Caterpillar shows no signs of stopping; year after year it delivers great results and its stock performance is solid. It's a cash machine and I don't see much reason why its growth should slow down significantly. The shares have been rising over the long term and major drawdowns aren't very common, so I'm thinking about buying even at the current price.

Looked at in context, though, this looks more like a classic overreaction to a complex research update than the start of a new downcycle. TurboQuant is an early stage algorithm aimed mainly at making inference more efficient on conventional DRAM, while the structural story in high bandwidth memory, the stacked chips bolted next to GPUs and critical for training large AI models, is still defined by tight supply, rising demand and full order books at the key suppliers.

Breakthrough technology sparks market panic

Google Research $GOOG has unveiled a new memory compression algorithm called TurboQuant , which researchers say can compress key cache memory used in large language models at least six times faster with up to eight times faster inference, without sacrificing accuracy .

The market reaction was immediate and dramatic. On Thursday, shares of the world's two largest memory chip makers, SK Hynix and Samsung $SSNLF, fell 6% and nearly 5% respectively in South Korean trading. Samsung Electronics closed down 4.71%, while SK Hynix fell 6.23%, pulling the South Korean benchmark KOSPI index down 3.22% .

A similar trend continued in the US markets, where shares of companies such as Micron Technology $MU, which fell 7% , and SanDisk $SNDK, which fell 6.8% . These moves followed declines in SanDisk and Micron shares in the US on Wednesday .

How TurboQuant works and why it scares investors

TurboQuant represents a revolutionary approach to solving one of AI's biggest bottlenecks - the enormous memory requirements during inference operations. TurboQuant is a compression method that achieves high model size reduction with zero loss of precision, making it ideal for supporting both key cache (KC) compression and vector search.

The technology works in two phases. The first phase uses PolarQuant, which thinks about mapping high-dimensional space differently. Instead of using standard Cartesian coordinates (X, Y, Z), PolarQuant converts vectors into polar coordinates consisting of a radius and a set of angles. The breakthrough lies in the geometry: after random rotation, the distribution of these angles becomes highly predictable and concentrated.

The second phase acts as a mathematical error corrector. Even with the efficiency of PolarQuant, a residual amount of error remains. TurboQuant applies a 1-bit quantized Johnson-Lindenstrauss (QJL) transformation to this residual data.

The actual market impact remains a question

Despite the immediate market reaction, analysts caution against exaggerated concerns. Ray Wang, a memory analyst at SemiAnalysis, said Google's research won't necessarily lead to the need for fewer chips. Cache values are "a key bottleneck that needs to be addressed for better models and hardware performance," he said. Wang said it will be "hard to avoid higher memory consumption" as a result of improving the performance of models .

It is also important to distinguish between different types of memory. It should be noted that compared to standard DRAM chips, this technology will have less impact on HBM (High Bandwidth Memory). TurboQuant is mainly used to optimize the inference of AI models, a phase that mostly requires only ordinary DRAM chips. However, HBM remains a necessity in the AI training phase.

According to a CNBC report, despite Thursday's stock drop, a perfect storm of factors continues to support the memory market over the long term. Significant demand coupled with supply shortages pushed memory prices to unprecedented levels and supported gains for Samsung, SK Hynix and Micron .

Structural fundamentals remain solid

It is also key to remember that TurboQuant is still only a research project. It is worth noting that TurboQuant has not yet been deployed on a larger scale; it is still a laboratory breakthrough at this time. This makes comparisons to something like DeepSeek, or even the fictional company Pied Piper, more difficult.

Data shows that the HBM market size will grow 58% to $54.6 billion in 2026, accounting for nearly 40% of the DRAM market. The sudden increase in demand has led to an imbalance between supply and demand. Despite Samsung, SK Hynix and Micron allocating 70% of their new/additional capacity to HBM, there remains a 50-60% capacity gap for HBM.

According to Wells Fargo analysts, the Google TurboQuant update could actually be a positive for memory companies. Although this kind of breakthrough might look negative for memory companies, the idea of the Jevons paradox suggests that the opposite can happen - making AI more efficient reduces costs, which can actually encourage much wider use and demand .

Structural drivers tied to AI infrastructure, supply constraints and tight HBM markets support a resilient long-term outlook. Investors should distinguish between short-term noise and fundamental trends anchored in persistent memory shortages and AI workload expansion .

For a long term investor, the question is not whether the story sounds exciting, but whether the current price already assumes too much. That means looking beyond headlines to growth rates, capacity build‑out, regulatory decisions and the rest of Lilly’s pipeline to see how much of future cash flow really comes from obesity and how much from other areas. Only vdetailní analysis of these moving parts can show whether Lilly is genuinely the best positioned big pharma name for the next 5 to 10 years, or whether expectations have pulled too far ahead of fundamentals.

Top points of analysis

• Lilly achieved record revenues of $65.2 billion in 2025 (+44% YoY), with tirzepathide (Mounjaro + Zepbound) contributing $36.5 billion, more than 56% of total revenues.

• Q4 2025 delivered revenue of $19.3 billion (+43% YoY), adjusted EPS of $7.54 (8.7% above consensus), management issued 2026 guidance for revenue of $80-83 billion and non-GAAP EPS of $33.50-35.00.

• Mounjaro (diabetes) generated roughly $23 billion in 2025 (+99% YoY), while Zepbound (obesity) added another $13.5 billion (+175% YoY), both figures well ahead of analyst estimates.

• The historic agreement with the Trump administration and Medicare brings GLP-1 drug coverage to millions of new patients at a copay (fixed amount) of $50 starting in July 2026, one of the biggest demand expansion catalysts in Lilly's history.

• Lilly has invested over $18 billion in manufacturing capacity (US, Ireland, Germany, China) since 2020, with the new $9 billion Indiana plant being the largest single investment in drug synthesis in US history.

• Valuation matches "premium growth" status: forward P/E of around 28-30× on 2026 EPS, analysts estimate growth of $1,260-$1,500.

What's changed: from mediocre pharmas to the most well-supplied obesity bet

Back in 2021, Lilly $LLY was a solid dividend title with a diversified portfolio (insulin, oncology, immunology, neuroscience) but no clear "next big thing". The breakthrough came with the results of the clinical program of tirzepathide, which was the first ever to combine both GIP and GLP-1 receptor agonism and achieve +20% weight reduction in obese patients in the SURMOUNT-1 trial - numbers that previous generation drugs (GLP-1 monotherapy like semaglutide) were unable to achieve.

Zepbound received FDA approval in November 2023 for obesity and Mounjaro was approved for diabetes back in 2022. Since then, Lilly has essentially been out of production. Demand has outstripped capacity, the stock has become one of the most followed titles on Wall Street, and management has reframed the entire industry around one thesis: obesity as a treatable chronic disease, not a cosmetic problem.

The deal with the Trump administration in November 2025 was a watershed moment not only from a medical perspective but also from an investment perspective. For the first time in history, Medicare agreed to cover GLP-1 drugs for obese patients (not just diabetics), opening access for an estimated tens of millions of previously uncovered patients in the US. Implementation is planned to begin in July 2026, when the "GLP-1 payment demonstration" model begins, and transition to a more permanent balance model in 2027.

What needs to work for this to work

• Mounjaro and Zepbound must maintain dominant market shares despite increasing pressure from Novo Nordisk and others.

• Oral GLP-1 orforglipron must get regulatory approval and establish itself as an alternative to injections.

• The Medicare deal must actually generate new patient volume and not just shift existing payments to lower prices.

• Manufacturing capacity must ramp up fast enough that Lilly does not lose share due to the unavailability of the drug.

• Retatrutide and other pipeline must deliver positive clinical data and perpetuate the "next big thing" story after tirzepatide.

How does that become money

1) Tirzepatide as a global bestseller - now and historically

Tirzepatide (Mounjaro + Zepbound) exceeded $24.8 billion in sales in the first 9 months of 2025, surpassing Keytruda (Merck's $MRK pembrolizumab ) as the world's best-selling drug in the period under review. This is an extraordinary fact: a drug that has been on the market for less than 3 years has reached the top of the global pharmaceutical ranking.

Projections to 2030 envisage tirzepathide sales of around USD 62 billion per year if the combination of Mounjaro (T2D) and Zepbound (obesity + new indications) continues to grow. This assumes maintaining about 58% market share in the US in GLP-1 prescribing, which Lilly currently holds, and expanding into new geographies (China, Japan, Europe).

2) Medicare deal as a demand gamechanger

Until November 2025, Medicare coverage of GLP-1 drugs for obese patients was minimal or non-existent - the 2003 Medicare law explicitly excluded drugs for "lifestyle" conditions. Trump's deal with Lilly and Novo Nordisk effectively circumvents this rule through a demonstration model: copay $50/month for eligible patients, Lilly agrees to a lower realized price, Medicare pays the rest.

If an estimated 10% of the Medicare population is eligible for GLP-1 coverage (and this number may continue to grow as indications expand), this is additional demand in the order of millions of patients who previously paid out of pocket or did not receive treatment at all. For Lilly, this means potentially hundreds of millions to units of billions of dollars of additional Medicare revenue per year, albeit at a lower unit cost. The exact balance (lower price vs. higher volume) remains to be seen in reality after July 2026, but the logic is clear: massive volume growth.

3) New indications for tirzepathide

Tirzepate is not just a cure for obesity and diabetes. Lilly is actively expanding indications, the most important being:

• Heart failure with preserved ejection fraction (HFpEF) - a huge unmet need in patients where there is not yet a good pharmacological therapy. Positive study results may add millions of potential patients.

• Sleep apnea - FDA recommends Zepbound as a treatment for obstructive sleep apnea in the obese - was approved in 2024, making Zepbound the first approved pharmacological treatment for this condition.

• MASH (NASH, metabolic liver disease) - tens of millions of patients in the US and Europe.

• Prediabetes, cardiovascular prevention - repeated studies show benefits of tirzepatide beyond weight reduction.

Each new approved indication expands the addressable population and reduces the risk of "single-indication" dependence. In particular, HFpEF and MASH are indications where other treatments are virtually non-existent, so Lilly could become a sole or first mover, giving a strong pricing position.

4) Orforglipron - a revolution via the tablet

The biggest debate in the GLP-1 space is whether an oral (tablet) drug can be effective. The current top products (Zepbound, Wegovy) are injectable, which puts off some potential patients and limits penetration in countries where patients refuse or cannot inject.

Lilly'sOrforglipron achieved -12.4% weight loss in 72 weeks in Phase 3 - a figure that is significantly higher than older oral GLP-1 candidates (for example, the semaglutide tablet Ozempic/Rybelsus achieves around -15% at a higher dose but with necessary food intake restrictions).

Nevertheless, orforglipron remains a potentially transformative product because:

• it's a tablet with no restrictions on food or water intake (unlike semaglutide tablet).

• It targets a segment that refuses or cannot inject.

• may dominate in developing countries and Asian markets where injections are less accepted.

Lilly plans a regulatory filing in 2026 and approval is also expected later this year. If the Medicare deal also covers oral GLP-1, orforglipron could catch a wave of new demand immediately after launch.

5) Retatrutide - the "nextgen" tirzepatide with historic numbers

If tirzepatide was a breakthrough, retatrutide is potentially a revolution. It is a triple agonist of GIP, GLP-1 and glucagon - the first such molecule in the clinical pipeline.

In the TRIUMPH-4 trial (phase 3), retatrutide achieved a mean weight reduction of -28.7% at 68 weeks, met all primary and key secondary endpoints, and produced results in pain and physical function in patients with musculoskeletal problems. This figure is higher than anything previously achieved in a clinical programme for the pharmacological treatment of obesity - by comparison, tirzepatide achieves around -20%, semaglutide (Wegova) around -15-17%.

In its Drugs to Watch 2026 report, Clarivate identified orforglipron and retatrutide as "defining next-generation GLP-1 candidates", highlighting the combination particularly in the context of "metabolic innovation beyond replicating existing GLP-1 therapies". If retatrutide makes it through the regulatory process and gets approval, Lilly will have products on the market in three efficacy generations: oral (orforglipron, -12%), injectable tirzepatide (-20%) and ultra-effective retatrutide (-28%). That's a portfolio that no other player in the industry has.

6) Production capacity as a strategic moat

One of the most underappreciated aspects of the Lilly story is manufacturing. GLP-1 drugs are chemically complex and manufacturing intensive - you can't just start "copying" them overnight. Lilly has invested over $18 billion in manufacturing capacity since 2020, with the new Indiana plant alone receiving a $9 billion investment - the largest-ever investment in drug synthesis in the US.

Parallel expansions are underway in Ireland (Limerick), Germany (Alzey), North Carolina and China (Suzhou, +$200 million). This infrastructure creates a real barrier to entry for competitors: even if someone were to announce a comparably effective drug today, it would take them years to build production capacity at a similar level. Lilly is effectively 'closing' the space before the big competitors get there.

The numbers that support this thesis

• Revenues: $34.1 billion (2023) → $45.0 billion (2024) → $65.2 billion (2025), a CAGR of roughly 38% over 2 years.

• Mounjaro: 5.2bn (2023) → 11.5bn (2024) → US$23.0bn (2025, +99% YoY).

• Zepbound: launch November 2023 → USD 4.9bn (2024) → USD 13.5bn (2025, +175% YoY).

• Q4 2025: revenue $19.3bn (+43% YoY), adjusted EPS $7.54 (+42% YoY), beating consensus by 8.7%.

• Q2 2025 gross margin: 85% (non-GAAP), operating margin around 44%, operating profit +63% YoY.

• R&D spending 2025: $13.3bn (+21% YoY), around 20% of sales.

• Outlook 2026: revenues USD 80-83bn (midpoint +25% YoY), non-GAAP EPS USD 33.50-35.00 (midpoint +35% YoY).

• Tirzepatid as the world's #1 bestseller: surpassed Keytruda with USD 24.8bn in the first 9 months of 2025.

• 58% market share in US GLP-1 prescribing (2025).

• Investment in manufacturing by 2020: over $18 billion.

Dividend and financial health

Lilly pays a conservative dividend - payout ratio is deliberately low, with the company prioritizing reinvestment in R&D and manufacturing capacity. The annual dividend is roughly around USD 5-6 per share, so the yield is only 0.5-0.6% at a price of around USD 989 - not a dividend story, but a growth story.

Financial health is solid: gross margin around 85% (pharma products are high margin after fixed costs), operating margin increasing towards 44-45% as sales grow faster than costs. R&D spending of $13.3 billion (20% of sales) is enormous in absolute terms, but necessary in the pharma industry to maintain pipeline.

Free cash flow is very strong and management is using it to:

• reinvestment in production and R&D.

• Acquisitions/partnerships (complementary pipeline).

• Moderate buybacks and dividends.

Valuation - what's included and what's not

Lilly is not a cheap stock even in the most optimistic scenario. Forward P/E on 2026 EPS (~$34) at a price of around $989 implies a multiple of around 29×, well above the S&P 500 average and typical pharmaceutical titles (P/E 15-20×). Thus, the market is paying a large premium for:

• Revenue Momentum: 44% revenue growth in 2025 is not the norm, but the exception.

• Pipeline: orforglipron, retatrutide and new indications of tirzepatide are priced in as likely positives.

• Structural position: dominance in the fastest growing segment of pharma.

To rerate (or maintain) multiples, companies with a P/E of 29× must regularly beat guidance, otherwise there is a "de-rating" effect where even a slight slowdown causes the stock to fall 20-30%.

A basic valuation framework for an investor:

• With EPS 2026 guidance midpoint of $34.25 and P/E 30× → implied value of ~$1,028 (slightly above price).

• At 2027 EPS (guidance around USD 45-50) and P/E 28× → implied value of USD 1,260-1,400.

• At a P/E compression of 20× (scenario where growth slows more significantly) and EPS of 34 USD → risk downside to ~680 USD.

Macro and market

The global obesity epidemic is a structural, not a cyclical trend: 1 billion adults globally now meet clinical criteria for obesity and the numbers are growing. Penetration of pharmacological treatments is still extremely low - estimated at less than 3-5% of eligible patients actually take the drug, although interest is growing. This means that Lilly is operating in a market that is still largely untouched.

The market for GLP-1 drugs is estimated at $157.5 billion today, with Lilly having a market share of around 58% in the US. At 25-30% market share in this market (Lilly vs. Novo Nordisk and new competition), this would imply revenues from obesity and diabetes alone in the tens of billions, which would be only one part of the overall portfolio.

Risks

1) Price pressure

The agreement with Medicare and Medicaid lowers the realized price of GLP-1 drugs relative to the listing price. Zepbound sells for a $50 copay to Medicare patients at a significantly lower price than commercial insurers pay. If similar pricing pressures spread to the commercial market, it could compress margins despite rising volume. This is a key issue for 2026-2027, when the Medicare model rolls out in full.

2) Competition

Novo Nordisk's semaglutide (Wegovy for obesity) is a direct competitor with proven cardiovascular efficacy (SELECT study) and a strong position especially in Europe. Amgen, AstraZeneca, Roche and a number of others have GLP-1 or GLP-1 combination candidates in the pipeline. If any of these candidates achieve efficacy comparable to retatrutide at a lower cost, Lilly's market share could be eroded.

3) Orforglipron expectation gap

While the -12.4% results in Phase 3 were clinically meaningful, the market expected 13-14% and the response was negative. If orforglipron achieves lower adoption after launch than management expects (because patients and physicians prefer the stronger injectable tirzepatide or future retatrutide), the contribution to sales may be lower than projections.

4) Manufacturing and supply chain risks

At $65 billion in revenue, tirzepathide accounts for over 56% of revenue - that's an extreme concentration per molecule and per production line. Any problem at the plants (contamination, natural disaster, regulatory footprint) would have an immediate impact on revenues. Geographic diversification of production (US, Ireland, Germany, China) reduces but does not eliminate this risk.

5) Political and regulatory risk

The Trump administration is now in favor of expanding access to GLP-1 drugs, but the political equation could change quickly. Potential "drug pricing negotiations" under the IRA (Inflation Reduction Act) or pressure on Medicaid rebate structures could affect realized prices in the U.S., a key market for Lilly.

Checklist of risks

• Significant compression of realized prices in the U.S. following the rollout of the Medicare model.

• Entry of a strong new competitor with better efficacy or a more favorable price tag.

• Lower than expected adoption of orforglipron after launch.

• Manufacturing incident or supply chain issue with tirzepatide capacity.

• Regulatory setback with retatrutide or new indications for tirzepatide.

• Policy change in the US around GLP-1 pricing or access.

Investment scenarios

Optimistic scenario

In the optimistic scenario, Lilly executes on all fronts. Tirzepate maintains dominant market share, Medicare deal generates massive new volume, orforglipron makes inroads as an oral alternative, and retatrutide gets approval for obesity by 2027-2028.

• 2027 revenues: $90-95 billion.

• Non-GAAP EPS 2027: around US$45-50.

• Valuation: P/E 32-35× → share price $1,440-1,750.

• Annual return from today's price: around 15-18% per year.

Realistic scenario

Lilly delivers solid numbers, but faces increasing competition, modest price compression from Medicare, and orforglipron reaches moderate adoption. Retatrutide is in registration or early launch phase.

• 2027 revenues: $80-87 billion.

• Non-GAAP EPS 2027: around $40-45.

• Valuation: P/E 28-32× → share price $120-1,440.

• Annual return from today's price: about 7-12% per year.

Pessimistic scenario

Medicare pricing pressure seeps into the commercial market, orforglipron disappoints, competitive pressure builds, and the pipeline hits a regulatory slowdown. Revenues stagnate around $75-80 billion.

• Non-GAAP EPS 2027: around USD 30-35.

• Valuation: P/E compression at 20-25× → share price of USD 600-875.

• Downside from today's price: -10% to -40%.

• This is a scenario where "premium valuation" hints at slowing growth.

What to watch next

• Mounjaro and Zepbound quarterly earnings - maintaining momentum is key to 2026 outlook.

• Details of Medicare GLP-1 model launch in July 2026, especially new patient volume vs. impact on realized prices.

• Regulatory filing and timeline for orforglipron approval (expected 2026).

• Phase 3 data on retatrutide for other indications and potential submission to FDA.

• Study results for HFpEF (heart failure) and MASH as new indications for tirzepatide.

• Competitor development - especially pipeline Novo Nordisk (cagrisema), Amgen and new entrants.

• Management comments on manufacturing capacity and ability to meet demand in 2026-2027.

• Developments in the policy environment around drug pricing in the US (IRA, Medicaid rebate).

What to take away from the article

• Lilly is the best-stocked pharma company in the obesity space today with tirzepatide dominance, a rich pipeline and massive manufacturing investment as a moat.

• The numbers for 2025 are extraordinary, $65.2 billion in revenue (+44%), but the outlook for 2026($80-83 billion) shows the story doesn't end there.

• The pipeline in the form of orforglipron (oral GLP-1) and retatrutide (-28.7% weight, triple agonist) gives Lilly the potential to dominate the obesity market over the next decade.

• The Medicare deal from July 2026 is the largest structural demand catalyst in the history of the GLP-1 segment.

• Valuation is challenging (forward P/E ~29×) but consistent with "high visibility growth" status. The main risk is P/E compression in a slowdown, not the fundamental breakup of the business.

• For an investor looking for a company with a visible 5-10 year growth story in one of the biggest healthcare trends of the generation, Eli Lilly remains the most compelling bet in the obesity space despite its premium valuation.

When the story overcomes reality

The year 2021 has been a dream come true for a certain category of stocks. Pandemic created a seemingly ideal environment for companies offering virtual fitness, telehealth, game engine or augmented reality technologies. Low interest rates reduced the discount rate on future earnings to a minimum, allowing extreme valuations for companies whose profits were years away. Investors were paying astronomical multiples for the mere promise of what could be.

Then came the turnaround. Rates began to rise, the pandemic premium began to fade, and the market began to demand concrete results instead of grand visions. The result has been declines unprecedented in history. Slump percentages are not mere statistics; they are a memento mori for anyone who buys purely on the basis of story without regard to valuation.

At the same time, it does not automatically mean that these companies have no investment value. The question that more and more analysts are asking today is different: is there a real catalyst for a comeback in any of these companies, or is it a value trap where the cheap price beckons but the fundamentals don't justify it?

Vuzix Corporation $VUZI

From the hype of AR glasses to reality

Vuzix Corporation is one of the pioneers of industrial smart glasses and augmented reality technologies for the enterprise segment. The company offers products such as the M-Series or Vuzix Shield that find applications in logistics, medical, defense, and industrial manufacturing. On paper, the business model is solid: it sells hardware to companies that integrate it into operational processes where it increases efficiency. The problem is that this vision has hardly been reflected in the numbers so far.

In April 2021, VUZI's share price hit an all-time high of around $32.43 apiece. Back then, the hype around AR technology was at its peak and investors believed that smart glasses were the next big hardware segment to replace smartphones in industrial settings. The reality is markedly different.

The company's quarterly revenue is around $1 million and the company reports negative EPS every quarter. Net losses exceed sales. The market capitalization today stands at $200 million, and the firm has no dividends and hedge funds hold zero stake in it. A negative EBITDA margin exceeding 450% is unique even by the standards of technology startups.

Where are the bright spots

There are some positive signs. Vuzix has entered into a strategic partnership with Taiwan's Quanta Computer, one of the world's largest ODM electronics manufacturers. In February 2026, it received new FCC and CE certifications for the LX1 model of smart glasses, opening the door for the company to further commercial deployment. The company holds over 246 patents and pending patents in the area of waveguide optics, a key technology for future next-generation AR headsets. The partnership with Garmin $GRMN to develop nanoscale optical projection systems also indicates that the company has customers in the premier league.

However, competition is strong. Meta Reality Labs, Apple Vision Pro, and new projects from Google are competing for the same segment, with multiples larger R&D budgets. It is structurally very difficult for a small company with revenues of a few million dollars to compete with these giants.

Overview of key metrics

Peloton Interactive $PTON

From pandemic star to giant drop

The story of Peloton is perhaps the most dramatic example of how quickly sentiment can turn. The company, which makes premium bikes and treadmills with a connected streaming ecosystem of trainers, experienced explosive growth during the pandemic. Investors have been snapping up shares of the company, whose revenue has been growing at triple-digit rates. At its peak in January 2021, the stock traded over $171. Today, they are trading around $4, down more than 97%.

The decline is due to a combination of factors that reinforce each other. Consumers stopped buying home fitness equipment after pandemic restrictions were eased. Demand for the expensive $1,500 bikes the company sells has plummeted. But in the meantime, the company invested massively in production and expanded capacity, leading to a huge inventory surplus. The company's operations swallowed up a cumulative total of over $2.7 billion between 2021 and 2024.

There were product sales at reduced prices, layoffs, CEO replacements, and founders leaving. Each of these moves dealt another blow to the stock price. A company that appeared to be the definitive winner of the pandemic fitness trend has turned into a cautionary tale for investors.

Is there a realistic recovery scenario?

Yes, there is, but it is very narrow. Peloton has focused in recent quarters on what it can influence: drastic cost cuts and cash flow optimization. In January 2026, the company laid off another 11% of its workforce. This brought it close to positive free cash flow for the first time, reaching $324 million in fiscal 2025. Net debt decreased 52% for the year to about $319 million.

On the other hand, fundamental issues remain. Subscriber numbers continue to decline and have reached a four-year low. Sales in FY2025 were $2.49 billion, down 7.77% year-over-year, and management projects a further decline to about $2.4 billion in FY2026. This is the fifth straight year of declining sales.

The company is expanding into the commercial segment, where it now offers bikes and treadmills for hotels, fitness centers and apartment complexes. Whether the move can stem the erosion of the customer base remains a question mark. The stock is trading at a price-to-sales ratio of around 0.7x today, but analysts say a low valuation without improving fundamentals is not an argument to buy.

Overview of key metrics

Unity Software $U

Game engine seeks comeback via AI advertising

Unity Software is the dominant game engine for mobile and indie games, powering over 70% of mobile games worldwide. The company went public in September 2020 and soon became a favorite of tech-oriented investors. In November 2021, the stock reached an all-time high of over $210. Today, they are around $18, down over 90%.

The road to the bottom has not been a straightforward one. Unity merged in 2022 with ironSource, a controversial mobile advertising platform that caused a massive exodus of developers. In 2023 came the so-called Runtime Fee, a fee charged to developers for each game installation. The community reacted with fury, and many studios began switching to the competing Unreal Engine or other alternatives. Eventually, under pressure, management had to abolish the Runtime Fee. But the reputational damage was palpable.

In May 2024, new CEO Matthew Bromberg, former COO of Zynga, stepped in. His first steps were clear: abolish the Runtime Fee, reduce the workforce by thousands of positions, rebuild trust with the developer community, and emphasize monetization through the AI-powered ad network Unity Vector.

Unity Vector as a key catalyst

Unity Vector is an AI-enhanced mobile ad network that has become a focal point for investors. In January 2026, it reported 72% higher ad revenue year-over-year. Analysts at Barclays $BCS have raised their 2026 growth estimate for the advertising segment to 17%. The Grow Solutions segment's total revenue reached $1.23 billion in 2025, and a potential run-rate of over $1 billion per quarter by the end of 2026 would be transformative for the company. More on Unity's advertising model can be found in their developer documentation.

On the other hand, the Q4 2025 and full fiscal year results were disappointing. Revenue for Q4 was $545 million, only 3% year-over-year, and the outlook for 2026 calls for only 5-7% growth, a dramatic drop from the historical 20% pace. The company is coming from a position of huge reputation loss and the rebuild is taking longer than expected.

The big structural threat is $APP, a direct competitor in mobile advertising that has been significantly outperforming Unity in efficiency in recent quarters. Losing mobile ad market share to Apple and Android games is a real risk that could slow or stop the rebound of Grow Solutions.

Overview of key metrics

Teladoc Health $TDOC

Telehealth giant that failed to deliver on its promises

Teladoc Health is the world leader in virtual healthcare. The company operates two main divisions: Integrated Care, which offers virtual medical consultations and chronic disease management, and BetterHelp, the world's largest online psychological counseling platform. As of February 2021, TDOC's stock price had reached over $308. Today, it is around $5.5, a drop of over 98% from its all-time high.

This dramatic drop is the result of a series of setbacks. The biggest was the acquisition of Livongo Health in 2020 for approximately $18.5 billion, one of the largest health-tech transactions in history. The integration did not succeed to the extent expected and the company subsequently had to book a huge goodwill impairment. Total write-downs over the years exceeded tens of billions of dollars, with the firm coming up with new losses quarter after quarter.

BetterHelp, which was considered the fastest growing part of the business, began to slow down significantly in 2024 and 2025. Revenues in Q4 2025 were down 7% year-over-year, and adjusted EBITDA for the entire segment was down 46% for 2025. The problem is the highly competitive online therapy market, where Hims & Hers $HIMS, Cerebral, and a number of other platforms are outbidding customers through aggressive marketing.

The struggle to stabilise

Management is trying to restructure on multiple fronts. In the Integrated Care segment, it projects 2026 revenue of about $1.61 billion, up slightly year-over-year. The company is expanding its preventive care offerings with the acquisition of Catapult Health in February 2025 for $65 million. Enrollment (number of registrations) in chronic programs (diabetes, hypertension) is also growing, where retention is higher and ARPU (Average Revenue Per User) is more favorable.

However, the overall market dynamics remain a structural problem. The expiring ACA subsidy (health care law) may reduce the number of insureds using telehealth. The outlook for Q1 2026 calls for revenues of $598 million to $620 million, well below the consensus of analysts who were expecting over $633 million. Overall guidance for FY2026 (EPS of -0.70 to -1.10) was weaker than market expectations. As a result, Leerink Partners lowered its February 2026 target price from $8.50 to $5.50.

The firm's market capitalization has fallen below $1 billion.

Overview of key metrics

Comparison Table: Four bubbles that popped

A summary of key data for all four firms:

Strategic view

Looking at these four companies as a whole reveals one key lesson about investing: valuation at the time of purchase is arguably the most important factor in long-term performance. All four firms had compelling businesses in real and growing market segments. The issue was not what the companies did, but what their shares sold for years ago.

At the height of the hype, investors were paying for valuations that implied decades of flawless execution without any problem. Once reality set in, it wasn't just revenues or profits that fell, but more importantly, reassessed valuations, causing declines that multiplied for both factors.

For conservative investors, these stocks are more of a cautionary tale than an opportunity. For speculative-oriented players with a high tolerance for risk and the ability to react quickly to changes in fundamentals, Unity $U in particular may offer asymmetric opportunities, but only with great respect for positional risk and stop-loss discipline.

What to watch next

• $VUZI: Orders from Quanta Computer and commercialization of the LX1; development of quarterly sales above $2 million would signal real scaling

• $PTON: Declining or stabilizing churn in sub-script base; new commercial segment results (hotels, fitness centers); free cash flow development in FY2026

• $U: Unity Vector's quarterly run-rate to end FY2026; whether it will reach $1 billion in annualized ad revenue; results compare to AppLovin in mobile advertising

• $TDOC: Evolution of BetterHelp enrollment after ACA subsidies expire; adoption of AI-enabled clinical programs at Integrated Care; whether management will deliver on top end guidance in 2026

Final lessons learned

These four companies share a common story: extreme valuations at the moment the story overtook reality, and a subsequent brutal return to basics. For investors, the lesson is clear. No thematic thesis, no structural trend or pandemic premium justifies ignoring valuations at the point of purchase.

The common denominator for those who can survive is discipline. The ability to reduce costs, generate positive cash flow and refocus the business model on segments with real demand. Companies that manage this may never return to their historical highs, but they can offer compelling returns from today's levels. Firms that fail to do this will remain a deterrent from the investing textbooks.

The deal is Merck’s third sizeable transaction in roughly a year and part of a broader buying spree worth more than 29 billion dollars. The company has already moved for Verona Pharma in a 10 billion dollar bet on respiratory medicine, acquired rare‑disease player SpringWorks for 3.9 billion dollars and is now adding Terns, all with the same strategic goal: building a pipeline strong enough to backfill the sales that will fade as Keytruda’s exclusivity runs out.

What Merck is actually buying for 6.7 billion

The entire transaction is based on one asset. TERN-701 is an experimental drug in clinical trials for the treatment of chronic myeloid leukemia (CML), a blood and bone marrow cancer in which leukemia cells grow uncontrollably.

The mechanism of TERN-701 differs from existing treatments. It is a highly selective allosteric inhibitor of BCR::ABL1 in oral form, i.e. in tablets. Unlike the older generation of drugs, it targets a different site of the protein that drives leukaemia, and the data from the first phase of clinical testing are exceptionally promising. At the ASH conference in December 2025, Terns reported that 64% of pretreated patients achieved a major molecular response after 24 weeks. This result was described as "unprecedented" in the oncology community for a group of patients for whom other treatments had previously failed.

$MRK is now managing the CARDINAL study, adding a new cohort in January 2026 to test the 500 mg once-daily dosing. Initial dose selection is expected in mid-2026, followed by a key interaction with the FDA regarding conditions for approval.

Is 6.7 billion too much or too little?

The view of the net price of an acquisition varies depending on how you approach the risk. Terns had about $1.4 billion in cash on the books, so the effective price for the drug and research team alone is closer to $5.3 billion. The premium over the stock's last closing price was just 6%, a very modest premium compared to standard pharma acquisitions. The market took notice: shares of Terns jumped 5.5% after the announcement, not tens of percent as is usual with large buyouts.

The problem is that TERN-701 is still in Phase 1/2 clinical testing. Historical data shows that drugs in this phase successfully make it through the entire FDA approval process less than a third of the time. Merck has paid for this drug as a future blockbuster, with years of clinical data, regulatory approval and ultimately the actual commercial launch still to come. A direct competitor in the CML market is Novartis' Scemblix $NVS, which is already approved and has shown a 67% molecular response rate in clinical testing. The difference in efficacy is minimal, the competition will be fierce.

The key question: will this save Merck from the fall of Keytruda?

The answer itself is no. No single acquisition of Merck by $MRK is enough to offset Keytruda's revenue. Keytruda's primary patent expires in the U.S. in December 2028, and in Europe in 2031. Bloomberg Intelligence estimates that the global patent is more likely to be extended to 2033, which would bring Merck extra revenue of about $22 billion.

Therefore, Merck $MRK is betting on a combination: extend the life cycle of Keytruda through new formulations and combination treatment protocols, while building a diversified oncology pipeline through acquisitions. In February 2026, the company announced the separation of its oncology business into a separate division with its own leadership and strategy. Terns Pharma fits into this plan as the hematology leg of the portfolio.

Scenarios for investors

Bullish: TERN-701 successfully passes the CARDINAL trial, the FDA approves the drug in late 2028/early 2029, and Merck launches commercial sales just as Keytruda revenues begin to decline. CML may be a rare disease, but blockbuster drugs for rare cancer diagnoses can generate billions even with a narrow patient population. In such a scenario, 6.7 billion paid today in 2030 may look like a good price tag.

Base case: TERN-701 will pass clinical trials, but commercial launch is slower than expected due to direct competition with Novartis' Scemblix. The drug will plateau at revenues in the range of $1 billion to $2 billion per year by 2031, justifying the acquisition as a reasonable diversification but not a strategic breakthrough.

Bearish: CARDINAL trial fails to show sufficient benefit over current treatment, FDA requires extensive additional studies or conditional approval with limited indication. Direct battle with Scemblix loses and Merck writes off a substantial portion of the 6.7 billion. In the context of an overall $29 billion acquisition strategy, such an outcome would increase pressure on management and MRK stock.

🟢 What exactly is it?

TurboQuant is a language LLM model from Google that aims to solve a technical problem: huge demands on memory usage and speed.

When you communicate with an AI (e.g., via ChatGPT or Gemini), the model needs to "remember" the context of the entire conversation. This "memory" is stored in the so-called KV Cache.

🛑 But here's the problem: This memory is incredibly space-hungry. The longer your conversation (longer context), the more memory (VRAM) the graphics card needs.

🟢 How does it work?

Think of it as compressing the conversation data so intelligently that the model can still work with it, even though it takes up a fraction of the space.

👉 6x less memory: That means where you previously needed 60 GB of memory, you now only need 10 GB.

👉 8x higher speed: Because the data is smaller, the chip can process it much faster. So instant responses from the AI.

🟢 How else can Google's TurboQuant help us?

• AI directly on mobile: Thanks to this, you'll soon see top models running directly on your phone without needing the internet (Local AI Inference).

• Huge context: You'll be able to load an entire book or thousands of lines of code and the AI will "remember" them without running out of memory.

• Cheaper operation: For companies like Google this means operating AI will be much cheaper, which could lead to better free versions for users.

🚨 Memory card companies are under pressure today.

The market is reacting positively to signs that Beijing is finally cracking down on the debilitating price wars that are devastating the profitability of China's biggest tech companies in the express delivery sector. $BABA and $MPNGY have been major beneficiaries of this change in regulatory approach.

The food delivery sector as a major winner

The State Council's top antitrust authority will launch an investigation into competitive practices among delivery platforms under China's Anti-Monopoly Law through local inspections, interviews and surveys, according to the State Administration for Market Regulation. The authorities aim to assess monopoly risks, restore market order and promote fair competition.

According to a January 2026 Bloomberg report, regulators aim to crack down on practices that distort the real economy and promote destructive competition. A commission official noted that while the delivery platform industry has played a key role in stimulating consumption, expanding employment and encouraging innovation, significant problems have emerged in recent years, including excessive subsidies, price wars and traffic control issues.

Economic motivation for intervention

The entire food delivery industry, according to a column in the Economic Daily, has fallen into a "vicious cycle" of losing money just to get attention, which ultimately puts a strain on the broader consumption recovery. Price wars go directly against the central government's efforts to boost consumption.

Meituan posted its first loss in nearly three years, reflecting the effects of a three-way battle with Alibaba Group Holding and $JD.com in China's weak consumer market. The company posted an adjusted net loss of 16 billion yuan ($2.3 billion) for the third quarter.

The situation illustrates the paradox of China's economy: while the government seeks to boost domestic consumption, aggressive competition in key sectors is paradoxically dampening consumption through unsustainable price wars. According to Asia Society analysis, Beijing faces the challenge of striking a balance between encouraging innovation and avoiding destructive competition.

Impact on market shares and the future

Analyst firm Morningstar projects that Meituan's share of the express delivery market will fall to 55% of gross transaction volume by 2027 from 73% in 2024. Alibaba's share should expand to 40% from 21% over the same period, and JD.com' s share should rise slightly to 6%.

According to Catherine Lim, an analyst at Bloomberg Intelligence, "increased control over competition will boost industry margins" and "limit subsidy-based expansion and increase compliance costs for new entrants."

The regulatory crackdown signals a fundamental shift in the Chinese authorities' approach to the technology sector. Whereas they previously tolerated aggressive competition as an engine of innovation, they now prioritise market stability and sustainable business models over growth at any cost.

The broader context of the Chinese economy

The intervention in the food delivery sector fits into a broader effort by Chinese regulators to stabilise an economy beset by deflationary pressures and weak domestic demand. According to a March 2026 CNBC report, Chinese policymakers have set an inflation target of "around 2%" for 2026, the lowest level in more than two decades, in an effort to boost domestic demand and curb aggressive price wars hitting many sectors.

Steven Leung, executive director of UOB Kay Hian in Hong Kong, expects the tone of regulators in curbing intense competition to "get stronger and stronger until the market finally gets the full message".

This regulatory crackdown may represent a key turning point for China's technology sector, which has long been dominated by a "growth at any cost" philosophy. Investors are now watching to see whether similar interventions will hit other sectors affected by price-destroying competition.

Yesterday I opened a new position in Rolls‑Royce Holdings. I had this stock on my watchlist for some time with a planned entry at 1170 GBP, which has now been reached. I entered with a position size of 1% of the portfolio and have set a target price of 1380 GBP.

(As usual, I wanted to include the company’s interactive chart for illustration, but it’s not available on Bulios. I’m therefore only attaching a generated image.)

Rolls‑Royce is one of the leading global players in aviation and the defense industry. It is best known for the Trent engines, power systems, and marine propulsion solutions.

Main advantages of Rolls‑Royce:

Strong recovery in civil aviation - The number of flights and the utilization of wide‑body aircraft are increasing, which raises demand for Trent engine maintenance — and service is a key source of recurring, high‑margin revenue for the company.

Large order backlog - A backlog exceeding £100 billion provides very good visibility of future revenues for several years ahead.

Exposure to the defense industry and new technologies - The company benefits from rising defense spending and is developing projects such as small modular reactors, which could be a significant growth segment in the future.

Key risks:

Higher indebtedness - The company still carries debt from the pandemic period, which limits financial flexibility.

Execution risk - Large and technologically complex projects carry the risk of delays or higher costs.

Industry cyclicality - The aerospace sector is sensitive to the economic cycle — an economic slowdown or supply‑chain issues can significantly slow growth.

Comparison with competitors:

Compared with companies like GE Aerospace, Safran or Honeywell, Rolls‑Royce has a strong position especially in the wide‑body aircraft segment and a higher share of revenues from aftermarket services. GE has broader diversification and Safran a stronger balance sheet, but Rolls‑Royce currently offers very interesting growth potential thanks to the ongoing post‑pandemic engine maintenance cycle.

For me, this is a well‑timed entry into a quality company with long‑term potential.

What do you think? Are you adding aerospace to your portfolio, or do you prefer to avoid this sector?

The English version of this post is available on my profile on www.etoro.com. If you want to follow me there or possibly copy my USD portfolio, I’d be very happy!

For 2026, management guided 1.452–1.462 billion dollars in revenue, 18–19% growth, with non-GAAP operating income of 165–175 million dollars and a compressed margin of 11–12%, reflecting deliberate reinvestment in AI product development rather than further near-term margin expansion. The margin guidance came in roughly 45 million dollars below analyst consensus and sent the stock down more than 20% in February, creating the valuation entry point the article sets up, while the strategic reason for the spending is a March 2026 announcement that AI agents can now sign up for and operate the platform autonomously alongside human teams - a structural platform shift not yet reflected in 2026 revenue guidance but one that could meaningfully accelerate enterprise expansion if the agentic workflow thesis plays out.

Top points of analysis

• Revenues for 2025 reached $1.23 billion, up 27% year-over-year, with the company posting a positive net income for the first time in its history - $32 million versus a loss of $1.9 million in 2023 - while gross margin held at an exceptional 89%.

• The outlook for 2026 calls for revenues of $1.452 billion to $1.462 billion (18% to 19% growth), operating profit in the $165 million to $175 million range, and adjusted free cash flow of $275 million to $290 million, a margin of 19% to 20%.

• The number of customers with annual contracts over $100,000 grew 45% to more than 1,750, and customers with contracts over $50,000 already account for 41% of total revenue, up from 36% in the prior year - the company is shifting from small teams to large enterprise customers.

• In March 2026, the firm launched an infrastructure for autonomous artificial agents that act and make decisions on their own instead of just designing - a fundamental shift from a passive tool to an active digital worker.

• Analysts are modeling $2 billion in revenue by 2028 at an average annual growth rate of around 23%, with fair value according to these models estimated to be in the neighborhood of $135 per share.

• Valuation after the stock dropped significantly in early 2026 - the market was disappointed by the slowdown in growth rate from 33% to an expected 18% to 19%.

Company performance

monday.com $MNDY was founded in Tel Aviv in 2012 under the original name daPulse, renamed monday.com in 2017, as a visual project management and team collaboration tool. Founders Roy Mann and Eran Zinman were at the birth of the idea that enterprise software doesn't have to be complex and unfriendly, but can function intuitively like a consumer app - and built a platform on this idea that today serves over 245,000 customers in virtually every industry around the world. The company went public on NASDAQ in 2021, during one of the worst sell-offs in technology stocks, but since then has systematically improved profitability and expanded its product portfolio toward enterprise customers.

The company makes its money on subscriptions - customers pay a monthly or annual fee depending on the number of users and the plan they choose, with higher plans unlocking advanced features, automation, analytics and artificial intelligence. The business model is therefore inherently recurring and predictable, with a natural tendency for the average contract value to grow over time as customers add users and migrate to higher tariffs. A key indicator of the health of this model is the so-called net customer retention rate, which is 116% for large customers with annual contracts over $50k - meaning that this group of customers spends on average 16% more each year than the previous year, purely through organic expansion without the need for new customer acquisition.

CEO and management

monday.com is a rare example where the company is run by a pair of co-founders as CEOs - Roy Mann and Eran Zinman have shared the CEO role since the company's inception. Mann is primarily responsible for product vision and customer experience, while Zinman is responsible for technology development and engineering - a role split model that is unusual but clearly workable in the case of monday.com. Their approach is characterized by an emphasis on simplicity and accessibility of technology for business users without a technical background, and it is this philosophy that they are now bringing to the world of AI - instead of an isolated AI tool, they are building a platform where intelligence permeates the entire workflow.

Products - where monday.com differentiates itself from the competition

monday.com is not a single product, but a platform of four interconnected product areas, each targeting a specific business function.

Project and work management remains the historical core where the company has grown. It's all about visual dashboards, timelines, reports and automation for teams that need to keep track of who is doing what and by when. The key differentiator is flexibility - monday.com doesn't prescribe one way of working for the customer, but gives them the tools to build the workflow themselves, without the help of a developer.

monday CRM is a newer product focused on customer relationship management, business process and marketing campaigns. In September 2025, the company added monday Campaigns to the CRM, a tool that allows marketing teams to create, run and optimize campaigns directly tied to CRM data and business results - in practice, this means the marketer can see which campaigns led to closed deals and can build on that. Net customer retention rates in the CRM area are above the platform-wide average, according to management, confirming that customers who try CRM stay with it and expand.

Monday Service is the youngest product area focused on internal service teams - IT help desk, HR support, facilities management and similar functions. It's a direct entry into a territory now dominated by ServiceNow and Zendesk, and the company is building it with AI as a foundation, not an add-on. The customer gets a service platform where AI automatically categorizes, assigns and resolves routine requests without human intervention, and the entire platform is connected to other monday.com products, making it a natural choice for a company that already uses monday.com elsewhere.

monday Dev targets development teams and their specific needs - version control, bug tracking, scheduling, and integration with technical tools like GitHub or Jira. It's the most competitive market in the portfolio, with Atlassian being an established giant, but monday.com is betting that developers in companies where other teams already use monday.com will reach for a tool that integrates with their entire company rather than running an isolated system.

Customers of

The strongest number in the entire investment story is not the revenue growth rate, but the net customer retention rate of 116% for the segment with annual contracts over $50K. In practice, this means that even if the company doesn't gain a single new customer from this group as of today, its revenue from this segment next year would be 16% higher than this year - purely by existing customers adding users, moving to higher tariffs, or expanding monday.com to other departments.

This segment accounts for 41% of total revenue and is growing faster than the rest of the customer base. Customers with an annual contract over $100,000 increased 45% to over 1,750, with the average value of their contract increasing year-over-year. If the company maintains a retention rate of 116% through 2026, and this segment accounts for, say, 45% of its projected $1.45 billion in revenue - or approximately $650 million - then approximately $104 million of incremental revenue will come in 2027 from organic expansion of existing customers alone, without a single new acquisition. That's the foundation on which the company builds its outlook, and the argument for why the customer base is more valuable than a simple total customer number indicates.

A key risk of this model is that retention rates may deteriorate if customers begin to consolidate tools or switch to competitors with more integrated offerings. But so far, the data speaks the opposite - customers who cross the $50k/year threshold are deeply integrated into the company's processes and leaving would be costly and organizationally complicated, which naturally increases loyalty.

Margins in the context of competition

The gross margin of 89% is among the absolute top in the application software segment and in itself says that once a customer is acquired, it costs the minimum to serve them. For comparison with key competitors:

monday.com has a gross margin comparable to Asana $ASAN and significantly higher than Salesforce $CRM or ServiceNow $NOW, but it also shows where the margin lies: the operating margin is still token, while Atlassian $TEAM or ServiceNow have it at 20-24%. The difference is not that monday.com has a structurally worse business - it's that it's still aggressively reinvesting in product development, sales expansion, and now AI infrastructure. If it manages to approach an operating margin of 15% to 20% within three years, which the outlook for 2026 with margins of 11% to 12% suggests is a realistic direction, it will compare significantly more favourably with its competitors - and the market will likely start to appreciate this before those numbers fully materialise.

Artificial intelligence as a platform rebuild, not just an extra feature

What differentiates monday.com's approach to AI from many competitors is the intention to rebuild the entire platform around autonomous action, not just add a chat window or a text summarization button. In March 2026, the company launched an infrastructure for artificial agents that aren't just advice or suggestions - they're autonomous digital workers capable of taking action, making decisions and completing tasks within the platform without the need for human approval of every move.

The platform is built on three principles. Artificial building blocks are the building blocks from which a user with no technical background can assemble intelligent automation of their workflow - for example, automatic task assignment based on the content of an email or alerts when a risk is detected in a project. The AI-enhanced products then mean that each of the four products - project management, CRM, service and development - receive specific intelligent features to meet the specific needs of those teams. The digital agents are then the top layer, where customers can build their own agent for repetitive processes - for example, an agent that produces reports every Friday, highlights deviations from the plan and suggests specific actions without having to be asked.

Sidekick, the entry point into AI on the platform, came out of beta in January 2026 and has become the main interface for working with AI tools across the platform. The company also announced a new partner rewards program at the February 2026 Partner Summit focused on selling AI products and specializing in AI implementations, building a distribution network for this new revenue stream.

Growth potential - where, how and by how much

The biggest source of growth is moving customers up the value ladder - from small teams to entire companies. There are now over 1,750 customers with annual contracts over $100k and growing at 45% per year, while the overall customer base is growing more slowly. If this pace holds, the large customer group will account for the vast majority of revenue within two years, while delivering higher margins, lower churn rates and a more stable revenue stream.

The second growth driver is the expansion of the product portfolio. A customer who has started project management is a natural candidate for a CRM, service or developer module - and each such extension increases the annual contract value without the need for costly new customer acquisition. At the 2025 earnings presentation, management emphasized that CRM saw a record number of new customers in the fourth quarter and that monday Service, while the youngest product, is attracting interest specifically from customers who are already using monday.com in other departments.

The third driver is artificial intelligence as a source of additional revenue, not just a marketing argument. The company is gradually rolling out premium AI features on higher tariffs and autonomous agents as a separate paid tier. If customers prove willing to pay for a digital worker to replace some of their manual administrative work, it could open up a revenue stream that is not yet fully visible in today's numbers. For 2028, analysts model $2 billion in revenue at an average annual growth rate of around 23%, with success in the enterprise and AI layer being key.

Financial performance and numbers

Revenues in 2025 are $1.23 billion, up 27% year-over-year, compared to $972 million in 2024. Gross margins hold at an exceptional 89%, among the highest in the entire application software segment, reflecting the fact that the subscriber software model has minimal incremental costs once a certain level of revenue is reached. Operating profit on an adjusted basis has increased significantly, and for 2026 the company expects operating profit of $165 million to $175 million, a margin of around 11% to 12%.

Net income for 2025 was $32 million, with 2024 ending with a token $1.9 million profit and 2023 still a loss of over $136 million. The transition from four years of losses to positive profits is a significant milestone, although the absolute numbers are still modest compared to sales. EPS for 2024 was $0.65 and is estimated to be significantly higher for 2025, while the outlook for 2026 with unadjusted numbers will depend on the pace of reinvestment in AI and corporate expansion.

Adjusted free cash flow for 2026 is estimated at $275 million to $290 million, a margin of 19% to 20%. This is a crucial number because it shows that the company is not just generating profit on paper through accounting adjustments, but is generating real cash that it can use to invest in products, share buybacks or acquisitions.

Balance sheet and company health

monday.com is in a comfortable position in terms of financial stability. Working capital exceeds $1.2 billion, a cash ratio of 2.19 indicates that the firm holds significantly more cash than current liabilities, and an Altman Z-score of over 3.7 is in the safe zone. The debt to equity ratio of 0.09 and minimal long-term debt say that the firm is not financed with external capital to a greater extent and has no structural financial stress.

This strong liquidity position gives management free rein - the firm can invest in AI infrastructure, hire talent, or make smaller acquisitions of complementary technologies without having to tap the debt market or dilute shareholders. For the investor, this means that the risk of existential pressure or involuntary restructuring is minimal and the story will be driven purely by commercial performance, not financial issues.

Valuation and what's included

After the stock's significant drop in early 2026, monday.com trades on a P/E of approximately 60 times and a price to earnings of 3.3 times, with an enterprise value of $2.4 billion, well below its market capitalization of $3.9 billion due to its strong cash position. Analytical models working with a 2028 outlook and $2 billion in revenue estimate a fair value significantly higher at approximately $135 per share, a potential upside of over 80% from current levels.

The gap between these estimates says that valuation is entirely dependent on assumptions about growth rates and whether the company can sustain 18% to 23% revenue growth while improving operating margins. The stock's decline after the fourth-quarter 2025 results was a reaction to a slight slowdown in the growth rate from 33% in previous years to an outlook of 18% to 19% - a signal to some investors that the expansion phase was peaking and the company was entering a more mature, slower stage. On the other hand, the outlook for $275 million to $290 million of adjusted free cash flow puts the valuation to cash flow at just over 10 times, which is attractive for a sophisticated software platform with 89% gross margin and recurring revenue from enterprise customers.

Investment scenarios

In a positive scenario, the firm maintains a growth rate of around 20% beyond 2026, AI agents become a paid premium feature that significantly increases average contract value, and customers with annual contracts over $100k surpass 3,000 by the end of 2027. Revenues reach $2bn in 2028, operating margins approach 15%, and the market assigns the firm a multiple equivalent to a mature premium software platform. The upside to today's price is significant in such a world.

In the base case, the company grows 18% to 19% annually as the outlook says AI products are successful, but more as a customer retention tool than a massive new revenue stream. Revenues in 2027 approach $1.7 billion, free cash flow grows to $350 million, and valuation gradually declines from today's revaluation to levels consistent with a quality mid-sized software company. Investors will profit from improving fundamentals rather than a dramatic rerating effect.

A negative scenario occurs if growth rates slow further below 15%, competitors like Microsoft or Salesforce more aggressively attack the same customer segments with integrated offerings, and AI agents prove to be a difficult feature to monetize. In such an environment, today's valuation at 60 times earnings would be difficult to defend, and the stock could face further downward revaluation.

What to take away from the article

• monday.com reported positive net income for the first time in its history, and the outlook for 2026, with revenues of over $1.45 billion and adjusted free cash flow of around $280 million, shows that the company is transitioning from growth speculation to the category of profitable software companies with predictable revenue streams.

• A key indicator of the health of the business is a net customer retention rate of 116% for large enterprise customers - this means that existing customers are spending an average of 16% more each year, purely through expansion with no new acquisitions, and this is the basis for sustainable compound growth.

• The firm's AI strategy is not just a marketing story - it launched an infrastructure for autonomous digital agents in March 2026, and the firm is building a new premium revenue stream behind it, the true size of which will become apparent in the 2026 and 2027 numbers.

• The valuation is less extreme after the stock's early 2026 plunge, but still assumes continued strong growth - at a price to free cash flow of around 10 times, the firm looks cheaper than at a P/E of 60 times, and it is this metric that is relevant for a software platform with recurring revenue.

• The biggest risk is not financial - the balance sheet is strong and debt is minimal - but competitive: Microsoft, Salesforce and Atlassian are playing on the same playing field with much deeper pockets and integrated ecosystems, and this is a factor that investors need to weigh.

• For an investor with a two- to three-year horizon who believes in the continued shift of enterprise work to cloud platforms and the monetization of AI in enterprise applications, monday.com may be an interesting position after the recent slump - but it requires monitoring the growth rate of enterprise customers and the first clear signals of the success of an AI premium revenue stream.

Memory chips are among the less visible, yet absolutely key building blocks of modern AI infrastructure. Every GPU accelerator, every server rack in a data center, and every large language model training requires massive amounts of fast, high-capacity memory. This is where DRAM and NAND chip manufacturers come into play, whose products today determine how fast and how efficiently AI systems can process data.

The year 2025 was a watershed year for the memory sector. After the protracted overcapacity of 2022 and 2023, which pushed memory chip prices well below the cost of production, came a structural turnaround. Huge investments in AI infrastructure by hyperscalers like Microsoft $MSFT, Amazon $AMZN, and Google $GOOG have created demand for a special category of memory called high-bandwidth memory, or HBM. These chips, which are mounted directly next to GPU accelerators using 3D stacking technology, allow data to be transferred extremely quickly and have become a critical bottleneck in the entire AI computing chain.

Analysts estimate that the HBM memory market will reach approximately $35 to $38 billion in 2025, and projections for 2026 call for a further jump to approximately $55 to $58 billion. By 2028, the market could exceed $100 billion, which would be larger than the entire DRAM market in 2024. The dramatic change in the situation is evidenced by the fact that HBM capacity is sold out for all key manufacturers for the entire year 2026. The environment in which manufacturers find themselves can be described as a supercycle, a period of structurally higher demand that goes beyond the traditional sector cycle. Let's take a look at the four companies that are benefiting most from this trend and that will shape this strategic sector in the future.

Micron Technology $MU

From commodity producer to tier one player

Micron Technology is an American memory chip manufacturer headquartered in Boise, Idaho, and the only major US company in the DRAM and NAND segment. This in itself makes it a strategically important entity at a time when geopolitical tensions are increasing government interest in diversifying supply chains away from Asian manufacturers. Over the past two years, Micron has transformed itself from a highly cyclically volatile commodity firm to a key supplier of AI infrastructure.

Results for its fiscal second quarter, which ended in February 2026, beat even the most optimistic analyst estimates. Revenue reached $23.86 billion, up 196% year-over-year. Gross margin on a non-GAAP basis jumped to a record 74.9%, up from 36.8% in the same period a year ago. Earnings per share of $12.20 significantly beat consensus. A key factor was the announcement that all of HBM's capacity is sold out by the end of 2026. For more on the company's results and outlook, check out the Flash News on Bulios.

HBM4 and the technology edge

Micron entered 2026 in the strongest competitive position in its history. The company has moved into high-volume production of HBM4, the latest generation of high-bandwidth memory, and has begun shipping samples with transfer rates of up to 11 Gbps. HBM3E, which is the basis for Nvidia's Blackwell series accelerators, delivers 1.2 terabytes per second of throughput while consuming 30% less power than the previous generation. The company also launched its Cloud Memory Business Unit, which aims to provide hyperscalers and cloud providers with tailored HBM solutions, shifting its model from a pure component supplier to an integrated solution provider. Capital expenditures for fiscal 2026 have been raised to $20 billion.

Key metrics

Industry cyclicality remains a risk. If the pace of AI investment slows, or if producers begin to aggressively add capacity, pricing dynamics may normalize faster than the market expects. Micron $MU has gone through very deep dips in profitability in the past, and again, the level of today's margins reflects an exceptional supply and demand side situation. The last time we saw this weakness was in 2024 when $MU stock fell as much as 60% from its peak.

SanDisk Corporation $SNDK

A spin-off with perfect timing

SanDisk went public as a separate company in February 2025 after parent Western Digital $WDC completed the spin-off of its NAND flash memory division. The timing of this move was extremely fortunate from an investor perspective. The flash memory market was in the midst of a strong recovery from a protracted overcapacity, and a global shortage of NAND chips for AI infrastructure had caused prices to rise dramatically. SanDisk thus went public just as its segment was beginning to experience one of the strongest booms in history.

The company is focusing on three key segments:

• enterprise storage for data centers

• client SSDs for PCs and tablets

• consumer flash memory.

For data centers, SanDisk offers PCIe Gen5 drives and is developing the BiCS8 QLC Stargate solution, which combines higher capacity with lower manufacturing costs. More about the product portfolio can be found on the company's official website.

Rocketing growth and fundamental issues

In the second fiscal quarter of 2026, which ended in early January, SanDisk reported revenues in excess of $3 billion, a 61% year-over-year increase. Revenue from the data center segment grew 64%, driven by interest from AI infrastructure builders, semi-custom customers and technology companies deploying AI at scale. Earnings per share of $6.20 were five times the previous year's result. The outlook for the third fiscal quarter calls for revenue of $4.40 billion to $4.80 billion, which would represent an extraordinary jump.

SanDisk shares have risen more than 1,500% since going public, adding approximately 147% this year alone. Valuation at approximately 4.7 times sales is above-par for a cyclical memory business and reflects the market's belief that the NAND flash market has transformed from an end-user commodity to a strategic asset in the AI era. A key structural catalyst is also the partnership with Japan's Kioxia in the form of a shared manufacturing venture, which differentiates SanDisk from direct competitors by sharing manufacturing costs while maintaining its own product portfolio.

Key Performance Indicators

The risk is the above-trend valuation for a cyclical business, the dependence on continued pricing power in the NAND segment, and the fact that SanDisk does not manufacture HBM compared to Micron, SK Hynix or Samsung, and thus does not benefit from the very hottest part of the AI memory boom. Still, it's a key player with a clear data center transition strategy is a very interesting bet on the continued digitization of infrastructure.

Samsung Electronics $SSUN.F

The giant with the broadest global reach

Samsung Electronics is a unique entity in the context of the storage industry. It is a company that combines leadership in DRAM and NAND manufacturing with extensive diversification into consumer electronics, displays, smartphones and contract chip manufacturing (Foundry). On the one hand, this diversification is a source of stability, while on the other hand it causes memory results to be diluted to some extent by other segments in the overall corporate numbers. More on the company's structure can be found at Samsung Global Newsroom.

The year 2025 was a different year for Samsung in the memory segment. While competition in the form of SK Hynix benefited from its dominant position in HBM right from the start of the AI boom, Samsung struggled as its HBM chips failed to pass qualification tests at Nvidia $NVDA, a key customer. The result has been a drop in HBM market share from around 41% in Q2 2024 to around 17% in Q2 2025. However, this situation has brought about a vigorous management response.

Large turnaround in the second half of 2025

In the third quarter of 2025, Samsung announced that its HBM3E chips had successfully passed qualification tests at Nvidia and shipments for mass production had begun. Sales of HBM3E, and SSDs led to record quarterly sales of 26.7 trillion Korean won for the memory division. In the fourth quarter of 2025, this record was broken again: the DS division achieved consolidated sales of 44 trillion won, or approximately $30.8 billion, with year-on-year growth of more than 46%. The DS division's operating profit for the fourth quarter reached 16.4 trillion won, and the operating profit of the entire company for Q4 2025 reached 20 trillion won, tripling the 2024 figure.

In addition, Samsung confirmed that pre-sales of its HBM4 capacity for 2026 are completely exhausted, and started shipping HBM4 samples with a transmission speed of 11.7 Gbps to key customers. Analysts at banks such as Goldman Sachs $GS and Bank of America $BAC predict that Samsung's operating profit for 2026 could surpass 100 trillion Korean won, more than double the 2025 level, assuming memory prices remain high.

Key metrics

Fragmentation of the business remains a structural risk. Improvement in the memory segment may be dampened by weaker performance in Foundry or DX (consumer electronics), especially in a price competitive environment in the smartphone market. The technology gap with SK Hynix in HBM persists and Samsung will have to prove it can close it in the HBM4 generation.

SK Hynix $HY9H.F

AI memory market leader

SK Hynix is a South Korean memory chip manufacturer and currently the clear dominant player in the HBM memory market. The company started focusing on this segment before its competitors, building capacity and technological know-how at a time when other manufacturers still preferred conventional DRAM. This strategic foresight has paid off in measurable ways.

According to Counterpoint Research data, SK Hynix held a 62% share of the HBM market in Q2 2025, and NVIDIA is its key customer, taking approximately 90% of production. The firm then posted a record operating profit of 11.4 trillion won in the third quarter of 2025, up 62% year-on-year. For the full year 2025, SK Hynix surpassed Samsung in operating profit for the first time ever, with total annual operating profit reaching 47.2 trillion won. Bank of America $BAC named SK Hynix as the Top Pick of the entire memory industry for 2026, and Goldman Sachs confirmed that the company will maintain its dominance in HBM3 and HBM3E at least through 2026 with a total share exceeding 50%.

HBM4 and retention of leadership

SK Hynix was the first manufacturer to complete HBM4 development and announced a 40% improvement in power efficiency at 10 Gbps transfer rates. Series production is expected to follow immediately after completion of customer qualification. The company also announced an increase in infrastructure investment for 2026 by more than four times the amount originally planned, and is building a new M15X plant expected to be operational in mid-2027. The combination of technology leadership, capacity sold for the full year 2026, and the structural advantage of an exclusive relationship with Nvidia puts the company in an extremely strong position.

SK Hynix also confirmed that customers are starting to place pre-orders not only for HBM but also for standard DRAM and NAND products for 2026, as capacity constraints on the HBM side of production are reducing available production for traditional memory types as well.

Key indicators

High dependence on Nvidia as a key customer is a risk. If Nvidia diversifies suppliers or if its growth rate slows, SK Hynix would feel the impact very quickly. Another risk is geopolitical exposure as a South Korean manufacturer dependent on Asian supply chains in an environment of heightened tensions.

Comparison of players and market dynamics

The four firms occupy distinct market positions, differing in both technology profile and exposure to different types of memory. SK Hynix is the clear leader of the HBM segment with a dominant share and the most trusted relationship with Nvidia. Samsung is the largest firm in terms of total memory sales, but loses points for its slower entry into the HBM generation and fragmentation of the business. Micron is the US player with the most aggressive margin growth and strategic position as the only US alternative to Asian dominance. SanDisk is a pure-play NAND company with no HBM exposure, betting that flash storage for AI data centers is experiencing its own structural boom.

All four companies share a common trend: HBM capacity is sold out for all of 2026 and demand from hyperscalers and AI chipmakers remains extremely strong. Bank of America talks of a supercycle similar to the boom of the 1990s and forecasts global DRAM and NAND sales growth of 51% and 45% year-on-year, respectively, in 2026. Moreover, Counterpoint Research and Goldman Sachs warn that the structural shortage of HBM capacity is likely to persist through 2027.

A strategic view

The memory sector has gone through repeated boom and bust cycles over the past two decades, and investors who have experienced these cycles are justifiably cautious. But today's supercycle has one key peculiarity: demand is being driven not by consumer products like PCs or smartphones, which are highly sensitive to the economic cycle, but by hyperscalers' investments in AI infrastructure. These investments are multi-year strategic decisions with huge barriers to backward movement.

Each of the four companies has a different risk profile. Of the four firms, SK Hynix has the most direct exposure to the HBM boom and the lowest valuation relative to 2026 earnings estimates. For investors looking for a direct bet on the AI memory cycle with a relatively compressed valuation, SK Hynix is therefore the cleanest play. The downside is limited accessibility for Western-oriented investors, as the stock trades primarily on the Korean exchange, while there are ADR certificates tradable in foreign markets.

Micron is the US alternative with a growing market share in HBM and the strongest absolute margins. After a series of record quarters, the market has repriced the company significantly higher, but forward valuations remain conservative if the supercycle continues. Samsung offers a combination of memory exposure with diversification into other technology segments, which reduces volatility but also dilutes the net gain from the HBM boom.

SanDisk is most interesting from an investment story perspective as a turnaround and spinoff story combined with the NAND supercycle, but is also the highest valuation firm with no HBM exposure. Investors should expect that any slowdown in AI investment or normalization of NAND pricing would have a very large and rapid impact on the stock.

What to watch next

• The pace of hyperscaler (Microsoft, Amazon, Google, Meta) investment in AI infrastructure in 2026 is the most important macro indicator for the sector.

• Samsung's qualification for HBM4 with Nvidia and other customers is a key catalyst for reassessing Samsung's HBM market share.

• SK Hynix's (M15X from 2027) and Micron's (Singapore, US$2.5bn investment) production capacity ramp-ups may impact market supply-side and price dynamics.

• The transition to HBM4 and HBM4E changes winners and losers in each generation. The companies that reach high volume production first will lock in a premium.

• Pricing trends for traditional DRAM and NAND outside the HBM segment will be key to the results of Samsung and SanDisk, whose mix is more heavily dependent on traditional memory types.

• Geopolitical factors, particularly US-China relations and their impact on semiconductor export controls, may affect supply chains and redistribute market shares.

• Edge AI and consumer electronics as a new wave of memory demand in the AI-PC and AI-smartphone segment in 2026 and 2027.

Summary

The memory sector has entered a structurally new cycle that differs from previous booms in one key characteristic: demand is being driven by AI infrastructure investments that have multi-year horizons and huge barriers to backward movement. SK Hynix, Micron, Samsung and SanDisk form an oligopoly that controls this market, whose individual members compete in a technology race for dominance in the HBM generation that directly determines whether their products end up in the most prestigious AI servers or are pushed into less profitable segments.

At the same time, it's important to remember that every supercycle will eventually end or at least transform into a more normal environment. Manufacturing capacity is ramped up, new plants are built, and historically the memory industry tends to overshoot on both sides of the cycle.

While Wall Street is selling MU after record results, the geopolitical crisis in the Persian Gulf is quietly rewriting the rules of the memory market for the next 5 years.

Ras Laffan – The Flash That Changed Everything

On March 2, 2026, Iranian drones and missiles struck the Ras Laffan industrial complex in Qatar – the world’s largest LNG terminal. Flashes from the explosions were visible from Doha. Within 48 hours it became clear that the world had lost 33% of global helium production.

Most investors heard the word "helium" and moved on to the next headline. Mistake. Helium isn’t just for birthday balloons. It is an irreplaceable component in the manufacture of every modern chip – without helium EUV lithography stops, and without EUV there is no Nvidia Blackwell, HBM or DDR5.

QatarEnergy CEO Saad Al‑Kaabi said it directly on March 19: “Extensive damage to production facilities will require up to five years of repairs and force us to declare a long‑term force majeure.”

Five years. Even if the war ended tomorrow.

Why the Repair Takes So Long

This is not about politics or willpower. It’s about physics and industrial reality.

LNG trains 4 and 6 in Ras Laffan – the heart of Qatari helium production – were destroyed by direct hits. Their original construction cost $26 billion and took 7 years. Key components – cryogenic heat exchangers and turbines – are made worldwide by only 2–3 companies (Air Products, Linde, Chart Industries) with lead times of 18–24 months. Helium recovery units are physically integrated into LNG trains – without a functional train there is no helium as a by‑product.

Realistic recovery timeline, even with immediate peace:

Phase — Time horizonDiagnostics and repair planning 3–6 monthsDelivery of critical components 12–24 monthsConstruction reconstruction 24–48 monthsTesting and commissioning 6–12 monthsFull capacity 2029–2031

The global helium deficit for 2026–2030 is therefore structural, not temporary. The world market lost 30% of supply and it won’t return in two or three years.

$SSNLF SAMSUNG and $HY9H.F SK Hynix: A Ticking Time Bomb

Now to the crux — why this crisis will hit the Korean competitors and Micron very differently.

SK Hynix, the current leader in the HBM market with a 62% share, sources 64.7% of all its helium from Qatar. Samsung is similar – dependence on Ras Laffan exceeds 60%. Both companies have stockpiles for roughly 6 months – i.e., about until September 2026.

SK Hynix did implement its own HeRS recycling system, but it captures only 18.6% of consumption. The rest must come from elsewhere – and spot supply elsewhere currently costs roughly double the normal price, with analysts forecasting a rise to $2,000/MCF (from pre‑conflict $500).

Moreover: shipping liquefied helium from the U.S. to Korea by tanker is not logistically trivial. Liquid helium evaporates – trans‑Pacific transport implies enormous losses and requires specialized cryogenic tankers with limited capacity. Samsung can buy capacity from the U.S., but at a price that will dramatically compress margins.

Timeline of the Korean crisis:

September 2026: SK Hynix and Samsung inventories critically low

October–December 2026: Forced production cuts in DRAM and HBM of 15–35%

2027–2028: Structurally constrained capacity, premium spot helium prices

Micron: A Natural Advantage Built into the Map

Micron Technology’s main fabs are in Boise, Idaho – in the heart of the U.S., the world’s largest helium producer (43% of global production, fully operational).

While the Korean competition built dependence on cheap Qatari helium, Micron quietly built a supply chain based on domestic U.S. production. Air Products and Linde have long‑term contracts with Micron as an established customer – Samsung, which will show up as an emergency spot buyer, will get whatever remains at triple the price.

But the most important number isn’t in the supply chain — it’s in recycling. Micron’s fabs recycle 80–90% of all helium. Effective external spot consumption is therefore 5–10× lower than the Korean competition. Micron simply doesn’t need as much helium for the shortage to threaten it.

An Investment Asymmetry the Market Ignores

After MU’s record Q2 results (revenue $23.86 billion, EPS $12.20, guidance Q3 $33.5 billion) the market sold the stock. Reasons? Higher capex, worries about a cyclical peak, macro uncertainty.

Yet the fundamental picture is exactly the opposite:

Micron at ~$400 trades at 5.3× forward P/E (FY27 EPS consensus $76–80, Barclays estimates $100+). That is less than half the historical average and roughly a third of Nvidia’s valuation.

The helium crisis doesn’t arrive as a risk for Micron — it arrives as a market‑share catalyst:

SK Hynix must cut production from September 2026 → 12–22% of the HBM market is missing

Nvidia, ~90% dependent on SK Hynix, has no option but to shift orders to Micron

Micron already has a 5‑year SCA contract with Nvidia – emergency volume expansion = premium pricing

MU’s U.S. manufacturing capacity is fully operational, without restrictions, unaffected by the helium crisis

Conclusion: A Quiet Revolution in the Memory Market

Ras Laffan was not just an industrial catastrophe. It was a tectonic shift in the balance of power of the memory industry – and it is unlikely to be fully reversed by the time Micron brings HBM4 into mass production, signs new SCA contracts, and possibly reaches a 25–30% share of the HBM market instead of today’s 21%.

The market doesn’t see this yet. Fitch, BNP Paribas and TrendForce see it. QatarEnergy’s CEO confirmed it. Repairs will take 3–5 years even with immediate peace.

Micron at $400 with a forward P/E of ~5× and a structural competitive advantage over the next 5 years is a story worth the attention of any investor with a medium‑term horizon.

The financial ambition behind the launch is significant. CEO Rene Haas told investors ARM targets 25 billion dollars in total annual revenue and CFO Jason Child confirmed the AGI CPU carries a roughly 50 percent gross margin, with analysts projecting the chip product line alone could contribute 15 billion dollars in annual revenue by 2031 as Haas forecasts a fourfold increase in CPU demand driven by agentic AI workloads requiring continuous inference rather than periodic query responses. The move is a direct bet that entering hardware will expand Arm's addressable market to customers who never engaged with its IP licensing model, though it also places the company in direct competition with longtime licensees like Nvidia, Qualcomm and AMD, who attended the launch and offered endorsements while knowing they now share a customer pool with their foundational supplier.

What AGI CPUs can really do

The chip is built on Neoverse V3 cores and offers up to 136 compute cores per processor. Arm claims performance over 2x higher per rack compared to classic x86 CPUs from Intel $INTC or AMD $AMD, and with significantly lower power consumption. A single air-cooled rack holds up to 64 AGI CPUs, for a total of over 8,700 cores - a compact configuration that targets power-constrained data centers.

The economic argument is also important: Mohamed Awad, head of Arm's cloud AI business, has calculated that deploying AGI CPUs can save up to $10 billion in building a single large AI datacenter at a cost of around $50 billion. At a time when hyperscalers like Meta $META, Microsoft $MSFT or Amazon $AMZN are planning AI infrastructure investments in the hundreds of billions, this is an argument that executives are hearing.

Chip production is entrusted to TSMC $TSM on a 3nm process - the same technology that underpins Apple's most advanced chips $AAPL or Nvidia's $NVDA. Arm has already received test samples that work as expected, and mass production is planned for the second half of 2026.

Meta as the number one partner

Arm is not alone behind the development of the AGI CPU. Meta Platforms has become a major partner and co-developed the chip directly with Arm. The goal is specific: Meta wants to deploy the AGI CPU alongside its own MTIA chips to more efficiently manage the large-scale AI systems powering Instagram, Facebook, WhatsApp or Llamas.

But the customer base for the new chip goes beyond a single company. Early customers include OpenAI, Cloudflare, SAP and SK Telecom, with Arm in talks with other hyperscalers. On the server manufacturer side, Arm is working with Lenovo, Quanta Computer and Supermicro, which are preparing complete server systems based on AGI CPUs.

A breakthrough in a business that has been around for 35 years

The move from a pure licensing model to selling its own chips is a strategic earthquake for Arm. Until now, the company has lived off royalties, a percentage of each chip sold by customers. That model delivered steady revenue without manufacturing risk, but it also limited maximum growth potential - Arm earned a fraction of the value its technology generated in chips from Apple, Nvidia or Qualcomm.

The new model brings higher margins, but also higher risk. For the first time, Arm is in direct competition with its own customers. Intel, AMD and, in a sense, Nvidia are now competitors of the company whose design instructions they also license. Haas doesn't hide this paradox - he talks about it openly and says the market is big enough for everyone.

The numbers and outlook for investors

Arm has consistently beaten analysts' estimatesin recent quarters. In the third quarter of fiscal year 2026, it reported revenue of $1.24 billion, up 26% year-over-year, and it was the fourth consecutive quarter with billions in revenue. For the full fiscal year, Wall Street expects net income of $1.75 per share on revenue of $4.91 billion.

AGI CPU hasn't yet fully factored the potential of the new segment into those estimates. Arm says the custom chips will add "billions of dollars" to annual revenue - a specific number is not yet in question, as mass shipments don't start until the second half of 2026. The next chip in the new family is expected to be 12 to 18 months out, and Arm plans a regular iteration cycle similar to Nvidia's with its GPU architectures.

For investors, the AGI CPU signals a structural change in Arm's valuation story. The company is moving from being a mere licensing company with a limited revenue ceiling to becoming a direct player in the most lucrative AI hardware segment. The question remains how key licensing customers - Apple, Qualcomm or Nvidia - will react to the fact that their supplier has turned into a competitor.

Scenarios and projections: where the numbers are heading

The basic Wall Street consensus is working with revenues of $4.91 billion for the current fiscal year and $6 billion in 2027. By 2030, analysts are projecting revenues of around $12 billion on average, with more optimistic models working with a figure of over $15 billion. AGI CPU does not yet fully incorporate the potential of the new segment into these estimates - mass shipments only start in the second half of 2026, and the first full fiscal year with a chip on sale will not be until 2027.

The key structural argument comes from data centers. Arm has more than doubled revenue from this segment year-over-year in recent quarters, and analysts estimate that Arm-based processors could power up to 50% of new CPU installations at the largest hyperscalers by the end of 2026. Server CPU royalty revenues are growing at a rate of over 76% annually and are expected to reach approximately $4 billion from this segment alone by fiscal 2031, according to analyst estimates.

Net income is growing even faster than revenues. From $792 million in 2025, analysts model a jump to $1.8 billion in 2026 and $2.3 billion in 2027, while EBITDA is expected to approach $6 billion by 2030, which would put Arm among the most profitable semiconductor companies in the world.

A consensus of 27 analyst houses puts the 12-month target price at $156, with the highest bull case going as high as $170. But Wells Fargo warned in February 2026 that the consensus estimate of royalty earnings growing 27% annually may be too optimistic - and this is where AGI CPU enters the picture as a variable that skews existing models in both directions. If the chip sees rapid adoption by hyperscalers, estimates will have to be rewritten upwards. If customers react to Arm's direct competition by limiting licensing, the equation may reverse.

Three scenarios for the next 18 months look like this:

Bullish: Meta, OpenAI and other AGI CPU customers deploy quickly, Arm reports first billion-dollar revenue from direct chip sales in H1 2027, royalty revenue simultaneously grows due to dominance in datacenters, and the stock approaches over $180.

Base case: AGI CPU deployments are gradual, with the first significant revenue contribution coming in FY 2028, revenues arriving at consensus around $6 billion in 2027, and the stock trading in the $140-160 range.

Bearish: Key customers like Qualcomm or Nvidia react to Arm's entry into direct chip sales by limiting investment in new licenses, transition costs for AGI CPUs are higher than expected, and market adoption remains slow - sales disappoint consensus and the stock tests levels below $120.

Is anyone following $GRAB?

After days of searching for a bottom, it looks like the company might finally be reversing the trend. On paper this company has a good chance to grow, but we’ll see what reality brings.

Do you have this stock in your portfolio as well? Did you buy more in the last few days (during the biggest drop)?

The first name on the risk factors list is Microsoft, the very partner that made OpenAI what it is. Microsoft has invested 13 billion dollars since 2019, holds a 27 percent stake in the for-profit entity valued at roughly 135 billion dollars, and supplies "a substantial portion of our financing and compute," according to the document, which explicitly warns that if Microsoft modifies or terminates its commercial partnership, OpenAI's business, prospects and financial condition could be "adversely affected." Beyond the Microsoft dependency, the document flags a projected 14 billion dollar net loss in 2026 driven by infrastructure buildout, model training and research hiring, legal exposure from three lawsuits by Elon Musk and xAI with a first hearing in April, at least 14 user class actions in the U.S., concentration risk at TSMC for chip supply and the unusual governance structure as a public benefit corporation overseen by the OpenAI Foundation.

Seven years, 13 billion and a complicated relationship

Microsoft $MSFT entered OpenAI in 2019 with its first billion dollars, back when it was still a non-profit research project. Today, the total investment exceeds $13 billion, and after the October 2025 deal, Microsoft holds a 27% stake in the new for-profit OpenAI entity, OpenAI Group PBC, worth roughly $135 billion.

This interdependence takes a concrete form: Microsoft provides OpenAI' s Azure cloud infrastructure, secures the rights to distribute APIs to third parties, and de facto determines what hardware GPT-4, o1 or their successors run on. As an added bonus, OpenAI has committed to purchase $250 billion worth of additional Azure services as part of the renegotiation of the October 2025 terms. In other words: dependency has deepened, not weakened.

OpenAI's prospectus document states in black and white - its operating results depend on its ability to build partner relationships outside of Microsoft. And it warns that if the tech giant were to reduce or end its collaboration, the impact would be substantial for the company. For investors considering an IPO, it's a signal that can't be overlooked.

110 billion and still not enough

In February 2026, OpenAI closed the largest private funding round in the history of the tech industry to that point. The total value of the round reached $110 billion at a valuation of $730 billion. Major investors:

• Amazon $AMZN - $50 billion

• SoftBank $SFTBY - $30 billion

• Nvidia $NVDA - $30 billion

Still, OpenAI is working hard to secure an additional $10 billion from a wider range of investors, including venture capital and sovereign funds. The reason is simple: an internal document estimates total spending on computing infrastructure by 2030 at a staggering $665 billion. So even a record funding round is not enough to cover the full cost of maintaining technological leadership.

TSMC $TSM : the risk no one wants to say out loud

Alongside Microsoft, OpenAI names a second major systemic risk in the document: the Taiwanese chipmaker TSMC. TSMC makes the vast majority of the world's most advanced AI chips, including those from Nvidia that run OpenAI's models. Any disruption to production caused by the conflict in the Taiwan Strait would directly impact the company's ability to scale its models.

In doing so, OpenAI acknowledges that its infrastructure sovereignty has clear geographic limits. At a time when geopolitical tensions around Taiwan remain one of the key systemic risks to the entire technology sector, this mention in the IPO document is more important than it first appears.

Elon Musk and the legal front

The internal document also mentions ongoing litigation as an operational risk. Elon Musk, the co-founder of OpenAI who left the organization and then publicly attacked it through his AI firm xAI, has led a series of lawsuits challenging the legitimacy of OpenAI's transformation into a for-profit company. Litigation burdens management and can affect the brand's pre-IPO reputation - just when OpenAI needs to look stable and credible.

In addition, OpenAI operates as a so-called public benefit corporation under the oversight of a nonprofit foundation, which brings specific regulatory constraints that prospective shareholders must consider.

What the numbers say

Despite all the risks, the business results are impressive. Revenues for 2025 have reached $13.1 billion, the platform has over 900 million weekly active users, and ChatGPT has transformed from a technological curiosity to a global work tool in less than three years. Meanwhile, annualized revenue is heading towards $20 billion by the end of 2026.

At the same time, however, the cost of developing and running the models is growing faster than revenue. OpenAI does not anticipate profitability until 2030. Record funding rounds are therefore not a signal of strength in the traditional sense - they are a reflection of the enormous capital intensity of the AI frontier as an industry.

What IPOs mean for the market

OpenAI plans to file documents with regulators in the second half of 2026, with the IPO itself expected in the fourth quarter. If the valuation reaches one trillion dollars, it will be one of the largest stock market debuts in history - comparable to the IPOs of Saudi Aramco or Alibaba.

For investors, this opens up direct exposure to an AI industry leader that has so far only been available to select institutional players. But the risk-factor document provides a sobering reminder: OpenAI is not just a ChatGPT growth story. It's a story about a company that depends on one technology partner, one chip manufacturing site, and a constant flow of outside capital. Those who want to enter the IPO will have to accept these risks as part of a bet on the most ambitious technology project of the 21st century.

The investment case hinges on three unresolved questions. First, the U.S. business: management guided North America revenue down 1–3% for full year 2026 and mid-single digits in Q1, with the recovery path tied to exiting the markdown cycle and launching new product innovation like the ShowZero fabric technology, but without a permanent CEO the execution risk is real. Second, the CEO search: interim co-CEOs Meghan Frank and André Maestrini are running the business while the board searches for a permanent chief, with founder Chip Wilson running a parallel proxy fight criticising the board's strategic direction, creating leadership uncertainty at exactly the wrong moment. Third, the valuation: with 1.8 billion dollars in cash, a 1.2 billion dollar buyback authorisation, 42-plus percent gross margins even under pressure and China growing at 20 percent in 2026, the current multiple prices in a structurally impaired brand, but if the U.S. stabilises and a strong CEO arrives, the rerating potential is significant.

Top points of analysis

• Q4 FY2025 (through Feb 2026): revenue $3.64B (+1% YoY), EPS $5.01 (beat consensus by 4.6%), but net income down 18.4% YoY due to margin.

• Full year FY2025: revenue $11.1B (+5% YoY), diluted EPS $13.26 (-9% YoY) - the first EPS decline in the firm's history as a publicly traded company.

• Outlook for 2026: revenue $11.35-11.5 billion (+2-4% total), but North America is expected to decline 1-3%, international markets are expected to grow +20% or more.

• CEO Calvin McDonald left Jan. 31, 2026; company is looking for a successor; interim leadership includes CFO Meghan Frank and CCO André Maestrini; founder Chip Wilson is pushing for strategic changes behind the scenes.

• Valuation at a decade low: trailing P/E ~12×, forward P/E on EPS guidance of $12.10-12.30 (~13×), while the firm's historical valuation in "normal times" was 30-40×.

• China pulls as only clear growth driver: comparable sales +11.5% in FY2025, plan to reach 200 stores in mainland China as second largest market.

• Gross margin fell 550 bps (basis points) YoY in Q4 FY2025 due to significant markdowns, expect another 250 bps compression in operating margin in FY2026 due to tariffs on Asian imports.

• Analyst expectations: 61 analysts, median price target $205, range $170-295, preponderance of "Hold" ratings (30), only 3 Buy, 0 Sell; stock around $160 implies consensus upside of +28%.

What has changed

lululemon $LULU has built one of the rarest business models in retail: a premium athleisure brand with minimal discounts, high customer loyalty, gross profit margins of over 55-58%, and the ability to open stores that instantly became destinations. Between 2019 and 2023, revenue nearly tripled from $3.3 billion to $9.6 billion, EPS shot up over 250%, and the stock peaked above $490.

Then came the double whammy. The first was the product. lululemon made a series of "missteps" - failed collections, inappropriate color spectrum, merchandise that didn't resonate with the target customer, and in parallel began pushing volume through markdowns (price reductions to clear out stock), slowly eroding the premium brand perception that had been built up over years as the company's biggest "moat." The second blow was macro: the US consumer in the higher income bracket became more discerning (words of McDonald's CEO himself), purchase frequency dropped, and the emergence of competitors (Alo Yoga, Vuori, On Running) began to take share in a segment where lululemon had dominated.

At the end of 2025 came the resignation of CEO Calvin McDonald - a signal that the problems were not just temporary. The company is now operating with interim leadership, searching for a successor, while also having to deal with pressure from founder Chip Wilson, who is calling for structural changes to strategy and the board. All of this gives the situation the feel of a classic retail turnaround: the fundamentals are sound (brand, margins, international expansion), but the short-term picture is unpleasant and the uncertainty around management adds another layer of risk for investors.

What needs to work for a return to the original story

• A new CEO needs to come in quickly, have credibility with the market and a clear strategy for US recovery.

• US comparable sales must stop declining by H2 2026 and start growing by the end of 2027.

• Markdowns must recede and gross margin must stabilize or improve.

• China must maintain double-digit growth despite Chinese consumer volatility.

• Tariffs (30% on Chinese imports, 10% on other Asian) must be absorbed through renegotiation with suppliers or shifting production, without dramatically impacting prices.

How does this become money

1) US: a problem that can be solved by product

The US generates approximately 72-73% of lululemon' s sales today, but this figure needs to be read in context: even the "troubled" US is a business with sales of over $8 billion a year with gross margins still above 50%. So the problem is not existential, but one of margin and reputation.

The way back is through full-price selling (selling at full price without discounts) - this is an explicit management priority for FY2026. If markdowns can be reduced, gross margin could stabilise or even improve on the 550bp lost this year. In parallel, the company is working on a product reset - returning the collection to what lululemon customers love: functionality, detail, clear identity. This is not a project for the quarter, but if the new CEO comes up with a clear product vision, sentiment in the US could turn around quickly.

2) China as a second home market

International expansion - especially China - is one of the strongest arguments for investment. lululemon is seen in China as a premium aspirational brand for the emerging middle class who want a fitness lifestyle and are willing to pay for quality. Comparable sales grew +11.5% in FY2025 in an environment where other foreign brands have faced a significant decline in the Chinese consumer.

The company is targeting 200 stores in mainland China (around 130-140 in 2025) and expanding into secondary cities. If it reaches 200 stores at a similar average revenue per store as today, China could generate over US$1bn in annual revenue by 2027-2028 - around 9-10% of total revenue, up from around 4-5% today. This would be a fundamental shift for valuations as China establishes itself as a true second growth pillar.

3) Male segment and digital - achievable goals?

ThePower of Three x2 strategy targeted a doubling of male revenue and digital revenue vs. 2021 by FY2026. Digital is around 40% of total revenue (good), but the male segment grew slower than the plan assumed. Still, the men's segment is an interesting opportunity: male penetration is dramatically lower than women, brand awareness is growing, and products like ABC Pant (All the Balls Comfortable) are building a loyal base.

If the new CEO comes up with a clearer male strategy and product planning, the male segment could become in 3-5 years what the female segment was 10 years ago. That's a big story, but for now, it's more of an option than a certainty.

4) Membership and community moat

lululemon has one of the strongest "community" strategies in retail: ambassador programs, fitness events, runs, yoga, collaboration with instructors. This isn't just marketing - it's a barrier to entry that's difficult for pure online or DTC (direct-to-consumer) players to replicate. If a company maintains this community identity even in turbulence and stops disrupting it with excessive discounting, chances are that once the product "clicks", customers will return faster than with other brands.

The numbers that support this thesis

Box - key numbers FY2021-FY2025

• Shares: top $490 ( 2023) → bottom around $150 ( 2026), down -67%.

• Trailing P/E: historical average 30-40× → today ~12× (decade low).

• Forward P/E on guidance 2026 EPS midpoint $12.20 at $160 price: ~13×.

• China comparable sales FY2025: +11.5%, 2026 guidance: +20-25%.

• S. America Q4 FY2025: -6% YoY in sales, guidance 2026 S. America: -1 to -3%.

• Gross margin Q4 FY2025: compression -550 bps YoY (impact of markdowns and tariffs).

• Share buyback: 1.4m shares repurchased in Q4 at average price of $188, new program approved for additional $1bn.

• Inventory: $1.7 billion (+18% YoY) - higher inventory is a short-term risk, but signals the company is preparing for a sales turnaround.

Dividend and financial health

lululemon does not currently pay a dividend - this is purely a growth and capital allocation story. Shareholder value is coming through:

• Organic sales and margin growth.

• Aggressive share buyback (share repurchase) - program raised $1 billion in December 2025.

• Investment in expansion (new stores, China, Europe).

Company's financial health remains solid despite margin pressure:

• Gross margin still above 55% (vs. retail average of 30-40%).

• Operating margin around 22% (Q4 FY2025), still well above sector average despite compression.

• Relatively low debt, strong ability to generate free cash flow (free cash after capex).

Buybacks at prices of $160-190 at a historical average of $300+ are extremely attractive from a capital allocation perspective if management believes in a turnaround - and they seem to.

Valuation - what's in and what's out

This is the crux of the "generational buy" argument. lululemon has never, since going public, traded at as low a multiple as it does today. Forward P/E of ~13 times on a company that:

• has a gross margin of 55%

• generates an operating margin of 22%

• is growing at double-digit rates internationally

• Holds one of the strongest consumer brands in the athleisure.

By comparison, Nike trades at a forward P/E of ~23×, Adidas ~22×, both of which have worse margins and weaker brand positioning in the premium athleisure segment.

The market fully discounts in price (accounting for worst case scenario):

• Continued US decline with no clear turnaround

• sustained margin issues from tariffs

• Risk of leadership vacuum and failure to find a CEO

• pressure from founder Wilson to destabilize the board

What the market, in turn, does not reflect or undervalues in the price:

• China as a second home market with potential for $1+ billion in revenue by 2028

• The ability of a new CEO to bring about the reset that the market has seen in, for example, Nike after new leadership takes over

• Structural premium brand positioning that remains stronger than direct competitors even after markdowns

• buyback at historically low prices

Valuation scenarios:

• With EPS recovery at $15-16 (FY2027) and P/E 18× (mid-case scenario) → price of $270-288.

• With EPS of USD 15 and P/E of 22× (re-rating in a clear US turnaround) → price of USD 330.

• At EPS of $12 (no turnaround) and P/E of 13× (no re-rating) → price of $156 (virtually today's price = "bottom").

Analyst estimates: median $205, range $170-295. At a price of $160, even the most conservative target ($175) goes up +9%, and the median +28%.

Macro and market

Athleisure (athleisure wear) is a structural trend, not a cyclical fluctuation. The lines between athleisure, everyday wear and workwear (workwear) are constantly blurring, and lululemon is better positioned than almost anyone else on this trend. The global sportswear market is estimated to be over $200 billion and growing at around 6-7% per year.

The macro risk for lululemon is specific: it targets the higher income segment of consumers, which is traditionally more resilient to economic fluctuations. The slowdown in the US is therefore not the result of a recession, but of a specific brand and product problem that management itself acknowledges. This is paradoxically good news - structural problems are hard to solve, but product and brand problems have precedents for successful turnarounds (see Nike, Adidas in past crises).

Risks

1) Leadership vacuum

The company has been operating without a permanent CEO since January 31, 2026. The search for a successor may take 6-12 months, interim management does not have a full mandate to make major decisions, and investors are nervous about uncertainty. If the new CEO comes in weak, with a unclear vision, or if the selection is controversial (Wilson vs. board pressure), sentiment may continue to fall.

2) US decline with no visible bottom

Guidance 2026 projects North American revenue declines of 1-3% - the third year in a row. If comparable sales do not improve at the end of 2026, the market will start to consider whether this is a structural erosion of the story or just a temporary reset. This is the biggest bear argument: that lululemon has lost its "must have" status with the American consumer.

3) Margins and tariffs

Tariffs of 30% on Chinese imports and 10% on other Asian goods are squeezing margins, which are compressed by 550 bps in Q4 FY2025. FY2026 guidance expects another 250 bps compression in operating margin. Some of this may be offset by production shifting, renegotiation with suppliers or slight price increases, but not all of it.

4) Inventory and markdown risk

Inventories of $1.7bn (+18% YoY) are well above historical norms. Failure to sell inventory at full price will continue to put pressure on markdowns and margins, further hindering the turnaround on the gross margin line.

5) China as a double-edged sword

China is the biggest driver of international growth today, but also the riskiest market. The Chinese consumer is volatile, geopolitical risks are high, and foreign brands can quickly lose popularity to domestic competitors (Anta Sports, Li-Ning). lululemon has held its position as a premium aspirational brand for now, but one political upset or PR incident could change this.

Checklist of risks

• No new CEO by mid-2026 or the arrival of a controversial candidate.

• US comparable sales not falling to -3% or worse in Q1-Q2 2026 with no signal of a turnaround.

• Gross margin falls below 53% due to a combination of tariffs and markdowns.

• China comparable sales slow to single digits due to macro fluctuations.

• Inventories of $1.7bn force further markdowns and worsen margins in H1 2026.

• Chip Wilson pressure escalates into proxy battle (shareholder vote fight) and destabilizes management.

• Increased competition from Alo Yoga, Vuori, On Running in core women's segment.

Investment scenarios

Optimistic scenario

New CEO arrives by late summer 2026, is a strong candidate with retail credibility. U.S. comparable sales will turn around in the second half of 2026, full-price sales will return and gross margins will begin to recover. China surpasses 200 stores by 2027 and generates over US$1bn. EPS will return to growth trajectory towards US$15-16 in FY2027.

• Valuation at a P/E of 20-22× → price of USD 300-350.

• Return from today's price of $160: +88-119% over 2 years.

Realistic scenario

New CEO arrives, US comparable sales stabilize (not turning around yet), China continues to grow. EPS remains under pressure in FY2026 (~$12) but improves to ~$14 in FY2027. P/E slightly re-rating to 16-18×.

• FY2027 price: USD 225-252.

• Return from today's price: +41-58% over 2 years.

Pessimistic scenario

CEO search drags on, US comparable sales continue to decline in 2027, tariffs and markdowns continue to squeeze margins, China slows to single digits. EPS falls to ~$10-11 and P/E remains compressed at 12-13x.

• Price: $120-143.

• Downside from today's price: -10 to -25%.

• It will be a "value trap" scenario: the company looks cheap, but earnings are falling and valuation does not warrant re-rating.

What to watch next

• Appointment of a new permanent CEO - who, when, what experience, how the market reacts.

• Q1 FY2026 results (circa June 2026) - first signal of whether US comparable sales are starting to stabilise.

• Gross margin developments - is the company returning to full-price selling or are markdowns continuing?

• China: pace of new store openings, comparable sales for Q1 and Q2 2026.

• Management comments on inventory - needs to decline towards normal US$1.2-1.3bn.

• Share buyback activity - at prices below USD 180, this is a strong signal of management confidence.

• Chip Wilson pressure development - escalation or retreat, implications for board composition.

• Competition: how fast Alo Yoga, Vuori and On Running are growing in the women's premium segment.

What to take away from the article

• lululemon is a premium company with 55% gross margin, strong international growth and one of the best brands in athleisure, trading at an all-time low of 12-13x forward earnings.

• At the heart of the problem is a product and leadership reset in the US - a structurally sound company facing a temporary loss of step, not an existential threat.

• China is a key differentiator: 11.5% comparable sales growth, 200-store plan, potential to become a $1B+ per year business by 2028.

• A new CEO is the number one catalyst: the right candidate can deliver a re-rating from today's 13× to 18-22×, meaning +80-120% upside as EPS stabilizes.

• The biggest risk is not bankruptcy or losing the business, but the "value trap" - a scenario where the company looks cheap but earnings continue to fall and re-rating doesn't occur.

• For the long-term investor (3-5 year horizon) who believes in the strength of the lululemon brand and international expansion, this is an asymmetric bet: limited downside at today's valuation, potentially significant upside at a successful turnaround.

Personally, I think Muddy Waters Research just wants to profit from a decline in the price of $SOFI, which is why their report was so strongly negative. SoFi is a regulated bank overseen by US authorities, including the Fed, so the scope for major accounting manipulation should be fairly limited. Even if certain shortcomings were found, they probably wouldn't be on the scale the report suggests.

A positive sign is also the CEO buying shares shortly after the report was published. To me, that shows management's confidence in the company's fundamentals. The CEO likely wouldn't have invested their own money in the stock if they knew of material problems.

This is just my opinion on the current situation, and I'd appreciate other viewpoints or any counterarguments.

At first glance, a double-digit dividend yield looks like an ideal investment for anyone looking for regular passive income. If a company offers a yield of 15%, 20% or even 25%, it seems like a significantly better opportunity than traditional dividend stocks with yields of around 2-4%. But the reality is usually more complicated. In fact, an extremely high dividend yield is often not the result of a company's generosity, but the result of a significant decline in share price. The market is essentially telling the investor that there is a real risk of a future reduction or complete cancellation of the payout.

The mechanism is relatively simple. The dividend yield is calculated as the ratio of the annual dividend to the current share price. If the share price falls significantly while the dividend remains unchanged for the time being, the yield automatically jumps to visually attractive levels. It is this phenomenon, known as the dividend trap, that poses one of the greatest risks to unsophisticated dividend investors. Instead of a stable income, the investor often sees a dividend cut and a further fall in price.

In the following review, we look at three specific stocks that currently offer yields well above 15% and where it is important to understand what is really behind the number.

PIMCO Dynamic Income Fund $PDI

Yield: 15.61%

The PIMCO Dynamic Income Fund is aclosed-end bond fund managed by PIMCO, one of the largest bond portfolio managers in the world. The fund invests in a broad range of bond instruments including mortgage-backed securities (MBS), investment- and speculative-grade corporate bonds, developed and developing country sovereign bonds, and securitized assets. It is the breadth of coverage and active portfolio management that are the main arguments that PIMCO Fund presents to the investing public.

The fund's current dividend yield is around 15%, making it one of the highest yields in the entire closed-end bond fund segment. The fund pays dividends monthly, which is particularly attractive to many income-oriented investors. The fund has a market capitalization of over $6.4 billion.

Where's the catch

The problem lies in how the fund generates its high yield. The use of leverage, which is approximately 35% for $PDI, plays a significant role. The fund essentially borrows money with which it buys more bonds, thereby increasing the overall return of the portfolio. In a low interest rate environment this approach has worked well, but in a higher interest rate environment the cost of borrowed capital increases significantly and leverage begins to work against the fund.

Another problem is the payout structure. A significant portion of the fund's distributions do not represent actual interest income, but what is known as Return of Capital (ROC). This means that the fund pays investors partly their own money instead of the actual returns earned. According to available data, ROC has accounted for up to 65% of total payouts in past periods. While this practice is not illegal, it gradually erodes the net asset value of the fund, meaning that the investor recovers his own capital and the illusion of high returns is partly misleading.

On top of this comes a high expense ratio, which for PDI is around 6%. By comparison, conventional bond ETFs operate with expenses below 0.5%. In addition, the fund has traded at a premium to NAV for most of its history, meaning that investors have paid more for the underlying assets than they are actually worth. The combination of leverage, high costs, return of capital and premium to NAV thus creates an environment in which an investor's real return may be significantly lower than the 15% optical dividend yield suggests.

Icahn Enterprises $IEP

Yield: 26.5%

Icahn Enterprises is a diversified conglomerate controlled by legendary activist investor Carl Icahn. The firm operates in a number of segments including energy (through its stake in CVR Energy $CVI), automotive, metals, real estate, food packaging, and investment funds. It is structured as a limited partnership, which has specific tax implications for investors.

The current dividend yield is around 26.5%, a number that should put any investor on notice immediately. Shares are trading around $7.50 apiece, down about 95% from the highs of 2013.

The drop and the clash with short-sellers

The turning point for $IEP came in 2023, when short-seller (a firm that makes money on stock declines) Hindenburg Research published a critical report that questioned the valuation of the company's assets and highlighted the unsustainability of the then-current dividend. Hindenburg called the payout ratio unsupported by the firm's earnings and cash flow. The market reacted to the report with a massive sell-off.

Subsequently, the company actually proceeded to cut the dividend dramatically. It paid out $2 per share per quarter from 2019 to mid-2023. The dividend was first cut to $1 and later to the current $0.50 per quarter, a total decrease of 75% over two years. Despite this significant reduction, the payout ratio remains extremely high. In 2024, the company paid out $391 million in dividends, while its EBITDA (earnings before interest, taxes, depreciation and amortization) was just $156 million. Dividends thus represented 251% of EBITDA, a clearly unsustainable level.

Why the yield is deceptive

The high dividend yield $IEP is almost entirely the result of the massive drop in share price. The company generates negative earnings per share (EPS for the past year was approximately -$0.79) and pays a dividend that far exceeds its generated earnings. On top of that, the company has significant debt in excess of $7 billion, which, with a current market capitalization of around $5 billion, creates an unfavorable debt-to-book ratio. In addition, there are significant bond issues maturing in the next few years, which may further limit the scope for dividend payouts. The combination of declining net asset value (NAV reduced by $654 million in the most recent reporting period), negative earnings, and a history of two dividend cuts in a short period makes IEP a textbook example of a dividend trap.

ZIM Integrated Shipping $ZIM

Yield: 7.69%

ZIM Integrated Shipping Services is an Israeli container shipping company that operates a fleet of approximately 128 ships and offers shipping services on all major global trade lanes. The company operates primarily on a charter basis, meaning that it charters most of its fleet instead of owning it. This model allows it to respond flexibly to changes in demand, but also increases cost sensitivity during periods of lower freight rates.

ZIM's dividend yield has been around 13-16% in recent months, which is appealing on the face of it. However, it is currently "only" above 7% as the stock has appreciated significantly (108%) in the last 6 months. However, to understand what this number really means, an investor needs to look at the extreme cyclicality that is typical of container shipping.

A rollercoaster of dividends

The ZIM story illustrates why it is important for cyclical companies to look at the dividend in the context of the entire cycle. In 2021, when container rates hit all-time highs due to post-covenant demand and supply chain congestion, ZIM paid a dividend of $17 per share in a single year. This was followed by a payout in excess of $27 per share in 2022. These amounts, however, reflected quite extraordinary market conditions. Once freight rates normalized in 2023 and demand weakened, the company went into a loss and stopped the dividend altogether. There was another brief recovery in 2024 due to geopolitical tensions that disrupted shipping lanes, but again this proved temporary.

Current situation and takeovers

In February 2026, ZIM announced that it had entered into a takeover agreement with Germany's Hapag-Lloyd for $35 per share, a premium of 58% to the previous day's price. The total value of the transaction is approximately $4.2 billion. However, the shares are currently trading around $26, well below the offer price. The market is signaling that there are significant risks associated with completing the transaction. Complications include the need for approval from the Israeli government (which holds the so-called Golden Share), geopolitical concerns over the ownership structure of Hapag-Lloyd, which includes sovereign funds from Qatar and Saudi Arabia, and resistance from employees who have called a strike at the company's headquarters.

Management behaviour is also a worrying sign. CEO Eli Glickman sold 87% of his shares at prices around $28-29, 20% below Hapag-Lloyd's offer price. If insiders are selling at a discount to the acquisition price, they are clearly indicating that they attach a high probability to the transaction not going through. In addition, ZIM has suspended the release of its 2026 financial outlook, which further reduces the visibility of future earnings and dividends, increasing investor nervousness.

Common features and comparisons

All three titles share several key characteristics that should make investors cautious:

• A visually high dividend yield that is largely the result of price declines, not strong income generation.

• A history of reduced or extremely volatile dividend payouts. $IEP has cut the dividend twice, $ZIM has completely stopped and reinstated it, and $PDI is partially paying out of capital returns.

• Structural risks specific to each title: high leverage and costs for PDI, negative earnings and declining NAV for IEP, extreme cyclicality and uncertainty around acquisition for ZIM.

• Payout ratios that are not sustainable over the long term at current levels in either case.

By comparison, companies like Johnson & Johnson $JNJ, Procter & Gamble $PG or Coca-Cola $KO offer dividend yields of around 2-3%, but with a history of decades of uninterrupted payout increases. It is this predictability and stability that tends to be ultimately more valuable to long-term dividend investors than a one-time high yield that can disappear at any time.

A strategic view

Dividend yields above 15% should not tempt investors, but rather warn them. In practice, there are very few cases where a company or fund can sustain such a high payout over the long term without a reduction in yield or a further fall in price. Key metrics that investors should monitor before buying high yielding titles include:

• Payout ratio: if the company is paying out more than it is earning, the dividend is being funded by the principal or debt.

• NAV or book value development: a declining net asset value indicates that payouts are eroding capital.

• Dividend history: firms that have cut the dividend in the past are statistically more likely to cut it again.

• Income structure: for funds, it is important to distinguish between actual return and return on capital (ROC).

An investment rule of thumb often mentioned in this context is: if the yield looks too good to be true, it probably isn't. For titles with double-digit dividends, it is therefore essential to understand the reasons behind the yield and not to go by the number alone.

What to watch next

• $PDI: the evolution of Fed interest rates and their impact on the fund's cost of leverage, the ratio of capital returns to actual returns, and the evolution of the premium or discount to NAV.

• $IEP: next quarter's results, especially the development of NAV and EBITDA, maturities of bond issues in the coming years and any further dividend cuts.

• $ZIM: the progress of the approval process for the acquisition by Hapag-Lloyd, the Israeli government's position on Golden Share, the development of freight rates and the possible renewal of the financial outlook for 2026.

• General: track payout ratio, free cash flow yield and dividend payout history for any title.

Lessons learned

PIMCO Dynamic Income Fund, Icahn Enterprises and ZIM Integrated Shipping show that extremely high dividend yields often signal structural problems, not extraordinary investment opportunity. Each of these titles has specific reasons why the yield has reached current levels, but the common denominator is the unsustainability of the current payout model over the medium term.

For investors looking for stable dividend income, the key rule remains: the quality of the dividend is more important than the amount. A reliable payout with moderate growth over time generates a significantly better total return than a one-off attractive number that can disappear at any time. In dividend investing, more than anywhere else, patience and careful analysis pay more dividends than chasing the highest yield.

In his latest commentary on the investment environment , Fink delivers not reassuring news, but a clear message: those fleeing to cash today are making arguably the biggest mistake of their investing careers. The reason? Artificial intelligence will change the rules of the game before most investors realize it.

As Fink describes today's markets

Fink notes that over the past 20 years or so, the S&P 500 index has managed to appreciate more than eight times an investor's capital, despite financial crises, pandemics, inflation and geopolitical shocks. The key conclusion is simple: over the long term, time in the market has been more important than trying to hit the ideal timing of entry and exit.

At the same time, he points out that we are living in a "compressed" era, where what used to fill an entire decade happens in a few years. Energy shocks, the rebuilding of supply chains, the rise of trillion-dollar tech companies and the breakthrough of AI. This brings more unknowns, but according to Fink, it doesn't change the fundamental principle: the store of value in the long run is real assets and quality companies, not cash.

AI as a structural trend, not a bubble

Fink sees AI not as short-term hype, but as a structural change that will create enormous economic value, transform entire sectors from finance to healthcare to industry to media, and become a key axis of competition between the US and China.

He mentions that the eventual "victims" of AI, i.e. companies that fail to adapt, are part of capitalism, but that does not mean that the sector as a whole is a bubble. On the contrary, Fink repeatedly says that a bigger risk than an "AI bubble" is a situation where Western economies invest little and leave the technological leadership to China.

From an investor perspective, he is implicitly pointing to titles like Nvidia $NVDA, Microsoft $MSFT, Alphabet $GOOG, Meta $META or Amazon $AMZN, but also to less visible infrastructure players such as providers of datacenters, optical networks and power solutions for AI.

Why not exit the market, according to Fink

Fink's key argument rests on historical experience: volatility, geopolitics and recessionary fears have always been there, but technology disruptions generally reward those who are inside the market, not on the sidelines.

Specifically with AI, this means that those who have no exposure to the AI ecosystem in their portfolio, which includes chips, cloud, models and apps, risk missing out on a substantial portion of future appreciation. A long horizon, typical of pension strategies or long-term savings, is an advantage: an investor with such a horizon is less forced to panic in short-term downturns.

Fink doesn't say it doesn't matter what you buy. He says that fleeing to cash at a time when the technology trajectory is breaking down is an expensive mistake in the long run.

AI, inequality and why growth shouldn't benefit only the elites

Fink also points out that AI can further open the scissors between those who own capital and those who sell labour. Globalisation has hit manual and blue-collar professions hardest. AI can push analysts, administration, parts of IT or marketing in the same way. And if AI mainly benefits the shareholders of a few companies, this will inevitably lead to political and social pressure.

The solution he proposes is to involve the wider public in the benefits of AI through pension schemes and funds. From the perspective of BlackRock $BLK, this is of course also a business. But the idea has real substance: AI is becoming such a big growth engine that if it remains locked in the portfolios of wealthy investors only, the inequality problem will deepen and with it the risk of a regulatory backlash that slows down the entire sector.

How this can inform portfolio construction

If you want to translate Fink's insight into concrete investment logic, several practical points emerge.

The key is not to play short-term scenarios, but to have a long-term portfolio skeleton consisting of a global or US equity index like the S&P 500, supplemented by a focused AI infrastructure and leadership component.

Within AI, it then makes sense to distinguish layers of exposure. The base layer includes chip players like Nvidia $NVDA or AMD $AMD and AI server makers. The middle layer covers cloud and hyperscalers, i.e. Microsoft, Alphabet or Amazon. The application layer then represents companies that can monetize AI in specific products, from Meta and Salesforce to Adobe and smaller niche players.

Fink himself says that some AI projects will fail, and that's okay because that's capitalism. But as a whole, AI will shift the profitability and productivity of companies that integrate it sensibly. So he recommends treating volatility as a cost of entry into a long-term growth trend, not as a signal to flee the market.

The Bay Area move has both symbolic and strategic weight. Wing has now completed more than 750,000 deliveries globally, operates through partnerships with Walmart and DoorDash, delivers items in as little as 30 minutes and has extended service hours to nighttime in some markets using infrared navigation. Competing directly in the heartland of Silicon Valley, where Amazon is also scaling drone delivery, signals that Alphabet is ready to translate Wing from a proof-of-concept into a logistics platform with real commercial scale — extending the company's ecosystem into the physical world that Amazon, FedEx and UPS have long dominated.

Why the Bay Area

The San Francisco Bay Area is not a random choice. It's one of the most densely populated and technologically advanced places in the U.S., where consumers are accustomed to fast delivery and willing to test new services before the rest of the country. It's also an environment Wing is familiar with - the company not only started here, but starting in 2024, DoorDash is building a logistics infrastructure in the Mission District designed specifically to work with drone systems.

The Bay Area also presents a regulatory challenge. Dense developments, wind corridors from the Bay, high-rise buildings and complex airspace all place high demands on navigation systems. Successfully navigating this environment would open the door for Alphabet $GOOG to dozens of other U.S. metropolitan areas where, until now, drone delivery has hit technical limits.

How delivery works

Wing drones are lightweight, fully automated machines designed to fly directly to a customer's home. Key parameters:

• Carrying capacity of approximately 2.3kg - ideal for food, medicine, smaller consumer goods or restaurant food.

• Travel speed up to 105 km/h.

• Range of approximately 10 kilometres from the dispensing point.

• Standard delivery time within 30 minutes of order.

The drone does not land - it lowers the delivery on a rope directly into the garden or front of the house, bypassing obstacles such as fences, cars or pets. The whole process is autonomous and does not require a courier.

Partners

Alphabet Wing in the US works with two key players covering different market segments:

• Walmart $WMT - groceries, over-the-counter drugs, cleaning products and household consumables. In January 2026, Alphabet announced the expansion of the collaboration to 150 additional Walmart stores. In total, the network is expected to cover over 270 locations by 2027, potentially reaching up to 40 million customers across the US.

• DoorDash $DASH - delivering prepared meals from restaurant chains like Wendy's, Panera Bread or other platform partners. Here, Wing acts as the logistics engine, while DoorDash handles the ordering system and customer relationship.

The combination of the two partners gives Alphabet access to two of the largest quick delivery categories in the U.S. - groceries and food delivery - without having to build its own customer platform from scratch.

Innovation: drone + ground robot

One of the most interesting experiments Wing is currently testing is a collaboration with Serve Robotics, a specialist in autonomous ground delivery robots. The principle is simple but ground-breaking: a Serve ground robot picks up the food directly from the restaurant, moves to a predetermined location and hands the delivery to a Wing drone, which delivers it to the customer by air.

This is the first real connection of two autonomous delivery layers in a commercial operation - from the curb to the air. In doing so, Alphabet is solving one of the key problems in drone logistics: how to get a shipment from source to drone without involving human labor. If this integration proves successful in the Bay Area, it opens the way to fully automated delivery corridors in cities.

What this means for Alphabet and its investors

Wing is proof that Alphabet can take "moonshot" projects into real business - the same path that Waymo took in autonomous driving. Division X, from which Wing emerged, has historically attracted criticism for the high costs and unclear commercial future of its projects. Wing is beginning to refute this criticism with numbers.

The drone delivery market is estimated to be worth about $1.5 billion in 2026, and analysts reckon it could reach between $7 billion and $27 billion by 2031, depending on the pace of regulatory easing. Alphabet holds the technological edge in this segment thanks to integration with Google's AI infrastructure, advanced computer navigation and flight in complex environments - exactly what the dense Bay Area development will test to the fullest.

For now, Wing doesn't represent a standalone fundamental valuation factor for Alphabet stock, but it reinforces a broader narrative: Alphabet is not just a search-dependent advertising firm, but a tech conglomerate capable of building the physical infrastructure of the future. At a time when investors are watching Alphabet diversify its revenue beyond advertising, every Wing success story reinforces this narrative.

The financial picture is mixed. Cash rose sharply to 167.9 million dollars after 234 million in equity issuances, giving Red Cat runway to keep scaling, but the net loss widened to 72.1 million dollars and the accumulated deficit stands at 196.8 million with no formal 2026 guidance provided, contributing to shares falling roughly 11 percent after the earnings release. About 73 percent of 2025 revenue came from U.S. government contracts, meaning concentration risk is high and any delay or repricing in the SRR programme or Drone Dominance competition could significantly shift the near term revenue picture in either direction.

Top points of the analysis

• The company's 2025 revenue of $40.7 million, up 161% year-on-year, is driven primarily by a massive expansion of Army orders for the Black Widow system and its first international orders through the NATO catalogue.

• The company won the US Army's key short-range drone tender in November 2024, displacing the then-dominant Skydio - the July 2025 contract has been progressively expanded to around $35 million and has become the backbone of current revenues.

• The Black Widow system was added to the NATO purchasing agency catalogue in September 2025, opening the way for orders from allied militaries without the need for individual tenders - the first order for 100 units through NATO came later that year.

• The company is still loss-making: the net loss for 2025 is approximately $72 million, gross margins are in negative territory, and the accumulated deficit has exceeded $196 million, with management warning that additional capital may be needed.

• The valuation is aggressive - a market value of about $1.4 billion on $40 million in sales implies a market value to sales ratio of over 33, a bet purely on future growth and the company's ability to turn government contracts into real profitability.

• Regulatory winds are blowing in the sails: a federal law restricting the use of drones with foreign components (mainly of Chinese origin) took effect immediately upon its promulgation, significantly narrowing the pool of eligible contractors for U.S. government and security contracts.

Company introduction

Red Cat Holdings $RCAT is a U.S. technology firm based in San Juan, Puerto Rico that focuses on the development and manufacture of advanced drones and autonomous systems for military and security applications. The company was formed as a holding structure and through successive acquisitions has assembled a portfolio of specialised technology companies - key ones being Teal Drones as the main manufacturing division, BlueOps specialising in maritime autonomous systems and FlightWave focused on commercial and reconnaissance applications. With 115 employees, it is a small company, but one that has expanded its manufacturing footprint dramatically in the last two years - operating a total of 254,000 square feet of manufacturing and development space at the end of 2025, up from a fraction of that capacity at the start of 2024.

The company is transforming from a research-oriented startup structure to an industrial defense manufacturer starting in 2024 - and this transformation is both the biggest opportunity and the biggest risk. On the one hand, military contracts provide a clear revenue stream and validate that products work under real military deployment conditions. On the other hand, the transition from prototypes to mass production is costly, technically challenging, and historically many promising defense startups have failed in this transition.

CEO

The company is led by Jeffrey Thompson, founder and CEO. Thompson built Red Cat from the ground up, took it through the tough covid years, several pivots of business strategy, and ultimately transformed it into the form that put it in the spotlight of the U.S. Department of Defense. His approach combines technological enthusiasm with the pragmatism of a government customer - the company, under his leadership, gambled that it would design drones based directly on feedback from soldiers in the field, not on what looks good in a lab environment. The result is the Black Widow, a system designed to be operated by a soldier without extensive technical training, and that was exactly one of the key arguments in winning the Army's tender.

The company states: Thompson is an experienced serial entrepreneur (he has founded several successful startups) in the technology space who has had several successful company stories - from startup to sale or going public. He was behind the creation of EdgeNet, a privately held Internet service provider, and co-founded Towerstream, a wireless carrier that went public on the NASDAQ in 2007. Then in 2018, he founded Red Cat, which focuses on the rapidly growing market for drones - both commercial and recreational.

Products - what makes them interesting and why the military buys them

Red Cat isn't a one-product company - it's building a portfolio of systems for different deployment scenarios, each targeting a specific military need.

The Black Widow is the company's flagship product and the reason Red Cat is even being talked about today. It is a small unmanned reconnaissance system weighing less than 1.4 kilograms, capable of flying for over 45 minutes, and equipped with electro-optical and infrared sensors for day and night operations. The key advantage is not the technical performance alone, but a combination of factors: the system is entirely US-origin (no Chinese components), certified for government use, easy to transport and operate in the field, and does not require specialized technical training for the operator. It was selected by the Army as one of two standardized contractors under the Short Survey Program, which assured the company of recurring government contracts for the longer term.

https://www.youtube.com/embed/DYjTRZ1GS_k?rel=1

The FANG is a new type of FPV category drone - it is a fast, cheap and easily replaceable system designed for attack or ambush operations, inspired by the experience of the war in Ukraine where this type of drone proved highly effective and difficult to resist. The company introduced it in 2025 and immediately partnered with AeroVironment, one of the largest US military drone manufacturers, to integrate the FANG system as a payload into its P550 platform. This is a key step - FANG is moving from being a standalone product to being part of a broader ecosystem where major players in the defence industry have chosen Red Cat as their preferred partner for this category.

https://www.youtube.com/embed/7-k0YqkxfCQ?rel=1

Maritime autonomous vessels are the youngest area of development. At Innovation Day in February 2026, the company live demonstrated an unmanned vessel working with a Black Widow drone in an operational scenario, demonstrating the ability to coordinate air and maritime elements in a single mission. The move is strategically important: it expands the firm's addressable market beyond purely airborne operations and continues a global trend where naval forces are looking to autonomous systems for reconnaissance, coastal surveillance and shallow-water operations.

Who's buying, who can buy and how big is the market

Red Cat's customers today are overwhelmingly government entities. Approximately 73% of 2025 sales came from U.S. government contracts, primarily from the military. The key contract is an expanded Black Widow production program for the U.S. Army worth approximately $35 million, and the company is actively competing in the next round of the Gauntlet, the outcome of which could significantly increase future orders.

International customers are beginning to form the second mainstay of sales. The company received an order for Black Widow systems from an allied military in the Asia-Pacific region in December 2025 - the specific country was not disclosed for security reasons - and this is the second customer from that region in short order. The inclusion in the NATO purchasing agency's catalogue in September 2025 is groundbreaking from a commercial perspective: instead of individual tenders in each member country, allies can simply order through a central catalogue, dramatically reducing the time from interest to order.

Potential customers include the full spectrum of government and security entities. On the military side, it is the militaries of allied countries that are looking for certified and safe drones without Chinese components - in addition, the FY25 NDAA prohibits them from purchasing Chinese systems in many cases, narrowing the pool of eligible suppliers significantly. On the security side, these include border guards, police and emergency services looking for drone systems for terrain reconnaissance, personnel searches or critical infrastructure security. The company does not yet operate much in this segment, but the technology and certifications are transferable.

Market and regulatory opportunity

The global market for short- and medium-range military drones is growing faster than the overall defense budget, driven by the experience in Ukraine, where unmanned systems have become a standard part of infantry weaponry, and the growing demand for low-cost, easily replenishable aerial reconnaissance assets. The United States is the largest market, but its demand goes hand in hand with alliance commitments - what the military standardizes for its own use inevitably spreads to the armaments of allies seeking interoperability.

The regulatory environment has become a direct business catalyst. A law passed as part of the 2025 defence budget restricts US government entities from purchasing drones containing components from manufacturers originating in countries designated as security threats, in practice primarily Chinese components. Implementation of this law began immediately, eliminating a number of previously popular systems from government procurement - and Red Cat, as a company with an entirely American manufacturing chain, is one of the few able to fill that gap. As a result, analysts talk about the firm being not just a beneficiary of organic market growth, but a beneficiary of a regulatory "market turnaround" where the supply of eligible suppliers is declining faster than new capacity is being built.

Financial performance - growth without profit

Red Cat's numbers are a combination of exciting and disturbing. Revenues grew 161% to $40.7 million in 2025, with record results in the fourth quarter alone and a full-year growth rate of an astonishing 646% year-over-year in the third quarter thanks to the ramp-up of the Army contract. Production capacity grew 520%, the company expanded three manufacturing facilities, and total square footage reached 254,000 square feet.

On the profitability side, the picture is not so favorable. Gross margins are negative, the company is literally selling below cost of production, while massive spending on research, development and scaling up production pushed the operating loss to nearly $39 million for the most recent reporting period. The cumulative deficit has exceeded $196 million, and management explicitly states in the risk factors that the company may need additional external financing. That's a reality that investors need to consider as much a part of the story as exciting contracts and media attention.

Free cash flow is zero, margins are negative, and return on assets is deep in the red. These metrics are not unusual for a pre-profit company in an expansion phase, but they tell a clear story: if the negative gross margin can be reversed and operating costs brought under control, the company will be a different investment than it is today. Analysts estimate that if revenue growth rates are maintained and gross margins normalize to industry levels of around 20% to 30%, the firm could reach operating profitability sometime in the 2027 horizon, but that depends on continuity of government contracts and success in other tenders.

Balance sheet and liquidity

Paradoxically, Red Cat's current strength is liquidity. A quick ratio of over 11 and a cash ratio of over 11 tell us that the company has significantly more liquidity than its current liabilities. An Altman Z-score of over 66 is statistically high in the safe zone - not because the firm is operationally sound, but because it holds a large proportion of cash and equity relative to liabilities. The firm has likely done a capital raise in the last two years that makes the balance sheet look healthy, but that liquidity will be gradually consumed to cover operating losses if sales and margins don't grow fast enough.

Debt levels are very low - a debt-to-equity ratio of 0.05 and minimal long-term debt tell us that the company is not yet financing itself with external capital to any great extent. Interest coverage of over 10 times confirms that the existing debt burden is not a problem. Thus, the main financial risk is not the debt spiral but the gradual depletion of the cash reserve if profitability cannot be achieved within a reasonable time horizon.

Valuation - the story of the future at today's price

Red Cat trades on a market value to sales ratio of over 33, a market value of over $1.4 billion on sales of $40 million - a valuation that reflects not today's business but a speculation on what the business can be. At today's valuation, the investor is paying on the assumption that the business will become a significantly larger and profitable supplier to the U.S. and allied militaries, that it will win more tenders, that international expansion through the NATO catalog will generate recurring revenues, and that gross margins will move into positive numbers.

If these assumptions come to pass and sales reach, say, $150 million to $200 million over two years with a gross margin of around 20%, the stock could look cheap in hindsight at current valuations. But if contracts hit delays, the Gauntlet tender process goes badly wrong or margins don't move out of negative numbers, today's market value will be difficult to defend.

What might surprise the company and what it plans to do

The biggest potential catalyst is the outcome of the Gauntlet tender. This is an expansion of the Army's short reconnaissance program into more mass production, and the winner or winners will win contracts worth potentially hundreds of times the current volume. Red Cat is actively competing in this competition and has the advantage of an existing supplier and a proven product, but the outcome is uncertain and competition is fierce.

The partnership with AeroVironment $AVAV to integrate the FANG system into the P550 platform is a second strong catalyst. AeroVironment is one of the most respected U.S. military drone manufacturers with extensive government relationships, and the AeroVironment contract brings Red Cat legitimacy, distribution and access to a customer base it would have spent years building its way to independently. If this collaboration grows into recurring contracts, the FANG platform could create a second significant revenue stream alongside Black Widow.

Expansion into maritime autonomous systems is the third pillar of the plan. The company is investing in the development of unmanned vessels and their integration with airborne systems, targeting a completely new market where the US Army and allied navies are only in the early stages of defining requirements and standards. Those who enter early with a proven product can have a big advantage in standards development - and Red Cat, with a live demonstration in February 2026, has done just that.

Risks that cannot be overlooked

For all the exciting dynamics, Red Cat has downsides that cannot be downplayed. Independent research in 2025 warned that the company may have misled investors about the true size of the military contract and that some components may have been sourced from Chinese production - just at a time when the company is presenting its "Americanness" as a key advantage. The company has denied these allegations, but it is a warning signal to investors that management communications need to be monitored with a critical eye.

Revenue concentration is another risk - 73% of revenue from a single customer (the US government) means that a change in defence budget priorities, a delay in approval or the loss of a single programme would leave a significant mark on the firm's results. The company is aware of this risk and international expansion through the NATO catalogue is a direct response to the need to diversify the customer base, but this diversification is just beginning.

Investment scenarios

In a positive scenario, the company wins or succeeds in the Gauntlet programme, military orders expand to volumes in the hundreds of millions of dollars, international orders through the NATO catalogue become regular and gross margins reach 20% as production scales. Revenues in 2027 exceed $200 million, the company is approaching operating profitability, and investors are willing to pay a multiple in the 4 to 6 times revenue range for a proven defense industrial business - in such a scenario, today's share price may look like a favorable entry point.

In the base case scenario, the company is growing, but at a slower rate. The Gauntlet delivers tighter-than-expected orders, international expansion runs into bureaucratic delays, and gross margins hover slightly above zero. Revenues in 2027 are around $100 million to $120 million, the company remains loss-making on a net income basis, but the operating loss is narrowing and the balance sheet is still solid. Valuation is difficult to justify in such a world and the stock is likely to oscillate without a significant move up or down.

A negative scenario occurs if the company loses a key tender, military contracts hit defense budget cuts, or the transition from prototypes to mass production proves more costly and slower than management planned. In such an environment, liquidity would begin to decline significantly, the need for a new capital raise would dilute shareholders, and a market value of over a billion dollars would be difficult to defend on sales in the tens of millions with no prospect of profit.

What to take away from the article

• Red Cat is a pure bet on the future growth of the defense drone business, not on today's fundamentals - the company is growing at 161% per year, but is selling below cost of production, does not have a positive gross margin, and has an accumulated loss of over $196 million.

• The key asset is not today's profits, but its position in the military ecosystem: a winner in the U.S. Army's short reconnaissance program, inclusion in the NATO catalog, a partnership with AeroVironment, and an expanding portfolio of systems for air and naval operations create a business position that takes years to build.

• A regulatory law restricting Chinese parts in U.S. government drones is a direct catalyst - and a company with an all-American manufacturing chain is one of the few eligible suppliers in an environment where demand exceeds the available supply of certified products.

• A valuation of over 33 times sales has no support in today's numbers - it's speculation on the company becoming an industrial manufacturer with billions in sales and solid margins, and that speculation may or may not come to fruition.

• For an investor with a high risk tolerance and a long horizon, Red Cat may be an interesting headline in the defense and autonomous systems space - but it requires ongoing monitoring of tender results, gross margin trends and cash position as the most important indicators of progress.

A Monday morning that will go down in the textbooks

Monday, March 23, 2026. 7:04 a.m. Trump publishes a post on Truth Social: the US and Iran are holding productive talks. Strikes on Iranian power plants are postponed for five days.

Within six minutes, by 7:10, the S&P 500 had jumped 240 points. The market capitalization of US equities rose by $2 trillion.

Twenty-seven minutes later Iran denied everything. “No contact with the US took place,” Tehran says.

By 8:00 the S&P 500 had erased half of the gain. $1 trillion in market cap disappeared.

Total swing: $3 trillion in 56 minutes. Just in the S&P 500.

This is not a normal market. This is a market that lives and dies by every post from one man.

What actually happened and why it matters

A minute-by-minute chronology of the wildest morning of the year

The weekend brought an escalation that terrified markets. Trump gave Iran a 48-hour ultimatum—open the Strait of Hormuz or face strikes on your power plants. Iran responded with fresh attacks in the Persian Gulf. Monday morning opened with futures in the red, Asian markets falling, and the KOSPI halting trading.

And then the post came.

7:04: Trump writes about “productive talks” and postpones strikes for five days. Futures explode. Oil collapses from $113 to below $84 per barrel in a single move. S&P 500 futures +240 points, market cap +$2 trillion in six minutes.

7:31: Iran denies the post. “No contact with the US took place.” Oil rebounds. Markets give back some gains.

8:00: S&P 500 is 120 points below the peak of Monday’s move. $1 trillion in market cap gone.

End result for the day: markets still closed significantly higher than on Friday. Dow futures +2.3%, S&P 500 futures +2%, Nasdaq futures +1.9%. The report of postponed strikes was enough to overpower Iran’s denial.

Why oil fell below $84 and what it means

Oil is currently the best barometer of geopolitical tension. Brent traded above $113 before Trump’s post, a historically extreme level. After the post it immediately fell by roughly $29, more than 25%.

That is a huge move in a single asset from a single social-media post.

But note: Goldman Sachs simultaneously raised its oil price targets. Analysts expect Brent to average $85 per barrel in 2026, higher than before the war. Short-term relief hasn’t changed the structural story. Hormuz remains potentially unstable. Iran denies talks. Five days is a very short pause.

Stagflation risk: bonds lost $2.5 trillion

While equities grabbed dramatic attention, an even more frightening number came from the bond market. Since the start of the war with Iran, bonds have lost $2.5 trillion in value—a move analysts compare to 2022.

Stagflation, a combination of high inflation and weak growth, is the worst scenario for central banks. The Fed can’t cut rates because of inflation. But it can’t raise them without choking the economy. It’s a trap with no clean solution.

Gold: the biggest weekly drop since 1983

The paradox of this Monday: at the moment it seemed war tensions were easing, gold plunged more than 4%, with the GLD gold ETF down over 3%. Gold futures traded around $4,388 per ounce.

Why? Because gold had been the main safe haven against geopolitical risk in recent weeks. Once even a hint of de-escalation arrived, investors started selling gold and moving capital back into risk assets. The weekly drop in gold was the largest since 1983.

The social-media market and what it means for me as an investor

This Monday taught me, or rather reminded me, one of the most important investment lessons of recent years.

$3 trillion in 56 minutes. One Truth Social post = one older gentleman with a phone.

This is not a market where the best fundamental analysis always wins. It’s a market where the winner is the one who correctly guesses what Trump will write next. That’s not a strategy, it’s roulette.

What does this mean for me practically? Two things.

1) In an environment of such volatility, positions that require minute-by-minute monitoring are unsustainable. A portfolio built for peace of mind has a huge advantage in this environment. Quality companies with real profits will survive every geopolitical tweet.

2) This volatility creates opportunities—but only for those who are prepared. If you know what you want to buy and at what price, these crazes are your best friend. If you react emotionally to every move, they are your worst enemy.

The social-media market is the new reality. Either you learn to operate in it, or it will continuously surprise you.

What to take away and what to watch out for?

Five days is not peace - Postponing strikes for five days is a diplomatic maneuver, not a structural shift. Iran denies talks. Hormuz remains problematic. The geopolitical premium in oil’s price hasn’t disappeared—just temporarily retreated. If talks don’t progress, we’ll be back to square one in a week.

Bonds are sending a more serious signal than stocks - A $2.5 trillion loss in bonds over the duration of the conflict indicates that professional investors still believe in a stagflation scenario. The equity rally may be short-lived; the bond market is playing a longer game.

Volatility is opportunity, not the enemy - If you have a list of quality companies you want to own at the right price, these environments are your best friends. That’s when assets trade at discounts you wouldn’t see in normal times.

$3 trillion in 56 minutes. The biggest weekly drop in gold since 1983. Bonds down $2.5 trillion. And five days of uncertainty ahead.

This Monday showed what world we invest in today. It’s a world where fundamentals still matter, but where a single social-media post can be stronger than the quarterly results of hundreds of companies combined.

How do you personally react to such swings?

In the stock market, investors naturally focus on companies with a "Buy" or "Strong Buy" recommendation. But they pay much less attention to the opposite end of the spectrum, companies that analysts actively recommend selling. Yet these are the very titles that can provide investors with valuable lessons about what fundamental issues can significantly damage an investment thesis, even for companies with billions of dollars in market capitalization.

"Sell" recommendations are still relatively rare on Wall Street. Most analysts lean toward neutral or positive ratings because negative ratings can damage relationships with corporate management and make access to information difficult. Thus, if a company is in the "Sell" or even "Strong Sell" category , it usually signals that analysts have identified fundamental structural risks that outweigh any potential improvement.

Moreover, the market environment this year has not been friendly to companies with uncertain futures. Higher interest rates, geopolitical tensions and a shift of capital towards quality and profitable companies are creating a mix in which the market is less tolerant of weak fundamentals. Companies that fail to generate stable cash flow or face structural problems in their business model are coming under increasing pressure.

Graham Holdings $GHC

Profile and business model

Graham Holdings is a diversified holding company that was formed from the former The Washington Post Company. The company now operates in the areas of education through Kaplan, television broadcasting, healthcare, manufacturing, and automobile dealerships. The market capitalization is around $4.7 billion and the stock is trading near $1,040 apiece.

Why analysts are hesitant

Although Graham Holdings may look like a value-attractive title at first glance, with a P/E ratio of around 16, the company's conglomerate structure poses a fundamental problem. Diversifying into such distinct segments makes it difficult for investors to accurately value each part of the business. The education segment through Kaplan, while delivering steady revenue, faces increasing competition from online education platforms and pressure on traditional standardized test preparation models.

Thus, while the company is undervalued by the Fair Price Index, this valuation is calculated from DCF and relative valuations. Thus, the analysts' outlook is much tougher for this company and the final verdict, which is negative for the company, looks accordingly. It is thus a beautiful example that even what may appear at first glance to be an interesting investment may not be so.

Revenues for 2025 showed a decline of nearly 6% year-over-year to $213 million in the most recent quarter. Net income for the full year was down 35% to $28 million. In addition, the company announced a $500 million unsecured bond issue in November 2025, increasing the debt burden on its balance sheet.

For NAV-focused investors, Graham Holdings remains a difficult investment thesis where individual segments are difficult to value and the overall trend points to stagnation rather than growth.

Although the stock has hit an all-time high this year, it is now 15.5% away from it.

Polestar Automotive $PSNY

Swedish challenger to Tesla in trouble

Polestar is a Swedish manufacturer of premium electric vehicles that went public through a SPAC transaction in 2022. The company operates under the umbrella of China's Geely Group and offers a range of models from the Polestar 2 sedan to the Polestar 3 and 4 SUVs to the upcoming Polestar 5 sport grand tourer. At first glance, it's an attractive story of a premium EV manufacturer with a Volvo design background, but the reality is quite the opposite.

The SPAC transaction is a way to take a private company public without a traditional IPO, and works on the principle of the so-called "empty company". First, a group of investors or executives form a SPAC, which is a company that has no business of its own and its only goal is to buy another company in the future. This SPAC then goes public and raises capital from the investors, which is held in a trust. Investors at this stage are essentially not betting on a particular firm, but on management's ability to find a quality acquisition target.

Deep structural problems

The reality is much harsher. The consensus rating from analysts is "Sell" with an average target price of $15 per share, with the current price around $16.5. Barclays $BCS reiterated an "Underweight" recommendation in January 2026, and Cantor Fitzgerald downgraded the rating to "Underweight" in February after being disappointed with the company's 2026 outlook. The main problem is the massive cash burn, which has amounted to negative operating cash flow of $1.13 billion over the past 12 months.

Revenues for 2024 are down 14% to $2.03 billion and losses have widened 73% to $2.05 billion. The company's gross margin is negative, specifically -32.7%, meaning Polestar is losing money on every vehicle sold. The company had to do a 1:30 reverse stock split in December 2025 to avoid being delisted from Nasdaq. At the same time, it has repeatedly raised new capital injections, most recently $400 million from Sumitomo Mitsui and $300 million from BBVA and Natixis, to maintain liquidity. The firm's Altman Z-Score of minus 4.77 clearly signals a high risk of bankruptcy.

Alexander's $ALX

New York REIT under pressure

Alexander's is a REIT owning five properties in the New York metropolitan area. The best known of these is the building at 731 Lexington Avenue, which houses Bloomberg's headquarters. The firm is managed through Vornado Realty Trust and has a market capitalization of about $1.22 billion. The dividend yield of 7.52% may look tempting at first glance.

Declining fundamentals and a concentrated portfolio

The analyst consensus at Alexander's is "Sell" with a target price of $190, down about 22% from current levels around $240. Financial results for 2025 confirm analysts' concerns. Revenue fell nearly 6% to $213 million, net income fell 35% to $28 million, and FFO (Funds From Operations) for the fourth quarter fell from $20.8 million to $12.5 million year-over-year.

The main risk is the extreme concentration of the portfolio. The company owns only five properties and its income is heavily dependent on a few key tenants. Dividend coverage has recently declined to approximately 90%, which raises questions about the sustainability of the current dividend level given the continued trend of declining earnings. In March 2026, the company announced the sale of the Rego Park I property to Northwell Health for $202 million, which, while providing one-time income, will further narrow an already very small portfolio.

JinkoSolar $JKS

Chinese solar giant in price war

JinkoSolar is one of the largest solar module manufacturers in the world, with an integrated annual capacity of 130 GW. The company is based in Shangrau, China, and its ADR (American Depository Receipt. This is a financial instrument that allows investors to trade shares of foreign companies on a U.S. exchange without those companies having to be directly listed in the U.S.) are traded on the NYSE. Despite being one of the technology leaders in the solar industry, the analyst consensus is a clear "Sell" with an average target price of $20.33.

Overcapacity and margin pressure

The solar industry is going through a period of massive overcapacity, which has driven panel prices to historic lows. For JinkoSolar, this means a dramatic impact on profitability. Revenues for 2024 are down 22% to CNY92.3 billion and profits have virtually evaporated, falling more than 98% to just CNY55 million. The firm's ROE is negative at minus 16.6%, clearly showing that the firm is destroying shareholder value in the current environment.

Another problem is the US and European tariffs and trade restrictions that are complicating Chinese solar manufacturers' exports to key markets. While JinkoSolar is responding by building manufacturing capacity outside of China, these investments are further weighing on the company's balance sheet.

The debt-to-equity ratio of 1.38 is quite high for a manufacturing firm, and a net debt position of minus $2.56 billion limits financial flexibility. Despite a dividend yield of around 4.7%, the investment thesis remains weighed down by the risk of continued pressure on solar panel prices and uncertainty over supply normalization in the industry.

Comparison of key metrics

Strategic view

When looking at this selection of companies, it's important to note that a "Sell" recommendation from Wall Street doesn't always mean the stock will immediately fall. Rather, it is a signal that analysts are identifying significant risks that they believe outweigh the potential returns.

Each of these companies faces a different type of problem.

Polestar represents the most extreme case, where the very viability of the company is uncertain due to negative margins and massive cash burn. JinkoSolar faces a cyclical overcapacity problem in the solar industry, but one that may last much longer than investors expect. Alexander's is an example of a concentrated REIT portfolio where declining fundamentals are eroding the dividend thesis. Graham Holdings suffers from conglomerate discounting, the absence of a clear growth catalyst, and fragmented attention.

What these companies have in common is that none of them can offer investors a compelling growth story in the current environment. At a time when capital is migrating towards high quality, growing and profitable companies, it is the absence of a positive catalyst that poses the greatest risk to valuation.

What to watch next

• Graham Holdings will report next earnings in May 2026. Investors will be watching to see if the revenue decline can be arrested and if the company indicates any structural changes to the portfolio.

• For Polestar, watch for cash flow developments and the company's ability to achieve at least positive gross margins. Any delay in the ramp-up of Polestar 3 and 4 models could trigger additional capital needs.

• Alexander's is completing the sale of Rego Park I. Watch how the company handles the proceeds from the transaction and whether it affects dividend policy.

• JinkoSolar publishes Q4 2025 results in the last week of March. The key will be whether the margin decline can be arrested and what guidance management sets for 2026.

Conclusion

Stocks with a "Sell" recommendation from Wall Street represent a valuable warning to investors, not an automatic sell signal. Each of the listed titles faces a specific set of challenges: from negative margins and cash burn to cyclical overcapacity and stagnation in a saturated market. Investors should pay close attention to whether there is a real catalyst that could reverse the negative trend.

At the same time, some of these companies may surprise with positive developments in the future. The solar industry has historically moved in cycles, and capacity normalization may return JinkoSolar to profitability, and the likes of Polestar may benefit from the eventual consolidation of the EV market. But the key is to distinguish between mere outlook and fundamental reality. You can analyze each of these stocks yourself using the tools on Bulios according to your own metrics.

Key motivation: on Tesla’s Q4 earnings call, Musk admitted that TSMC, Samsung and Micron have hit the ceiling of their expansion and his companies will surpass them in 3–4 years – so Terafab is, in his view, a necessity rather than an option. That’s a reasonable argument, but at the same time this is a project where Musk has no background in chip manufacturing, he hasn’t released any production timeline and the announcement so far lacks a clear start date – which is nothing new for him.

Into that vacuum, the Trump administration announced the most ambitious public‑private maritime insurance program since the Cold War. The U.S. International Development Finance Corporation named Chubb as lead underwriter for a 20 billion dollar war‑risk reinsurance facility designed to restore insurability for ships willing to transit the strait. Chubb will set pricing and terms, issue policies and handle all claims for qualifying vessels, covering war hull risk, war protection and indemnity and cargo, while the DFC will coordinate a consortium of American reinsurers and define the government eligibility criteria ships must meet before accessing coverage. The scheme is unprecedented in modern peacetime: Washington has decided the geopolitical risk is too large and too systemic for private insurance markets alone to absorb, so a government guarantee sits behind every policy that Chubb writes.

Why the Strait of Hormuz is so critical

The Straitof Hormuz is a narrow sea lane between the Persian Gulf and the open ocean through which roughly 20-25% of the world's seaborne oil traffic flows. Tankers from Saudi Arabia, Iraq, the United Arab Emirates, Kuwait and elsewhere pass through it and it is considered the most important 'energy neck' of the planet.

The US-Israeli conflict with Iran has changed the situation dramatically:

• Iran has declared the strait closed and has been attacking or intimidating ships attempting to pass through since March 4, 2026.

• The US attack on an Iranian war tanker off Sri Lanka further escalated the situation.

• The U.S. further attacked Iran's Kharg oil export terminal, Iran's key export hub for oil.

The result is transport paralysis, oil price spikes and a threat that goes far beyond the Middle East: those who need Gulf oil must pay significantly more or look for alternative sources and routes.

How wartime marine insurance works

Standard marine insurance covers the usual risks - accidents, weather, piracy. War risk is explicitly excluded from these policies and must be bought separately, usually at significantly higher premiums.

In calm times, war insurance premiums for Hormuz passage are a minimal affair. Today it is different:

• Gulf and Strait war insurance premiums have skyrocketed and are effectively unaffordable for many ships.

• Without insurance coverage , the shipowner, the bank financing the ship, and the cargo carrier bear the risk of hundreds of millions of dollars in losses in the event of an attack or seizure.

• Insurance and reinsurance companies such as Lloyd's of London do offer insurance, but with increasingly stringent conditions and rising prices.

The result: even if the war doesn't physically close the Straits, commercial ships without affordable insurance aren't going anywhere. Thus, the insurability of shipping is as important as the physical security of the passage itself.

What Chubb and DFC offer

The U.S. government agency DFC (U.S. International Development Finance Corporation), in partnership with the Treasury Department, announced a $20 billion marine reinsurance plan in early March to make insurance coverage for ships in the Straits effectively available again. Chubb $CB has been selected as the lead underwriter - it will issue the policies, assume the risk and manage all claims.

The structure of the program looks like this:

• Chubb, as direct underwriter, issues policies for eligible ships.

• Chubb is backed by a consortium of US reinsurers as a second layer of risk.

• behind the consortium is DFC as the final safety net, covering losses up to $20 billion on an ongoing basis.

Coverage includes:

• war hull and engine room riskinsurance.

• War P&I insurance (protection and indemnity, covers third party liability, crew, etc.).

• war cargo insurance (war cargo insurance).

The whole scheme is a public-private partnership: the government brings a volume of guarantee that the private market could not handle on its own, and Chubb brings insurance expertise and contractual infrastructure.

Terms and uncertainties

The fundamental question is what Chubb and the DFC mean by "certain conditions" for accessing insurance. The information available so far suggests that:

• ships must meet eligibility criteria set by the US government.

• the US government plans to provide sea escorts for qualified ships.

• The program will initially focus mainly on tanks and cargo ships carrying energy and commercial goods.

The ambiguity around the conditions is an investment issue: if the pool of eligible ships is too narrow or the conditions too stringent, the program may not effectively unlock passage through the Strait, even if it is formally available. One source says bluntly that the program will work "only after the Strait opens" - i.e., that insurance coverage alone will not allow passage unless military tensions ease.

Why the Trump administration is betting on insurance coverage

On 3 March 2026, President Trump publicly pledged that the US would provide political risk cover for ships carrying oil and gas from the Middle East at a "very reasonable price". This was in direct response to the escalation of the conflict and the beginning of the shipping paralysis.

From Washington's point of view, this approach makes strategic sense:

• Restoring the flow of oil pushes down energy prices and eases inflationary pressures at home.

• Insurance guarantees are cheaper than a direct military escort of each ship.

• With a U.S.-government guarantee, Washington seeks to maintain the flow of shipments without the need for a definitive military solution to the conflict with Iran.

It is also a geopolitical signal: the US is signaling that it will not let Iran effectively shut down the global oil market without immediately escalating a direct military conflict.

Impact on markets and the energy sector

The Hormuz crisis is one of the most significant market risk factors today:

• Oil prices have risen sharply following the closure of the Strait, and analysts speak of the potential to reach levels not seen since the oil shocks of the 1970s.

• Energy firms with production outside the Gulf - particularly US oil and gas producers(ExxonMobil $XOM , Chevron $CVX , ConocoPhillips $COP , Pioneer) - are indirectly advantaged by the crisis.

• In contrast, Asian economies heavily dependent on Middle East oil (Japan, South Korea, China) are under significant pressure.

• Shipping companies and tankers such as Frontline $FRO , Euronav or Nordic Tankers $NAT are balancing between extremely high war premiums and potentially very high transit costs in times of crisis.

It is also an extremely exposed project for Chubb itself. As a major underwriter, it bears direct insurance risk, albeit covered by a government reinsurance program. If there are massive losses (billions of dollars worth of ships sunk or seized), the level of compensation and the speed of claims payments will be a major test for the entire structure.

What to watch next

• Whether there will actually be a military lull or a diplomatic breakthrough that actually opens the Strait - without that, insurance cover is just ready-made infrastructure with no use.

• How quickly Chubb and the DFC will publish the specific terms of access to the insurance policy and who will be eligible.

• how oil prices evolve - if the Strait opens and supplies are released, prices could plummet, affecting the entire energy sector.

• whether the scheme will attract other large insurance and reinsurance companies as added partners, which would increase the capacity of the scheme and the credibility of the cover.

The Strait of Hormuz has been closed or heavily restricted multiple times in history - during the Iran-Iraq war in the 1980s, or in various escalations in recent years - but each time it has eventually reopened. The question is not just whether it will happen again, but how quickly and at what cost - to the oil market, to global trade and to the insurance companies that have just bet a total of $20 billion on the survival of this passage.

Cybersecurity company $S has in recent months significantly strengthened its strategy through partnerships focused on artificial intelligence.

The latest step is an expanded collaboration with $NET, which links Cloudflare's global network and edge infrastructure data with SentinelOne's Singularity AI platform.

This integration is intended to enable more advanced real-time threat detection and create a unified security ecosystem that combines network data, endpoint protection and AI analytics. In other words, it's not just a product upgrade but a move toward a platform model of cybersecurity.

Strategic significance: a shift toward “platformization”

The partnership with Cloudflare is important primarily from a positioning perspective. SentinelOne is not trying to compete solely as a standalone endpoint security vendor, but as part of a broader infrastructure.

This trend mirrors a broader development in the industry, where players like Palo Alto Networks or Microsoft dominate by building integrated platforms. For smaller firms, it's therefore key to "plug into" existing ecosystems rather than try to build everything independently.

From an investment point of view, this means reduced risk of product isolation and a higher likelihood of long-term adoption by enterprise customers.

AI as the main growth driver

SentinelOne's entire strategy rests on the belief that the future of cybersecurity is AI-native. The company is developing products like Singularity and Purple AI that automate detection and response to attacks without the need for manual intervention.

Growth in this segment is also supported by structural trends—an increasing complexity of attacks and an expanding "attack surface" due to cloud computing and generative AI. The cybersecurity market is expected to grow at a double-digit pace, and AI plays a key role in product differentiation.

Moreover, the partnership with Cloudflare adds another layer of data (network telemetry), which improves the quality of AI models and their ability to detect threats.

Investment interpretation: positive, but not without risks

From an investor's perspective, AI partnerships are clearly a positive signal. They indicate that SentinelOne is capable of integrating into a broader ecosystem and increasing the value of its product.

At the same time, partnerships alone are not "game changers." The market tends to react moderately because competitors are taking similar steps and AI is gradually becoming a standard rather than a differentiator.

Additionally, the company faces strong competition from major players and pressure on margins. Recent results showed a cautious profitability outlook, signaling that monetizing AI innovations is not immediate.

Key risk: commoditization of AI in cybersecurity

One of the biggest structural risks is the gradual commoditization of AI tools. Once AI capabilities become a standard component of products (e.g., at Microsoft), there may be downward pressure on prices and a reduction in the competitive advantage of smaller players.

This means that SentinelOne's long-term success will depend not only on technology but also on the ability to create a comprehensive platform and retain customers within its ecosystem.

Conclusion: evolution, not revolution

SentinelOne's AI partnerships represent an important step in the right direction, but not a fundamental breakthrough. The company is shifting from a purely product-focused firm to a platform player, which is necessary in the current environment.

For investors, this means a mildly positive signal—confirmation of the right strategy and the ability to keep pace with the market. At the same time, key questions remain about competition, monetization of AI and the sustainability of growth in an environment where AI is rapidly becoming the standard.

The direct link between the framework and Big Tech lobbying is visible in the numbers. Meta, Alphabet, Nvidia, Amazon, Microsoft and others collectively spent more than 100 million dollars influencing U.S. government AI policy in 2025, the first time that threshold was ever crossed, and the results have been tangible: relaxed chip export controls toward China, fast tracked data center permitting and now a federal framework that explicitly tells Congress not to burden AI developers. Critics, however, point out that the framework leaves some of the most contested questions, including how copyright law applies to AI training data, to the courts rather than Congress, and warn that without clear liability rules the resulting legal vacuum could create uncertainty that ultimately slows enterprise AI adoption and hurts the very companies the policy aims to protect.

What the framework specifically proposes

The White House has structured the framework into six pillars:

• Protecting Children and Empowering Parents: Congress should create better tools to manage children's digital presence, especially in the context of AI content and interactions.

• Preventing censorship and protecting free speech: the state must not force technology providers to block or edit content based on a political or ideological agenda.

• Data centre construction: simplify permitting processes and allow on-site power generation to accelerate the expansion of AI infrastructure.

• Fighting AI fraud: strengthening legal tools against AI deepfakes, identity theft and fraudulent schemes using AI.

• Fostering innovation and US dominance in AI: removing unnecessary barriers, access to testing environments, and accelerating AI deployment across sectors.

• Education and AI-ready workforce: investing in retraining and new jobs in the AI economy.

The most important point is federal preemption of state laws: Congress should prohibit states from regulating the development of AI models, prevent them from holding developers accountable for third-party misuse of their models, and replace the "patchwork of fifty misaligned rules" with a single national standard.

Copyright: the biggest unresolved issue

One of the most closely watched areas is the framework's stance on training AI on copyrighted content. The administration's position is clear: training AI models on copyrighted material does not, in its view, infringe copyright. But the Framework also says that Congress should not intervene in this dispute and should leave the decision to the courts.

This is in direct conflict with Senator Blackburn's current Senate "TRUMP AMERICA AI Act" proposal, which would make training on protected content a fair use violation. For companies like OpenAI, Google $GOOG (Gemini), Anthropic, Meta $META (Llama), and Microsoft $MSFT, this issue is existential: if the courts or a later law determined that training on text, images, and videos without the authors' consent is illegal, these companies would either have to pay massive licensing fees or completely rethink what data they train models on.

The framework also supports the creation of collective licensing platforms where rights holders could negotiate with AI firms as a whole without the risk of antitrust lawsuits. This could lead to a compromise where AI firms pay collective fees for access to content but are not exposed to thousands of individual lawsuits.

What this means for specific companies

Nvidia $NVDA is the biggest direct beneficiary of relaxed AI regulation as a whole. The faster AI data centers grow and the less AI development is hampered by regulation, the more GPUs Nvidia will sell. The streamlining of data center construction and federal preemption of state rules directly removes barriers that may have hindered Nvidia customers from expanding capacity.

Microsoft and OpenAI directly benefit from the framework in two areas: the end of fragmented regulation reduces compliance costs across states, and the stance on copyright (letting the courts decide, not legislatively prohibiting training) gives them room to continue their existing access to training data. Microsoft in 2025 has invested heavily in lobbying for just these positions.

Google, Meta, and Amazon benefit from preemption of state laws that nullifies the situation where they had to pursue dozens of different regulatory environments. Meta in particular has welcomed the stance on open-source model development, as the framework does not explicitly recommend restricting access to model development or imposing liability on developers for how third parties use their models.

Anthropic is in a more delicate position. The company has long profiled itself as a proponent of responsible AI development, and its affiliated group Public First Action has directly criticized the framework as hollow and unresponsive to the technology's actual risks. At the same time, Anthropic lost a contract with the Pentagon precisely because of its refusal to comply with all DoD requirements - in an environment where the White House says "less regulation," Anthropic's position as a "responsible" player is more complicated.

The impact on valuations and numbers of AI companies

As such, the framework is not law and has yet to pass Congress, so the direct impact on companies' financial results is indirect. But it is important for AI sector valuations for several reasons:

• It reduces the regulatory risk premium: investors in AI firms had some risk built in that strict regulation (like the EU AI Act) may limit AI deployment and monetization. If the US goes in the opposite direction, this premium shrinks.

• Accelerates data center construction: simplifying the permitting and power situation for data centers directly opens up space for faster CAPEX growth for cloud players and higher demand for GPUs and infrastructure.

• Eliminates compliance costs: companies operating in multiple states don't have to pay for compliance with 50 different regulatory frameworks.

• Opens up commercial AI in regulated sectors: sector-specific regulation (healthcare via FDA, finance via SEC, etc.) is more predictable and will allow AI firms to enter areas such as healthcare, insurance or financial advisory more quickly.

On the negative side, the ambiguity around copyright remains as a "hanging sword" over the whole sector: until the courts decide, companies live in legal uncertainty about the validity of their training data. If a future ruling determines that training on protected content without consent is illegal, it could lead to billions in damages and forced changes in data strategies for some companies.

What to watch next

For investors watching the AI sector, the following signals are key:

• How quickly Congress translates the framework into concrete legislation and whether it can get it done before the election.

• How the courts rule in ongoing cases around copyright and AI training - particularly the cases against OpenAI, Meta and Google.

• whether simplifying data center construction will lead to accelerated CAPEX for AWS $AMZN, Azure, Google Cloud and Nvidia.

• how the framework will affect the international competitiveness of US AI firms vis-à-vis the EU, which has taken the opposite approach - tighter regulation via the AI Act.

The most important variable is whether Trump's light regulatory approach will actually help US AI firms in the global race with China, and whether today's framework will become a workable law or remain a political signal with no real legislative force.

If we look at the real estate sector, VICI Properties is among the most interesting companies you can find in the US. For those who don’t know the firm, it’s a REIT that owns the land and buildings of the most famous casinos in Las Vegas, such as Caesars Palace, MGM Grand, and The Venetian. It’s important to understand that VICI does not operate gambling and doesn’t risk money at the gaming tables. It acts as an owner that leases premises long-term to large operators and essentially just collects rent reliably.

However, the current market situation is not exactly favorable for real estate funds. The main reason is high interest rates, which generally make financing new purchases and debt more expensive for REITs. Additionally, when rates are high, investors often flee dividend stocks to government bonds that offer a similar yield with almost no risk. Given the current tensions in Iran and across the Middle East, we can’t expect rates to fall soon. Oil prices remain elevated, which keeps inflation "sticky" and gives central banks little reason to loosen monetary policy. And I personally think this environment will stay with us at least through this year.

Despite these macroeconomic pressures, VICI has one big advantage: the quality of their lease agreements. These are the so-called "triple-net leases," where the tenant pays not only rent but also all taxes, insurance and maintenance. Moreover, their contracts are structured so they are partially protected against inflation, so the company’s revenues naturally grow with rising prices. Looking at the chart, it’s also visible that the stock is currently at a fairly strong support. The trend of how the price has developed in recent years and bounces off certain levels is quite clear and stable.

From my point of view, this is a great long-term opportunity to add a quality REIT with a nice dividend to a portfolio.

I don’t currently own any other REIT companies, but as additional "anchors" of my portfolio I also have $BRK-B and $WM.

What about you? Do you own any REITs?

Super Micro said it is cooperating with the investigation, suspended Liaw and placed another employee on administrative leave while terminating a contractor, and stressed that the conduct described in the indictment "contravenes the company's policies and compliance controls". The stock fell as much as 27 percent on the news, adding to a long list of reputational shocks that have hit the company since its accounting controversy in 2024, and raises fresh questions for investors about whether the compliance and governance weaknesses that allowed this alleged scheme to reach 2.5 billion dollars in diverted sales are isolated individual failures or something more structural.

What exactly does the US government claim

The indictment accuses Liaw, as well as Ruei-Tsang "Steven" Chang and Ting-Wei "Willy" Sun, of conspiring to smuggle U.S. AI servers into China without the required licenses from the Commerce Department. Super Micro is one of the key manufacturers of servers equipped with Nvidia chips, which are prohibited from being exported to China without government approval for national security reasons.

According to the indictment, the defendants used a combination of:

• false documents claiming the servers were destined for other customers.

• fake "dummy" servers to deceive inspectors during inspections.

• complex transshipment routes through third countries to conceal the true destination of shipments.

The DOJ reports that through this scheme, Super Micro generated at least $2.5 billion in revenue between 2024 and 2025.

Super Micro $SMCIResponse

Super Micro issued a statement emphasizing that the company itself is not a defendant in the indictment and that it is cooperating with authorities. Two employees named in the case have been suspended, and the contractor has been fired.

The company claims that the actions of these individuals violated company policies and compliance systems. However, this is a phrase that investors hear repeatedly with Super Micro - the company also went through a forced suspension of results over accounting irregularities and the threat of delisting from the stock market in recent years before it stabilised its situation.

Why Super Micro is in the spotlight again

Super Micro is one of the largest AI server manufacturers in the United States, and its products power the infrastructure of large hyperscalers and AI companies. It's this role that makes it an attractive target for those looking to bypass US export controls, as its servers equipped with Nvidia chips are among the most powerful platforms available for training and running AI models.

The case fits into a broader trend: the US government has stepped up its fight against illegal AI technology exports to China, and indictments for smuggling Nvidia GPUs are now relatively common. But so far, most of the cases have involved outside players - resellers, shippers and brokers. The Super Micro $SMCI case is more sensitive because the accused are people with direct ties to the manufacturer itself, and one of them is even a co-founder of the company.

Impact on Nvidia and the AI server ecosystem

While Nvidia $NVDA is not charged in the case and is cooperating with the authorities, the case inevitably casts a shadow on it as well. Super Micro is one of its key partners and distribution channels for GPUs into AI servers, and any tightening of controls or shutting Super Micro out of part of the market will affect Nvidia's chip flows.

At the same time, Nvidia itself is under pressure over export controls: CEO Jensen Huang has repeatedly criticized the restrictive policy as economically damaging, estimating that in a single quarter Nvidia lost approximately $2.5 billion in H20 revenue due to the ban on advanced chip sales to China. While the Super Micro case shows that chips are making their way into China despite the controls, it is also a reminder to Nvidia that allegations of "leakage" of its GPUs through third parties may increase political pressure for even tighter regulation of its distribution channels.

Impact on stocks and investment risks

The immediate market reaction - a 14.6% drop in Super Micro - reflects a combination of concerns:

• Legal risk: even if the company as a whole is not charged, investigations may bring additional costs, penalties or restrictions.

• Reputational risk: For a company that has only recently settled an accounting case, each new scandal is a major blow to customer and partner confidence.

• Operational risk: temporary staff suspensions and possible widespread investigations can disrupt business operations.

• Regulatory risk: the US Department of Commerce may tighten the conditions under which Super Micro operates or exports.

Super Micro is in a delicate position: it is a key link in the AI infrastructure supply chain, so any restriction on its operations would affect customers who rely on its servers. If the investigation widens or the company loses certifications and licenses, it could also affect the performance of existing contracts.

Wider context: AI chip smuggling as a growing issue

The Super Micro case is not unique in the US justice space. In the past year, multiple groups have been charged or arrested for illegally exporting Nvidia GPUs to China:

• Operation Gatekeeper in late 2025 uncovered a network of AI chip smugglers and led to the seizure of more than $50 million worth of GPUs.

• In November 2025, the DOJ charged three Chinese nationals with smuggling Nvidia chips through Malaysia and Thailand.

• In August 2025, two Chinese nationals were arrested in Los Angeles for exporting Nvidia GPUs through a network of transshipment companies.

The pattern is similar in all cases: false documents, third-country brokerage firms, and efforts to conceal the final destination of the shipment in China. This suggests that demand for US AI chips in China is so strong that it motivates systematic and well-organized attempts to circumvent controls, and that the US government has intensified its monitoring of these flows.

What to watch next

For investors watching both Super Micro (SMCI) and Nvidia (NVDA), the following questions are key:

• Whether the Justice Department will extend the indictment to the company itself or other employees.

• How Super Micro will operationally manage the loss of two key employees and the potential reputational impact on business relationships.

• whether the case will prompt U.S. authorities to tighten export licensing terms for AI server makers.

• what signal it sends to the entire AI server sector regarding compliance and supply chain controls.

Super Micro has been through several crises over the past two years and has managed to bounce back each time due to strong demand for AI infrastructure. This time, however, the challenge is qualitatively different: it's not just about accounting, but about allegations that people running the company knowingly orchestrated the circumvention of US export laws, with proceeds from the scheme exceeding $2.5 billion. How much this affects the trust of partners, customers and regulators will be a crucial question for the future of Super Micro.

Even after rate normalisation, it has had several years of exceptional results, double-digit EPS, very high margins and is now starting to return capital to shareholders in the form of dividends and other payouts without sacrificing the safety of its balance sheet or the potential for further growth. For the investor looking for a combination of undervaluation, a strong balance sheet and exposure to the mid-size tanker cycle, this is a name that may pleasantly surprise in the years ahead if the tight market holds longer than the market is pricing in today.

Top points of analysis

• The firm's revenues have fallen from a peak of US$1.47bn in 2023 to around US$0.95bn in 2025, reflecting the normalisation of tanker rates - yet net margins remain around 37% and the firm generates strong cash flow even in a 'worse' year.

• The firm trades at a valuation of around 6 times trailing P/E and an EV/EBITDA below 4, with the data suggesting a current fair value of $78.92 per share - the stock is therefore roughly 24% undervalued relative to fundamentals.

• The balance sheet is one of the strongest in the sector: debt-to-equity of just 0.03, debt-to-assets of 0.02 and an Altman Z-score of over 8.7 - the company has virtually no net debt and is sitting on a record cash position, with enterprise value well below market cap.

• The orderbook of mid-size tankers is below 6% of the existing fleet near historic lows, roughly a quarter of the fleet will reach 20 years of age by 2026, and the yards have capacity filled 90% by vessel types other than tankers - creating a structurally tight market for the next few years.

• The dividend yield is around 3% with a very low payout ratio, and the company has room to supplement the base dividend with special payouts or buybacks in good years due to its strong FCF and net balance sheet.

• The firm belongs to the Suezmax and Aframax/LR2 tanker segment, where it has achieved the best spot rates in the last 15 years - averaging USD 26.5k/day for Suezmaxes and USD 38.8k/day for Aframaxes - illustrating the strength of the business strategy and the tightness of the market.

Company performance

Teekay Tankers $TNK was formed in 2007 as a spun-off tanker "pure play" from the broader Teekay Corporation group, which since the 1970s has evolved from a regional shipper to a global carrier of crude oil, LNG and other energy commodities. The aim was to give investors direct exposure to the volatile but potentially highly profitable market for the ocean transportation of crude oil and products without mixing with the group's less cyclical segments. Today, TNK is based in Hamilton, Bermuda, is listed on the NYSE and is a major player in the mid-sized tanker segment.

In terms of fleet, TNK's main focus is on Suezmax and Aframax/LR2 tankers, ships of around 80-160k DWT that are more flexible than giant VLCCs and serve a wide range of routes between key export regions and customers. These vessels carry crude oil and refined products between the US, Latin America, West Africa, Russia, the Middle East, Europe and Asia, and changes in geopolitics and sanctions in recent years have significantly lengthened some routes and increased tonne-miles. Additionally, TNK operates ship-to-ship transfer services, particularly in the U.S. Gulf and Caribbean, which generate fee income and increase fleet utilization.

https://www.youtube.com/embed/4MszJuCJAyw?rel=1

What this business really profits from is primarily the spot and short-term charter market. Teekay Tankers keeps much of its fleet "open" to spot and short contracts, allowing it to quickly capitalize on periods of high rates and regularly report top mid-size spot rates.

In terms of its financial profile, TNK has changed significantly in recent years: it has transformed from a company that carried higher debt in earlier cycles to one with minimal debt, a high equity ratio and record cash. Ratios such as debt-to-assets of around 0.02, debt-to-equity of around 0.03 and Altman Z-score of over 8 indicate an exceptionally strong balance sheet, especially for a sector as cyclical as tankers. Management is leveraging this strength for disciplined capital allocation: instead of aggressively ordering new ships through the cycle, TNK is selectively upgrading the fleet, reducing structural risk while opening up room for shareholder payouts.

The market and the cycle in which TNK plays

The global tanker market is defined by a balance between fleet size and the demand for oil and product transportation in tonne-miles. In recent years, these parameters have swung in a direction that favours tanker companies: demand for oil carriage has stabilised after covide, geopolitics has lengthened some routes, while at the same time the orderbook of new ships has remained very low. Teekay has repeatedly stressed in his market updates that mid-size tanker market fundamentals remain positive, although short-term rate fluctuations are inevitable.

For the mid-size segment, which is TNK's focus, it is key that the orderbook is at a level below 6% of the existing fleet. This means that there will not be a large amount of new capacity entering the market in the next few years, even if companies start ordering new ships today. At the same time, around 11% of the mid-size fleet is over 20 years old and another 14% will reach that age between 2024 and 2026, meaning that almost a quarter of the fleet will soon face either scrapping or costly upgrades due to emissions and safety requirements.

The situation in shipyards is also an important detail. According to data presented by TNK, the global forward cover of shipyards is around 3.5 years, with around 90% of the vessels ordered being non-tanker types. Shipyards thus have orders for containerships, LNG and other segments and do not have unlimited capacity to benefit from any late "rush" for tankers. The combination of a minimal orderbook, an aging fleet and busy yards creates an environment in which even a relatively moderate rate of growth in tonne-mile demand can keep rates above the long-term average for a longer period than is typical of past cyclical periods.

Growth potential - where can it come from and how big can it be

The growth potential for TNK is not in being a classic growth company, but in a combination of three things: a low valuation, a strong market and room for capital returns.

The first source of growth is valuation alone. With earnings per share of around $10 in 2025 and a price that implies roughly 6-7 times trailing P/E, TNK is valued lower than many other cyclical companies with similar or worse balance sheets. If earnings settle in the range of, say, $6-8 per share after rate normalization, and the market is willing to pay 8-10 times earnings for a more stable business, that implies room for tens of percent share price appreciation versus today, purely from a re-pricing multiple.

The second source of growth is the extended mid-size tanker cycle. A low orderbook of under 6% of the fleet and aging ships (roughly a quarter of the fleet in the mid-size segment will reach 20 years of age by 2026) suggest that the favorable environment for rates may last longer than just one or two seasons. Even if rates fall from extreme peaks, they may still remain above levels that correspond to the "old normal", which would allow TNK to hold margins higher than the market currently models.

The third source is capital allocation. A strong balance sheet and cash position give management the ability to combine several actions: continue to renew the fleet at attractive prices, return capital to shareholders through dividends and buybacks, and possibly take advantage of weakening competitors to drive acquisitive growth. If these steps can be taken during a period when valuations are still low, TNK can increase value per share not only through earnings, but also through NAV per share growth and share count reduction.

Management and CEO

The current management of Teekay Tankers is led by Kenneth Hvid, a long-time manager of the entire Teekay group. He has more than two decades in various roles at Teekay - from managing specific segments to the entire group - and thus has experience of several complete tanker cycles, including the high rate period before the 2008 crisis, the subsequent downturn and the current super-cycle.

His approach manifests itself mainly in three areas. The first is debt and risk. After previous years when parts of the group were more leveraged, TNK has moved under his leadership to a structurally low debt position, with debt-to-asset and debt-to-equity ratios well below the sector average. The second area is disciplined capex - the company has not used the recent strong years to aggressively order large numbers of new ships, but rather to selectively renew and modernise its fleet. The third area is the flexibility of capital returns: the dividend was only reinstated when the balance sheet could bear it, and is now set to be extended in the future with special payouts or buybacks if results are good and valuations are low.

Importantly for the investor, this is management that has seen both sides of the cycle and seems to prefer long-term capital return over short-term "chasing" fleet size or dividend yield. In an environment where it is easy to succumb to the temptation to go into debt to order new ships, this is a plus rather than a minus.

The market in which TNK operates - outlook and expectations

Teekay Tankers operates in the global marine transportation market for crude oil and products, with its centre of gravity in the midstream segment. Demand for these services is driven by the volume of crude oil transported, the length of routes and the structure of trade flows. Sanctions on Russian crude oil, growth in exports from the US and Middle East and changes in refining capacity around the world have led to the extension of a number of routes in recent years, thereby increasing tonne-miles.

The medium-term outlook for tanker shipping demand is rather positive, although not linear, according to Teekay and independent analysis. Global oil demand is expected to hover close to current levels in the coming years, possibly rising slightly depending on economic developments and energy policy. At the same time, it will not be easy to scale up the fleet quickly with new ships, as shipyards have contracted production for several years ahead and most slots are occupied by other vessel types.

TNK itself has repeatedly stated in its market updates that the fundamentals for the mid-sized tanker market remain positive. Management anticipates that while rates will remain volatile between quarters, the low orderbook and aging fleet will support solid average rates over the next few years. In addition, in Q4 2025 Teekay indicated that it expects rates to remain strong in Q1 2026 due to seasonal factors and continued tonne-mile demand, which combined with the state of the fleet suggests that it does not view recent strong quarters as a one-off anomaly.

What may surprise the company in the future

The scope for pleasant surprises in the years ahead lies mainly in what TNK can do beyond the baseline scenario.

• It can accelerate capital returns if it maintains high free cash flow and low debt.

  • That could mean a more stable/higher ordinary dividend, more frequent special dividends, or more aggressive buybacks in periods when the P/E remains low.

• Can profitably renew and optimize the fleet.

  • By selling older ships at times when the value of second-hand vessels is elevated, and buying younger units when opportunities arise, TNK can increase value per share even without growth in the number of ships.

• It can take advantage of any weakening of competition.

  • If some of the less disciplined players come under pressure in the next cycle due to debt, Teekay's clean balance sheet gives it the opportunity to buy quality ships or entire portfolios in forced sales.

• It can benefit from the fact that regulation and ESG pressures will accelerate the phasing out of old ships faster than the market now expects.

  • If older units are scrapped earlier due to emissions standards, supply tensions may intensify, which would support rates and profitability for longer.

Results and figures

On a results level, Teekay Tankers has had a very strong couple of years. It has sales of around $1.47 billion in 2023, around $1.23 billion in 2024 and around $0.95 billion in 2025, reflecting the normalisation of rates after an extremely strong period, but still levels well above historical averages. Gross margins are around 27-28%, operating margins around 27% and net margins for 2025 around 37%, which is very attractive for a capital intensive, cyclical business.

Net income for 2025 was about $351 million, compared to $404 million in 2024 and $520 million in 2023. EPS for 2025 is around $10 per share, after two previous years of $11-15, with a relatively stable share count of about 34.5 million. This, coupled with very low debt, leads to a valuation of around 6 times trailing earnings, which remains conservative even with expected profitability normalization.

The cash flow picture is similarly strong. TNK has generated robust operating cash flow in recent years, which, combined with relatively moderate capex, has led to high free cash flow and double-digit FCF growth over time. This has allowed it to virtually eliminate net debt, fund its fleet renewal program, and at the same time begin regular payouts to shareholders. In Q4 2025, the company reported a record cash position and emphasized that its capital strategy will continue to combine prudent fleet investment with disciplined capital returns to shareholders as valuations and market conditions permit.

Dividend and capital allocation

Dividend at TNK is an add-on to the cyclical story, not its main axis. After a period when the priority was clearly on reducing debt and stabilising the balance sheet, the company has restored the dividend in recent years and gradually increased it. It currently pays about $0.25 per share per year, which at the current price represents a dividend yield of about 3%.

The payout ratio is very low given the high EPS, so the base dividend represents only a small portion of earnings, leaving the firm with wide leeway. Moreover, Teekay is not dogmatically focused on just regular quarterly payouts - in periods of exceptionally strong performance, management has the option to supplement the base dividend with special payouts.

Capital allocation is guided by several priorities: maintaining a very strong balance sheet, prioritizing returns over growth at all costs, continuously rejuvenating the fleet, and only then increasing direct payouts to shareholders. This means that in a period of high rates, an investor need not expect an extreme dividend yield of 10-15%, but rather a combination of a reasonable base dividend, potential special payouts and growth in the value of the company through an improving fleet and balance sheet. In a weaker market, TNK is then likely to take the dividend flexibly and possibly adjust it to protect the balance sheet - but the main difference from the past is that today's balance sheet gives it much more room to manoeuvre.

Balance sheet and valuation

Teekay Tankers' balance sheet is one of the strongest in the sector. The debt to assets ratio is around 2%, the debt to equity ratio is only 0.03 and the Altman Z-score of around 8.7 is well above the point where an investor would need to worry about solvency or restructuring. Moreover, the company holds a large amount of cash, so its enterprise value is well below its market capitalization, which in practice means that shareholders have both fleet value and a liquidity "cushion".

From a valuation perspective, the stock trades at roughly 6-7 times trailing P/E, with EV/EBITDA around 3-4 times and P/B near 1.1. Analysts in the "deep value" community point out that with this combination of balance sheet and profitability, it is still an undervalued title, even though some of the good news is already in the price. The analyst consensus sees room for further upside in TNK's price target, though it warns of cyclical risks and the possibility of earnings normalizing to lower levels in the coming years.

The key argument for undervaluation is therefore the difference between how the market values current profits (as highly cyclical and temporary) and how long the fundamentals suggest the favourable conditions of an ageing fleet and low orderbook can last. If it turns out that the "new normal" of rates is higher than in the past, and that TNK retains some of today's margins longer than the market expects, the stock has the potential to combine rerating multiples with continued high cash flow generation.

Investment Scenarios

In a positive scenario, the tanker market for the mid-size segment will remain tight for most of the next decade. The orderbook will remain constrained, an aging fleet will accelerate the scrapping of old ships, tonne-miles will be supported by geopolitical factors and global oil demand will remain close to current levels. In such an environment, Teekay Tankers will maintain a portion of today's rates, continue to generate high FCF, upgrade the fleet and increase payouts to shareholders; the stock has room for a combination of solid dividend yield and significant capital appreciation.

In the base case, rates gradually normalize to levels that are below today's levels but still above the long-term average as the orderbook and aging fleet keep the market in a relatively healthy balance. TNK's EPS will fall from today's double digits to mid-single to low double digits, but the company will retain flexibility, a reasonable dividend, and the ability to continue fleet renewal due to its strong balance sheet. In such a world, the stock could move from "deep value" status to a valuation closer to mid-cycle multiples without an investor having to rely on extreme scenarios.

A downside scenario would combine a steeper decline in oil demand, changes in export flows that shorten routes and reduce tonne-miles, and a potential freeing up of yard capacity leading to a wave of new tanker orders. Rates could fall near or below levels that cover fully loaded cash breakeven, EPS would thin significantly, and the company would have to realign the dividend and pace of investment. A strong balance sheet should allow it to weather such a period in better shape than weaker competitors, but for shareholders it would be a classic cyclical "drawdown" where undervaluation could persist for some time.

What to take away from the article

• Teekay Tankers is a pure play on the mid-size tanker cycle, not a dividend title or a defensive position - investors need to be prepared for cyclical volatility as earnings will move significantly over time as rates move.

• The company is entering the rate normalization phase from an exceptionally strong position: near-zero debt, record cash, an Altman Z-score of over 8, and an enterprise value below market cap are a rare combination for the tanker sector, giving the investor a safety margin that is not the norm for cyclical stocks.

• Valuations around 6 times trailing P/E and EV/EBITDA below 4 may look like a trap as the market factors in rapid earnings normalisation - but the key question is whether an orderbook below 6%, an ageing fleet and busy yards will keep rates above the "old normal" for longer than consensus models.

• According to the data, the stock is currently undervalued by about 24% against a fair value of $78.92, and this discount is not due to poor business quality but cyclical pessimism - a material difference for a value investor with a longer horizon.

• Management, led by Kenneth Hvid, has so far shown that it prefers capital discipline to expansion at any cost - that's a plus, but it's also true that the main "test" will come when rates actually fall and the company has to demonstrate its ability to counter-cyclically allocate capital.

• For an investor who understands the tanker cycle and is willing to follow the evolution of rates, orderbook and capital allocation as closely as quarterly earnings, TNK may be an interesting medium-term bet on a combination of re-rating multiples and continued FCF - but for an investor looking for stable passive income or a defensive position, it will be too cyclical a profile.

Personally, I find this sector quite attractive, but I haven't invested in it yet and I'm looking for a suitable stock. So far I'm most interested in $RKLB and $ASTS. SpaceX could also be interesting once it goes public.

I don't understand this sector very well yet, so I'd appreciate your opinions, tips, or warnings about any potential risks.

The energy market in 2026 is going through one of the most turbulent periods since the oil shock of the 1970s. Coordinated US and Israeli military strikes on Iranian infrastructure in late February triggered an avalanche of events that completely redrew the global oil market map. The closure of the Strait of Hormuz, through which around 20% of the world's oil production normally flows, drove prices to levels that analysts were talking about as an unlikely extreme scenario just six weeks ago.

According to the latest data from the International Energy Agency, Brent futures reached over $114 per barrel in the short term during March 2026, while the EIA revised its March outlook for the average Brent price for the full year 2026 to $78.84 per barrel, up from its original estimate of $57.69. At the end of the trading session on 19 March, Brent was trading around USD 108 and WTI near USD 96 per barrel.

The war premium, which Goldman Sachs $GS estimates at around $18 per barrel, remains firmly anchored in prices and will most likely be present throughout the first half of 2026.

This macroeconomic context creates an exceptional environment for oil producers and refiners outside the Gulf. US energy companies, which are benefiting from higher oil prices without direct exposure to the logistical turmoil in the Strait of Hormuz, are now enjoying increased investor interest.

Occidental Petroleum $OXY

Company profile and market position

Occidental Petroleum is one of the largest oil and gas producers in the United States with a strong concentration in the Permian Basin, the most productive oil basin in the world. At the same time, the company is building a unique strategy combining traditional production with Direct Air Capture (DAC) carbon capture and storage technologies, making it a player with an asymmetric profile towards different energy transition scenarios. Revenues for fiscal 2025 reached $21.6 billion.

The most significant move of 2025 was the sale of OxyChem's chemical division, completed on January 2, 2026, which immediately reduced the company's debt by $5.8 billion. This brought total debt down to $15 billion, down from nearly $40 billion in 2019 after the Anadarko acquisition. This structural shift significantly improved the company's financial resilience and flexibility just as the energy market entered a phase of extreme volatility.

Key financial indicators

Breakeven oil price - the minimum price per barrel of oil where the company is covering all of its costs

Berkshire Hathaway and the geopolitical premium

The presence of Berkshire Hathaway as the largest shareholder remains a key stabilizing factor for investors. Warren Buffett has repeatedly reaffirmed this position, buying up shares of $OXY on every major downturn. This institutional anchor significantly limits downside volatility and signals long-term confidence in the firm's business model even in an environment of price shocks.

In addition, Occidental is benefiting from one of the current oil shocks. While Gulf producers face direct production constraints caused by the closure of the Strait of Hormuz, US production in the Permian Basin continues uninterrupted. Meanwhile, a breakeven price below $40 per barrel means that even with a significant cooling of the geopolitical premium, the company will remain strongly profitable.

The STRATOS project and the long-term outlook

The second pillar of OXY's long-term strategy is CO2 capture technology through the STRATOS project, the world's first commercial DAC station, which is expected to be fully operational by mid-2026. This investment is strategically interesting for two reasons: first, it opens up access to premium-priced carbon neutrality credits from large corporate customers, and second, it builds a platform that can make OXY a relevant player even in the tougher regulatory environment of the future. Analysts point to the launch of STRATOS as a key catalyst for potential share repricing in the second half of the year.

Shares of $OXY have already gained 45% this year.

Phillips 66 $PSX

From a single refiner to a diversified energy player

Phillips 66 has undergone one of the most significant transformations in its industry over the past two years. The company, traditionally viewed as a pure refining colossus dependent on volatile crack spreads (the difference between the price of crude oil (input) and the price of refined products like gasoline or diesel (output)), has been systematically repositioning its business model toward a more diversified and stable source of revenue. In the words of CEO Mark Lashier, 2025 was "a year of transformation" and the fourth quarter results confirm this assessment.

Revenue for the full year 2025 was $132.19 billion, with net income up more than 108% year-on-year to $4.39 billion. In Q4 2025, the company reported adjusted earnings per share of $2.47 versus the consensus estimate of $2.14. Shares of $PSX have appreciated approximately 24% since the beginning of 2026 and are now trading at an all-time high of just under $180.

Key Financial Highlights

Refinery recovery and structural improvements

The refining segment, which was the company's biggest weakness a year ago, is undergoing a significant turnaround. Phillips 66 achieved a record net product yield (88%) in Q4 2025 and operated at 99% of refining capacity. A key factor is the ability to process Canadian heavy crude at a significant discount to the WTI benchmark, and the closure of the Strait of Hormuz will widen this differential further in favor of US refiners.

The Midstream segment complements the refining business with steady revenues from NGL transportation and fractionation. The acquisition of a 50% stake in WRB and the takeover of Coastal Bend have strengthened the company's competitive position, and management has signaled EBITDA guidance of $4.5 billion by the end of 2027. Goldman Sachs $GS recently raised the target price to $186.

LA refinery closure as a strategic move

One controversial but strategically logical move was the plan to close the Los Angeles refinery. Although the company booked $239 million of accelerated depreciation in Q4 2025, it is getting rid of one of the most costly and regulatory-intensive assets in the portfolio. The result is a portfolio with higher average margins and lower environmental liabilities, consistent with the strategy to optimize capital allocation to the most profitable segments.

ConocoPhillips $COP

World's largest independent oil and gas producer

ConocoPhillips is the world's largest independent oil and gas company by production and is one of the few players in the industry whose portfolio diversification includes both premium Améric assets (Permian Basin, Eagle Ford, Bakken) and a global presence in Canada, Norway, Qatar and other regions. The Marathon Oil integration, completed in 2024, added high-quality oil reserves and management reports achieving double the originally planned synergies in the acquired regions.

Fourth quarter 2025 delivered mixed results, with revenue of $14.19 billion, adjusted EPS of $1.02 (slightly below consensus of $1.08) and operating cash flow of $4.3 billion. The main negative factor was lower average realized oil prices, USD 42.46/bbl vs. USD 52.37 in the same period in 2024. Production, on the other hand, surprised positively: total output in Q4 2025 reached 2,320 Mboed (thousand barrels per day), up 137 Mboed year-on-year.

Key financial indicators

Cost reduction program and path to free cash flow

Management's most significant commitment for 2026 is to reduce capital expenditures and operating expenses by more than $1 billion compared to 2025. CEO Ryan Lance has repeatedly called this a "margin improvement initiative" and the company has already shown measurable progress. Well productivity in the Permian is up 8% year-over-year in 2025 in terms of oil production per foot of well length, and 7% in the Eagle Ford. These internal efficiencies allow the company to maintain production at a lower CapEx.

Four large projects, Willow in Alaska (scheduled to begin production in 2029), LNG projects in Qatar, the Surmont expansion in Canada and the Permian Basin development, are the backbone of the plan to double free cash flow by the end of the decade. Management estimates incremental free cash flow from these projects of $7 billion by 2029 compared to a 2025 base year. The Willow project, approximately 50% complete, alone should add approximately $4 billion of annual free cash flow.

The current geopolitical premium in oil prices significantly improves the near-term outlook. With Brent prices above USD 90 per barrel, COP is generating significantly stronger cash flow than the original forecasts for 2026. At the same time, the company has one of the lowest breakeven points in the industry, at around USD 40 per barrel, giving it exceptional resilience even in the event of a rapid price decline following de-escalation of the conflict.

Comparison of key indicators

An overview of key metrics for a quick comparison of companies:

Strategic view

All three companies are direct beneficiaries of increased prices due to the current geopolitical shock, but for different reasons and with different risk profiles.

Occidental $OXY offers the purest exposure to oil price movements in the DAC technology space. The drastic reduction in debt eliminates the existential risk that accompanied the company after the Anadarko acquisition and opens up the space to return capital more aggressively to shareholders. An investor who believes in a continued war premium in oil prices or a future regulatory premium for carbon emissions will find an asymmetric risk/reward ratio in $OXY.

Phillips 66 $PSX provides a different type of exposure. As a refining and midstream company, it benefits from a wide differential between the price of heavy oil and the resulting products, and the ability to process Canadian heavy oil from discounted sources is a distinct competitive advantage in the current environment. A diversified model with five segments (refining, midstream, chemicals, marketing and renewable fuels) reduces dependence on a single price driver and stabilizes earnings. The combination of record efficiency, strong cash flow and strict capital discipline makes $PSX an attractive choice for investors seeking energy exposure with lower commodity volatility.

ConocoPhillips $COP represents the "strength and discipline" category. It is the company with the broadest global diversification, the deepest inventory of drilling locations in the United States, and a clear plan to double free cash flow by the end of the decade. The goal of returning 45% of cash flow from operations to shareholders in 2026 while maintaining an ambitious investment program is an exceptional commitment in the industry. Lower near-term sensitivity to oil price movements (versus $OXY) remains a key risk as some production comes from regions with lower realized prices.

What to watch next

Key factors to focus on in the coming months:

• The length and intensity of the Hormuz crisis: each week the Strait is closed means approximately 10 billion barrels less oil transported. The length of the disruption will determine whether oil companies benefit from a short-term shock or a longer-term restructuring of supply chains.

• Launching STRATOS ($OXY): the first DAC station is expected to begin operations in mid-2026. Its operational results will test whether OXY can monetize carbon credits at commercial scale.

• Refining Margins and Crack Spreads ($PSX): Crack spreads (the difference between the price of crude oil and the resulting petroleum products) remain a key indicator for Phillips 66. In an environment of supply disruptions from the Middle East, they could continue to rise.

• Willow development ($COP): the Alaska project represents the company's largest single generator of future cash flow. Any regulatory or environmental complications could rewrite the outlook.

• OPEC+ moves and production easing: the eight OPEC+ members agreed in March 2026 to add 206k b/d from April. If the geopolitical premium subsides faster than the market expects, there could be a rapid correction in oil prices and thus energy company valuations.

• De-escalation of conflict and opening of the Strait of Hormuz: Israeli Prime Minister Netanyahu's March 19 statements about helping to reopen the Strait signal a possible de-escalation that could significantly reduce the war premium in oil prices. Investors should watch these diplomatic negotiations very closely.

Is the era of the oil giants beginning?

The year 2026 reminds investors that the energy sector remains an indispensable part of the global economic system and that geopolitical events could completely redraw the investment map in a matter of weeks. Occidental Petroleum, Phillips 66 and ConocoPhillips are three companies with different business models and risk profiles, but they share one common characteristic: a distinct ability to generate free cash flow even in a lower oil price environment.

A key factor in the future course of all three will be the length and intensity of the geopolitical disruption in the Persian Gulf and the ability of management to maintain capital discipline no matter which way oil prices go.