owes on the liabilities.

This is why insurance can be powerful inside Brookfield.

The capital is long term. Brookfield has a large asset management platform. BAM can manage the insurance assets. Brookfield can invest in private credit, infrastructure debt, real estate debt, and other income-producing assets where it believes it has an advantage.

At the end of 2025, Brookfield Wealth Solutions had $143B of insurance assets and $1.9B of annualized cash flows. In Q1 2026, after the Just Group acquisition, insurance assets increased to about $180B.

The business targets a 15% return on equity.

The opportunity is clear. If Brookfield can scale insurance assets and invest them safely, Wealth Solutions becomes another large source of cash flow.

The risk is also clear. Private equity-backed insurance has attracted more scrutiny. Regulators are watching duration matching, liquidity, related-party activity, and how much insurance capital gets invested into affiliated private funds.

Our read is that Brookfield’s insurance business looks more conservative than many headlines suggest. The liabilities are long duration and more predictable than demand deposits. The company also says it holds large amounts of liquid assets.

The risk still deserves attention because this business depends on trust, regulation, and asset quality.

The operating businesses are lower-margin and more capital intensive than asset management. They also give Brookfield something many asset managers do not have: operating knowledge.

Brookfield buys assets, runs them, improves cash flows, refinances when appropriate, sells mature assets, and redeploys the proceeds.

That operating culture is the part of Brookfield that investors either love or struggle to underwrite.

How Brookfield makes money

Brookfield makes money through recurring fees, investment income, insurance spread, distributions, carried interest, and asset sales.

The asset management fee stream is the easiest to understand. If BAM manages $614B of fee-bearing capital, it charges fees on that capital. As fee-bearing capital grows, fee-related earnings should grow too, assuming fee rates and margins hold.

Carried interest is less predictable. Brookfield earns carry when fund returns pass agreed thresholds. This can become very large, but it depends on asset sales, fund performance, market liquidity, and timing.

Insurance spread is another earnings stream. Brookfield Wealth Solutions takes in retirement and insurance capital, owes a cost on that capital, and invests the assets to earn a higher return. The spread is profit.

Operating distributions come from Brookfield’s stakes in BIP, BEP, BBU, and real estate. These cash flows are often linked to contracted or inflation-linked revenues.

Asset sales are the fifth source.

Brookfield buys assets, improves them, and sells mature assets when the market price is attractive. In Q1 2026, Brookfield executed $17B of asset sales across the business. The company said substantially all sales were completed at or above carrying values.

This is one of the most important tests of the Brookfield story.

Skeptics often argue that private asset managers may overstate asset values. Brookfield’s answer is simple: when it sells assets, it aims to sell them at or above the values carried in its books.

That does not remove the need for skepticism. It gives investors a real test to monitor.

Why Brookfield is hard to copy

Brookfield’s moat does not come from one asset, one fund, or one famous deal.

It comes from the combination of scale, trust, operating knowledge, long-duration capital, and the ability to move capital across different parts of the group. That is why Brookfield is difficult to analyze. The same structure that creates the advantage also creates the complexity.

The BE Invested Labs 7 Powers review rates Brookfield’s moat as Wide and its moat trend as Strengthening. We agree with that assessment, with one condition: the moat depends on Brookfield continuing to allocate capital well.

If BAM’s returns weaken for a long period, the whole structure becomes less powerful.

Here is how the 7 Powers framework applies to Brookfield.

The strongest powers are scale, switching costs, brand, counter-positioning, and process power.

Scale matters because Brookfield can do transactions that smaller firms cannot fund or operate. A $30B infrastructure strategy changes the types of assets Brookfield can pursue and the type of counterparties it can sit across from.

Switching costs matter because Brookfield’s customers are pension funds, sovereign wealth funds, insurance companies, endowments, foundations, family offices, and other institutions that think in decades. If Brookfield performs, those investors often come back for the next fund.

Brand matters because the assets are complex. A government selling infrastructure, a company looking for a partner, or a pension plan allocating billions wants a manager that can survive a bad cycle. Brookfield’s history helps it raise capital even when headlines around private markets are negative.

Counter-positioning matters because Brookfield sits between 2 worlds. It has the investment platform of a large asset manager and the operating culture of a company that owns physical assets. That is rare. A financial sponsor can copy the language, but building the operating depth takes years. An industrial company can own assets, but raising and recycling third-party institutional capital is a different discipline.

Process power may be the most important one. Brookfield’s best deals are usually not based on buying something and waiting for a higher multiple. The best examples involve improving cash flows. Westinghouse, Clarios, toll roads, ports, rail, renewable power, and infrastructure assets all show the same pattern: Brookfield wants to buy assets where operational decisions can change the economics.

This is also where the risk sits.

The moat is only useful if Brookfield keeps using it properly. Scale can become a burden. Complexity can hide mistakes. Internal capital movement can create conflicts. Large funds can make it harder to find high-return deals. The more capital Brookfield manages, the more discipline it needs.

That is why we do not treat the moat as a free pass.

Brookfield has earned a premium because the track record is real. It still needs to prove that the same method can work at a much larger scale, inside a more complex structure, with higher interest rates and more scrutiny on private markets.

That is the Brookfield trade-off: a wide moat, a trusted management team, and a structure that still requires investors to keep asking hard questions.

Famous investment moves

The moat becomes easier to understand when you look at the deals.

Brookfield’s reputation was built through transactions where the company bought complicated assets, improved operations, and returned capital through sales, refinancings, or distributions.

Oaktree

The Oaktree deal was one of the most important moves in Brookfield’s modern history.

Brookfield bought a majority stake in Oaktree in 2019. Oaktree brought distressed credit, opportunistic credit, and the Howard Marks and Bruce Karsh reputation into the Brookfield group.

That deal mattered because credit is now one of the largest growth areas in alternatives. Oaktree also gave Brookfield deeper access to private credit, distressed debt, and dislocation opportunities.

In 2025, Brookfield agreed to acquire the remaining Oaktree stake for about $3B, making Oaktree even more central to Brookfield’s credit strategy.

For investors, Oaktree matters because it gives Brookfield another way to profit when markets become stressed.

Westinghouse

Brookfield acquired Westinghouse out of bankruptcy in 2018.

The thesis was that Westinghouse’s fixed-price nuclear construction problems had hidden a valuable nuclear services business. The core business served the installed nuclear fleet with fuel, maintenance, and services.

According to the research pack, Brookfield worked with the Westinghouse team, made add-on acquisitions, improved cash flows, and later sold the business at an implied enterprise value of around $8B. The reported result was roughly 6x invested capital and an IRR around 60%.

That is the Brookfield method in one example: buy when a good business is hidden inside a bad situation, fix the structure, improve cash flows, then sell when the market appreciates the asset.

Clarios

Brookfield acquired Clarios from Johnson Controls in 2019.

Clarios makes automotive batteries. The business had a long operating history, large scale, and exposure to replacement battery demand. The BE Invested research pack describes it as one of Brookfield’s most documented operational transformations.

Brookfield improved operations, reduced debt, and increased cash generation. The research pack says Brookfield later refinanced the business and received a $4.5B special distribution while still owning the equity interest.

That matters because it shows the business can return capital before a full exit.

Infrastructure assets

Brookfield has also created value through rail, ports, toll roads, pipelines, data infrastructure, telecom towers, and renewable power.

The research pack gives examples such as Australian rail expansion, Chilean toll roads, and global toll road operations where Brookfield created value through capacity additions, traffic growth, contract wins, and better cost management.

These examples are less glamorous than AI or semiconductors. They are exactly the type of assets Brookfield likes.

They produce cash. They require operating skill. They can use debt responsibly if cash flows are predictable. They can become more valuable when cash flows grow.

The financial picture

Brookfield’s financials require care because GAAP earnings do not fully explain the business.

The company focuses heavily on Distributable Earnings, or DE. Brookfield uses DE to show the earnings available for distribution to shareholders, built from asset management earnings, wealth solutions earnings, operating business distributions, realized carried interest, and disposition gains.

This metric is useful because Brookfield is a capital allocation business.

It is also a management-defined non-IFRS metric. Investors need to understand what is included, what is excluded, and how much judgment sits inside the number.

Here are the latest headline numbers we are using:

The financial picture is strong if you accept Brookfield’s DE framework.

Brookfield has growing fee-related earnings, a large insurance asset base, large deployable capital, and active asset sales. It is also repurchasing shares when management believes the stock trades below intrinsic value.

In Q1 2026, Brookfield repurchased $470M of BN shares at an average price of $41 and $575M of BAM shares.

Management said the BN repurchases were done at roughly a 40% discount to its view of intrinsic value at quarter end, which management estimated at $66 per share.

That is an important data point.

Our BE Invested Labs model is more conservative. It estimates fair value at $48. The difference comes from Brookfield management valuing the business from inside the system and our model applying a higher discount for complexity, leverage, and credit risk.

The bull case

The bull case is built on 5 points.

1. Brookfield has a long record of capital allocation

Brookfield says shareholders have earned 19% annual compound returns over more than 30 years.

That record matters. Few companies can point to multiple decades of strong returns through different interest rate regimes, crises, real estate cycles, credit cycles, and market crashes.

Bruce Flatt and the Brookfield team have earned the right to be taken seriously.

2. The assets are in demand

Brookfield focuses on assets tied to infrastructure, energy, data centers, transmission, real estate, credit, renewable power, nuclear services, and essential operating businesses.

The AI infrastructure theme is especially important now.

AI needs data centers. Data centers need power. Power needs generation, storage, transmission, land, cooling, financing, and operators that can work across governments, utilities, hyperscalers, and capital markets.

Brookfield has exposure across several of those pieces.

This is why management keeps talking about digitalization, decarbonization, and deglobalization. The language can sound corporate, but the underlying asset need is real.

The world needs more power, more grid capacity, more data infrastructure, more private capital, and more complex financing.

Brookfield has built a business around those needs.

3. BAM is still growing

Brookfield Asset Management is the center of the thesis.

Fee-bearing capital reached $614B in Q1 2026, up 12% year over year. Fee-related earnings increased 11% in the quarter.

That is important because the market often fears that large asset managers eventually run out of runway. Brookfield’s current numbers suggest the platform is still growing.

BAM also benefits from repeat capital.

Large pension funds, sovereign wealth funds, insurance companies, and institutions tend to re-commit to managers that perform well and treat them fairly. This creates long-term relationships that are difficult for smaller managers to copy.

4. Insurance can add another earnings base

Brookfield Wealth Solutions is becoming more important.

After Just Group, insurance assets increased to about $180B. Brookfield wants to combine BN and BNT into a larger insurance and investment organization.

The bull case is that insurance gives Brookfield a large base of long-duration capital. BAM can manage the assets. BWS can earn spread. BN can use the combined capital base to support growth.

This could make Brookfield look more like a mix of asset manager, insurer, and real-asset investor.

If it works, the market may value BN with less of a conglomerate discount.

5. Asset sales support the marks

Brookfield had a record year for asset sales in 2025, and Q1 2026 stayed active with $17B of asset sales.

The company says substantially all Q1 sales were completed at or above carrying values.

For a private asset manager, this matters. The quality of marks is tested when assets are sold. If Brookfield keeps selling assets at or above carrying values, the bear argument around inflated private marks becomes harder to defend.

This is one of the most important monitoring points for BN.

The bear case

The bear case is not about Brookfield being a bad business.

It is about the risks that come with the model.

1. The structure is extremely complex

Brookfield has asset management, insurance, operating businesses, listed affiliates, private funds, internal capital commitments, direct investments, public stakes, carried interest, and real estate holdings.

Capital can move between entities. One Brookfield entity can manage capital for another Brookfield entity. BN can invest into BAM funds. BWS can allocate assets to Brookfield-managed strategies. Operating businesses can use capital sourced from Brookfield’s broader system.

This structure can create advantages.

It can also make the business difficult to analyze from the outside.

Public investors have to trust that capital is being allocated fairly, reported properly, and managed with long-term discipline.

2. Distributable Earnings needs trust

DE is the main metric for BN.

It is also a non-IFRS metric. Brookfield defines it. Brookfield reconciles it. Brookfield uses it to explain the business.

We understand why DE matters. GAAP earnings can be noisy for a business that owns assets, sells assets, consolidates some businesses, equity accounts others, and holds unrealized gains.

Still, investors should not outsource judgment.

The useful way to read BN is to track DE, fee-related earnings, realized carry, asset sales, cash flow, liquidity, and debt together.

DE alone is too clean for a company this complex.

3. The debt story needs careful interpretation

Brookfield’s consolidated debt looks massive.

The key defense is that much of it is non-recourse, asset-level financing. That means a problem at one asset should not automatically become a claim on the parent company.

That structure has worked over time.

The risk is a systemic credit event. If refinancing markets freeze, commercial real estate values fall, and lenders demand more equity, Brookfield may have to choose between walking away from assets, injecting capital, or accepting lower returns.

Brookfield has significant liquidity and long-dated corporate debt. Q1 2026 disclosures show $188B of deployable capital and a 15-year weighted-average term on corporate debt.

That helps.

The risk does not disappear. The business is still sensitive to rates, credit spreads, refinancing markets, and asset values.

4. Real estate has already shown stress

Brookfield’s real estate business has had public office defaults in recent years.

The company’s response is that a large property portfolio will always have weaker assets during stressed markets. It is also shifting its real estate exposure toward what it considers the best assets and reducing some cyclicality over time.

Q1 2026 data showed strong occupancy in super core and core plus assets, and a successful $1.9B refinancing of Two Manhattan West.

That is a positive sign.

The real estate issue is still worth watching. Office, retail, refinancing rates, tenant demand, and asset sales all matter.

5. The whole thesis depends on BAM’s future returns

This is the biggest risk.

Brookfield’s insurance profits depend on how the capital is invested. BN’s retained cash flows are reinvested into the system. Carried interest depends on fund performance. BAM’s fee growth depends on investors trusting the platform and committing more capital.

If BAM stops producing strong returns, the entire system slows down.

That is why BN is ultimately a bet on capital allocation.

Brookfield can own the right asset classes, attract the right clients, and raise the right funds. The value still depends on buying well, financing well, improving operations, and selling well.

Valuation

The BE Invested Labs platform rated BN as Neutral.

The fair value estimate is $48 per share.

The platform’s conclusion is not bearish.

It says the current price does not offer a large enough margin of safety for the risks.

This is where our view lands too.

Brookfield is a high-quality business with a long runway, strong management, and exposure to real assets that should matter more over time. It also requires more trust than the average public company.

If the stock trades at a large discount to a conservative view of intrinsic value, the complexity becomes easier to accept. If it trades close to fair value, the investor is being paid less for taking on that complexity.

That is the difference between liking the business and liking the price.

What we are watching

Brookfield is measurable if you focus on the right items.

For us, the most important numbers are fee-related earnings, DE before realizations, asset sales versus carrying value, and parent-level liquidity.

If those remain strong, the Brookfield thesis is intact.

If those weaken together, the bear case gets louder.

Our view

Brookfield is one of the rare companies where the business quality is clear and the structure still demands caution.

We respect the track record. We respect Bruce Flatt. We respect the owner-operator culture. We respect the fact that large institutions keep giving Brookfield more capital. We also respect the fact that public investors should not blindly accept management-defined metrics from a complex investment company.

The BE Invested Labs research platform lands on Neutral, with a $48 fair value estimate.

Our qualitative view is more positive on the long-term business than the rating may sound. Brookfield is probably one of the better ways to get exposure to real assets, private markets, infrastructure, energy demand, AI infrastructure, and long-term capital allocation.

The valuation view is more restrained.

BN is not a simple compounder. It is a complex compounder.

The business can work very well for patient investors if BAM keeps producing strong returns and management keeps simplifying the structure. The same complexity can punish investors if credit markets turn, real estate stress grows, or the market loses trust in DE and private asset marks.

That is the Brookfield debate.

The company is good. The current price needs more margin of safety.

Our answer: close, but not enough to call it compelling.

Try BE Invested Labs

Use BE Invested Labs to generate structured company research, valuation notes, bull and bear cases, and AI audio briefings.

Subscribe now

Subscribe to BE Invested Labs for more business breakdowns like this.

— Buyce & Emmanuel, BE Invested Labs

References and data notes:

1. BE Invested Labs Master Synthesis report

2. BN official website

Disclaimers & Disclosures

AI Transparency: BE Invested Labs uses AI to aggregate financial data and format reports. The AI structures the information. Users determine their own investment thesis, risk management, and market outlook. We audit the system outputs to reduce AI errors. We use these exact reports to manage our own capital.

Legal Disclaimer: This platform provides educational and informational content. BE Invested Labs is not a registered investment advisor and does not provide financial, investment, tax, or legal advice.

The Problem

Published numbers of data centers don’t add up

Every announce of new data center creates a fascade of growing AI economy where investors pump in capital in these ventures. Open any technology publication today and you will find announcements of staggering scale of a half-trillion-dollar campus in Texas, a two-gigawatt facility the size of part of Manhattan, server farms that will consume more electricity than entire cities. The data centre industry has never looked more ambitious or more crowded with press releases.

But look past the headlines and a troubling pattern emerges. Analysts at Sightline Climate are currently tracking 190 GW of capacity across 777 large data centres and AI factories announced since January 2024. Out of the facilities due to come online in 2026, only around 5 GW is confirmed under active construction. The typical construction timelines can be 12–18 months. 1

In other words, the majority of the world’s most-discussed infrastructure boom exists, for now, primarily on paper.

"The market is not cooling. It is overcommitted and underdelivered. Projects are being leased before construction starts, but grid delays and permitting hold-ups are pushing delivery further out. This isn't a demand problem. It is a development bottleneck."

- Colby Cox, Managing Director Americas, DC Byte.

DC Byte's 2025 Global Data Centre Index quantifies the divergence with striking precision: committed supply has increased more than six times since 2019, while live capacity has grown at a far slower pace. This is not a rounding error but there is a structural disconnect between what the industry declaration and what is built2.

Agitate

Reasons that keeps this gap growing

To understand why so many proposed data centers and capex exist on paper. This is due to a number of factors. There is a disconnect during the press release and the construction of new data centers. These are not isolated problems. They tend to arrive together and each problem compounds on the other.

1. Grid Connection: It takes up to 12-18 months to construct new data centers. Connecting them to the electrical grid currently takes five to seven years. Hence there is a backlog of data centers that can not be connnected to national grids. Grid-power capacity in most major US markets is already committed through 20303

2. Construction permit:Between March 2024 and 2025, 16 data centre developments were delayed or denied due to permitting restrictions. Permitting accounted for 29% of all milestone change requests for projects under construction, according to PJM grid operator data.[³4]

3. Equipment Shortages: Supply chain issues for electrical back-up generators, cooling system account for the delays. Electrical sub-station transformer lead times averaged around 140 weeks in 2023, rose to 150 weeks in 2025, and now exceed 160 weeks in 2026. Data center constuctors are competing directly with utilities providers for the same limited pool of electrical components.[⁴]

4. Community oppostion: Most residences living close to data centers, heavily oppose these huge construction despite the promise of jobs. At least 188 local activitist opposition groups are now active across 40 US states. Data centre project cancellations more than quadrupled from 6 in 2024 to 25 in 2025. Public concern over water use, grid strain, and environmental impact has become a genuine material risk.56

5. Capital Constraints: This might be surprise. These hyperscale can not afford to expontially build these data centers with debt. New data center deal volume fell more than 40% between Q3 and Q4 of 2025. Out of the 240 GW planned, only one-third is actually under construction. Even hyperscalers have begun cancelling or pausing leases, with Microsoft alone withdrawing "a couple of hundred megawatts" of commitments.7

6. Strategic Facade: Many announcements are intentionally speculative: some data center developers launch multiple projects in parallel and build only those that secure power and permits fastest. Announcements serve investor relations and competitive positioning as much as they do operational planning. ^[1]

7. Environmental Concerns: Data centers present severe environmental challenges across energy, water, and local ecosystems. Large data centers can consume up to 5 million gallons of water daily equivalent to the footprint of a town of 50,000 people. Sound pollution of continuous humming from rooftop HVAC systems and massive industrial fans can reach internal levels of 96 decibels. Externally, this noise creates a persistent low-frequency hum that disrupts sleep of residence. E-waste acceleration with the hardware lifecycles of cloud computing add a hidden layer of chemical and physical waste to the environment such as the PFAS “forever chemicals“ accumulate permanently in the local soil and water.

The track record bears the truth

Goldman Sachs Research, drawing on granular facility-level data from Aterio that incorporates satellite imagery and permitting progress, found that historically only about 72% of data centres scheduled for activation within the following four quarters actually come online on time. The further out a project is scheduled, the worse that ratio becomes.8

Bessemer Venture Partners, points out that out of the 110 data centre projects slated to come online in 2025, more than a quarter were delayed due to power, permitting, and construction constraints. ^[3]

"What looks viable on paper usually breaks down on power, permitting or labor, and sometimes all three."

— Brennan Church, Director of engineering, procurement and construction at data center developer Hut, New York Build Conference 2026 ^[5]

The problem is not limited to any single geography. Across Europe the equivalent strain is also exist, particularly in London, Dublin, and Frankfurt, where low vacancy rates, planning restrictions, and constrained power availability are slowing delivery. In Southeast Asia, the volume of live capacity is falling behind the rate of announcements even in high-growth markets. ^[2]

Construction costs are rising on top of all this: Turner & Townsend’s 2025 Data Centre Construction Cost Index shows a 5.5% increase per watt for traditional cloud data centres, with AI-specific facilities facing cost variables that are still too early to benchmark globally. The three most expensive markets — Tokyo, Singapore, and Osaka now cost upwards of $14 per watt to build. ^[9^]

Solution

Power certainty is the Primary filter criterion

Grid capacity has displaced fibre connectivity, land cost, and tax incentives as the first filter in serious site selection. DC Byte’s regional analysis shows that markets with predictable power supply and coordinated planning frameworks deliver capacity more consistently, with smaller gaps between announced, committed, and live supply. Norway’s qualified supply grew at a 43% five-year CAGR since 2019; Finland added more than 1,400 MW of IT load across 2023 and 2024 alone both markets benefiting from stable grid environments. ^[2]

Sites offering power access within 18 to 36 months are among the most sought-after assets in the market. Data center developers who can demonstrate grid certainty, not just construction plans, command premium positioning with anchor tenants. 10

On-Site power is a strategic approach

Hyperscalers are increasingly bypassing the public grid entirely for AI training capacity. Hybrid systems combining batteries with gas turbines, solar-plus-storage. In Google’s case, investment in nuclear reactor partnerships are becoming mainstream. The US Department of Energy issued a directive in October 2025 proposing federal jurisdiction over new loads above 20 MW, aiming to create standardised, faster approval pathways for projects willing to accept occasional curtailment during grid stress events. ^[4, 11^]

Execution capability is now a competitive moat

Bain & Company’s 2030 global data centre forecast identifies a decisive shift underway: the early scramble of AI-driven demand is giving way to a more disciplined, execution-focused phase. Data center developers who have already secured power, land, and permits are in a position to lead; those who have not may be forced to pause or substantially rework plans as grid timelines extend. [12]

“As demand continues to grow, the real differentiator is no longer how much capacity is announced, but how much can actually be delivered. Power availability, planning certainty, and realistic timelines are now what separate markets that scale from those that stall.”— Siddharth Muzumdar, SVP of Research, DC Byte ^[2]

Customers particularly large enterprises and cloud buyers are increasingly prioritising partners who can offer contractual certainty around timelines and activation, not just headline capacity numbers. This is reshaping how capital is deployed: investors are applying more rigorous deliverability, durability, and adaptability screens before committing, rather than treating announced pipeline as a reliable proxy for future supply. ^[2, 13^]

For Investors and Analysts: Read Beneath the Announcement

No one know with certainty if we are in an AI bubble compared with the DotCom bubble where there were speculative investing, reckless capex expenditure, investing in business model that were not profitable such as notable unprofitable “paper” business Pet.com. We at BE Invested, believe that if there will be an AI bubble in the future. This will be initial indication point.

For investors, a practical implication for anyone tracking this space whether allocating capital, procuring capacity, or modelling energy demand is that announced figures require heavy discounting and due diligence. Goldman Sachs’ research cautions explicitly that ambitious construction schedules carry meaningful uncertainty in both directions, and notes that elevated capital spending could compress timelines while delays and cancellations could reduce activated capacity well below current projections. ^[8]

The ITIF, writing in April 2026, offered the clearest summary of the gap: Out of 240 GW of planned construction, only one-third is actually being built. The Stargate project where OpenAI’s flagship $500 billion commitment in Texas. Stargate itself is stalled amid disputes between partners. Announcements, in the current environment, are better understood as options than as commitments. ^[7]

The data centre industry will almost certainly build a remarkable amount of new infrastructure over the next five years. McKinsey projects $6.7 trillion in cumulative global capital expenditure between 2025 and 2030 to meet demand, with $5.2 trillion dedicated to AI-specific infrastructure. ^[13] But the gap between what gets announced and what gets built will remain large and understanding why that gap exists is the prerequisite for navigating the AI infrastructure era with any degree of analytical clarity.

If you enjoy B.E Invested Labs, it would mean the world to me if you invited friends to subscribe and read with us. If you refer friends, you will receive benefits that give you special access to B.E Invested Labs. Also visit beinvestedlabs.com for your stock research.

How to participate

1. Share B.E Invested Labs. When you use the referral link below, or the “Share” button on any post, you'll get credit for any new subscribers. Simply send the link in a text, email, or share it on social media with friends.

Refer a friend

2. Earn benefits. When more friends use your referral link to subscribe (free or paid), you’ll receive special benefits.

• Get a 1 month comp for 3 referrals

• Get a 3 month comp for 5 referrals

• Get a 6 month comp for 25 referrals

Visit the leaderboard

To learn more, check out Substack’s FAQ.

Thank you for helping get the word out about B.E Invested Labs!

The bull case sounds simple at first: the company is profitable, growing, using AI to improve credit decisions, and trading at a low valuation compared with many fintech peers.

The bear case is just as direct: Pagaya still depends on consumer credit, securitization markets, institutional funding, partner concentration, and risk-retention assets.

Both sides have evidence.

That is why this business is worth breaking down properly.

This article was also written as a response to Deacon Brantley’s Flashpoint article, “The Next Visa: A Deep Dive Into Pagaya.” Deacon made a strong argument that the market may be treating Pagaya like a risky lender when the business is closer to a fee-based credit network.

We agree with part of that framing.

We also think the Visa comparison needs limits.

Pagaya may be a credit network. It is also a credit network operating inside consumer lending and asset-backed securitization markets.

That distinction matters.

The 30-second view

Pagaya is an AI-powered credit platform.

It works with banks, fintech lenders, auto finance providers, and point-of-sale financing partners. When a borrower fails a lender’s first credit screen, Pagaya can take a second look using its AI model.

If Pagaya’s system sees an acceptable risk-return profile, the borrower may still receive an offer. The lending partner keeps the customer relationship. Institutional investors fund the exposure, often through asset-backed securitizations. Pagaya earns fees from the transaction flow.

The opportunity is clear. Pagaya may be building a capital-light credit network that sits between lenders and institutional capital.

The risk is also clear. The model depends on credit performance, funding markets, partner flow, and investor appetite for consumer credit.

Our current view is constructive, but strict.

Pagaya is improving fast enough to deserve a serious re-rating discussion. The company still needs a credit-cycle discount until FRLPC, partner concentration, retained-risk exposure, and ABS funding prove durable under stress.

If you want more stock breakdowns like this, subscribe to BE Invested Labs.

Subscribe now

What Pagaya actually does

Pagaya sits inside the consumer credit process.

A borrower applies for a loan or financing product through a lending partner. The partner uses its own underwriting model first. If the borrower does not pass that model, Pagaya can evaluate the application again.

That second look is the business.

Pagaya uses machine learning, repayment data, and its credit models to decide whether the borrower can be approved at an acceptable risk level. If the answer is yes, the loan can be originated by the partner and funded through Pagaya’s capital network.

Pagaya connects 3 sides:

1. Lending partners that want more approved customers.

2. Borrowers who may be rejected by traditional models.

3. Institutional investors looking for consumer credit assets.

The company makes money when credit flows through that network.

That structure is attractive because Pagaya does not need to spend like a consumer app to acquire borrowers directly. It plugs into partners that already have customer demand.

That is the B2B2C logic.

Pagaya gets access to demand through its partners. Partners get more approvals without taking the full balance-sheet risk. Institutional investors get access to credit assets.

When this works, Pagaya can look like a platform. When funding markets tighten, it can look much closer to a credit-cycle business.

Why the market may be misreading PGY

Pagaya still carries SPAC-era baggage.

That matters. Many investors got burned by growth fintech names that promised a lot, lost money, and then got hit by higher interest rates.

Pagaya was part of that broader group. The market does not easily forget.

The company is now trying to prove that the old version of the story is outdated.

The numbers have improved. Pagaya reported its fifth consecutive quarter of GAAP profitability in Q1 2026. It also raised full-year 2026 net income guidance.

That changes the discussion. The question is no longer whether Pagaya can produce any profit at all. The question is whether the profit is durable enough to deserve a higher multiple.

Q1 2026 scorecard

Pagaya’s Q1 2026 report was strong on the surface.

The clean positive signal is profit growth.

Revenue grew 10%. Adjusted EBITDA grew 18%. Operating income grew 68%. That means revenue is growing faster than some major cost lines. For a platform-style business, that is exactly what investors want to see.

The more cautious signal is FRLPC margin.

FRLPC increased 5%, but FRLPC as a percentage of network volume fell to 4.6%. Pagaya said this reflected asset class mix, new partner contributions, tighter ABS pricing, higher cost of capital, and market conditions.

That is important. If volume grows but Pagaya earns less per dollar of volume, the quality of growth can weaken. If FRLPC stabilises while volume grows, the thesis gets cleaner. For Pagaya, FRLPC is one of the most important numbers to track.

If you know someone watching PGY, fintech, AI lending, or consumer credit, share this breakdown with them.

Share

The number management wants investors to notice

Management highlighted that network volume grew 23% excluding SFR.

That is useful. It tells us that the core business looks stronger when Single-Family Rental is removed from the comparison.

It also needs context.

Headline network volume grew 9%. Ex-SFR growth was 23%. That gap tells us SFR affected the comparison and should not be ignored.

Our BE Invested Labs earnings-call sentiment report flagged this as a narrative shift. In earlier periods, SFR had been part of the broader diversification story. In Q1 2026, the main narrative focused more on Auto and Point-of-Sale growth, while SFR moved into the background.

That does not mean management is hiding a disaster.

It means investors should separate headline growth, ex-SFR growth, and the quality of growth by vertical.

The best version of the bull case does not need SFR to be ignored. It needs personal loans, auto, and POS to grow with healthy economics.

Guidance is stronger, but the test is still ahead

Pagaya raised its full-year 2026 outlook.

This is a meaningful step.

Pagaya is guiding for full-year GAAP net income, not only adjusted EBITDA.

That matters because the market has been skeptical of many fintech companies that rely heavily on adjusted metrics.

Still, guidance is not proof of durability.

Management can control costs and risk posture. It cannot fully control unemployment, interest rates, ABS spreads, securitization demand, or consumer credit stress.

The next few quarters should be judged by quality:

Is growth broad across products?

Is FRLPC stable?

Are ABS deals clearing at workable spreads?

Are credit losses within expectation?

Are retained interests holding their value?

Is GAAP net income still growing without too much dependence on adjustments?

Those are the questions that matter.

Why the Visa comparison works

Deacon’s Visa framing is useful because it forces investors to think about category.

Visa does not lend money to cardholders. It connects banks, merchants, and cardholders, then earns fees across the network.

Pagaya wants a similar position in credit origination.

It wants to sit between lenders and institutional capital. It wants to approve more borrowers, route exposure to funding partners, and earn fees from the flow.

That is a good business if it works.

The model has the ingredients investors usually like:

• Low direct customer acquisition cost.

• Partner distribution.

• Data accumulation.

• Fee-based revenue.

• Operating leverage.

• More volume flowing through one infrastructure layer.

That is why PGY can look mispriced.

If the market treats Pagaya like a distressed lender, the valuation can stay low. If the market starts treating Pagaya as a profitable credit network, the multiple can expand.

That is the bull case in one sentence.

Where the Visa comparison breaks

Visa does not depend on securitizing consumer loans.

Pagaya does.

That is the main difference.

Pagaya’s business still depends on institutional investors wanting consumer credit exposure. It also depends on securitization markets staying open at acceptable economics.

If ABS investors pull back, Pagaya’s network gets harder to run.

If consumer defaults rise, retained assets can be marked down.

If funding spreads widen, FRLPC can compress.

If a large partner reduces volume, the network effect can look weaker than expected.

This is why PGY should not be analysed with one label.

The business is not a traditional lender. It is also not Visa.

A better category is this:

Pagaya is an AI credit network whose value depends on underwriting quality, funding access, partner depth, and credit-cycle durability.

That is the more precise thesis.

Cash conversion is the strongest pushback against the bear case

One of the best points in Pagaya’s favour is cash conversion.

Our BE Invested Labs forensic accounting report found that in FY2025 Pagaya generated $238.6M of operating cash flow against $81.4M of GAAP net income.

That is about $2.93 of operating cash flow for every $1 of reported earnings.

That matters.

A common bear argument is that Pagaya’s profitability may be heavily shaped by accounting marks, fair value adjustments, securitization structures, or retained interests.

Cash flow does not answer every question, but it makes the “accounting profit only” argument harder.

The business is producing cash.

The caution is that Pagaya also uses cash and capital to support risk-retention assets and securitization-linked investments. So cash conversion should be read alongside the balance sheet, not in isolation.

The clean view:

Cash generation has improved. The balance sheet still needs careful reading.

The Buffett test: yellow, not green

Our Warren Buffett-style financial statement report gave Pagaya a yellow-light verdict across balance sheet, earnings quality, and cash generation.

That is the right tone.

Pagaya is improving, but it is not yet the clean, predictable compounder that a strict Buffett-style investor would call obvious.

The positives are clear:

• Revenue has scaled.

• GAAP profitability has returned.

• Cash generation improved.

• Gross margins have recovered above 40%.

• The model is less asset-heavy than a traditional lender.

The concerns are also real:

• Cash flow has been volatile historically.

• Owner earnings only recently turned positive.

• Debt and retained-risk exposure matter.

• Dilution remains a factor.

• The balance sheet is not a true fortress yet.

• The model has not been fully proven through a severe consumer downturn.

This does not kill the thesis.

It defines the margin of safety problem.

Pagaya may be cheap. It is not simple.

The moat: process power first

Our 7 Powers report is useful here because it avoids overclaiming.

Pagaya’s strongest power today is process power.

The company has built an AI-driven process for evaluating credit, connecting partner flow with institutional capital, and scaling decisions through technology. If the model keeps improving with more data, that process can become harder to replicate.

Other powers are still developing:

This is a better view than simply saying Pagaya has a strong moat.

The moat is real enough to study.

It is not proven enough to take for granted.

The key issue is credit stress. A credit moat needs to survive bad vintages, not only strong quarters.

Partner concentration is still real

Pagaya often talks about its broad partner network.

That is fair. The company says it has more than 30 lending partners and has evaluated more than $3.7T in applications since inception.

The risk is concentration.

Our BE Invested Labs business model report flagged that the top 5 lending partners accounted for about 61% of total network volume in Q1 2025. It also flagged that the top 5 ABS funding partners accounted for 47.5% of total ABS funding.

Those numbers matter.

A company can have many partners and still depend heavily on a few.

This is one of the most important risks to watch. If the top few lending partners reduce volume, change strategy, renegotiate economics, or build more internally, Pagaya’s growth can slow quickly.

The same applies to funding partners.

If a few major ABS investors pull back, the funding side of the platform can become more expensive or less reliable.

The bull case needs more evidence of broader volume contribution over time.

Funding execution is improving

Pagaya has been active in ABS markets.

In Q1 2026, the company raised $2.1B of ABS funding across 4 transactions despite market volatility. It also received its first AAA Fitch rating on a $368M PAID resecuritization.

After Q1, Pagaya closed RPM 2026-3, a $600M upsized auto ABS transaction. The company said it was its largest auto ABS deal to date and its third fully pre-funded auto ABS transaction of 2026. Pagaya also said it had raised $1.5B of pre-funded auto ABS in 2026 and more than $38B across 90 ABS transactions since 2018.

That supports the bull case.

Funding demand is not theoretical. Pagaya continues to raise capital through its structures.

The risk is that funding quality can change before volume does.

ABS access should be watched through 3 lenses:

1. Can Pagaya keep issuing deals?

2. At what spread and structure?

3. How much risk does Pagaya need to retain to get deals done?

The first answer is currently positive. The second and third will decide how attractive the model remains.

Product expansion: Upgrade Flex Pay matters

Pagaya also announced an expansion of its partnership with Upgrade into Flex Pay, Upgrade’s BNPL product.

This is relevant because it supports the diversification thesis.

Pagaya has been trying to expand beyond personal loans into auto and point-of-sale financing. The Upgrade Flex Pay announcement gives another proof point that Pagaya’s partner relationships can move into new products.

That matters because one-product dependence would make the business more fragile.

The caution is that new verticals can carry different economics. Growth in a new asset class is useful only if underwriting quality, funding demand, and FRLPC economics hold.

For now, the Upgrade expansion is a positive signal. It does not settle the full thesis.

Management tone: confident, but careful

The Q1 management message was confident.

The company is pointing to 5 consecutive quarters of GAAP profitability, operating leverage, product expansion, and stronger funding execution.

That is the right message to give after years of investor skepticism.

Our earnings-call sentiment report still found caution underneath the confidence.

Management is talking about tighter risk controls, higher borrower income profiles, higher cost of capital, and tighter ABS pricing. Those comments matter because they show the company is adapting to a tougher credit environment rather than ignoring it.

That is good risk management.

It also confirms that the environment is not easy.

The CFO transition is worth watching too. Evangelos Perros is stepping down as CFO and Jon Dobres is taking over, while Perros remains as a strategic executive adviser through the end of 2026.

A CFO transition is not automatically a red flag.

But after a period of improving profitability, investors should watch whether reporting discipline, guidance quality, and capital allocation remain consistent.

The Upstart comparison

The hardest question for Pagaya is Upstart.

Upstart also had an AI credit story. It also promised better underwriting. It also looked powerful when capital was available and loan demand was strong.

Then the cycle changed.

Funding pressure, credit concerns, and macro uncertainty damaged the thesis.

Pagaya has differences that matter.

Its partner base appears broader. Its asset mix includes personal loans, auto, and point-of-sale financing. It has active securitization programs. It has reported 5 consecutive quarters of GAAP profitability. It has shown stronger cash conversion recently.

Those differences are meaningful.

The shared risk is still credit.

AI underwriting gets judged during stress. A model can look strong when borrowers are employed, capital markets are open, and losses are manageable. The real test comes when unemployment rises, funding spreads widen, and institutional buyers become more selective.

Pagaya has not yet proven the model through a full-scale consumer credit downturn as a mature public company.

That is the open question.

Valuation: cheap for a reason, still worth studying

At the latest price we checked, PGY traded around $15.42.

Our BE Invested Labs reports used a similar reference price and showed the following valuation picture:

Those are not demanding multiples if Pagaya keeps growing profitably.

Our neutral BE Invested Labs synthesis estimated fair value around $22.50 per share using a blend of DCF and peer-relative valuation. That was not a price target promise. It was a framework.

A bullish scenario can produce much higher values if the market starts treating Pagaya as a scalable credit network.

A bearish scenario can still take the stock lower if credit losses rise, ABS markets tighten, FRLPC compresses, or retained interests get marked down.

The valuation is attractive enough to study.

The risk is large enough to demand discipline.

The bull case

The bull case rests on 6 points.

1. The market may be using the wrong category

If Pagaya is treated as a distressed lender, the low multiple makes sense. If it earns trust as a profitable credit network, the multiple can expand.

2. GAAP profitability is now visible

Pagaya has reported 5 consecutive quarters of GAAP profitability. Management guided FY2026 GAAP net income of $110M to $160M.

That moves the discussion from promise to execution.

3. Operating leverage is showing

Revenue grew 10% in Q1 2026. Adjusted EBITDA grew 18%. Operating income grew 68%.

That is a strong profit conversion signal.

4. Cash conversion supports earnings quality

FY2025 operating cash flow was much higher than GAAP net income in our forensic report. That supports the idea that recent profitability is not only accounting presentation.

5. Funding execution remains active

Pagaya continues to issue ABS deals, including the $600M upsized auto ABS transaction after Q1.

Funding access is not theoretical.

6. Product expansion is real

Auto and POS are becoming more central to the story. The Upgrade Flex Pay expansion supports the idea that Pagaya can grow beyond one product line.

The bear case

The bear case also has 6 points.

1. Credit markets still drive the model

Pagaya depends on consumer credit performance and institutional demand for credit assets. A weaker credit cycle can pressure both sides.

2. FRLPC can compress

Q1 FRLPC margin fell to 4.6%. The decline was not dramatic, but the direction matters. If the company grows volume at weaker unit economics, the quality of growth changes.

3. Partner concentration is high

The top 5 lending partners represented about 61% of network volume in Q1 2025, according to our business-model report. That limits how far investors should push the diversification argument.

4. Funding concentration matters too

The top 5 ABS funding partners represented 47.5% of ABS funding in the same report. A broad funding network is useful, but large funding partners still matter.

5. AI underwriting is hard to verify externally

Investors can see outcomes. They cannot fully inspect the model. One weak vintage can damage trust quickly.

6. Regulation can change the economics

AI in lending faces scrutiny around fairness, bias, explainability, consumer protection, and data privacy. Regulation could increase compliance costs or limit how models are used.

What could break the moat

The moat breaks if the process stops producing better outcomes.

That can happen in several ways.

AI underwriting becomes easier to copy. Competitors catch up. Traditional lenders build more internally. Data access becomes restricted. Regulation limits model flexibility. Partner churn rises. ABS investors demand much higher yields. Credit performance deteriorates.

The most damaging version would be a credit downturn that exposes poor underwriting and weakens investor confidence at the same time.

That would attack both sides of Pagaya’s network.

Lenders would send less flow or demand better terms. Investors would demand more yield or reduce funding. Pagaya would face lower volume, lower FRLPC, and weaker retained-interest values.

That is the real bear case.

What we would watch next

For us, the next 4 quarters matter more than the story.

Here is the BE Invested Labs watchlist:

This is the work investors need to do.

Not deciding whether PGY is “the next Visa” or “the next Upstart.”

The work is tracking whether the numbers confirm the category the market is being asked to accept.

How we would frame Pagaya today

Pagaya is a profitable AI credit network with improving operating metrics, active ABS funding, and a valuation that may be too low if the business keeps executing.

It is also tied to consumer credit, capital markets, partner concentration, and retained-risk exposure.

That combination creates both the opportunity and the danger.

Deacon is right to challenge the lazy version of the market view. Pagaya is no longer just another SPAC-era fintech promising future profits. The company is producing GAAP net income, raising guidance, issuing ABS deals, and expanding product relationships.

The stricter view is that the final test has not arrived.

A tougher credit cycle will show whether Pagaya’s underwriting, funding network, partner relationships, and retained-risk structure deserve a higher multiple.

Until then, PGY sits in a difficult but useful category:

A business that may be better than its market reputation, but still needs to prove that its model works when conditions get worse.

Try the research process yourself

Pagaya is exactly the kind of company we built BE Invested Labs to analyse.

The story is attractive. The numbers are improving. The risks are real. The accounting is not simple. The valuation depends heavily on assumptions.

That is where structured research matters.

You can test the platform with PGY or another company you already follow:

Visit BE Invested Labs

Run the stock through the research flow. Compare the bull case and bear case. Check the valuation assumptions. Look at the financials. Ask whether the thesis still makes sense after the risks are visible.

That is the point of BE Invested Labs: Clearer research before stronger conviction.

Final view

Pagaya deserves serious attention.

The company is profitable, cash generative, and growing in areas that matter. It has active funding access, improving operating margins, and a business model that could be worth more than the market currently gives it credit for.

The risk side is just as important. Pagaya still depends on ABS markets, consumer credit health, partner flow, funding demand, retained-risk performance, and management discipline.

Our view is constructive, but strict.

PGY becomes more compelling if 4 things happen:

1. GAAP profitability continues.

2. FRLPC margin stabilises.

3. Partner and funding concentration improve.

4. ABS execution remains healthy through tougher conditions.

If those hold, the market may need to rethink Pagaya. If they break, the discount may be justified.

That is the thesis.

Pagaya is not a simple AI story. It is a test of whether AI underwriting, partner distribution, and institutional funding can create a durable credit network through a full cycle. That test is still open.

Reader question

What would you focus on first with Pagaya?

1. Valuation

2. Credit risk

3. ABS funding

4. Partner concentration

5. AI underwriting quality

6. FRLPC margin

Drop your answer in the comments. We may use the strongest questions for a follow-up breakdown.

Leave a comment

Sources used

Pagaya Q1 2026 earnings release and shareholder materials.

Pagaya 2025 annual report and risk disclosures.

Pagaya May 2026 auto ABS announcement.

Pagaya June 2026 Upgrade Flex Pay partnership announcement.

BE Invested Labs PGY research reports: business model, earnings-call sentiment, forensic accounting, moat analysis, Buffett-style financial statement review, and valuation synthesis.

Flashpoint by Deacon Brantley: “The Next Visa: A Deep Dive Into Pagaya.”

Disclaimers & Disclosures

AI Transparency: B.E Invested Labs uses AI to aggregate financial data and format reports. The AI structures the information. Users determine their own investment thesis, risk management, and market outlook. We audit the system outputs to reduce AI errors. We use these exact reports to manage our own capital.

Legal Disclaimer: This platform provides educational and informational content. B.E Invested Labs is not a registered investment advisor and does not provide financial, investment, tax, or legal advice.

Subscribe now

Share BE Invested Labs

ASML is one of those companies that sounds boring until you understand what it controls.

The company makes lithography machines. These machines print circuit patterns onto silicon wafers. Without lithography, there are no advanced chips. Without advanced chips, there is no serious AI infrastructure, no high-end GPUs, no advanced smartphones, no leading-edge data centers, and no continued push toward smaller, faster, more efficient semiconductors.

That is the simple thesis.

ASML sits in one of the most important parts of the semiconductor manufacturing chain. Its EUV machines are used to produce the most advanced chips in the world, including the chips needed for AI, high-performance computing, advanced memory, and next-generation logic nodes.

The BE Invested Labs ASML report gives the company an Outperform rating, with a probability-weighted fair value estimate of $2,150 per share. At the report price of $1,899.48, that implies around 13.2% upside over a 12 to 24 month horizon. The bull case reaches $2,550. The bear case falls to $1,450.

That range matters.

ASML is a very high-quality company. It is also expensive. The investment debate is not whether ASML is important. The debate is whether today’s valuation still leaves enough room for future returns.

If you want business breakdowns like this in your inbox, subscribe to BE Invested Labs. We break down companies through history, numbers, valuation, risk, and long-term business quality.

Subscribe now

From a leaky shed to the center of the chip industry

ASML started in 1984 as a joint venture between Philips and ASM International. It began life in a leaky shed near Philips’ buildings in Eindhoven, the Netherlands.

That detail matters because ASML was not born as an obvious winner. It was a small Dutch lithography company entering a market with bigger competitors, limited customers, and expensive research needs. ASML’s own history says the company had few customers in the late 1980s and could not stand on its own feet. ASMI withdrew, Philips was cutting costs, and the young company’s future was uncertain.

So before ASML became a European technology giant, it nearly failed.

The first real escape came in the 1990s with the PAS 5500 platform. That system helped ASML win important customers, improve productivity, and move toward profitability. In 1995, ASML became a public company listed in Amsterdam and New York. Philips sold half its stake at the IPO and later sold the rest.

The early 2000s then gave ASML another push. In 2001, the company introduced TWINSCAN, a dual-stage lithography system that could expose one wafer while measuring and aligning another. That improved productivity for customers. In 2003, ASML introduced its first immersion lithography machine. By 2007, it had shipped an immersion system with a numerical aperture of 1.35, the highest in the industry at the time.

This period matters for investors because ASML did not win by accident. The company kept solving specific customer problems: resolution, productivity, throughput, cost per chip, and manufacturing accuracy.

Those wins made ASML stronger before EUV became commercially real.

The EUV bet

EUV stands for extreme ultraviolet. ASML’s EUV systems use light with a wavelength of 13.5 nanometers, which is almost in the x-ray range. That shorter wavelength allows chipmakers to print extremely small and complex layers on advanced chips.

This technology took more than 2 decades to develop. ASML says it invested more than €6 billion in EUV R&D over 17 years. The company also acquired Cymer in 2013 to strengthen its EUV light source capability.

The technical problem was brutal. EUV light is absorbed by air, so the system must operate in a vacuum. The light source fires laser pulses at droplets of tin up to 50,000 times per second. The wafer stage positions wafers within a quarter of a nanometer and checks or adjusts 20,000 times per second.

That is why EUV became a graveyard for competitors.

Many companies explored EUV. ASML kept going. In 2006, it shipped the first EUV demo tools. In 2010, it shipped the first pre-production EUV system. In 2013, it shipped the first production EUV system. In 2016, customers began ordering the NXE:3400 in higher volumes. At the start of 2020, ASML celebrated the 100th EUV system shipment.

The timing is important. ASML did not suddenly become central to AI because of Nvidia headlines in 2023 or 2024. The company had already spent decades building the technology that AI now depends on.

There is a temptation to tell the EUV story as a simple genius bet that obviously paid off. That is too clean. EUV was expensive, slow, technically painful, and commercially uncertain for years. ASML needed deep customer support to keep pushing. In 2012, Intel, TSMC, and Samsung joined ASML’s customer co-investment program to help fund EUV and other next-generation work.

That was the important signal.

ASML’s biggest customers were willing to fund the company because their own futures depended on the technology.

Why EUV changed the economics

A normal equipment supplier sells machines and waits for the next order cycle.

ASML’s EUV business has different economics because customers build multi-billion-dollar fabs around its tools. Once a leading-edge chipmaker commits to an ASML-based process, the relationship can last for decades through service, software, upgrades, field options, and future tool generations.

The machine sale is only the beginning.

ASML calls its service and upgrade revenue Installed Base Management. In Q1 2026, Installed Base Management sales were €2.488 billion, up from €2.134 billion in Q4 2025. That is a large business inside the business. It matters because every machine shipped can create future service and upgrade revenue.

This is part of why ASML deserves a premium valuation.

The company sells some of the most complex machines in the world to a small group of customers that cannot easily replace it. TSMC, Samsung, Intel, SK Hynix, Micron, and other major chipmakers do not buy ASML systems casually. They buy them because their roadmaps require them.

The customer list is concentrated, which creates risk. The same concentration also shows how hard the market is to enter. Only a small number of companies can afford leading-edge fabs. Those companies need the best lithography tools available.

At the leading edge, ASML is the supplier.

High-NA EUV and the next technical record

The next phase is High-NA EUV.

ASML’s current NXE EUV systems use a numerical aperture of 0.33. The newer EXE systems use High-NA technology with a numerical aperture of 0.55. ASML says the EXE platform prints with 8 nm resolution and is designed to support sub-2 nm logic and leading-edge DRAM nodes.

The first High-NA EUV system was delivered in December 2023. ASML expects the platform to support process development and high-volume manufacturing in 2025 to 2026.

This matters because High-NA can increase ASML’s revenue per system and extend its leadership into future chip nodes. The BE Invested Labs report models High-NA as one of the main drivers of future growth and margin expansion. The report’s bull case assumes faster High-NA adoption, strong AI-driven demand, and operating margin expansion toward 45% by 2030.

The technical risk is still there. High-NA systems are expensive. Customers need to justify the economics. The semiconductor industry may use advanced packaging and chiplet designs to improve performance in ways that reduce pressure to shrink everything at the same pace.

Still, the commercial incentive remains clear. AI workloads require more compute. More compute requires more advanced chips. Advanced chips require advanced manufacturing. ASML’s tools sit inside that chain.

The Business Model: Two Revenue Streams, One Relationship

ASML’s revenue comes from two places.

Net System Sales is the headline number — selling new lithography machines to chipmakers. These are not simple transactions. A single fab build can involve multi-year planning cycles, technical integration, and machine orders that commit customers years in advance. ASML regularly receives large down payments before machines are even delivered, which is why the company can occasionally show strong cash generation even in quarters where formal revenue recognition is lower.

Installed Base Management is the second stream, and it is the one that gives ASML a quality that pure equipment suppliers rarely have. This covers service contracts, hardware upgrades, and software updates for the thousands of ASML systems already in operation. It is recurring, high-margin, and grows automatically as more machines get deployed. It also means that once a customer buys an ASML system, the relationship does not end at the purchase. It compounds for decades.

This is a meaningful distinction from a typical capital equipment company. ASML is not just selling machines. It is building a permanent, recurring revenue stream on top of every machine it has ever shipped.

The Moat: Why Nobody Can Catch Them

ASML did not become a monopoly by accident. It took decades of accumulated advantage to reach a position where no competitor can credibly enter their core market.

Scale is a barrier nobody can replicate. The R&D required to develop EUV technology cost billions over multiple decades. ASML is the only company that spread that cost across the entire global market for advanced chips. A new entrant would need to spend similarly — and then win customers away from a supplier with 30+ years of integration into their fabs. That is not a business decision any competitor has been willing to make. Nikon and Canon tried, and fell behind, not because they lacked ambition, but because the physics and the capital requirements were simply too hard.

Switching costs are among the highest of any industry. Swapping out a lithography supplier is not a software upgrade. It means redesigning fab processes, retraining engineers, renegotiating supply chains, and accepting years of yield risk during the transition. Customers like TSMC and Samsung have built their most advanced production lines entirely around ASML’s technology. They do not swap suppliers. They deepen the relationship.

The supply chain is cornered. ASML sources light sources from a single German supplier, Trumpf. Their optics come primarily from Zeiss, where ASML owns a 24.9% stake. They have effectively secured exclusive or near-exclusive access to the components needed to build these machines. A competitor could not simply set up a production line — the raw inputs are controlled.

The installed base creates compounding lock-in. Every system shipped adds to a recurring revenue base. Every upgrade cycle reinforces integration. Every new fab build starts from the assumption that ASML machines will be inside it.

The BE Invested report rates ASML’s moat as Wide, with a trend of Strengthening. The introduction of High-NA EUV extends the technological frontier further out of reach for any would-be competitor. The next generation is more complex, more expensive, and more deeply integrated into customer roadmaps. The moat does not just hold — it widens with every product cycle.

The numbers before and after the inflection

ASML was growing consistently even before the AI era reshaped demand.

Revenue moved from €18.6 billion in 2021 to €35.3 billion in 2025, a 5-year CAGR of 18.8% in USD terms. The one soft year was 2024, where revenue dipped slightly as customers worked through inventory. Then 2025 came in at $37.67 billion, up 24.1%.

Gross margin has held above 50% through supply chain disruptions, inflation, and two years of softer chip demand. It currently sits at 52.6% on a trailing twelve-month basis. Operating margin is 36.0%, up from 30.7% three years ago. Net income for the trailing twelve months is $10.01 billion.

The peer group — Applied Materials, Lam Research, KLA Corporation, and Tokyo Electron — averages TTM revenue growth of 5.8%. ASML is growing at 13.2% TTM.

The balance sheet is genuinely strong. ASML holds $8.38 billion in cash against $2.71 billion in debt, making it a net cash position of $5.67 billion. Free cash flow is $8.97 billion. The company returned approximately €8.5 billion to shareholders in 2025, through a combination of dividends and buybacks. It could repay its entire debt load from a single quarter’s net income.

ROIC sits at 26.1% against a WACC of roughly 8-11%. That 1,500 to 1,800 basis point spread is the quantitative proof that the moat is real.

The bull case: AI has changed the demand structure

The specific argument here is not just that AI demand is growing. It is that AI requires leading-edge chips at a volume and complexity the world has never produced before, and every one of those chips can only be manufactured using EUV.

Hyperscalers — Microsoft, Google, Amazon, Meta — have made capital expenditure commitments measured in hundreds of billions of dollars per year. That spending goes into data centres, training clusters, and inference infrastructure. All of it runs on advanced silicon. All of that silicon requires EUV.

The semiconductor sovereignty dynamic adds another layer. The US CHIPS Act, Europe’s Chips Act, Japan’s subsidies for TSMC’s domestic fab, and Intel’s domestic expansion are all government-funded programmes to build new advanced manufacturing capacity. Every new fab being built in the western world is a new ASML customer.

High-NA EUV is just starting its adoption cycle. The first EXE systems are shipping now. Every major chipmaker will need to transition to this platform to stay competitive at the leading edge. That cycle will run for years and generate revenue at $350-380 million per machine.

CEO Christophe Fouquet, on recent earnings calls, has described demand as “unprecedented” and confirmed the company is sold out for 2026. The company’s problem, he says, is not finding demand. It is building enough machines to meet it.

Management’s 2030 guidance — €44 to €60 billion in revenue — implies a continuation of the growth rate they’ve been delivering for a decade.

The bear case: premium multiples and real risks

ASML trades at 63.7x trailing earnings and 57.2x EV/EBITDA. Peers trade at 31.5x and 22.1x respectively. ASML commands roughly a 100% premium to peers on both metrics.

That premium is defensible, as explained above. A monopoly on the only technology that can produce leading-edge chips justifies pricing well above commodity equipment suppliers. But the premium also means any negative surprise hits hard. There is no valuation cushion here.

China is the live risk. China currently accounts for roughly 26% of ASML’s revenue, predominantly from older DUV machines. ASML has been barred from shipping EUV to China since 2019 due to Dutch and US export controls. If those controls expand further to include DUV systems, the revenue impact would be immediate and significant. Management acknowledges this as the most actively monitored risk in the business.

Geopolitics in Taiwan is the tail risk. TSMC is ASML’s largest customer. TSMC’s primary manufacturing is in Taiwan. A military or political crisis affecting Taiwan would cascade directly through ASML’s order book. This is a low-probability, high-severity scenario that the market intermittently reprices. It has not gone away.

The terminal risk is the end of Moore’s Law economics. EUV machines cost over $200 million. High-NA systems are approaching $380 million. At some point, the cost of transitioning to a new process node may no longer be justified by the performance gain. If chipmakers shift to alternative approaches like advanced packaging and chiplets to improve performance rather than shrinking transistors further, demand for the most advanced EUV systems could flatten. This is a long-duration risk, not an imminent one, but long-term holders should keep it in mind.

Customer concentration is real. Over 80% of revenue comes from 3 customers. A serious strategic or financial problem at TSMC, Samsung, or Intel would show up immediately in ASML’s order book.

The bear case price target from our engine is $1,450, representing 23.7% downside from current levels. It requires a confluence of negative events — recession, export restriction escalation, and multiple compression — to materialise simultaneously. Possible. Not probable.

Forensic: what the accounting tells us

Our engine’s forensic review found no signs of earnings manipulation. Cash conversion is strong: operating cash flow of $10.53 billion against net income of $10.01 billion gives an OCF/Net Income ratio of 1.05x, confirming that reported profits are translating directly into cash.

One yellow flag worth watching: ASML capitalises a portion of development costs under IFRS. In the first half of 2025, €473.9 million was capitalised. This is a compliant accounting practice, but it boosts current-period net income by removing certain R&D costs from the income statement. If the rate of capitalisation accelerates sharply relative to total R&D spend, it would be worth investigating.

Working capital can swing significantly due to large customer down payments and the timing of machine deliveries. In 2023, net profit grew 27% while operating cash flow fell 31% — a divergence caused by working capital movements, not a business problem. That has since normalised.

No significant off-balance sheet liabilities. Inventory days are elevated at approximately 218-319 days, reflecting the complex production cycle, but this is structural. ASML often collects down payments before delivery, giving it a cash position that runs ahead of its production commitments.

What the BE Invested Labs engine concluded

Our engine initiated coverage of ASML with an Outperform rating and a probability-weighted fair value of $2,150 per share.

Current price at time of report: $1,899.48. Implied upside: 13.2%.

The 3 scenarios:

The DCF model assumes 15% revenue CAGR through 2030, a terminal growth rate of 3.5%, and a WACC of 8-11%. Management’s own guidance for 2030 is €44 to €60 billion in revenue, with gross margins of 56-60%. Those targets look achievable given the trajectory.

The bull case for margin expansion is grounded in product mix. High-NA EUV systems, at $350-380 million per unit, carry gross margins expected north of 60%. As those units grow as a share of revenue, they mechanically pull the corporate average up from 52.6% toward the 56-60% range management is guiding. Installed Base Management also carries margins above the corporate average, and that segment is growing faster than total revenue.

Management and governance

Christophe Fouquet became ASML CEO in April 2024 after more than 15 years inside the company. He previously led key technology and business areas, including the EUV business. That gives ASML leadership continuity after the Peter Wennink era.

This matters because ASML is not a normal sales organization. The CEO needs to understand physics, customers, geopolitics, supply chains, government restrictions, long-term R&D, and capital allocation.

The BE Invested Labs report gives Fouquet an A- management assessment. The main reason is his technical background and long operating history within ASML. His shorter time as CEO means investors should keep watching execution, but the transition looks orderly.

Governance looks clean overall. ASML has a Dutch two-tier board structure, with a Board of Management and an independent Supervisory Board. The report found no major governance red flags such as dual-class shares, poison pills, or problematic related-party transactions.

Capital allocation also looks sensible. ASML invests heavily in R&D, maintains a strong balance sheet, pays dividends, and repurchases shares. In 2025, it returned €8.5 billion to shareholders.

For a company with this much technical risk and geopolitical pressure, that financial strength matters.

Our view

We hold ASML in our own portfolios. ASML is one of the few companies in the world where, if you understand the business, the investment case is not complicated. It is the only supplier of the technology that makes advanced chips possible. The demand for those chips is growing because AI is growing. AI is not going away.

Our BE Invested Labs engine rates it Outperform, with a probability-weighted target of $2,150.

Current price is $1,899. That is 13.2% upside to fair value, with a bull case of $2,550 if AI adoption accelerates faster than expected. The bear case of $1,450 requires specific negative events to coincide.

The entry zones from our report are around the $1,750 to $1,900 range for a measured position, with more aggressive adding below $1,600. We see the current price as a fair entry for a long-duration compounder, with particular strength if there is any macro-driven pullback.

One governance note worth holding: total insider ownership at ASML is estimated at roughly 0.02% of the company. There is no meaningful open-market buying by the executive team. The absence of a large personal stake does not impair our thesis — management’s actions have created immense shareholder value — but it is a data point we note for anyone who values owner-operator alignment as a signal.

The metrics to track quarterly:

• EUV system bookings: above €7B per quarter is a thesis upgrade signal; below €4B for 2 consecutive quarters is a concern

• Gross margin: sustained above 54% is a positive signal; sustained below 48% raises questions about mix and pricing

• Inventory days: currently 218-319 days; watch for a sustained trend toward 300+ days

• China revenue percentage: currently approximately 26%; any abrupt policy-driven drop signals immediate downside risk

• High-NA order announcements: firm orders from all 3 key customers (TSMC, Samsung, Intel) would be a meaningful thesis upgrade

• ROIC: currently 26.1%; a sustained drop below 25% would flag deterioration

The company that started in a shed in Veldhoven, that nearly went bankrupt pursuing a technology most competitors abandoned, that had to be rescued by its own customers in 2012, is now the single most important supplier in the global semiconductor supply chain.

That is not a narrative. That is forty years of compounding decisions that turned out to be right.

Try the BE Invested Labs platform here: beinvestedlabs.com

If research like this is useful, subscribe to BE Invested Labs.

Subscribe now

— Buyce & Emmanuel, BE Invested Labs

Disclaimers & Disclosures

AI Transparency: B.E Invested Labs uses AI to aggregate financial data and format reports. The AI structures the information. Users determine their own investment thesis, risk management, and market outlook. We audit the system outputs to reduce AI errors. We use these exact reports to manage our own capital.

Legal Disclaimer: This platform provides educational and informational content. B.E Invested Labs is not a registered investment advisor and does not provide financial, investment, tax, or legal advice.

References and data notes

1. ASML Q1 2026 results were cross-checked against ASML’s official press release: €8.8 billion net sales, 53.0% gross margin, €2.8 billion net income, Q2 2026 sales guidance of €8.4 billion to €9.0 billion, and full-year 2026 sales guidance of €36 billion to €40 billion.

2. ASML 2025 full-year figures were cross-checked against the official ASML 2025 Annual Report page: €32.7 billion net sales, 52.8% gross margin, €9.6 billion net income, €24.73 basic EPS, 535 total systems sold, including 48 EUV systems.

3. ASML EUV and High-NA technology descriptions were cross-checked against ASML’s official EUV product page.

4. ASML 2030 revenue and gross margin targets were cross-checked against ASML Investor Day materials and Reuters reporting on ASML’s 2024 Investor Day.

Subscribe now

A single stock idea can take over your evening.

You start with one company. Maybe it is ASML, Nvidia, Brookfield, Tesla, or a smaller stock someone mentioned online. You want to understand the business properly before putting money behind it.

Then the tabs start multiplying: annual report, investor presentation, earnings call, analyst article, YouTube breakdown, news update, valuation spreadsheet, macro view, one bullish thread, one bearish thread, and one more video because now you are unsure again.

By the end, you have more information than when you started. The decision still feels unclear.

That is the problem we kept running into.

We are Buyce Ampomah and Emmanuel Yeboah, two engineers, investors, and lifelong learners. BE Invested Labs started from our own frustration with stock research. Financial information is everywhere, but clear research is still hard to build from scattered sources.

That gap is where BE Invested Labs began.

Buyce grew up in Italy, studied in England, spent a few more years there, and later moved to the Netherlands to work for ASML. That decision shaped a lot of his professional life. Through ASML projects, he spent years working in Taiwan, Korea, and Ireland.

Emmanuel grew up in Ghana, moved to Belgium, and later came to the Netherlands for university. He interned at Philips, worked on major airport systems at Vanderlande, moved into medical technology at Philips, and later joined ASML.

We both studied electrical and electronic engineering. We both work in the semiconductor industry. We both care about systems, microcontrollers, process, and the small details that make complex technology work.

That background affects how we invest.

Engineering teaches you to check inputs, test assumptions, and respect process. Investing asks for the same discipline. A business can sound exciting while the balance sheet tells a different story. A stock can look cheap while the market is pricing in a serious risk. A company can grow fast while the valuation leaves little room for mistakes.

The hard part is connecting the pieces.

The first version was rough

Our first solution was private and basic.

Spreadsheets. Research notes. Prompts. Valuation templates. Checklists.

We wanted a repeatable way to answer the same questions before buying a stock:

How does the business make money?
How strong are the fundamentals?
What is the industry doing?
What does valuation suggest?
What are the main risks?
What would make us change our mind?

Buyce then tested a more serious version: 18 AI agents, each analysing a company from a different angle. Valuation. Fundamentals. Sentiment. Macro. Risk.

The output was useful. The process was still clunky.

Running 18 separate prompts to research one company created another version of the same problem. Too many scattered outputs. Too much manual work. Too much time lost before reaching a clear view.

So we kept building.

The first actual version of BE Invested Labs was a plain white website with input fields, report options, and a chat box. It looked like a fancy Excel sheet.

Still, it proved something important.

A single starting point could produce a structured research flow.

You could enter a company name, ticker, industry, and a few details. From there, the system could generate research sections, valuation work, charts, and reports from one place.

That was when the idea became clearer.

BE Invested Labs could become the research platform we wanted as investors.

Share

What BE Invested Labs does

BE Invested Labs is a dedicated research platform built to help investors understand a specific stock or industry with more clarity and conviction.

You can try the platform here: BE Invested Labs

It brings together valuation tools, interactive charts, automated reports, financial concepts, and audio briefings in one interface.

The goal is simple: reduce tab fatigue and make research easier to act on.

If you are researching a stock, BE Invested Labs can help you organise the work. It gives you a structured written report, compares key metrics, explains valuation assumptions, and makes the business easier to understand.

If you are studying an industry, it can help you review the main drivers, risks, competitors, and market context.

If you are learning a finance concept, the Academy section can explain it in plain language.

ROIC. Earnings yield. Free cash flow yield. DCF. Index weighting. Free float. Yield curve inversion.

The concept can become a written lesson, a Word-style report, or a podcast-style script you can listen to while walking, commuting, training, or moving between work and family life.

BE Invested

That matters because people learn differently.

Some people want the formulas. Some people need examples. Some people prefer to listen first, then read later. Some people need the simple explanation before the deeper version makes sense.

We are building for those people.

Professionals. Students. Parents. Founders. Busy workers. Curious investors.

People who want to understand what they own before putting money behind it.

Subscribe now

Why we are sharing the build

BE Invested Labs does not make decisions for you.

It gives you a cleaner research process so you can make better decisions yourself.

That distinction matters to us. We want investors to see the numbers, understand the assumptions, compare the risks, and build their own view.

This Substack will follow that same mission.

We will share company breakdowns, product updates, investing lessons, valuation examples, founder notes, and the thinking behind BE Invested Labs as we build it in public.

We are still early. The product is improving. The Academy is growing. The reports are getting better. The audio briefings are being tested. Every user, mistake, and piece of feedback is teaching us something.

If you care about investing, business, technology, and clearer research, you are welcome here.

Test the platform with one company you’re researching this week. We’d genuinely like to know what helps, what feels missing, and what should be improved: https://beinvestedlabs.com/

Subscribe now

Share

A question for you

What slows you down the most when researching a stock?

1. Finding reliable information

2. Understanding the numbers

3. Valuing the business

4. Deciding when to buy

5. Staying objective after you already like the company

Drop your answer in the comments. We are building BE Invested Labs around problems investors actually face.

Leave a comment

Share B.E Invested Labs

Welcome to BE Invested Labs.

My father, like most men of his generation, spent his leisure time reading physical newspapers. One afternoon in the early 2000s in Ghana, he called me over and told me about a man he had just read about. A man whose money worked for him even while he slept.

That man was Warren Buffett.

My father had been reading about Buffett’s investment in The Coca-Cola Company. To him, the story was astonishing. The idea that someone could own part of a business and continue earning from it for decades felt almost unreal.

For context, this was Ghana, West Africa, at a time when investing culture looked nothing like what people experience today through apps, brokerages, and financial content online. Back then, ownership in my environment was almost entirely associated with land. Families bought plots to preserve wealth, secure inheritance, and protect themselves from future uncertainty. Land was tangible. You could walk on it, fence it, build on it, and pass it to your children.

But land ownership also carried fear.

I grew up hearing brutal stories about land disputes and litigation. Families fought over boundaries, inheritance claims, and ownership rights. Some people lost their savings. Others lost relationships. In extreme cases, people lost their lives over contested property.

So when my father described a world where ownership could exist through businesses rather than physical land, something shifted in me. For the first time, I understood that wealth was not built only through labor. It could also be built through ownership.

That idea never left me.

The Problem: A System Built to Overwhelm You

At the time, I was a teenager with no access to stock markets, no investing mentors, and no roadmap. Most people around me were understandably focused on surviving immediate economic realities first. Investing in global businesses was not a common conversation.

Yet the fascination stayed. When my mother moved with my younger brother and me to Belgium, I carried it with me. At nineteen, I wrote in my diary that one day I wanted to own businesses, invest in companies, and build enterprises. It sounds ambitious, perhaps even naive. But I could not let it go.

Ironically, despite that fascination, I never pursued finance academically. Engineering called me instead. Mathematics, physics, systems, and computation stimulated my mind in a way few things could. Engineering taught me to think structurally: inputs and outputs, constraints and optimization, feedback loops and stability, failure modes and resilience.

Years later, I realized investing was not entirely different. A company is also a system. Management allocates resources the way engineers allocate energy or computational capacity. Poor leadership introduces instability. Strong organizational architecture creates resilience and long-term compounding.

Once I started working and earning a salary, I finally began acting on the ideas that had lived in my mind for years. And I made mistakes costly ones.

I bought stocks without understanding the underlying business or management incentives. I invested emotionally, buying when markets were euphoric and panicking when volatility arrived. I bought a home before building proper financial reserves. Like many beginners, I confused activity with strategy.

The Agitation: The Modern Investor Is Being Set Up to Fail

My early mistakes were honest ones. But here is what I eventually realized: the system itself is working against retail investors.

Today’s investing environment is noisier, more distracting, and more psychologically manipulative than at any point in history. Meme stocks. Day trading apps engineered around dopamine. Gamified interfaces that nudge users toward impulsive trades. Financial Twitter full of conflicting opinions. Reddit threads masquerading as research.

On one side, there are the institutional tools Bloomberg, Capital IQ , yahoo finance overwhelmingly complex, text-heavy, and priced at thousands of dollars per year. They were never built for ordinary investors.

On the other side, there is the noise clickbait headlines, speculative hype, and the relentless pressure of FOMO. That fear of missing out that pushes people into positions they do not understand, at prices they cannot justify, for reasons they cannot articulate.

The result? A generation of investors who genuinely want to build long-term wealth but find themselves paralyzed between complexity and chaos. I know this feeling intimately. I remember having twenty browser tabs open simultaneously, cross-referencing SEC filings, analyst notes, earnings transcripts, and macroeconomic reports and still feeling uncertain.

That is not research this causes exhaustion.

Disciplined, long-term investing understanding competitive moats, risk management, valuation discipline, and compounding requires exactly the kind of deep, structured thinking that the modern information environment actively punishes. Patience is boring. Rigor is slow. And in a world optimized for speed and stimulation, boring and slow rarely win attention.

The Solution: BE Invested Labs

Out of those circumstances the frustration, the tab overload, the noise my co-founder and I built BE Invested Labs.

The premise is straightforward: investing should be deeply rigorous and intuitively accessible at the same time. It should remove the cognitive biases, the FOMO, and the exhausting burden of fragmented research and replace them with clarity, confidence, and conviction.

Here is what that looks like in practice.

BE Invested Labs gives you instant access to 18 specialized AI analysts Forensic Accounting, Economic Moat Analysis, Valuation Discipline, and more operating 24/7 on any stock you want to research. Rather than raw data dumps or conflicting opinions, you receive structured, institutional-grade output in seconds:

• Scorecards that grade each opportunity on a rigorous 9-point scale, mathematically flagging value traps before you put a single euro at risk. The system acts as an automated Red Team desk cross-referencing eight fundamental parameters to separate high-conviction setups from dangerous speculation.

• Audio-first research podcasts featuring two AI analysts one Bull, one Bear debating the merits of up to five stocks simultaneously, designed to fit into your morning commute rather than demand another evening at your desk.

• Institutional-grade synthesis reports that translate millions of data points into decisions you can actually act on — without the 150-page filings.

• A free integrated Academy that grows alongside your portfolio, so the platform is not just a utility but a curriculum. The more you use it, the more your investing knowledge compounds alongside your capital.

You can explore the platform at beinvestedlabs.com.

What Ownership Really Means

Looking back now, I realise that newspaper article my father shared with me did far more than introduce me to investing. It changed how I saw ownership. And I believe ownership changes how people see the world.

Ownership changes your relationship with time it teaches patience over consumption. It changes your relationship with work you stop thinking only about income and begin thinking about the systems and competitive advantages that produce value beyond your direct labor. It changes your relationship with society you begin observing how businesses function, how incentives shape behaviour, and how capital flows through economies.

Most importantly, ownership expands imagination.

For a young boy growing up in Ghana, the realisation that he could someday participate in the global economic system not merely as a worker, but as an owner was transformative. That dream still lives in me today.

And perhaps that is the real reason my co-founder and I built BE Invested Labs. Because somewhere between my upbringing, engineering career, and investing in global markets, I discovered a simple but life-changing idea:

Wealth is not only earned. It can also be designed, owned, and compounded.

We built the platform to make sure that idea is no longer reserved for the few.

Try BE Invested Labs at:

Visit BE Invested labs

Disclaimers & Disclosures

Transparency on AI utilisation: We utilises advanced artificial intelligence and synthetic strictly as data-aggregation and formatting force multipliers. AI does not dictate our thesis or that of the users of BE Invested platforms, manage our risk, or formulate our macroeconomic outlook. Every piece of intelligence is rigorously audited, curated, and synthesised by our founder to minimise hallucinations. We have skin in the game; this is the exact intelligence we use to position our own capital.

This is Educational and informational purposes only. Not financial, investment, tax, or legal advice. BE Invested Labs is not a registered investment advisor.

Subscribe now

Growing up in age of Nintendo DS2, and Nikon digital camera, having a SanDisk memory stick is a must for using these devices. However, we are looking at SanDisk from the lens of investors.

A few days ago, we launched BE Invested Labs with a simple mission: help everyday investors move from scattered market information to structured research.

Not more noise.
Not another dashboard.
Not another place where you open ten tabs, read five opinions, check two YouTube videos, and somehow end up less clear than when you started.

The goal is different.

The goal is to take a stock, structure the research, pressure-test the thesis, show both the bull case and the bear case, and help the investor understand what is really driving the opportunity.

Today, we are using Sandisk Corporation, ticker SNDK, as a live example. And it is a perfect case study.

Because Sandisk is exactly the kind of stock that can confuse investors. On one side, the story is powerful. AI data centers need more storage. Enterprise SSD demand is surging. NAND flash pricing has tightened. Revenue has exploded. Margins have expanded. The balance sheet has improved. The stock has gone almost vertical.

On the other side, memory is one of the most cyclical areas in semiconductors. These businesses can look unbeatable at the top of the cycle and broken at the bottom. Profitability can swing violently. Pricing can reverse. Valuation multiples can compress. Investors who buy the narrative late can get punished, even if the underlying company is genuinely strong.

That is why Sandisk is not a simple “AI winner” story.

It is more interesting than that.

It is a battle between AI momentum and cyclical gravity.

And this is exactly where structured research matters.

Why Sandisk Is Suddenly Back In The Market Conversation

Sandisk became a standalone public company again after its separation from Western Digital in February 2025. The company is a pure-play flash memory and storage business, focused on NAND flash memory products, including enterprise SSDs, consumer storage, and edge devices. Since the spin-off, Sandisk has moved from being a forgotten component inside a larger storage company to becoming one of the most talked-about AI infrastructure stocks in the market. Reuters and other outlets have reported that Sandisk’s shares have surged dramatically since the spin-off, supported by AI-driven demand for high-performance data storage. (Reuters).

The reason is simple: AI does not only need GPUs.

AI needs memory.
AI needs storage.
AI needs data to be stored, moved, retrieved, trained on, and processed at scale.

That is where NAND flash becomes important. NAND is non-volatile memory, meaning it stores data even when power is removed. It is the technology behind many SSDs and flash storage products. In the AI data centre world, high-performance SSDs are increasingly important because modern AI workloads need fast access to massive amounts of data.

Sandisk’s recent numbers show how quickly the market has changed. In its fiscal third quarter, the company reported revenue of $5.95 billion, far ahead of analyst expectations, with adjusted earnings of $23.41 per share. Reuters reported that the company guided for the following quarter to revenue between $7.75 billion and $8.25 billion, also ahead of consensus expectations. (Reuters)

That is not normal growth. That is a step-change. But the BE Invested Labs report did something important: it did not stop at the growth story.

It asked a harder question:

How much of this good news is already priced in?

What The BE Invested Engine Concluded

The BE Invested Labs institutional research engine initiated coverage on Sandisk with a Neutral rating. That may surprise people who only look at the revenue growth, the AI narrative, or the share price momentum.

But this is exactly why the report is useful. A shallow AI summary could easily say:

“Sandisk is benefiting from AI demand. Revenue is growing. The stock is strong.”

That is not research. That is a headline!

Our research engine went deeper. It concluded that Sandisk is executing extremely well, but that the stock’s valuation already reflects a near-perfect version of the future. The report estimated fair value around $1,550 per share, compared with a current price used in the report of $1,831.50, implying roughly 15.4% downside to the fair value estimate. The report also modelled a bull case around $2,200, a base case around $1,500, and a bear case around $900, showing a wide distribution of possible outcomes.

That wide range matters. It tells us the stock is not easy.

Sandisk could continue winning if AI demand remains strong, NAND pricing stays tight, customers sign long-term contracts, and margins remain elevated. But if the market begins to believe this is a cyclical peak rather than a permanent new level of profitability, the downside could be painful.

That is the point. A good research platform should not just confirm excitement.

It should structure uncertainty.

The Bull Case: AI Has Changed The Storage Market

The bull case for Sandisk is not weak. In fact, it is very compelling.

The core argument is that AI has created a structural change in storage demand. Large language models, AI agents, enterprise automation, cloud workloads, and hyperscale data centres require enormous amounts of high-speed storage. That means demand for enterprise SSDs and NAND flash products may remain stronger for longer than historical cycles would suggest.

Reuters reported that Sandisk has signed five long-term customer agreements, with three of those agreements worth at least $42 billion. These agreements are designed to help reduce the extreme boom-and-bust dynamics that have historically defined the memory business. CEO David Goeckeler told Reuters that the company wants more consistent and predictable economics, with contract structures that include financial commitments from customers. (Reuters)

That is important because memory companies have historically suffered from weak visibility. Prices rise, companies add capacity, supply eventually catches up, prices fall, margins collapse, and the cycle repeats. If Sandisk can genuinely move more of its revenue into long-term agreements with financial commitments, that could change the quality of the business.

The bull case rests on four pillars.

First, AI infrastructure demand remains strong. Cloud providers, hyperscalers, and enterprise customers continue spending heavily on data centres, and storage becomes a critical bottleneck.

Second, NAND supply remains disciplined. If competitors such as Samsung, SK Hynix, Micron, and Kioxia do not flood the market with new capacity, pricing can stay stronger for longer.

Third, Sandisk shifts more of its mix toward enterprise and data centre SSDs. This matters because enterprise products typically carry better margins than lower-end consumer storage.

Fourth, Sandisk’s balance sheet and capital return strategy improve shareholder outcomes. The company has announced a $6 billion share buyback, which could support earnings per share if executed at reasonable valuations. (Reuters)

In this version of the future, Sandisk is not simply a cyclical memory company. It becomes a premium AI infrastructure supplier with better pricing visibility, higher margins, and a more durable earnings base.

The Bear Case: Memory Cycles Have Not Disappeared

But here is where investors need to be careful.

The bear case is not that Sandisk is a bad company.

The bear case is that Sandisk is a very good cyclical company being valued like a structurally superior compounder.

There is a big difference.

The memory industry has a long history of booms and busts. When demand rises and pricing improves, margins expand rapidly. But when supply catches up, pricing can fall brutally. In capital-intensive industries, today’s shortage often becomes tomorrow’s oversupply because every competitor sees the same high prices and has an incentive to add capacity.

That is the old gravity of the memory business.

The BE Invested Labs report captured this tension well. It described Sandisk as having powerful AI-driven near-term momentum, but also highlighted that the business remains tied to NAND pricing, industry supply discipline, and capital intensity. The report flagged the risk that current profitability may represent peak-cycle conditions rather than a new permanent baseline.

This is why the valuation matters.

At the time of the BE Invested report, Sandisk was trading at valuation levels that looked stretched compared with traditional semiconductor memory peers. The report cited a trailing P/E above 62x and EV/EBITDA around 44.5x, while also noting that forward multiples depend heavily on whether analysts are right about sustained earnings power.

That creates a problem. If Sandisk keeps beating expectations, the stock can keep working. But if investors begin to question the durability of the cycle, the multiple can compress quickly.

This is one of the most dangerous situations in investing:

a real business, with real growth, at a price that already assumes too much goes right.

The Financial Turnaround Is Real

One of the strongest parts of the Sandisk story is that the financial turnaround is not imaginary.

This is not just a stock moving on vibes.

Revenue has genuinely accelerated. Profitability has improved. AI-related demand is showing up in reported numbers. The company’s fiscal third quarter revenue of $5.95 billion represented a major beat versus expectations, while adjusted earnings of $23.41 per share also beat analyst estimates by a wide margin. For the following quarter, Sandisk guided revenue between $7.75 billion and $8.25 billion, again above consensus. (Reuters)

The BE Invested report also highlighted the longer-term revenue recovery. It showed revenue falling sharply during the previous NAND downturn, before recovering through fiscal 2024 and fiscal 2025, and then accelerating strongly on a trailing twelve-month basis. This is the V-shaped recovery visible in the report’s financial section.

But this is also where investors must think carefully.

In a cyclical industry, the income statement can look terrible at the bottom and unbelievable at the top. The same company can swing from losses to record profits in a short period because pricing, utilization, and operating leverage all move together.

That is why the question is not only:

“Are the numbers good?”

The question is:

“Are these numbers sustainable?”

That is a more difficult question. And it is the question the market as a both weighing and voting machine is now trying to answer.

The Balance Sheet Gives Sandisk Flexibility

One area where the report was clearly positive is the balance sheet.

The BE Invested analysis described Sandisk’s balance sheet as a “fortress,” noting a significant cash position, minimal debt, and a net cash position after debt reduction. The report also highlighted management’s capital allocation decisions, including the planned buyback.

That matters because cyclical businesses need balance sheet strength.

When a cycle turns, weak balance sheets become dangerous. Companies with too much debt can be forced to cut investment, issue shares, sell assets, or refinance under pressure. Companies with strong balance sheets have more options. They can continue investing through downturns, repurchase shares when valuations are attractive, and avoid being forced into bad decisions.

Sandisk’s current balance sheet strength is therefore a real positive.

But again, balance sheet strength does not make valuation risk disappear. A strong balance sheet can help a company survive a downturn. It cannot guarantee that investors who paid too high a price will avoid losses.

Management Is Trying To Change The Industry Model

One of the most interesting parts of the Sandisk story is management’s attempt to reduce cyclicality through long-term agreements.

This is important because the historical memory model has been painful for shareholders. Customers often negotiate aggressively, supply/demand changes quickly, and prices can move sharply. Sandisk’s long-term supply agreements are an attempt to bring more stability to a business that has traditionally lacked it.

The company has said these agreements include financial commitments and structures intended to prevent customers from simply walking away when pricing weakens. Reuters reported Goeckeler’s comments that Sandisk is trying to escape the “boom-bust cycle” and create more consistent economics. (Reuters)

If this works, it could justify a higher valuation multiple.

But investors should not assume success too quickly.

The bear case is that long-term agreements help, but do not fully eliminate the commodity nature of NAND. If the market becomes oversupplied, customers may still push back, competitors may discount, and the market may rediscover that memory is still memory.

This is why the BE Invested report framed management’s narrative carefully. It recognised the strategic importance of long-term agreements, but did not accept the idea that Sandisk has completely escaped cyclicality.

The Moat: Strong Technology, Narrow Industry Protection

The report’s moat assessment was also useful.

It did not call Sandisk a wide-moat business. It rated the moat as narrow, despite recognising the company’s strong intellectual property, manufacturing scale, technology base, and customer relationships.

That may feel harsh, but it is reasonable.

Sandisk has real advantages. It has decades of flash memory experience, a recognised brand, a manufacturing partnership with Kioxia, deep customer relationships, and exposure to mission-critical storage demand.

But the industry structure is difficult.

There are well-funded competitors. Technology cycles move fast. Manufacturing is capital intensive. Pricing is volatile. And state-supported competition, especially from China, remains a long-term risk.

In other words, Sandisk may be a very strong player in a very hard industry.

That is not the same thing as being a dominant software company with recurring revenue, low capital intensity, and high switching costs.

This distinction matters because investors often confuse technological importance with investment quality.

A product can be essential, but the economics can still be cyclical.

A company can be strategically important, but the stock can still be overvalued.

A business can be real, but the market can still price it too aggressively.

That is exactly the kind of nuance BE Invested Labs is built to surface.

Valuation: The Hardest Part Of The Sandisk Story

The BE Invested report valued Sandisk using a mix of discounted cash flow analysis, peer comparison, and scenario analysis. The base DCF value was around $1,500 per share, with a probability-weighted value around $1,530 to $1,550 depending on the scenario mix. The report’s bull case was $2,200, while the bear case was $900.

Those numbers tell the story better than any headline.

The upside exists, but it is not massive relative to the downside.

In the bull case, AI demand remains strong, NAND pricing stays tight, margins remain elevated, and Sandisk uses cash flow to repurchase shares. In that world, the stock can still move higher.

In the base case, the AI demand surge continues for a while, but margins eventually normalise and the market begins to apply a more reasonable multiple.

In the bear case, the market realises this was another memory cycle. NAND prices fall, margins compress, the multiple de-rates, and investors who bought late experience serious downside.

That is the risk/reward issue.At lower prices, Sandisk could be a very attractive cyclical growth opportunity.

At stretched prices, it becomes a much harder decision.

This is why the report’s conclusion was Neutral rather than Bullish. It was not saying the business is weak. It was saying the price matters.

And price always matters.

What This Shows About BE Invested Labs

This is the real reason we wanted to publish this research.

Not because Sandisk is trending. Not because AI storage is exciting.
Not because we want to chase the hottest ticker.

We wanted to show the type of work BE Invested Labs is designed to do.

The platform does not simply summarize a stock. It builds a structured research framework around it. It looks at the business model, the moat, the bull case, the bear case, the financial trajectory, the balance sheet, management, valuation, and scenario risk.

Most importantly, it can hold two ideas at once.

Sandisk can be a strong company. Sandisk can also be expensive.

AI demand can be real. The memory cycle can also still matter.

Management can be right about structural improvement.
The market can still be overpaying for that improvement.

That is how investment research should work. Not based on FOMO or hype.

Our View

Our view is broadly aligned with the engine’s conclusion.

Sandisk is one of the most interesting AI infrastructure names in the market right now. The company is benefiting from real demand, real pricing power, and real financial momentum. Its recent results and guidance show that AI-related storage demand is not just a narrative; it is flowing through the numbers.

But the stock is also pricing in a lot of good news.

For investors, this creates a very specific problem. The question is not whether Sandisk is important. It clearly is. The question is whether the current price offers enough margin of safety given the cyclicality of NAND, the intensity of competition, and the risk that today’s extraordinary margins eventually normalise.

That is why Neutral makes sense. Not because the story is weak.

But because the valuation is demanding. If Sandisk continues executing, there may still be upside. But if the market begins to doubt the “AI super-cycle” narrative, the downside could be much larger than investors expect.

This is not a stock to approach casually. It requires monitoring. Specifically, investors should watch:

NAND contract pricing. Data centre revenue growth. Gross margin sustainability. Competitor capital expenditure. Long-term customer agreement quality. Buyback execution. Inventory levels. Management language around supply discipline.

Those are the signals that will tell us whether Sandisk is becoming a structurally better business, or whether we are simply witnessing a powerful cyclical peak.

Final Thought

Sandisk is a perfect example of why modern investors need better research tools.

The story is too important to ignore, but too complex to judge from headlines alone.

If you only look at the AI narrative, you may become too bullish.
If you only look at memory cyclicality, you may become too bearish.
If you only look at the stock chart, you may arrive too late.
If you only look at valuation multiples, you may miss the structural change.

The truth sits somewhere in the tension.

That is what BE Invested Labs is trying to help investors find.

Not certainty. Clarity. Because in markets, certainty is usually expensive. But clarity compounds. The BE Invested Labs platform is live now. You can generate AI-powered equity research reports, valuation analysis, and audio briefings on public companies in minutes.

Try the platform here:

BE Invested

And if you want more research notes like this, subscribe to BE Invested Labs.

Subscribe now

— Buyce & Emmanuel
BE Invested Labs

Disclaimer

This article is for educational and research purposes only. It is not financial advice, investment advice, or a recommendation to buy or sell any security. Always do your own research and consider speaking with a licensed financial adviser before making investment decisions.

References And Data Notes

1. BE Invested Labs internal Sandisk Corporation institutional research report, generated June 3, 2026. The report initiated coverage with a Neutral rating, estimated fair value around $1,550 per share, and scenario values of $2,200 bull case, $1,500 base case, and $900 bear case.

2. Reuters reporting was used to cross-check Sandisk’s fiscal Q3 2026 revenue of $5.95 billion, adjusted EPS of $23.41, Q4 revenue guidance of $7.75 billion to $8.25 billion, and Q4 adjusted EPS guidance of $30 to $33. (Reuters)

3. Reuters reporting was also used to cross-check Sandisk’s long-term customer agreements, including five long-term agreements and three third-quarter agreements worth at least $42 billion, as well as the announced $6 billion share buyback. (Reuters)

4. Investor’s Business Daily and MarketWatch reporting were used to cross-check Q3 revenue, adjusted earnings, Q4 guidance, data centre strength, and market reaction after the earnings release. (MarketWatch)

5. Reuters and WSJ reporting were used to cross-check the January 2026 quarter, including Q2 revenue of $3.03 billion, adjusted EPS of $6.20, Q3 guidance at the time, and the extension of the Kioxia joint venture through 2034. (Reuters)

6. Reuters reporting was used to cross-check Western Digital’s partial sale of its Sandisk stake after the spin-off and the broader post-spin-off ownership context. (Reuters)

Not as an idea anymore. Not as a screenshot. Not as a private prototype running quietly in the background. The web app is now live, and people can actually use it.

That alone is worth pausing on.

A few weeks ago, this was still a conversation between two co-founders trying to solve our own frustration with stock research. We were tired of opening too many tabs, reading conflicting opinions, checking financial statements, watching YouTube breakdowns, and still feeling like the process created more noise than clarity.

So we built BE Invested Labs around a simple belief:

Investors do not need more scattered information.
They need structured research.

Now the first users are starting to test that belief.

The first launch numbers

For a first launch window, those numbers matter to us. Not because they prove everything. They don’t. This is still early. But they show something important: people are not just visiting the site and leaving. They are using the product.

That distinction matters.

Traffic is nice. Usage is better.

A landing page can get visits because people are curious. But when people start generating reports, testing tickers, and creating podcasts, they are telling you something different. They are saying:

“I have a stock in mind.”
“I want to understand it better.”
“I want a research process that saves me time.”
“I want the analysis in a format I can actually use.”

That is the problem we are building for.

Generate your first report

The five most searched stocks so far

The early search data has already started telling a story.

The first one did not surprise us.

SMT, Scottish Mortgage Investment Trust, became the most searched ticker by far. That likely connects directly to our first public research note on SpaceX exposure through Scottish Mortgage, ARK, and other proxy vehicles.

That is an important signal.

It means content can drive product usage. A reader sees a research idea, becomes curious, visits the app, and tests the ticker for themselves. That is exactly the loop we want to build: research that does not end with reading, but continues into exploration.

NVO, Novo Nordisk, being the second most searched name is also interesting. It shows that users are not only looking at speculative technology names. They are also exploring high-quality global compounders in healthcare, obesity treatment, diabetes care, and long-term pharmaceutical growth.

PLTR, Palantir, being high on the list makes sense. It sits right at the centre of AI enthusiasm, government software, data infrastructure, valuation debate, and retail investor attention. It is exactly the kind of company where investors need more than a headline. They need a framework.

BE, Bloom Energy, tells a different story. That is a more speculative clean-energy and fuel-cell related name, and it shows users are willing to test higher-risk, higher-volatility ideas.

DCI, Donaldson, is perhaps the quietest but one of the most interesting names on the list. It is not the loudest stock on social media. It is not the typical retail hype ticker. That tells us some users are already using the platform to explore less obvious companies, which is exactly where structured research can become useful.

Leave a comment

What this tells us

The first launch data is small, but it already confirms three things.

First, investors are curious, but curiosity alone is not enough. The market is full of stories. SpaceX. AI. GLP-1 drugs. Clean energy. Industrial quality. The challenge is not finding a story. The challenge is understanding what is underneath it.

Second, content and product need to work together. Our SMT article did not simply live as a Substack post. It became a reason for users to test the platform. That is important for how we will build BE Invested Labs going forward. We do not want research to sit passively on a page. We want it to lead to better investigation.

Third, investors want different formats. The fact that 6 AI podcasts were created so early matters. Some people want to read a report. Others want to listen while driving, walking, working, or preparing for deeper study. Research should not be trapped in one format.

This is why BE Invested Labs includes both written reports and AI-generated audio briefings.

The goal is not simply to produce information.

The goal is to make research usable.

If you know someone who researches stocks but constantly gets lost between tabs, YouTube videos, analyst notes, Reddit threads, and financial statements, share this post with them. This is exactly who we are building BE Invested Labs for.

Share B.E Invested Labs

Why this matters for ordinary investors

Most ordinary investors are not lazy. They are overwhelmed.

They have access to more information than any generation before them, but access has not automatically created clarity. In many cases, it has done the opposite. More tabs. More opinions. More screenshots. More “top 5 stocks to buy now” videos. More confidence from people who may not have done the work.

That is dangerous.

Because investing is not only about having opinions. It is about building conviction from a process.

That is what we are trying to help with.

BE Invested Labs is not here to tell people what to buy blindly. It is here to help investors research better, compare ideas more clearly, and turn scattered information into structured thinking.

That is the mission.

What comes next

The platform is live now at:

https://beinvestedlabs.com/

Over the coming days, we will be watching how users interact with the product, which tickers they search, which reports they save, which podcast formats they use, and where the research experience still needs improvement.

This is not the finished line. It is the start line.

We will keep building in public. We will share what users are searching, what we are improving, what breaks, what surprises us, and what we learn.

If you want to test the platform, generate your own report, or create an AI podcast on a stock you are researching, the app is live now. Click the button below to try it.

BE Invested

And if you want to follow the company journey, research drops, product updates, and future investor tools, subscribe to BE Invested Labs and follow our LinkedIn page:

https://www.linkedin.com/company/be-invested-labs-ltd/

Final thought

The first few hundred visits do not make a company.

But they do make a signal.

And the signal is clear enough for us to keep going.

People are curious.
People are researching.
People are testing.
People want clarity.

Now our job is to keep building the system that gives it to them.
If this resonates, share it with one investor, founder, or friend who would find the platform useful.

Share

Subscribe now

Generate your first report

— Buyce & Emmanuel
BE Invested Labs

If the internet has become a free university for investors, the tuition is your time. We built BE Invested so the tuition would be paid by AI.

Not because investors need more noise. Not because the world needs another finance dashboard filled with blinking red and green numbers. But because ordinary investors are constantly being pushed into the deep end of the market with too much information, too little structure, and very little help in turning all of that data into clear thinking.

Few things create more noise than a hyped IPO, especially when the company involved is SpaceX.

Recent reporting around a potential SpaceX IPO has created exactly the kind of environment where investors can become emotional: huge valuation expectations, limited public financials, index inclusion speculation, private-market mark-ups, and the quiet fear that everyone else might get rich before you. That fear has a name: FOMO. And FOMO is rarely a good investment strategy.

So the question is not simply, “How do I buy SpaceX?” The better question is: how do I think about SpaceX exposure without losing my discipline?

That is why, ahead of the launch of BE Invested Labs, we wanted to bring one of our research ideas into the public: Scottish Mortgage Investment Trust, ticker SMT.L. Scottish Mortgage is one of the most interesting publicly listed vehicles for investors who want exposure to SpaceX before it becomes fully public. But it is not a clean SpaceX investment. It is not risk-free. And it is definitely not something to buy blindly because the word “SpaceX” is attached to it.

This is not financial advice. It is a research framework. The goal is not to tell you what to buy. The goal is to show you how to think.

Why This Matters Right Now

SpaceX has become one of the most anticipated potential public listings in modern market history. Recent reporting has suggested that the company could target a valuation around $1.75 trillion, with some commentary pointing to even higher market expectations. If that happens, SpaceX would not be just another IPO. It would become one of the largest public market debuts ever attempted.

But the higher the excitement, the more disciplined investors need to become. SpaceX is not only a rocket company anymore. The investment story has expanded far beyond launches and reusable boosters. The broader SpaceX ecosystem now touches satellite internet through Starlink, artificial intelligence through xAI-related narratives, and potentially even the long-term idea of orbital infrastructure and AI compute.

That sounds exciting. It also sounds expensive. And that is the problem.

When a company is private, ordinary investors do not usually get the same access as institutions, venture funds, insiders, and early backers. By the time the public gets invited to the party, the early returns have often already been captured by someone else. That does not mean the IPO will fail. SpaceX may still become an extraordinary public company. But if the IPO arrives at a very high valuation, retail investors could be walking into a situation where the story is already priced for perfection.

That is why indirect routes matter. Not because they are perfect, but because sometimes the better question is not, “Can I buy the rocket?” Sometimes the better question is, “Can I own part of the launchpad?”

Meet Scottish Mortgage Investment Trust

Scottish Mortgage Investment Trust sounds like it should be a sleepy financial product from another era. It is not.

Despite the name, Scottish Mortgage is not a mortgage company. It is not simply a Scottish fund buying local businesses. It is a global investment trust managed by Baillie Gifford, with a long history of backing transformational growth companies. Founded in 1909, the trust is more than a century old, but its portfolio is anything but old-fashioned.

Its mandate is to invest in companies that management believes can shape the future of the global economy. That includes public companies, but also private companies that most ordinary investors cannot access directly. This is the key point. Scottish Mortgage gives public market investors a listed route into a portfolio that includes both public technology leaders and private innovation companies.

In that sense, SMT sits somewhere between a traditional global equity fund and a venture-style growth portfolio. That is why it matters in the SpaceX conversation. Scottish Mortgage owns a large stake in SpaceX, but it also owns much more than SpaceX. And that distinction is important.

Track Record At A Glance

Scottish Mortgage’s recent performance shows why it continues to attract long-term growth investors despite its volatility. Over the last year, SMT delivered a strong share price return and significantly outperformed the FTSE All-World Index over the same period. But the longer-term picture is more complicated. Over a five-year period, SMT’s annualised Net Asset Value return has lagged the FTSE All-World Index, reflecting the painful reset that growth and technology stocks experienced during the 2022 to 2024 interest-rate cycle.

That is the nature of this trust. When long-duration growth is loved, SMT can look brilliant. When long-duration growth is punished, SMT can look broken. Investors need to understand both sides before attaching themselves emotionally to the upside story.

What makes SMT different from a normal global index fund is its aggressive exposure to private innovation companies. A large portion of the trust is invested in private businesses such as SpaceX, ByteDance, Stripe, Databricks, and Anthropic. These are the kinds of companies that ordinary investors usually cannot access unless they are institutions, venture capital funds, or ultra-high-net-worth individuals.

The fee is also relatively low for what the trust offers. SMT’s ongoing charges are around 0.34%, which is modest considering the private market access, research depth, and active management involved. But low fees do not remove risk. They simply mean investors are not paying hedge-fund-style fees to access this strategy.

Management Philosophy

Scottish Mortgage is run by people with a very particular worldview. Tom Slater and Lawrence Burns are not trying to beat the market by trading around quarterly earnings. Their approach is built around patient ownership, long-term compounding, and the belief that a small number of exceptional companies can create most of the value in a portfolio.

This is the same philosophical world that allowed Baillie Gifford to back companies like Amazon and Tesla early and hold them through extreme volatility. That approach can be powerful, but it can also be uncomfortable. Long-term investing sounds beautiful when prices are rising, but it becomes emotionally difficult when a favourite holding falls sharply, the market turns against growth stocks, and everyone begins asking whether the managers have lost their touch.

That is the tension inside SMT. It is built for investors who can tolerate volatility in exchange for exposure to companies that may compound for a decade or longer. That is not everybody, and it should not be sold as if it is.

Top Holdings: Why SMT Is More Than A SpaceX Proxy

As of the latest data we reviewed, Scottish Mortgage’s largest holdings included a mixture of public and private companies across semiconductors, AI, digital platforms, payments, e-commerce, and space. The list includes names such as SpaceX, TSMC, NVIDIA, ASML, ByteDance, Amazon, Stripe, MercadoLibre, Meta, and Anthropic.

This matters because many investors will hear “SMT owns SpaceX” and immediately reduce the trust to a SpaceX proxy. That is too simplistic. SpaceX is a major part of the story, but SMT is not just a rocket bet. It is a portfolio built around the idea that the next era of global growth will come from companies building the infrastructure of the future.

That includes advanced chips, AI models, global e-commerce, digital payments, social platforms, cloud infrastructure, private space exploration, and next-generation software. In other words, SMT is not just asking whether SpaceX will win. It is asking whether the companies building the next technological layer of the economy will continue to compound.

That is a bigger thesis, and it deserves to be analysed that way.

How SMT Gets You Into SpaceX

Scottish Mortgage invested in SpaceX years before ordinary investors could even dream of buying the company on a public exchange. That early access is the entire point.

Most retail investors cannot buy SpaceX directly today. There are secondary market platforms and specialist products that may offer access, but they often come with restrictions, higher minimums, liquidity limits, accreditation requirements, or valuation uncertainty. SMT offers a simpler route. You buy the listed trust on the London Stock Exchange, and inside that trust is a stake in SpaceX.

But here is the important part: Scottish Mortgage does not value SpaceX based on hype headlines alone. Recent reporting shows Baillie Gifford has defended its SpaceX valuation using verifiable transaction data rather than media speculation. The trust has reportedly valued SpaceX around $1.25 trillion, even while some IPO speculation has pointed toward a much higher figure.

That matters because if SMT is valuing SpaceX more conservatively than the rumoured IPO price, there may still be upside if the public market eventually confirms a higher valuation. But there is also risk. If the IPO prices lower than expected, gets delayed, or trades poorly after listing, SMT shareholders could feel that too.

So SMT is not a magic trick. It is an access vehicle with benefits and trade-offs.

The NAV Question: Discount Or Premium?

One of the most important parts of understanding Scottish Mortgage is the relationship between its share price and its Net Asset Value. Net Asset Value, or NAV, is the estimated value of everything the trust owns after liabilities. If SMT’s share price is below its NAV, it trades at a discount. If its share price is above NAV, it trades at a premium.

Historically, Scottish Mortgage has often traded at a discount to NAV, especially when investors were worried about private company valuations, higher interest rates, or the broader growth stock sell-off. That discount was part of the attraction because it meant investors could sometimes buy a portfolio of high-growth public and private companies for less than the stated value of the underlying assets.

But this is where the analysis needs to be current. Recently, SpaceX-related optimism has helped push SMT back toward, and at times above, NAV. That changes the equation. Buying SMT at a discount is different from buying SMT at a premium. At a discount, you may be getting a margin of safety. At a premium, you are paying extra because the market is already excited.

So investors must not simply ask, “Does SMT own SpaceX?” They must also ask, “Am I paying a fair price for SMT today?” That is the type of question that separates investing from chasing.

The Fine Print

Owning SpaceX through SMT is not pure upside. Private company valuations are harder to assess than public market valuations. They are not priced every second like listed shares. They depend on funding rounds, secondary transactions, internal valuation work, and independent valuation processes. That means the reported value can move sharply when new information appears.

There is also concentration risk. SpaceX has become a very large part of SMT’s portfolio. That is exciting when SpaceX is being marked up, but it can become painful if sentiment changes. Then there is ByteDance exposure. ByteDance owns TikTok, and TikTok remains politically sensitive, especially in the United States. So SMT is not only taking technology risk. It is also taking regulatory and geopolitical risk.

Finally, there is the emotional risk. SMT can fall hard. Investors who cannot handle volatility should not pretend they are long-term growth investors just because the upside story sounds attractive. That is how people get shaken out at the worst time.

ARKX And Other Space Funds: The Comparison

When investors search for SpaceX exposure, ARK usually enters the conversation. That makes sense. Cathie Wood and ARK Invest are strongly associated with disruptive innovation, electric vehicles, space, robotics, AI, and high-growth technology. So naturally, many investors assume ARK must be the cleanest route to SpaceX exposure.

But the details matter. ARK Space Exploration & Innovation ETF, known as ARKX, sounds like the obvious SpaceX play. In reality, it does not own SpaceX directly. It is a space and aerospace-themed ETF that invests in listed companies connected to space exploration, defence, satellites, and enabling technologies. That includes companies such as Rocket Lab, L3Harris, Kratos, AMD, and others.

Those businesses may benefit from growth in the broader space economy. But owning ARKX is not the same as owning SpaceX. That distinction is crucial. ARKX owns the surrounding theme. SMT owns the private-market asset.

ARK Venture Fund is different. It does own private SpaceX shares, and therefore comes closer to the direct exposure many investors are looking for. But it also comes with different trade-offs: higher fees, different liquidity terms, a shorter track record, and venture-style valuation risk. Based on recent reporting, SpaceX has represented a mid-teens percentage of ARK Venture Fund at times, but that figure can change. Cathie Wood has also indicated ARK could rebalance its SpaceX position after the company becomes public.

So the key lesson is simple: do not buy a fund because the marketing theme sounds right. Look inside the vehicle. The name on the wrapper is not enough.

Other Public Vehicles With SpaceX Exposure

Scottish Mortgage is not the only listed vehicle with SpaceX exposure. Baillie Gifford US Growth Trust also provides access to SpaceX, although with a smaller weighting than SMT. It may appeal to investors who want Baillie Gifford’s growth philosophy but with more of a US-focused portfolio.

Schiehallion Fund, another Baillie Gifford vehicle, is also connected to later-stage private companies and has been discussed as part of the broader SpaceX exposure conversation. Other UK-listed vehicles and private market funds may also hold SpaceX stakes, but many of them either have smaller exposure, different mandates, less liquidity, or a portfolio that is not primarily focused on high-growth innovation.

So again, the question is not simply, “Which fund owns SpaceX?” The better question is: “How much SpaceX exposure am I really getting, what else am I buying, what fees am I paying, and what risks am I accepting?” That is the adult version of the conversation.

SMT vs ARK: The Practical Difference

When comparing the different ways to get exposure to SpaceX, the details matter more than the hype. Scottish Mortgage Investment Trust feels like the most balanced option because it gives investors a meaningful direct stake in SpaceX, but not in isolation. You also get exposure to other long-term growth companies across AI, semiconductors, fintech, and digital infrastructure, including names like ASML, NVIDIA, TSMC, Stripe, ByteDance, and MercadoLibre.

But let’s not romanticise it. SMT is volatile. A large portion of its assets are in private companies, which makes valuation less transparent. Its exposure to ByteDance brings China and regulatory risk. And as we saw during the growth stock correction, this trust can underperform badly when the market turns against long-duration technology businesses.

ARKX, on the other hand, sounds like a SpaceX play, but it does not actually own SpaceX. It gives exposure to the broader space and defence theme through companies like Rocket Lab, L3Harris, and Kratos. That may benefit from growth in the space economy, but it is not the same as owning SpaceX.

ARK Venture Fund comes closer because it does own private SpaceX shares, but the fees are higher, liquidity is more limited, and the exact SpaceX weighting can change over time. So the view is straightforward: if you want the cleanest “SpaceX only” exposure, none of these are perfect. But if you want meaningful SpaceX exposure inside a broader long-term innovation portfolio, SMT looks like one of the most practical options. Imperfect, yes. Volatile, definitely. But far more balanced than chasing IPO FOMO without understanding what you are paying for.

Beyond Rockets: The Semiconductor And AI Angle

One of the reasons SMT is interesting is that the SpaceX story is only one layer. The broader portfolio gives exposure to the companies building the infrastructure behind AI, semiconductors, cloud computing, digital platforms, and global commerce. This is where SMT becomes more than a SpaceX proxy.

NVIDIA and TSMC are central to the AI and semiconductor supply chain. NVIDIA provides the chips and computing platforms powering much of the AI revolution, while TSMC manufactures many of the world’s most advanced semiconductors. ASML is also critical. Without ASML’s lithography machines, the most advanced chips in the world become much harder to produce. That makes ASML less flashy than some AI companies, but structurally important.

Anthropic gives SMT exposure to private AI model development. Stripe gives exposure to digital payments infrastructure. MercadoLibre gives exposure to Latin American e-commerce and fintech. ByteDance and Meta represent different versions of the digital attention economy.

So when you buy SMT, you are not simply asking whether rockets will reach Mars. You are asking whether the next generation of economic infrastructure will be built by a concentrated group of exceptional technology companies. That is a very different thesis, and it is why SMT can be attractive even if someone is not obsessed with SpaceX.

The Human Element: Tom Slater, Lawrence Burns And Baillie Gifford

There is another thing that separates Scottish Mortgage from an ETF: it is actively managed. That means you are not just buying a basket of stocks. You are buying the judgement of Baillie Gifford’s investment team.

Tom Slater has been closely associated with Scottish Mortgage’s long-term growth approach. Lawrence Burns also plays a key role in the trust’s management. Together, they represent a philosophy that is very different from short-term market timing. They are looking for outliers, the kind of companies where one or two winners can drive a large part of the portfolio’s long-term return.

That approach requires patience. It also requires a willingness to look wrong for long periods of time. And that is where investors need self-awareness. It is easy to say you believe in long-term investing when the chart is going up. It is much harder when the market disagrees with you for two years and the comment section starts calling your favourite fund a disaster.

SMT’s approach can produce extraordinary returns when it works. But it can also produce painful drawdowns. That is the bargain. You do not get venture-style upside without venture-style discomfort.

How We Would Frame The Decision

This is not financial advice. This is the framework we would use.

First, define your objective. Do you want pure SpaceX exposure, or do you want a diversified innovation portfolio with SpaceX as a major growth engine? If you want pure SpaceX exposure, SMT is not perfect because it owns many other companies. But if you want meaningful SpaceX exposure plus AI, semiconductors, digital platforms, fintech, and private market optionality, SMT becomes much more interesting.

Second, check the price you are paying. Is SMT trading at a discount or a premium to NAV? This matters because a great portfolio can become a poor investment if bought at the wrong price.

Third, understand the private market risk. Private holdings can be powerful, but they are less transparent. They can be marked up, but they can also be marked down. Fourth, compare fees and liquidity. ARKX gives daily ETF liquidity but no direct SpaceX ownership. ARK Venture Fund gives direct private-market exposure but with higher fees and less straightforward liquidity. SMT sits somewhere in the middle: listed, liquid, lower fee, but still exposed to private company valuation risk.

Finally, be honest about your temperament. If you cannot handle volatility, SMT may not be for you. If you buy because of SpaceX excitement and panic when the trust falls 25%, the problem is not the trust. The problem is the mismatch between the investment and your psychology.

Our View

At BE Invested Labs, we see SMT as one of the most pragmatic publicly listed ways for everyday investors to gain meaningful SpaceX exposure while also participating in other long-term growth themes such as AI, semiconductors, fintech, and digital infrastructure.

That does not mean SMT is a “buy at any price.” It is not. It means SMT deserves serious research because it gives ordinary investors access to a type of portfolio that would otherwise be difficult to build.

ARKX gives you the space theme, but not direct SpaceX ownership. ARK Venture Fund gives you private SpaceX exposure, but with higher costs, different liquidity terms, and changing weightings. SMT gives you a large SpaceX stake inside a wider innovation portfolio, managed by a team with a long-term growth philosophy.

That is why it stands out. But the key word is “framework.” Do not buy stories. Analyse structures. Because a good story does not retire you. Productive assets do.

What Happens Next — And Where BE Invested Comes In

We built BE Invested Labs because we kept seeing the same pattern. A friend would ask about the next hot stock. Someone would send a video about an IPO. A headline would go viral. Then suddenly everyone would be opening ten tabs, reading conflicting opinions, checking Reddit, watching YouTube, and still ending up more confused than when they started.

That is not research. That is modern financial noise.

Our platform is designed to turn that noise into structured analysis. With BE Invested Labs, users will be able to generate research reports on public companies, run valuation models, compare scenarios, save reports in a research library, and listen to AI-generated audio briefings.

The idea is simple: less scattered research, more structured thinking.

Ahead of our launch, this article is an example of the kind of work we want to bring to ordinary investors. Not hype. Not predictions dressed up as certainty. But structured analysis that helps you think better.

Our waitlist is open now:

BE Invested Waitlist

If you want early access, join the list. Our first users will help shape the product directly.

Closing Thoughts

We are living through a period where innovation moves faster than regulation, narratives move faster than earnings, and retail investors are often asked to make decisions with incomplete information. That is dangerous, because in markets, the loudest story is not always the best investment.

SpaceX may become one of the greatest public companies of the next decade. Or it may list at a valuation that already assumes too much perfection. We do not know yet. But we do know this: chasing hype is not a strategy. Research is.

By looking at SMT, ARKX, ARK Venture Fund, and other possible routes, we are not trying to crown one perfect winner. We are trying to understand the structure behind the exposure. Because the goal is not to be first into the hottest IPO. The goal is to build wealth with clarity.

And clarity compounds.

See you on launch day.

— Buyce & Emmanuel
BE Invested Labs

And if you want early access to the platform behind this analysis, join the waitlist before launch: BE Invested Waitlist

Disclaimer

This article is for educational and research purposes only. It is not financial advice, investment advice, or a recommendation to buy or sell any security. Always do your own research and consider speaking with a licensed financial adviser before making investment decisions.

References And Data Notes

1. Scottish Mortgage Investment Trust official factsheets, annual reports, and manager commentary were used for portfolio holdings, ongoing charges, private company exposure, performance data, and management information.

2. Reuters reporting was used for SpaceX IPO timing, potential valuation expectations, and possible fast index inclusion following new FTSE Russell rules. (Reuters)

3. The Times reporting was used for Scottish Mortgage’s SpaceX holding, portfolio weighting, Tom Slater’s commentary, share buyback activity, and recent investor optimism around the trust. (The Times)

4. BE Invested Labs internal SMT.L research report and the uploaded draft were used as the base research file for this article.