Microsoft announced $190 billion in capital expenditures (capex) for 2026, with $25 billion of that attributable to rising component costs alone. That is 61% more than the previous year and a figure that exceeded the analyst consensus by $35 billion. The market is thus faced with a simple yet difficult question: Is Microsoft building the world’s largest and most profitable AI infrastructure, or is it paying an astronomical entry fee into a race whose outcome no one yet knows?

Key Points of the Analysis

• Revenue in the third fiscal quarter of 2026 reached $82.9 billion, representing year-over-year growth of 18%. Operating income rose 20% to $38.4 billion, and net income rose 23% to $31.8 billion. These are figures that most tech companies would consider an exceptional year; Microsoft posted them in a single quarter.

• Azure and cloud services grew 40% year-over-year, and the entire Intelligent Cloud segment reported revenue of $34.7 billion, up 30%. This marks the fifth consecutive quarter of accelerating growth. Nevertheless, management acknowledged that growth is being hampered by capacity constraints, as demand exceeds the available infrastructure.

• The AI business exceeded an annual run rate of $37 billion, representing 123% year-over-year growth. By comparison, Palantir expects total revenue of approximately $7.6 billion this year. Microsoft generates more than four times that figure from its AI business alone.

• The number of paid Copilot seats has risen to more than 20 million, a 33% sequential increase from the 15 million reported in January 2026. The number of customers purchasing more than 50,000 seats has tripled over the past twelve months. The numbers are growing rapidly, but they still represent only 3.3% of the more than 450 million commercial M365 users.

• The commercial backlog reached $627 billion, up 99% year-over-year. Approximately 30% of this volume is expected to convert to revenue over the next twelve months. Such revenue visibility is unparalleled in the technology sector.

• Capex for 2026 was increased to $190 billion, of which $25 billion is attributable to rising component costs, not additional capacity. As a result, free cash flow fell 13% year-over-year in the last quarter, despite record operating profitability. This tension between operating results and the investment cycle is currently the main source of uncertainty.

• The trailing P/E ratio hovers around 22–24, while the forward P/E ratio is around 32. The consensus among 56 analysts is a “Strong Buy” with an average target price of over $561—roughly 60% higher than where the stock is trading today.

What changes have taken place at the company?

For most of its modern history, Microsoft $MSFT has been viewed as a high-quality but slow-growing tech giant—a company with dominance in Windows and Office, stable dividends, and predictable growth of around 10% per year. Satya Nadella began rewriting this narrative when he took over as CEO in 2014, betting on the cloud at a time when that was far from a given. Azure evolved from an internal project into the world’s second-largest cloud platform. But that was just the first act.

The second turning point came with the advent of generative AI and the partnership with OpenAI. Microsoft has invested a total of approximately $13 billion in OpenAI and was the first major tech company to integrate large language models directly into its most widely used products—Teams, Word, Excel, Outlook, and GitHub Copilot. This changed the fundamental logic of the business. Previously, Microsoft sold access to tools. Today, it sells access to tools plus an AI layer that uses these tools on behalf of the user or in collaboration with them, and charges for this separately.

A key symbol of this shift is Microsoft 365 Copilot, offered as a paid add-on to existing subscriptions for $30 per user per month. The premise is straightforward: enterprise customers are already paying for M365 on a large scale, and Copilot transforms this installed base into a business with higher average revenue per seat, while the AI infrastructure enables the product capabilities themselves. Microsoft has thus ceased to be merely a software company and has become the operator of one of the largest AI infrastructures in the world, with all the benefits and costs that come with it.

The third change is less visible but strategically crucial: Microsoft is shifting away from its reliance on OpenAI toward its own models. The company is developing seven proprietary AI models, its own AI hardware codenamed Project Solara, and the autonomous assistant Microsoft Scout, which is integrated across Microsoft 365. The goal is to reduce the concentration risk associated with the OpenAI partnership and, in the long term, to control its own AI stack from the silicon level all the way up to the application layer. This transformation from a distributor of third-party models to a developer of its own AI platform is the biggest strategic shift in the company’s history since its transition to the cloud.

How does the company make money?

Intelligent Cloud: $34.7 billion, +30% year-over-year

The Intelligent Cloud segment includes Azure and other cloud services, server products, GitHub, and Nuance cloud services. It is now the heart of the entire business and the focus of the vast majority of investments. Azure alone grew by 40% year-over-year in Q3 FY2026 on a constant-currency basis, marking the fifth consecutive quarter of acceleration. The business model is based on customers paying for computing power, storage, databases, and AI inference based on actual usage, with long-term contracts ensuring revenue predictability. Microsoft does not report margins for the segment separately, but the company’s overall operating margin of 46% suggests that the cloud business is driving profitability significantly above average. The main constraint on growth today is not demand but physical capacity; data centers are struggling to keep up with customer demand, and management has acknowledged that the company will remain capacity-constrained throughout 2026. This also makes a strong case for massive capital expenditures: without new infrastructure, Azure would be unable to accommodate the customers currently waiting in line.

Productivity and Business Processes: $35.0 billion, +17% year-over-year

This segment includes Microsoft 365 for both the commercial and consumer markets, LinkedIn, and Dynamics 365. Microsoft 365 Consumer Cloud grew 33% year-over-year, Dynamics 365 by 22%, and LinkedIn by 12%. This is the most stable part of the business with the highest revenue visibility; most revenue comes from recurring subscriptions with annual or multi-year contracts. This is where the company’s key monetization strategy is playing out: Copilot as a paid add-on for $30 per user per month. With 20 million paid seats, Copilot represents an annual run rate of approximately $7.2 billion—a figure that is growing by roughly $7 million per day. The segment is also important because LinkedIn and Dynamics 365 provide diversification beyond the purely cloud-based and productivity-focused business. LinkedIn, in particular, is benefiting from growing corporate demand for recruitment and B2B marketing tools in an environment where AI is transforming the job market.

More Personal Computing: $13.2 billion, −1% year-over-year

This segment includes Windows, Xbox, Surface, and Bing Search. It is a declining part of the business that interests the market the least when evaluating the company. Windows revenue is tied to the PC sales cycle, which remains under pressure. Bing, with its AI integration, saw an influx of users following the integration of ChatGPT, but it has not managed to significantly challenge Google’s $GOOG dominance in search. Xbox is an interesting case: the Game Pass subscription service is generating recurring revenue, but hardware sales are under pressure. Microsoft recently announced a price increase of $100 to $150 for Xbox consoles due to higher costs for memory and components. The segment as a whole does not significantly contribute to the company’s growth thesis, nor does it undermine it—its stable revenue contributes to diversification and reduces dependence on a single segment.

What is the main growth driver?

The main driver is Azure and the broader AI infrastructure business. It’s not just that Azure is growing the fastest of all three segments—it’s that it’s accelerating at a time when most cloud competitors are struggling to keep pace. Forty percent year-over-year growth in constant currency marks the fifth consecutive quarterly acceleration—a pattern that is virtually unprecedented for a platform of this size. Under normal circumstances, the larger the user base, the harder it is to maintain a high growth rate. Azure is defying this rule for now.

The reason is structural. Demand for AI computing power is growing exponentially, and Azure is one of only three platforms in the world—alongside AWS and Google Cloud—that can meet that demand at an enterprise scale. Microsoft’s AI business has surpassed an annual run rate of $37 billion, with year-over-year growth of 123%. This figure isn’t just about Copilot; it includes all AI services running on Azure, from inference and model training to specialized AI APIs. Companies building their own AI products need infrastructure. And Microsoft sells it to them.

The second driver, which is just gaining momentum, is Copilot in the Productivity and Business Processes segment. Seats grew sequentially by 33% in a single quarter, and the number of customers purchasing more than 50,000 seats tripled over the course of a year. The wave of large-scale enterprise deployments is therefore just beginning. The economics of the model are key here; Microsoft doesn’t need to convince customers to switch to a new platform. It only needs to convince existing M365 users to pay extra for Copilot. The sales cycle is dramatically shorter, and customer acquisition costs are a fraction of what it would cost to build a new customer base from scratch.

The third driver, which is less frequently discussed, is the commercial backlog as an indicator of future revenue. Commercial remaining performance obligations reached $627 billion, up 99% year-over-year, with approximately 30% expected to convert to revenue over the next twelve months. This means that approximately $188 billion in revenue is already contractually secured for the next twelve months, even before a single new order is received. Such visibility gives management the leeway to invest aggressively without risking that revenue will not follow.

Products and Upcoming Strategy

The centerpiece of the product strategy is the shift from tools to agents. Microsoft has stopped building software that waits for user instructions and has started building software that acts autonomously. Copilot was the first layer: an AI assistant integrated into existing products that helps with writing, summarizing, and analyzing. Agentic AI is the second layer: autonomous systems that take over entire workflows without ongoing human input. The rapid adoption of Copilot Studio and the measurable return on investment for specialized agents suggest that autonomous AI workflows will become a larger source of revenue than the general-purpose Copilot assistant. Microsoft itself predicts that agentic AI revenue will surpass traditional Copilot revenue by the second quarter of fiscal year 2027.

The hardware layer is the second pillar of the strategy, one that investors often overlook. Microsoft is developing its own AI hardware under the codename Project Solara and the autonomous assistant Microsoft Scout, which is integrated across Microsoft 365. The goal is to reduce dependence on Nvidia chips, control inference costs, and, in the long term, build its own silicon stack—similar to Apple $AAPL in the consumer electronics sector or Amazon $AMZN with its Trainium chips. This is a long-term bet, and results won’t materialize for at least two to three years, but the strategic intent is clear: a company that controls both hardware and software structurally enjoys higher margins and less dependence on the supply chain.

The third pillar is quantum computing, which does not yet contribute to revenue but could redefine the future of computing power. Microsoft has developed the Majorana 2 chip and set a goal of achieving scalable quantum computing by 2029. This area is highly speculative and the timeline is uncertain, but Microsoft is one of the few companies with both the resources and the research base to achieve a real breakthrough. For today’s investor, this is more of a free option than a core part of the investment thesis.

The fourth and most immediately important element is the multi-model strategy. Microsoft is consciously moving away from exclusive dependence on OpenAI. The company is developing seven proprietary AI models while also integrating models from Google, Meta $META, and other providers into its products. In doing so, Microsoft is shifting into the role of an AI coordinator—a platform that selects, combines, and deploys the best models according to customer needs—rather than a dependent distribution channel for a single provider. For customers, this is an argument for vendor lock-in with Microsoft rather than with a specific model. For Microsoft, this serves as a safeguard in case OpenAI loses its technological leadership or changes the terms of the partnership.

Market Share and Market Position

Microsoft does not operate in a single market; it operates in several simultaneously and holds a different position in each. This diversification is one of the company’s greatest structural advantages, but it also complicates simple comparisons with competitors.

In enterprise cloud infrastructure, Microsoft is the second-largest player in the world behind Amazon’s AWS. Azure holds approximately 22–25% of the global cloud infrastructure market, and this share has been growing slightly over the past two years at the expense of smaller players. Microsoft’s key advantage over Amazon is not the technology itself, but integration. A company that already pays for Microsoft 365, Teams, and Windows has a natural incentive to add Azure—everything is in a single ecosystem, under a single contract, with a familiar sales contact. This integration creates a moat that is difficult to cross, even if a competitor offers a lower price for computing power.

In productivity software, Microsoft’s position is even stronger. Microsoft 365 is used by 99% of Fortune 500 companies, and its total commercial user base exceeds 450 million. That’s a starting position that competitors like Google Workspace can only dream of. The strength here does not lie in a better product; it lies in the costs of migration. Migrating an entire organization from Outlook to Gmail, from Teams to Meet, and from SharePoint to Google Drive is a project that takes months, costs millions of dollars, and requires retraining employees. Most companies simply won’t take this step unless Microsoft falls significantly behind.

In the AI space, Microsoft’s position is the newest, but it’s gaining ground the fastest. With an AI business boasting an annual run rate of $37 billion and 123% year-over-year growth, Microsoft is positioned as the world’s largest commercial AI distribution channel. No other company sells AI capabilities to such a broad range of enterprise customers. Copilot isn’t just a product; it’s a distribution lever that transforms the existing installed base into AI revenue without the need to build a new customer base from scratch.

The only segment where Microsoft is maintaining rather than building its position is search. Bing, with its AI integration, has seen an influx of users but has failed to significantly challenge Google’s dominance. The search market remains more of a complementary business for Microsoft than a strategic priority, and with the rise of AI assistants that provide direct answers instead of referring users to links, the entire search category is undergoing a transformation whose outcome is still unknown.

Main Competitors

Microsoft does not face a single dominant rival; it faces various competitors in each segment. While its position is not existentially threatened in any of them, there is real pressure in each.

Amazon, the main rival in the cloud

AWS remains the world’s largest cloud platform, with approximately 30–32% of the global market share, compared to Azure’s 22–25%. Amazon has a lead over Microsoft in the pure-play cloud infrastructure business—companies that do not need integration with Microsoft 365 and are looking solely for computing power, storage, and databases often choose AWS as their default option. Amazon has guided capex of approximately $200 billion for 2026, Alphabet $175 to $185 billion, and Meta $115 to $135 billion, bringing the combined hyperscaler capex to $680 to $720 billion. The race in AI infrastructure is therefore a collective bet by the entire industry, not just Microsoft. Amazon also has a head start with its own AI silicon; the Trainium2 chips deployed in the Rainier project represent the first true alternative to Nvidia GPUs on a large scale.

Alphabet, a Rival in Both Cloud and Productivity

Google Cloud is the third-largest cloud platform, with approximately 11–12% of the global market share and the fastest growth rate among the three hyperscalers. In the AI space, Alphabet is a direct competitor; its Gemini models rival both GPT-4 and Microsoft’s proprietary models, and Google Workspace is challenging Microsoft 365, particularly in the mid-market segment. Alphabet also has a structural advantage in the form of search, which generates data and distribution that Microsoft lacks. The rivalry between Azure and Google Cloud is likely to intensify over the next two years, as Google aggressively invests in enterprise sales, an area where it has historically lagged behind.

Salesforce and ServiceNow: Rivals in Enterprise AI Agents

This rivalry is less discussed but strategically important. Microsoft is building autonomous AI agents integrated across Microsoft 365, and Salesforce $CRM, with its Agentforce platform, and ServiceNow, with its own AI workflow tools, are targeting exactly the same space. The difference is that Salesforce and ServiceNow $NOW come from the CRM and ITSM worlds, while Microsoft comes from productivity and the cloud. The clash will occur when enterprise customers decide on which platform to build their AI agent infrastructure—and that decision is only just beginning to take shape.

OpenAI: Partner and Potential Rival

OpenAI is currently a partner, but the dynamics of the relationship are changing. As OpenAI grows and has access to $110 billion in funding following its latest funding round, Microsoft will need to develop its AI capabilities independently of this partnership. OpenAI is gradually shifting toward enterprise products—ChatGPT Enterprise, its own API, and tools for businesses—and is no longer just a model that Microsoft distributes. If OpenAI builds a sufficiently strong direct distribution channel, some of the value that Microsoft currently derives from the exclusivity of the partnership could shift back to OpenAI. This is one of the most underestimated risks of the entire investment thesis.

The Company’s Valuation and Where It Is Undervalued

What the Market Is Valuing Correctly

The market has correctly identified that the $190 billion in capex is a real drag on free cash flow in the short term. Free cash flow fell 13% year-over-year despite record operating profitability, and this divergence between operating results and actually available cash is not an accounting anomaly but a real consequence of the investment cycle. The market is also correctly pricing in the risk of concentration surrounding the OpenAI partnership and regulatory pressure in Europe, where the EU has designated Amazon and Microsoft’s cloud services as systemically important providers subject to stricter digital competition rules. These risks are real, and the discount they imply is justified.

What the market is likely underestimating

The first underestimated factor is the economics of Copilot adoption in the context of the existing customer base. The argument is straightforward: enterprise customers pay for M365 on a large scale, and Copilot transforms this installed base into a business with higher average revenue per seat. With a penetration rate of 3.3% of the total commercial base, Copilot is still in its infancy. If adoption reaches just 15% within three years, that would mean approximately 67 million paid seats—more than triple the current number—with an annual run rate exceeding $24 billion from this single product alone. The market is aware of these figures, but it appears to be discounting them more aggressively than the current adoption rate warrants.

The second underestimated factor is the quality of the backlog as a leading indicator. Commercial remaining performance obligations reached $627 billion, with year-over-year growth of 99%, and approximately 30% is expected to be converted into revenue over the next twelve months. Approximately $188 billion in revenue is therefore already contractually secured today, even before a single new order is received. This visibility dramatically reduces the risk of revenue surprises, but the market is largely ignoring it amid the capex debate.

What Is Really in Dispute

The central point of contention is not the quality of the business, but the return on capex. For every dollar of annual Copilot revenue, Microsoft is spending $26 on infrastructure this year. This figure isn’t a problem in itself—the infrastructure serves the entire Azure business, not just Copilot—but it illustrates just how much of the value lies in the future. If Azure maintains 35–40% growth even after capacity constraints are lifted and Copilot adoption continues at its current pace, the capex will retroactively prove to be a sound investment. If demand slows or competition drives down prices, that same capex will turn out to be an overpriced ticket to a race with no clear winner.

The market currently does not know which of these scenarios is correct, and it is precisely this uncertainty—not weak fundamentals—that explains the 37% discount to the analyst consensus.

NUMBERS

Key Results: Q3 FY2026 vs. Q3 FY2025

Segment revenue Q3 FY2026

Key Subsegment Metrics

Microsoft 365 Consumer Cloud grew 33% year-over-year, Dynamics 365 by 22%, and LinkedIn by 12%. Azure and cloud services grew 40% year-over-year, while analysts had expected 38.8% to 39.3%.

Capital Expenditures and Balance Sheet

Capex for 2026 was increased to $190 billion, of which approximately $25 billion represents higher component costs rather than additional capacity. This represents a 61% year-over-year increase compared to 2025. Microsoft holds one of the few AAA ratings in the world, has virtually no net debt, and generates operating income exceeding $38 billion per quarter; the company therefore finances its Capex from its own revenues, not through debt or stock dilution.

Full-Year Outlook for FY2026

Management guided for continued acceleration of Azure in the second half of calendar year 2026 following the lifting of capacity constraints. For Q4 FY2026, management expects revenue for the More Personal Computing segment to be between $11.8 billion and $12.3 billion, with Azure growth of 39% to 40% on a constant-currency basis. Full-year revenue is expected to exceed $320 billion if this pace is maintained.

Valuation and Fundamentals

Microsoft’s P/E ratio ranges from 22 to 24, with a forward P/E of around 32. At first glance, this appears to be a reasonable valuation for a company with this growth profile; compared to Palantir, with a forward P/E of over 100, or pure-play AI companies trading at tens of times revenue, Microsoft looks almost cheap. But this comparison is misleading. The proper context is Microsoft’s own historical average and the implicit assumptions currently built into its price.

Before the AI boom took off, Microsoft traded at a forward P/E of around 25 to 28—significantly lower than today. Today’s forward P/E of 32 therefore includes a premium for the AI story, but paradoxically, this premium is lower than one would expect for a company whose AI business is growing 123% year-over-year. The reason lies precisely in the capex debate: the market adds a premium for growth but simultaneously applies a discount for the uncertainty surrounding the return on investment. The result is a valuation that is neither obviously cheap nor obviously expensive, but depends almost exclusively on a single assumption: whether capex generates an adequate return.

What exactly is factored into today’s price of around $352? The market implicitly assumes continued Azure growth in the 30–35% range over the next two to three years, a gradual easing of capacity constraints in the second half of 2026, Copilot adoption growing at a steady pace without any dramatic acceleration, and free cash flow stabilizing after the capex cycle peaks sometime in 2027. In other words, the price reflects a solid but not exceptional scenario.

The analyst consensus among 56 analysts is “Strong Buy” with an average target price of over $561, which is approximately 60% above the current price. Such a large gap between the market price and the analyst consensus is unusual for a company of this size and quality. Either the market sees risks that analysts are underestimating—particularly capex returns and regulatory pressure—or this represents one of the biggest value-creation opportunities in the S&P 500 index. Meanwhile, analysts are balancing their enthusiasm for the AI tailwinds for Azure with a wave of recent target price cuts stemming from caution regarding margins and the level of capital expenditures.

A comparison with competitors adds another layer of context. AWS, as part of Amazon, trades at higher earnings multiples than Microsoft, even though Amazon as a whole generates lower margins. Google Cloud, as part of Alphabet, offers a similar cloud story at a lower valuation for the entire company, but without the same depth of enterprise integration. Microsoft thus sits in the middle—it is neither the cheapest nor the most expensive of the hyperscalers, but it has the clearest narrative on how AI revenue is growing across its entire product portfolio simultaneously.

The key conclusion of the valuation analysis is simple: today’s price does not assume failure, but it does not assume exceptional success either. It assumes an average scenario. If Copilot adoption accelerates beyond consensus and Azure capacity becomes available sooner than the market expects, today’s price is significantly undervalued. If the capex cycle lasts longer and returns turn out to be lower than management promises, today’s forward P/E of 32 is still too high for a company with constrained free cash flow.

Risks

Capex payback is the biggest and most immediate risk

The $190 billion in capex represents a 61% year-over-year increase and exceeded the analyst consensus by approximately $35 billion. The problem isn’t the size of the investment itself; Microsoft is funding it from its own cash flow without debt or dilution. The problem lies in the timing of the return on investment. For every dollar of annual Copilot revenue, Microsoft is spending $26 on infrastructure this year. If AI demand slows, if hyperscalers begin squeezing margins through a price war, or if enterprise customers consolidate the number of Copilot seats after initial pilots, revenue could slow while the cost base continues to grow. This combination would put simultaneous pressure on both revenue and margins—the worst-case scenario for a stock trading at a forward P/E of 32.

Copilot adoption may fall short in both speed and depth

Twenty million paid seats sounds impressive, but in the context of the entire installed base, that represents a penetration rate of 3.3%. A bigger problem than the absolute number is the question of who these customers are. Serious challenges and functional issues with Copilot were still being reported as early as the beginning of 2026, with some enterprise customers failing to find sufficient productivity value for widespread deployment following initial pilots. If it turns out that Copilot works exceptionally well for a narrow group of power users but fails to deliver an adequate return on investment for the average employee, the total addressable market is significantly smaller than the 450 million commercial M365 users would suggest.

Regulatory pressure in Europe is mounting

The EU has designated Amazon and Microsoft’s cloud services as system providers subject to stricter digital competition rules under the DMA. This entails obligations regarding interoperability, data portability, and a ban on certain bundling practices. Furthermore, Italy has launched an investigation into Microsoft 365’s pricing practices. For a company whose competitive advantage rests precisely on the integration and bundling of products into a single ecosystem, the regulatory dismantling of these ties poses an existential threat to part of its business model. European business accounts for approximately a quarter of Microsoft’s total revenue, so the regulatory risk here is far from marginal.

The OpenAI partnership is shifting from an advantage to a vulnerability

Microsoft has long benefited from exclusive access to OpenAI’s most advanced models as a key differentiator from the competition. Now that OpenAI has secured $110 billion in funding, the company is building its own distribution channels and reducing its dependence on Microsoft as its primary distributor. If OpenAI builds a sufficiently strong direct enterprise sales channel, some of the value that Microsoft currently derives from this partnership will shift back to OpenAI. At the same time, any deterioration in the relationship or renegotiation of terms would impact Azure AI revenues, which form a key part of the growth thesis.

Legal Risk and Reputational Pressure

A class-action lawsuit has been filed against Microsoft by investors who allege that the company misled investors regarding AI, Copilot, and capital allocation. This litigation is in its early stages and its outcome is uncertain, but the very existence of the lawsuit adds legal uncertainty at a time when the company is undergoing the largest strategic transformation in its history. In this context, reputational risk is at least as serious as financial exposure.

Investment Scenarios

Optimistic Scenario

For this scenario to materialize, several factors would need to align simultaneously. Azure capacity constraints would ease sooner than the market expects—in the second half of 2026—and growth would accelerate back above 40% even without artificial supply-side constraints. Copilot adoption would exceed consensus and reach 10 to 15% penetration of the commercial M365 base by the end of 2027, representing approximately 45 to 67 million paid seats. At a price of $30 per user per month, this would imply an annual Copilot run rate of $16 to $24 billion, more than double the current level. The Capex cycle peaks in 2026, and free cash flow returns to or exceeds historical levels in 2027. Regulatory pressure in Europe remains a nuisance but does not disrupt the underlying business model. In this scenario, today’s analyst price targets of around $560 are achievable within twelve to eighteen months, and continued growth opens up potential for prices to rise even beyond that level.

Realistic Scenario

The most likely outcome assumes continued strong but gradually normalizing growth. Azure maintains a growth rate of 30 to 35% throughout 2027, capacity constraints are gradually easing, and management adheres to its guidance of acceleration in the second half of 2026. A commercial backlog of $627 billion ensures revenue visibility, which dramatically reduces the risk of a negative revenue surprise. Copilot adoption is growing at a solid pace; seats will double to approximately 40 million by the end of 2027, but will not exceed consensus estimates. The Capex cycle will weigh on free cash flow in 2026 and partially in 2027, after which the situation will stabilize. Gross margins will begin to recover starting in 2027, as AI infrastructure achieves economies of scale and software optimization delivers the promised efficiencies. In this scenario, the stock trades in the $420 to $500 range over a twelve-month horizon, meaning a partial recovery of today’s discount relative to the analyst consensus, but without a dramatic rerating.

Pessimistic Scenario

In a negative scenario, Azure growth would slow to below 25% in 2027 due to a combination of capacity issues, price competition among hyperscalers, and weaker-than-expected AI demand from enterprise customers. Copilot adoption stagnates at less than 5% penetration of the commercial M365 base, as enterprise customers, following initial pilots, do not find sufficient productivity value to justify widespread deployment. The $190 billion in Capex will prove to be an overpriced ticket to a race with an uncertain outcome, and depreciation will begin to put significant pressure on reported profits. Regulatory interventions in Europe are disrupting the bundling model, and investor lawsuits are adding legal costs and reputational pressure. In this scenario, the stock returns to the $280–$320 range—levels at which the forward P/E falls to its historical average before the AI premium. Unlike weaker companies, however, Microsoft remains a highly profitable company in this scenario as well, with an AAA rating, a strong dividend, and a dominant market position; thus, this is not an existential risk, but rather a painful revaluation of the AI premium.

What to Watch Next

• Azure’s growth rate in Q4 FY2026 and the first quarter of FY2027. Management has signaled an acceleration following the lifting of capacity constraints in the second half of 2026. If Azure does not accelerate even after the supply bottleneck is removed, it means that part of the demand was one-time, and the structural growth rate is lower than the company suggests.

• The number of paid Copilot seats and the retention rate after the first annual contracts expire. Seats rose from 15 million to 20 million in a single quarter, but the key factor is not the addition of new seats, but whether enterprise customers renew and expand them after a year, or consolidate and cancel them.

• Free cash flow trends and capex guidance. Free cash flow fell 13% year-over-year despite record operating profitability. Stabilization of capex below $40 billion per quarter will be the first sign that the investment cycle is peaking and cash flow is returning to historical levels.

• Regulatory developments in Europe. The key question will be whether the EU and national regulators will resort to structural measures to limit Microsoft’s bundling practices or stick to fines. Structural intervention into the integration of Azure, Microsoft 365, and Teams would be far more significant for the business model than a one-time fine.

• The pace of reducing dependence on OpenAI. Microsoft is developing seven proprietary AI models and has its own AI hardware. The faster these proprietary models replace OpenAI in products like Copilot and Azure AI, the lower the concentration risk. Also monitor whether OpenAI is building a direct enterprise distribution channel; each of its direct customers is a potential customer who does not need Azure as an intermediary.

Healthcare is often referred to as a classic defensive sector. People need medications, diagnostics, and surgeries regardless of where the economic cycle stands, and demand for healthcare is therefore among the most stable in the entire market. In 2026, however, this theory is being challenged by a reality in which the entire sector is lagging behind the broader S&P 500 index, and within the sector, there are huge differences between individual segments.

The main driver of this year’s developments is a combination of two opposing forces. On the one hand, a wave of mergers and acquisitions in biotechnology has gained full momentum, driven by the looming patent cliff facing major pharmaceutical companies. Major players must replace revenue from drugs whose patent protection is soon to expire, and the fastest way to do so is by acquiring smaller, innovative companies. On the other hand, medical device manufacturers have come under pressure due to high valuations, sensitivity to interest rates, and concerns about the impact of a new generation of weight-loss drugs on the volume of certain surgical procedures.

Health Care Select Sector SPDR Fund $XLV

The XLV fund is the simplest and cheapest way to gain exposure to the entire U.S. healthcare sector. It tracks the healthcare component of the S&P 500 index and, using a market-capitalization-weighted methodology, holds the sixty largest companies in the sector. With an expense ratio (i.e., annual management fee) of just 0.08% per year, it ranks among the lowest-cost sector funds overall and manages approximately $38 billion in assets.

The portfolio’s structure is heavily concentrated in its largest holdings. Eli Lilly ($LLY ) alone accounts for about 16% of the fund, Johnson & Johnson ($JNJ ) for roughly 10%, AbbVie ( $ABBV ) and UnitedHealth ($UNH ) for just under 7% each, and Merck ( $MRK ) for about 5%. The top ten holdings together account for nearly 59% of the entire fund, so although XLV appears to be a broad-based sector fund, its performance is in fact largely determined by a handful of giants.

In terms of segments, pharmaceuticals make up about 38% of the portfolio, healthcare providers 19%, biotechnology 18%, and medical equipment roughly 16%.

The fund is trading around $155 per share and is slightly down year-to-date, while the broader market is rising. Over the past twelve months, however, it has managed to gain about 14%, suggesting that the worst phase of last year’s regulatory uncertainty may be behind us. In 2025, the sector was weighed down by a combination of political pressure on drug prices and uncertainty surrounding tariff policies on pharmaceutical imports.

Eli Lilly and its GLP-1-based weight-loss drug program remain the fund’s key driver. The market is closely watching at what price the company will be able to monetize this new generation of products, as the fund’s largest position largely determines its overall trajectory. In contrast, UnitedHealth $UNH has rebounded significantly this year, gaining about 23% since the start of the year after last year was marred by a cyberattack and an investigation by regulatory authorities.

The XLV fund is particularly attractive due to its simplicity and low costs. It offers broad exposure to a sector that benefits from long-term demographic trends, and its underperformance relative to the market this year can also be viewed as a relative valuation discount compared to red-hot tech stocks. Risks remain in the political environment surrounding drug prices and the fund’s high concentration in just a few names, where a stumble by Eli Lilly or other giants has a disproportionately large impact on overall performance.

SPDR S&P Biotech ETF $XBI

While XLV represents a cautious and concentrated approach, XBI is its exact opposite. This fund tracks the biotechnology segment of the market using the so-called equal-weight methodology. This means that each of the 157 companies in the portfolio has roughly the same weight, and no single stock exceeds a few percent of the fund. The top ten holdings together account for only about 16% of assets.

This structure has significant implications for the fund’s performance. A small clinical-stage biotech company carries virtually the same weight in the portfolio as an established giant. Thus, the successful approval of a drug or positive clinical trial results from a smaller company can drive the entire fund, while a failure can have the opposite effect. XBI is therefore significantly more volatile than broader healthcare funds and acts as a direct bet on the entire biotechnology ecosystem, including smaller and riskier names. Annual expenses amount to 0.35%.

Biotechnology is this year’s winner within the healthcare sector. The sector is being driven by an exceptionally strong wave of mergers and acquisitions, fueled by the looming patent cliff. Major pharmaceutical companies face the risk of losing hundreds of billions of dollars in revenue due to expiring patents, and the fastest way to fill this gap is through the acquisition of smaller, innovative companies. This year alone, biotechnology transactions have exceeded $100 billion across more than 200 deals, and full-year estimates range between $140 billion and $160 billion.

In terms of performance, this has made the XBI one of the strongest sector funds. Over the past twelve months, the fund has gained roughly 40%, more than double the category average, and is trading near its annual highs around $140. By comparison, its annual low was below $81, which illustrates just how dramatic the sector’s turnaround has been. However, investors should be aware that, precisely because of its balanced structure and exposure to clinical-stage companies, the fund could fall just as quickly as it rose should market sentiment deteriorate.

Furthermore, following its sharp rise, the fund has shown signs of technical fatigue and short-term consolidation in recent weeks. This in itself does not necessarily signal the end of the uptrend.

iShares Biotechnology ETF $IBB

The IBB fund targets the same sector as the XBI but takes a completely different approach. Instead of equal-weighting, it uses a market-capitalization-weighted methodology with caps, whereby the five largest holdings can account for a maximum of 8% and the rest 4%. The result is a portfolio dominated by the largest and most profitable biotechnology companies. The fund is managed by BlackRock $BLK under the iShares brand, and its annual expense ratio is 0.44%, which is slightly higher than that of its competitor, the XBI.

The top ten holdings together account for about 47% of the fund. In terms of structure, nearly 90% consists of pure-play biotechnology companies, supplemented by a smaller share of companies in the scientific instruments and pharmaceuticals sectors.

Thanks to its concentration in large, profitable companies with already approved drugs and stable revenues, the IBB performs significantly more steadily than the XBI, and its profile is closer to that of a large-cap pharmaceutical fund. When a small company achieves a breakthrough, the IBB is barely affected. Conversely, when a single smaller stock underperforms, the fund remains relatively stable thanks to the big names in its portfolio. Over the past twelve months, the IBB has gained roughly 34%, slightly less than the more aggressive XBI, which precisely reflects the difference between the two strategies.

The choice between IBB and XBI is thus essentially a choice of risk appetite. IBB is a more conservative way to bet on the biotech recovery, while XBI offers greater leverage on the sector’s overall growth at the cost of significantly higher volatility.

iShares U.S. Medical Devices ETF $IHI

While biotech has shone this year, medical device manufacturers are experiencing the exact opposite. The IHI fund, which tracks the U.S. medical equipment manufacturers’ index, has lost approximately 21% since the start of the year and ranks among the weakest segments of the entire healthcare sector. It is trading around $49 per share, near its annual low. The fund is managed by BlackRock, holds roughly fifty companies, and has an annual expense ratio of 0.38%.

The portfolio is heavily concentrated in a few large companies. The top ten holdings account for roughly 74% of the fund, making IHI the most concentrated of the four funds in terms of individual stocks.

Prior to the decline, the medical device sector was trading at high valuations with a price-to-earnings ratio of around thirty, and was therefore sensitive to a higher interest rate environment, which reduces the present value of future earnings for growth companies. Added to this were concerns about the impact of a new generation of weight-loss drugs on the number of certain procedures, particularly in the fields of cardiology and obesity care. Although this is a defensive sector with a relatively low beta of around 0.86, this year has shown that even a defensive nature does not protect against a significant correction if the market changes its view on valuations.

The segment benefits from long-term structural trends, such as an aging population, the rising prevalence of chronic diseases, and the gradual expansion of robotic surgery and AI-assisted diagnostics. If interest rates continue to fall and concerns about the impact of weight-loss drugs prove to be overblown, this year’s decline may, in hindsight, appear to have been an opportunity.

Fund Comparison

Strategic Perspective

In terms of their role in the portfolio, the individual funds can be viewed as follows:

• XLV as a core, low-cost, and relatively defensive exposure to the entire sector, suitable for long-term investors who do not want to bet on a single specific segment.

• XBI as a more aggressive complement for investors who believe the biotech supercycle will continue and can tolerate high volatility in exchange for higher potential.

• IBB as a more stable way to gain biotech exposure, anchored in large, profitable companies.

• IHI as a contrarian bet for those who consider this year’s decline in medical device manufacturers to be exaggerated and believe in a return to long-term structural trends.

Importantly, thematically narrow funds such as XBI, IBB, and IHI are significantly more sensitive to changes in the market cycle than the broadly diversified XLV. This year is a prime example of this, as the performance spread between the best and worst performers among this quartet exceeds tens of percentage points, even though all four funds belong to the same sector.

The healthcare sector in 2026 shows that the sector’s reputation as a defensive one does not mean its performance is boring or uniform. Beneath the surface, one of the biggest rifts between segments in recent years is unfolding, with biotechnology benefiting from a record wave of acquisitions, while medical device manufacturers are undergoing a painful revaluation. Therefore, the choice of a specific ETF this year will have a much greater impact on the outcome than the decision to invest in healthcare as a whole.

This gives the investment thesis a clear outline, but also creates a clear tension. On the one hand, Alphabet has contracted demand that it cannot keep up with; according to management’s own statements, cloud revenue would be higher if the company had more computing capacity. On the other hand, that very capacity costs a fortune. Management has raised its capital expenditure plan for 2026 to $180–190 billion and has indicated that it will grow significantly again in 2027. The question addressed in this article, therefore, is not whether demand exists, as that is evidenced by customer contracts. The question is how much of that $462 billion will translate into actual revenue, how quickly, and whether the return on investment will justify one of the most expensive investment waves in the history of the technology sector.

Key Points of the Analysis

• The cloud has gone from being a long-standing burden to the fastest-growing and, more recently, highly profitable segment. Google Cloud’s operating profit rose to $6.6 billion—a threefold increase year-over-year—and its operating margin expanded from 17.8 percent to 32.9 percent.

• The backlog reached approximately $462 billion, nearly doubling quarter-over-quarter from roughly $240 billion. According to the company, roughly half of this is expected to be converted into revenue within 24 months, meaning more than $230 billion in already contracted cloud revenue is scheduled to be recognized by mid-2028. Part of this volume also consists of sales of the company’s own hardware, specifically TPU chips.

• Despite concerns, search has accelerated, not slowed down. Revenue in the Search segment grew 19 percent year-over-year; according to the company, search queries are at record levels, and AI-powered features such as AI Overviews and AI Mode are increasing search usage rather than cannibalizing it, according to management.

• Overall results were strong, but net income was inflated by a one-time gain. In the first quarter of 2026, revenue rose 22 percent to $109.9 billion, and net income rose 81 percent to approximately $62.6 billion. However, this figure included an unrealized gain of about $37.7 billion from the revaluation of non-traded equity investments. After adjusting for this, earnings per share fell just short of estimates.

• The main cost of this growth is capital expenditures. The plan for 2026 was raised to $180–190 billion, and the company promises another significant increase for 2027. Over the past twelve months, capital expenditures totaled around $110 billion, which squeezed free cash flow down to roughly $64 billion despite operating cash flow exceeding $170 billion.

• Subscriptions and the developer ecosystem are growing. The number of paid subscriptions reached 350 million, driven by the Gemini app, YouTube, and Google One, and the first quarter of 2026 was, according to the company, the strongest quarter in history for consumer AI plans. Alphabet’s models process around 16 to 19 billion tokens per minute via a direct interface, which is roughly six times the year-over-year figure.

• Despite the record price, the valuation remains relatively modest. The stock is trading around $368 with a market capitalization of about $4.5 trillion, and the forward price-to-earnings ratio ranges roughly between 26 and 29. The consensus among more than sixty analysts is a strong “buy” with an average price target of around $430 to $433.

What Has Changed at the Company

Two years ago, the prevailing narrative surrounding Alphabet $GOOG was more defensive. The market feared that AI-powered chatbots, particularly those from OpenAI, would siphon off its most valuable search queries and disrupt the advertising machine that had sustained the company for decades. At the time, Google Cloud was a smaller, slower-growing, and long-loss-making segment that served mainly as proof that the company could do more than just advertising.

This perception has shifted in two ways. First, it turned out that Search does not stand in the way of artificial intelligence, but rather absorbs it. According to the company, the AI Overviews and AI Mode features increase both the number and complexity of search queries, and as a result, Search accelerated to 19 percent growth instead of declining. Second, Google Cloud has transformed from a money-losing experiment into a profit machine with a margin approaching thirty-three percent and an order book the market has never seen from any competitor.

Underlying all of this is a structural shift. Alphabet has ceased to be purely an advertising company with ambitions in the cloud and has become a company that owns the entire artificial intelligence ecosystem—from its own TPU chips and Gemini models to end-user products in search, office software, the cloud, and autonomous mobility. It is precisely this vertical integration that is today the bulls’ main argument and, at the same time, the source of the company’s largest costs.

How does it make money from this?

Alphabet monetizes artificial intelligence in three distinct ways, which vary in maturity and profitability.

The first and most mature avenue is advertising. Search and YouTube remain the cash cow of the entire company, and artificial intelligence acts as a multiplier here, increasing the number of queries and improving targeting and ad formats. This is the part of the business that funds everything else.

The second avenue is the cloud. Companies rent computing power from Google Cloud and purchase access to Gemini models and tools such as Vertex AI and BigQuery. A key shift is that this segment is no longer subsidized by the rest of the company but generates its own revenue with a margin exceeding thirty percent. Furthermore, a portion of the orders consists of direct sales of TPU chips, which expands the ways in which the company profits from the demand for computing power.

The third revenue stream comes from subscriptions and end-user applications. The Gemini app, Google One, and paid AI plans generate recurring revenue from consumers, while Gemini Enterprise targets enterprise customers, among whom the number of paid active users grew by 40 percent quarter-over-quarter.

Beyond these three lies the “Other Bets” segment, which includes Waymo’s autonomous mobility business. That segment is currently burning through cash; in the first quarter of 2026, it reported revenue of just $411 million and an operating loss of $2.1 billion. It is purely a bet on the future, not a current source of profit.

What Is the Main Growth Driver

Google Cloud is clearly the dominant growth driver today. It is growing faster than any other major provider, its profitability is rising sharply, and its order book suggests that this pace isn’t slowing down anytime soon. Crucially, the limiting factor is not demand but the supply of computing capacity—which, from an investment perspective, is a welcome problem as long as the company can finance it.

Search remains the largest and most profitable driver in absolute terms, but its role has shifted from a threat back to a stable pillar. A 19 percent growth rate is exceptional for such a large user base and refutes the notion of cannibalization.

Waymo, meanwhile, is a potential growth driver. It does not contribute to current results, but it represents the greatest source of potential upward revaluation, similar to Baidu’s Apollo Go, with the difference that Waymo operates in higher-priced Western markets and is technologically perceived as a leading player in autonomous mobility.

Products and Upcoming Strategy

In terms of models, the company is banking on the Gemini series, whose newer versions power both the consumer app and the enterprise offering, as well as the search function itself. The strategy is for the same underlying model to be used across all products, which reduces costs and increases utilization. The company also reports that over the past five years, it has reduced the cost of a basic AI response by more than thirty percent and search latency by more than thirty-five percent, which is crucial for the economics of the entire model.

A key asset in its infrastructure is its proprietary TPU chip. This allows Alphabet to offer computing power more cost-effectively and with less reliance on supplies from NVIDIA, and the company now sells it directly, as evidenced by the inclusion of hardware sales in its order book. In addition to TPUs, it also offers graphics processing units (GPUs), thus meeting both types of demand.

Beyond its core business, the company is strengthening its energy infrastructure for data centers, as evidenced by acquisitions in the energy sector, and continues to develop Waymo. Waymo, on the other hand, is currently recalling nearly 4,000 vehicles due to a software issue, a reminder that autonomous mobility remains challenging both operationally and from a regulatory standpoint.

Market Position and Market Share

In the cloud, the situation is interesting precisely because of the contrast between market share and growth dynamics. According to Synergy Research, Amazon AWS holds roughly 30 percent of the global cloud infrastructure market, Microsoft Azure around 25 percent, and Google Cloud about 13 percent. Alphabet is thus still the smallest of the big three, but it is growing the fastest and has roughly doubled its market share over the past two years from approximately 6 percent.

Moreover, its growth rate is accelerating, while that of its competitors is stagnating or growing more slowly. Google Cloud has gradually accelerated from 32 percent in the second quarter of 2025, through 34 and 48 percent, to 63 percent in the first quarter of 2026. Azure, on the other hand, has remained in the 39–40 percent range, and AWS—though it accelerated to 28 percent, marking its fastest growth in roughly four years—remains the slowest of the three in terms of growth rate.

In search, Google remains globally dominant, with a long-term market share exceeding 80 to 90 percent in many regions, although exact figures vary and the definition of the market is becoming blurred with the advent of AI assistants. Importantly, unlike Baidu, whose domestic search business is facing a structural decline, Google continues to grow in search.

Main Competitors

In the cloud, the main rivals are Amazon AWS $AMZN and Microsoft Azure $MSFT. AWS benefits from being the largest in absolute terms and from accounting for the majority of Amazon’s operating profit; like Alphabet, it is betting on its own chips—in its case, Trainium. Azure benefits from deep integration into Microsoft’s enterprise software ecosystem and its partnership with OpenAI, giving it a strong position with enterprise customers and in the Copilot tool.

In AI models and end-user applications, Alphabet’s main competitors are primarily OpenAI and Anthropic, followed by Meta $META with its open models, and in China, a separate ecosystem that includes Baidu $BIDU. Competitive pressure in this sector is intense and is also reflected in a brain drain, with several key researchers from Google DeepMind recently moving to competitors.

In autonomous mobility, Waymo is viewed as the technology leader in Western markets, facing competition from Tesla and, in China, Baidu Apollo Go, Pony.ai, and WeRide. The market is still in its infancy, and scaling, regulation, and unit economics will be the deciding factors.

Comparison of the Big Three in Cloud Computing

Note on the table: The figures are not fully comparable because each company reports cloud revenue in a differently defined segment. Google Cloud also includes Workspace; Azure is reported within a broader segment, and its standalone growth rate is reported in constant currencies; and AWS is the most purely infrastructure-focused. The table should therefore be read as a rough comparison of trends, not as an exact measurement.

What the Market Is Valuing Correctly—and What It May Not

The market today appears to correctly recognize that Alphabet is the clear winner in the current phase of the cloud race and that search is not dead. Following the first-quarter results, the stock rose significantly, and concerns about cannibalization subsided. The stock price also reflects the strength of the advertising core business and improved cloud margins.

What the market may still be underestimating is the value of the order book. The contracted volume of around $462 billion represents unrecognized demand, and if it translates into revenue faster and with higher margins than the market expects, there remains room for a revaluation even after the stock’s sharp rise. The market also assigns a relatively low value to Waymo.

Conversely, it is possible that the market is underestimating the risk to return on investment. Capital expenditures of $180 to $190 billion annually, with the promise of further growth, are unprecedented and are squeezing free cash flow. If demand for artificial intelligence were to cool off before the capacity pays for itself, today’s enthusiasm could quickly turn. The market is betting that returns will come, but it’s a bet, not a certainty.

Figures Supporting the Thesis

• Total revenue in Q1 2026: $109.9 billion, 22 percent year-over-year growth.

• Search revenue: 19 percent year-over-year growth, with record search volume.

• Google Cloud: revenue over $20 billion, 63 percent growth, operating profit of $6.6 billion, and a margin of 32.9 percent.

• Cloud contract revenue: approximately $462 billion, nearly double quarter-over-quarter, with about half to be recognized within 24 months.

• Net income: approximately $62.6 billion, earnings per share of $5.11, but with an unrealized gain of about $37.7 billion from the revaluation of equity investments.

• Generative AI products: growth of around 800 percent year-over-year; Gemini Enterprise up 40 percent quarter-over-quarter in paid active users.

• Paid subscriptions: 350 million, the strongest quarter ever for consumer plans.

• Capital expenditures: $180–190 billion planned for 2026; $35.7 billion in Q1; further significant growth promised for 2027.

• Free cash flow for the last twelve months: approximately $64 billion, with operating cash flow exceeding $170 billion and investments around $110 billion.

• Dividend raised by 5 percent to $0.22 per quarter, yielding around 0.24 percent.

Valuation

Alphabet’s valuation must be viewed through two lenses. Reported net income for the first quarter was significantly inflated by unrealized gains from the revaluation of investments totaling nearly $38 billion, so earnings per share overstate actual operating performance. Adjusted earnings, on the other hand, were largely in line with estimates or slightly below them.

However, at a price of around $368 and a forward price-to-earnings ratio of roughly between 26 and 29, the stock—despite its record high and 106 percent growth over the past twelve months—does not appear extremely expensive for a company with 22 percent revenue growth, margins exceeding 30 percent, and exceptional return on capital. A price-to-earnings-to-growth ratio of around 1.8 suggests a reasonable balance between price and growth momentum.

What is priced in? Priced in is victory in the cloud race, the resilience of the search business, and confidence that massive investments will yield returns. Conversely, the price does not reflect a scenario in which Waymo would become a significant asset in its own right, nor does it reflect the order book being fully aligned with profit margins.

What would lead to further upward revaluation? Continued acceleration in the cloud sector with improving margins, tangible evidence of a return on capital investments, and, potentially, a standalone valuation of Waymo. Conversely, a decline in valuation would be driven by a slowdown in demand for artificial intelligence, regulatory intervention in search, or the discovery that margins on large cloud contracts are lower than the market hopes.

Risks

The biggest specific risk is the return on capital investments. Annual expenditures of $180 to $190 billion, with the promise of further growth, represent a massive bet. If demand cools or if margins on large cloud contracts disappoint, free cash flow—which has already been significantly squeezed by these investments—could come under further pressure.

Regulatory risk at Alphabet is persistent and multifaceted. In the U.S., antitrust pressure on search and potential remedies continues; in the U.K., the regulator is implementing specific rules of conduct for search; and similar pressure exists in the European Union. Any intervention in the way Google offers search and advertising strikes directly at the company’s cash cow.

Competitive risk in the models remains real. OpenAI and Anthropic are pushing the limits of the models’ capabilities, and the exodus of key researchers from Google DeepMind shows that the battle for talent is fierce. If Alphabet were to lose its lead in model quality, the rationale for its entire vertical integration strategy would be undermined.

Waymo bears the operational and reputational risk, as evidenced by the recall of nearly 4,000 vehicles due to a software issue. Autonomous mobility is sensitive to safety incidents, which can slow both expansion and regulatory approval.

The risk associated with the composition of earnings is illustrated by the most recent quarter itself, when unrealized gains from the revaluation of equity investments increased reported earnings by tens of billions. This item is volatile and may have the opposite effect in other periods. On the other hand, debt levels are low and the balance sheet is very strong, so financial risk is minimal.

Investment Scenarios

Optimistic Scenario

For this scenario to materialize, Google Cloud would need to maintain growth in the range of fifty percent or more, cloud margins would continue to improve beyond 35 percent, the order book would convert to revenue faster than the company expects, and search would maintain double-digit growth. At the same time, evidence of a return on massive investments would emerge, or the market might begin to value Waymo separately. In that case, total revenue would grow by high double-digit percentages, margins would widen, and the stock would move toward the upper end of the analyst range—around $500 or higher—as the separately valued contributions of cloud and autonomous mobility would be added to the core business.

Realistic Scenario

The most likely outcome is continued strong but gradually slowing growth. The cloud continues to grow at an above-average rate, but the pace normalizes from extreme levels as the user base expands and capacity is added. Search maintains roughly double-digit growth, cloud margins remain high, and capital expenditures continue to put pressure on free cash flow, but the company can withstand this thanks to the strength of its core advertising business. Waymo continues to expand without making a significant contribution to profits. In this scenario, the stock would trade in a range around the analyst target price of $430, with a gradual upward shift if results confirm that the order book is converting into revenue.

Pessimistic Scenario

In a negative scenario, a slowdown in demand for artificial intelligence would coincide with disappointing margins on large cloud contracts and tough regulatory intervention in search, whether in the U.S., the U.K., or the European Union. In such a case, massive capital investments would become a burden, free cash flow would shrink further, and the market would begin to doubt the return on investment. The stock could correct back toward the lower end of its current range—well below $300. Unlike smaller and riskier stocks, however, Alphabet maintains an exceptionally strong balance sheet and profitable core business, so even in this scenario, it is more a matter of a correction than an existential threat.

What to Watch Next

• Monitor the pace at which the order book is converted into revenue. The key will be whether the roughly $230 billion earmarked for recognition within 24 months actually begins to show up in revenue—and at what margin.

• Monitor the trend in capital expenditures and free cash flow. The plan for 2027 was announced as significantly higher, and the market will be watching closely to see if returns keep pace with spending.

• Monitor the pace of cloud growth relative to the competition. If Google Cloud maintains its lead over Azure and AWS, it will reinforce the overall thesis, while a slowdown would signal that the advantage was temporary.

• Monitor regulatory developments surrounding search. Any decisions on remedies in antitrust cases or new rules in the United Kingdom and the European Union will directly impact the cash flow.

• Keep an eye on Waymo. The pace of expansion, unit economics, and how it handles safety incidents will determine whether the market assigns it standalone value.

For shareholders, the development raises questions about the pace and cost of Ford's manufacturing modernisation efforts. Rebuilding a human inspection workforce suggests the company is prioritising near-term quality assurance over automation savings, which could affect operating costs as it works to improve its historically troubled vehicle quality record.

Second, and more interesting to me, is the paragraph about the government. The Trump administration asked OpenAI to phase the rollout of the new GPT 5.6 model over safety concerns — the model is initially being provided only to selected partners in a limited preview, and the government is reportedly approving access on a customer‑by‑customer basis. This is a much more important signal than an IPO date. It means the most advanced AI models are beginning to be treated as strategic technology subject to state control, similar to chips or weapons. That changes the game for the whole sector — regulation ceases to be theoretical and becomes a real operational factor.

Their financial results for the third fiscal quarter delivered incredible numbers:

Revenue: $41.46 billion (a massive increase of 74% quarter-over-quarter and 346% year-over-year).

Gross margin: 84.9%, compared with just 39% in the same period last year. Even the best software companies could dream of a margin like this.

Net income: $28.86 billion, with net operating cash flow reaching $25.39 billion.

Q4 outlook: Management expects revenue of up to $50 billion with a gross margin around 86% and EPS of $31.00.

Why might this company no longer be the cyclical commodity firm we knew a few years ago? There are several reasons:

1. Strategic customer agreements (SCAs):

Micron has broken the traditional market model and signed 16 long-term contracts (effective from 2026 through the end of 2030) with giants in data centers, automotive, and consumer electronics. These contracts already cover roughly 20% of DRAM volume and one-third of NAND volume. Once all planned agreements are fully implemented, about half or more of the company’s total revenue will be allocated under these contracts.

2. Take-or-pay commitments with a guaranteed price floor:

Customers have fixed volumes they must purchase under the contracts. Crucially, the agreements define a price floor that ensures Micron an extremely robust gross margin, far exceeding the peaks of any previous cycles. Fourteen out of the 16 contracts alone guarantee cumulative minimum revenues of $100 billion. In addition, Micron will receive $22 billion in upfront cash deposits and financial commitments based on these contracts.

3. Structural supply deficit at least through 2027+:

Demand for memory driven by complex AI models far outpaces production capacity. Building new fabs (greenfield projects) takes years due to shortages of skilled labor, approval processes, and heavy energy infrastructure requirements. Moving to advanced nodes also yields lower natural bit-growth. The new generation of HBM4 memory (where Micron has already booked over $1 billion in revenue) consumes many times more manufacturing capacity and cleanroom space, creating enormous pressure on capacity for standard DRAM.

4. Expansion into high-margin, high-value segments:

Cloud and core data centers (CDBU) grew quarter-over-quarter by 78% and 103%, respectively, with margins between 83% and 87%. At the same time, a wave of physical AI is arriving. Autonomous vehicles at L2+ and above require more than 5× the memory of a typical vehicle. The biggest long-term catalyst, however, is humanoid robots — they carry up to 10× the memory capacity of L2+ vehicles, which will trigger a massive secular megacycle in the second half of this decade.

Given the structural supply deficit that will persist at least through 2027 (and likely beyond), the company is poised for a period of the highest profitability in its history.

For shareholders, the CFO transition is a notable development as Nike works through a period of slowing sales and a broader turnaround effort. How Denton shapes the company's financial strategy and cost discipline will be closely watched in the quarters ahead.

For shareholders, repeated price hikes on consumer hardware can help protect margins but also risk dampening demand, particularly in a competitive gaming market where price sensitivity among buyers tends to be high.

For investors, the shift suggests Apple may be consolidating its chip roadmap, though the reasons and any impact on product timing or sales have not been disclosed. The plan could still change before 2027.

In a separate but closely related move, the bank named Doug Petno and Troy Rohrbaugh as co-presidents, while longtime executive Marianne Lake is leaving the firm. The appointments effectively reopen the question of who will eventually succeed long-serving CEO Jamie Dimon, with Petno and Rohrbaugh now seen as the leading candidates.

How much more will you pay?

It’s not just a few pennies. Apple $AAPL quietly updated its price lists overnight from Wednesday to Thursday, and the results are significant:

• MacBook Air: up $200 to $1,299

• MacBook Pro: +$300, now priced at $1,999

• MacBook Neo (latest model): from $599 to $699

• Mac Studio: from $1,999 to $2,499

• iPad Air: up $150, new price $749

• iPad Pro: price increased by $200 to $1,199

• Base iPad: from $349 to $449

Smart speakers and set-top boxes have also gone up in price. The iPhone remains at its original price for now—but Apple has hinted that this may not last long.

Apple’s stock fell following the announcement, dragging down the entire U.S. market with it. The market clearly didn’t expect this to happen so soon.

Why is Apple doing this—and why now?

The official explanation is clear: it’s due to a dramatic increase in the price of memory chips. Apple stated in a press release that it “has never seen the price of a component rise so quickly and to such an extent” —and that the situation has become unsustainable.

Two types of components are key: RAM (working memory) and NAND flash (storage). Both are found in every MacBook and iPad, and both have risen in price dramatically. DRAM chip prices rose by 90 to 95% in the first quarter of this year compared to the previous quarter. NAND flash prices rose by 55 to 60%. These aren’t figures from a dystopian novel—they’re real market data from the analytics firm TrendForce.

Why such a dramatic increase? Because chip manufacturers— Samsung $SSNLF, SK Hynix $HY9H.F, and Micron $MU — have redirected 93% of their production to specialized high-bandwidth memory (HBM), which powers AI servers. Every wafer allocated for a data center GPU is a wafer missing from an iPad or MacBook. It’s a zero-sum game.

Tim Cook spoke about the situation just a few days ago in an interview with the Wall Street Journal. He described the price increases as “inevitable” and added that Apple had tried to absorb the increases for as long as possible—but its inventory of previously purchased chips has run out.

The AI boom as a catalyst for higher consumer electronics prices

The entire situation stems from a structural shift in the industry, not a temporary fluctuation. Hyperscalers— Microsoft $MSFT, Google $GOOG, Meta $META, Amazon $AMZN — plan to spend over $600 billion on AI infrastructure this year. A large portion of this money is going toward servers equipped with HBM chips, which are much more difficult to manufacture than standard memory.

A single HBM wafer consumes capacity equivalent to the production of two to three conventional DRAM modules. The result: memory supplies for consumer electronics have fallen to critically low levels, and manufacturers like Apple simply have nothing to draw from.

New production capacity from Micron, SK Hynix, or Samsung won’t come online until 2027 at the earliest. Until then, the market will be as tight as a drum.

"The consumer electronics industry is facing an unprecedented challenge. The rapid expansion of AI data centers has created an extraordinary surge in demand for memory and storage."

Apple, official statement on price increases, June 2026

Who will raise prices next?

Apple isn’t the first—and it certainly won’t be the last. Microsoft $MSFT raised prices on its Xbox Series X consoles by 20%; Dell $DELL and HP $HPQ have warned of rising laptop prices; Samsung and Sony $SONY are next in line.

Analyst David Naranjo of Counterpoint Research openly states that other PC and tablet manufacturers will follow Apple’s lead. Some will raise prices outright, while others will prefer to reduce discounts on entry-level models or quietly downgrade specifications.

Apple, however, has one key advantage: a loyal customer base that won’t easily abandon the platform. Dipanjan Chatterjee of the research firm Forrester sums it up bluntly:

“If anyone can survive a price hike with minimal consequences, it’s Apple. Their customers are loyal and will accept a higher price rather than switch to a competing product.”

Dipanjan Chatterjee, Vice President and Principal Analyst, Forrester

Other manufacturers—especially those with lower-priced models— are in a significantly worse position. For a mid-range Android phone, memory accounts for 15 to 20% of total production costs. With such a price increase for components , they have no room to maneuver.

The iPhone still costs the same. But for how long?

The price hikes haven’t spared the Mac and iPad. The iPhone has so far been spared—and that’s by design. Apple needs to protect the most important product in its portfolio, which forms the backbone of its entire ecosystem.

However, JPMorgan analysts estimate that the share of memory and storage in the iPhone’s total production costs could rise from today’s 10–15% to more than 45% by 2027. That’s a figure that simply cannot be absorbed into the margin.

Cook himself has hinted that the iPhone will not escape a price hike. The new iPhone 18 series is expected this fall—and it will be more expensive than its predecessor.

The chip crisis has thus ceased to be just a problem for data centers. It has entered living rooms, offices, and pockets. And electronics manufacturers are now passing it on where you’ll feel it the most—on the price tag at the register.

Few stocks have divided investors as sharply this year as Adobe. As recently as 2021, it was one of the market’s darlings, with a price exceeding $680 and a market capitalization of over $300 billion. Today, the same company is trading around $199 per share, and its market value has fallen below $80 billion. Yet this slump isn’t accompanied by any accounting disaster or revenue collapse. On the contrary, Adobe has just reported a record quarter and raised its full-year outlook. Nevertheless, the market is punishing it as a company in decline.

The reason is fear—specifically, the concern that generative AI will transform content creation to such an extent that professional software tools like Photoshop or Illustrator will no longer be a necessity for the average user. The market has begun to refer to this fear as “SaaS-mageddon”—that is, the concern that the rise of agent-based artificial intelligence will undermine the model in which software companies charge a fee for each user and each license seat. Adobe has become its symbol. This year’s layoff wave even cost the company’s CEO his job.

From market darling to fallen angel

The stock’s price performance over the past year is deeply concerning in and of itself. $ADBE shares have traded in a range of roughly $190 to $393 over the last twelve months and are currently hovering just above their annual low. Since the beginning of the year, the stock has lost approximately 43%; over the past twelve months, roughly 48%; and over a five-year period, around 66%. It is now more than 72% below its all-time high from November 2021, when the stock traded above $688.

This decline is largely part of a broader trend. At the beginning of this year, the software and cloud sectors were hit by a sell-off. The market began to reevaluate companies whose revenue relies on per-user licensing fees, as agent AI could theoretically reduce the number of licenses needed. Adobe fit perfectly into this category and became one of the most closely watched casualties of the entire trend.

Business Model: Three Pillars and Where Value Is Created

The company is built on three segments, two of which account for the vast majority of its business.

Digital Media

The core of the company. This includes Creative Cloud (Photoshop, Illustrator, Premiere, Lightroom, and the new generative tool Firefly) and Document Cloud (Acrobat, the PDF format, and Adobe Express). This segment generates subscription revenue from creative professionals, students, businesses, and everyday consumers. It is this segment that the market fears the most, because image and video creation is currently the most visible battleground for generative AI.

Digital Experience

A less visible but strategically crucial segment. This is a marketing and analytics platform (Adobe Experience Cloud, Adobe Analytics, GenStudio) for major brands. Adobe claims that 99% of Fortune 100 companies use its tools. This enterprise business has high switching costs and is far more difficult to replace with a simple AI tool than a consumer application.

Publishing & Advertising

The smallest and historically declining segment, which also resulted in a one-time goodwill write-down in this year’s financial statements. It is marginal to the overall picture.

The key metric for the entire business is ARR (annual recurring revenue, i.e., annual recurring subscription revenue). It is this metric that the market watches most closely, as it shows how quickly the stable subscriber base is growing.

Second-Quarter Results: Record Revenue, but Also Warning Signs

Adobe released its second-quarter fiscal results on June 11, and on paper, they look very strong. The company reported record revenue of $6.62 billion and raised its full-year outlook.

• Record quarterly revenue of $6.62 billion.

• Adjusted (non-GAAP) earnings per share of $5.96, compared to $5.06 a year ago.

• Total ARR at the end of the quarter was $27.10 billion, of which approximately $480 million came from the newly consolidated Semrush.

• ARR from “AI-first” products tripled year-over-year and exceeded $500 million.

• ARR for the GenStudio platform grew by more than 25% year-over-year.

• In a single quarter alone, the company repurchased $2.11 billion worth of its own shares.

Reported GAAP earnings per share were only $4.25, significantly lower than the adjusted figures. The difference is largely attributable to one-time items, specifically a goodwill impairment charge of approximately $70 million (Publishing & Advertising segment) and a litigation reserve of roughly $30 million. These items are non-recurring, and the company’s core operating performance remains exceptionally profitable. The adjusted operating margin hovers around 44.5%, while the GAAP operating margin is around 35%. When evaluating the company, it’s important to consider both figures.

The problem, however, lay elsewhere entirely. Management stated that it would intentionally slow ARR growth in the second half of the year. Moreover, just a few days after the earnings release, on June 15, CFO Dan Durn unexpectedly stepped down. Steve Day has taken over as interim CFO. All of this comes at a time when the company is also searching for a new CEO.

Adobe’s Leadership Under Threat

This leadership uncertainty is one of the main reasons for the market’s caution. Longtime CEO Shantanu Narayen, who led the company for 18 years and transformed it from a license seller into a subscription giant, announced in March that he would step down as CEO as soon as his successor is named. He will remain with the company as chairman of the board, and a special committee is leading the search for a new CEO, considering both internal and external candidates. No specific date or name has been announced yet.

Investors believe this is, at least in part, a response to their impatience with how quickly Adobe is turning its AI rhetoric into tangible results. The combination of the CEO search and the simultaneous replacement of the CFO means that the most important strategic transformation in the company’s history will be led by a transitional management team. This is a risk that must be taken seriously, regardless of how good the quarterly numbers look.

A Major Strategic Shift: Freemium, Firefly, and the Bet on AI

It is precisely this planned slowdown in ARR growth that is, in fact, at the heart of the new strategy—not a sign of weakness. Adobe has decided to aggressively push the freemium model—offering basic features for free to attract the broadest possible user base and then gradually convert them to paid subscriptions.

Management explained that roughly half of the reduced ARR outlook for the second half of the year stems from the postponement of planned Creative Cloud price adjustments, and the other half from the full-scale rollout of the freemium expansion. The strategy is clear. Adobe is consciously sacrificing short-term subscription growth in exchange for widening the top of the funnel at a time when AI is changing the way people discover and start using software. The initial signs look promising. Traffic to adobe.com has increased by roughly 40%, and according to the company, users who upgrade from the free to the paid version show higher activity than those who start paying right away.

The second pillar is AI monetization itself. Adobe reports that consumption of generative credits has tripled quarter-over-quarter, and the number of first-time Firefly product subscriptions has doubled. This is important because it suggests that customers are paying for AI within Adobe, rather than leaving Adobe because of it. Furthermore, the company has adopted a smart multi-model strategy. Instead of competing with every AI startup individually, it integrates models from OpenAI, Google $GOOG, and others into its applications, positioning itself as a middle layer that connects these models. At the Cannes Lions festival, the company announced new partnerships in the field of agent AI with Accenture $ACN, WPP, Microsoft $MSFT, and Anthropic.

The Semrush Acquisition and the Question of Growth

In April, Adobe completed the acquisition of Semrush for approximately $1.9 billion in cash ( $12 per share). Semrush is a platform for brand visibility in search—that is, a tool for SEO and so-called generative optimization—ensuring that brands are found not only on Google but also in the responses of AI assistants.

Adobe can create content (Creative Cloud) and measure its impact (Experience Cloud); Semrush fills the missing piece—namely, whether a brand is even discoverable in a world where more and more people are making purchases based on AI recommendations.

Semrush adds approximately $480 million to total ARR. At the same time, however, Adobe has lowered its organic ARR outlook, leading some on Wall Street to suggest that the company’s growth is now “more acquired than organic.”

Citi $C, for example, lowered its price target to $228, citing an implied shortfall in organic ARR of approximately $500 million. In other words: acquisitions are roughly offsetting the organic growth that the company temporarily sacrificed due to its transition to a freemium model. Whether this is a smart trade-off or a way to mask a slowdown will become clear over the next two or three quarters.

Competition and Concerns

The Bears’ Arguments

Today, generative AI can create an image, video, design, or marketing text in a matter of seconds from a single text prompt. There are a number of challengers. Canva targets non-professional users and recently relaunched its Affinity creative suite as a single free app. Figma $FIG (listed on the stock exchange since summer 2025) dominates user interface design. Added to this are purely AI-based tools such as Midjourney, Runway, and OpenAI’s video model Sora.

Why pay for a complex professional tool when a simpler AI product can produce a “good enough” result more cheaply? If this reasoning holds true, Adobe’s overall market share will shrink, and pressure on prices—as well as the number of paid subscriptions—will increase.

The Bulls’ Arguments

The opposing side argues that the market is confusing the ability to generate an image with the ability to complete professional work. A professional doesn’t want to switch between five applications in a browser just to generate an image, edit it, convert it to vector, and prepare it for print. Adobe’s value lies not in a single step, but in an integrated workflow. Added to this are several structural advantages. High switching costs for enterprise customers, standardized formats (PDF, PSD), Firefly training data with resolved copyright issues, and a commercial guarantee that the generated content won’t infringe on anyone’s rights. For major brands dealing with legal risks, this is a crucial difference compared to freely available models.

So far, reality has been on the bulls’ side in one specific regard. The numbers do not confirm any decline. Revenue is accelerating, margins are among the highest in the sector, and AI monetization is growing. So the question isn’t whether AI is hurting Adobe today, but whether it will hurt the company in a few years. No one knows the answer to that, and it is precisely this uncertainty that is currently priced into the stock.

Key Metrics at a Glance

Caution regarding the interpretation of ROE. A return on equity of around 63% looks phenomenal, but it is significantly inflated by aggressive share buybacks. Massive share buybacks reduce the book value of equity in the denominator, so the ratio rises mechanically without necessarily indicating proportionally higher operational efficiency.

Similarly, the low P/E ratio should be viewed in the context of one-time items and the expected slowdown, not as an automatic signal of undervaluation.

A Strategic Perspective

Adobe is today a textbook example of the battle between fundamentals and narrative. From a purely numerical perspective, it is a company with double-digit revenue growth, one of the highest margins in the sector, massive free cash flow, and a valuation at single-digit to low double-digit multiples of earnings. The market typically reserves such a combination for companies with structurally declining businesses, not for one that is currently raising its outlook.

On the other hand, there are three real risks that cannot be ignored. First, the existential question of AI: whether “good enough” tools will gain a foothold in the lower and middle segments of the market. Second, the company is in a transitional phase without a permanent CEO or CFO, precisely at a time of fundamental transformation. Third, the quality of growth—that is, the unfortunate coincidence where organic growth is temporarily slowing down and acquisitions are roughly replacing it.

This year’s Adobe story is essentially a bet on whose narrative is correct. Either the market is correctly anticipating a slow decline and the current price is fair to generous, or it is over-extrapolating fears and offering a high-quality company with an established ecosystem at an exceptionally low price. The answer won’t come from a single quarter, but rather from whether the freemium transition drives conversions, whether the new leadership stays the course, and whether AI monetization grows faster than AI erodes the user base.

P.S.: A video about Adobe will be released this Sunday on the Bulios YouTube channel, so be sure to subscribe so you don’t miss it: Bulios YouTube

However, the investment thesis explored in this article is not based solely on current figures, but on one specific opportunity that, according to the bulls, the market has not yet factored into the price. That opportunity is the Golden Dome missile defense project, with a total budget of around $185 billion, for which Palantir is collaborating with Anduril on the software layer. The key question is how much of this massive program will actually go to Palantir, how quickly, and whether today’s extreme valuation—with the stock trading at tens of times revenue—already factors in this hope, just as it does dozens of other optimistic assumptions. In other words, is the Golden Dome a hidden catalyst, or just another floor added to an already towering edifice of expectations?

Key Points of the Analysis

• Growth accelerated to its highest level since the company’s IPO. Revenue in the first quarter of 2026 rose 85 percent year-over-year, U.S. revenue by 104 percent to $1.28 billion, with the U.S. commercial segment up 133 percent to $595 million and the U.S. government segment up 84 percent to $687 million.

• The company is now solidly profitable, not just on an adjusted basis. Reported net income in the first quarter reached approximately $870 million, roughly four times the amount from the previous year; the adjusted operating margin was 60 percent, and adjusted free cash flow was $925 million with a margin of 57 percent. This sets Palantir apart from most fast-growing software companies, which are still striving to become profitable.

• Management has once again raised its full-year outlook. For 2026, the company expects revenue of $7.65 to $7.66 billion, representing growth of around 71 percent— which is ten percentage points higher than the previous estimate, and U.S. commercial revenue of over $3.22 billion, with growth of around 120 percent. The company also promises accounting operating and net income in every quarter of the year.

• Forward visibility has improved. Unreported contract commitments rose to approximately $4.45 billion from $1.9 billion a year ago, the U.S. commercial value of unfulfilled orders stood at around 4.92 billion, representing 112 percent growth, and the revenue retention rate from existing customers reached 150 percent.

• The centerpiece of the bullish thesis is the Golden Dome. The entire missile defense program has a budget of around $185 billion, and Palantir, together with Anduril, is providing the software “glue layer” designed to integrate satellites, radars, sensors, and weapon systems into a single command-and-control system. Importantly, Palantir’s Maven system has been designated by the Department of Defense as the reference program for combat command and control, giving the company a structural advantage.

• The valuation is extreme and is the main source of controversy. The stock is trading around $128 with a market capitalization of approximately $308 billion; its price-to-sales ratio is in the tens, and its price-to-earnings ratio exceeds 140. This is one of the highest valuations in the S&P 500 index.

What Has Changed at the Company

Palantir ($PLTR ) was long perceived as a slow-growing, secretive company with ties to intelligence agencies and the military, whose software, while powerful, was difficult to scale and dependent on the costly involvement of its own engineers. This image has fundamentally changed with the introduction of the AIP (Artificial Intelligence Platform), which has opened up the commercial market to the company in a way that was previously impossible.

The change has two dimensions. First, the company’s U.S. commercial business has accelerated dramatically, growing at a rate of over 100 percent and demonstrating that Palantir’s tools are no longer just for the government and large corporations, but are becoming a broader standard for deploying artificial intelligence into business processes. Second, the company has become profitable on an accounting basis and is generating strong cash flow, thereby dispelling its main historical criticism—namely, that it buys growth at the cost of persistent losses and extensive stock-based compensation.

Underlying this is also a strategic shift in the company’s narrative. CEO Alex Karp actively distinguishes Palantir from creators of foundational models, such as OpenAI or Anthropic, arguing that while token prices for these models are falling and the winners change every six months, Palantir sells the application and integration layer that actually turns models into usable tools. This thesis—that value is shifting from models to their deployment—is now at the heart of the entire narrative.

How This Turns Into Money

Palantir monetizes artificial intelligence through the sale of software platforms, not through the sale of computing power or models themselves. Specifically, there are three products. Gotham serves defense and intelligence; Foundry is an enterprise operating system for data; and AIP is a layer that embeds generative artificial intelligence into these platforms and enables its direct deployment into operational decisions.

The business model is based on the company first deploying its software as part of a short, intensive pilot project; if the customer sees value, the contract is quickly expanded. This is reflected in an exceptionally high revenue retention rate of 150 percent, meaning that existing customers spend significantly more annually than they did previously. Margins are very high, with the gross margin exceeding 80 percent and the adjusted operating margin reaching 60 percent, as this is recurring-revenue software with low incremental costs.

The government business operates similarly, but with longer-term and larger contracts. Examples include a multi-year military contract worth up to $10 billion and a fourfold increase in the use of the Maven system over twelve months. It is precisely the transition of pilot deployments in defense to long-term programs that turns technology into stable, recurring revenue.

What Is the Main Growth Driver

Today’s main driver is the U.S. commercial segment. It is growing the fastest, at a rate exceeding 100 percent, and is proof that Palantir’s offerings have a market far beyond the government sector. It is precisely this segment that has driven the most significant increase in the full-year outlook and is the main reason why the company has accelerated to the highest growth rate in its history.

The U.S. government sector is the second driver; while it is somewhat slower, it is more stable and offers greater long-term visibility. This sector also holds the greatest long-term potential in the form of mega-defense programs, of which Golden Dome is a part.

Outside the U.S., the company is growing more slowly and facing political and regulatory hurdles, particularly in Europe, where some customers and regulators are concerned about dependence on software controlled from the United States. International expansion is therefore not yet the main driver, but rather a complement with mixed results.

Products and Upcoming Strategy

The centerpiece of the product strategy is the AIP platform, around which the company is building an ecosystem through intensive customer events where it demonstrates specific use cases. The emphasis is shifting from selling individual tools to the idea that Palantir is an operational layer on which organizations run artificial intelligence across their processes.

In the defense sector, the key is the Maven Smart System, whose status as a reference program for combat command and control gives the company an advantage that competitors would take years to achieve. It is precisely this position that makes Palantir a natural candidate for the software layer of the Golden Dome program, where the general in charge of the program has described the software as the most important part of the entire system.

Strategically, the company is betting that artificial intelligence is becoming a commodity at the model level, while lasting value lies in data integration, in ontology—that is, in the way the software maps the organization’s reality—and in operational deployment. If this thesis is correct, Palantir’s lead in the integration layer is difficult to replicate.

Market Position and Share

Palantir does not operate in a single, easily definable market, but rather at the intersection of several categories, ranging from data analytics to artificial intelligence platforms to defense software. This complicates the measurement of market share, as there is no single competitor doing the same thing on the same scale.

In the defense and intelligence segment, the company holds an exceptionally strong—and in some areas, dominant—position, bolstered by the status of the Maven system as a benchmark program. In the commercial sphere, however, it is one of many players in the rapidly growing but fragmented enterprise artificial intelligence market, where it competes with both large cloud providers and smaller specialized firms.

The overall market in which Palantir operates is growing rapidly, as both businesses and governments are only just beginning to deploy artificial intelligence. This is a favorable backdrop for the company, but it also means that the size of the ultimate market and Palantir’s share of it are currently more a matter of estimation than measurement.

Main Competitors

In the commercial segment, Palantir faces pressure from two directions. On one hand, there are the major cloud providers—namely Microsoft $MSFT, Google Cloud $GOOG, and Amazon $AMZN—which offer their own tools for deploying artificial intelligence and have deep relationships with enterprise customers. On the other hand, there are creators of foundational models such as OpenAI and Anthropic, against whom Palantir positions itself, but who are also moving toward the application layer.

In the defense sector, competitors include traditional contractors such as Lockheed Martin $LMT, RTX $RTX, and Northrop Grumman $NOC, which serve as the primary hardware suppliers for the Golden Dome program, while Palantir and Anduril provide the software. The relationship is therefore complementary rather than directly competitive, although there is competition over the allocation of the software component.

A key competitive risk is the argument that barriers to entry in AI software are falling as foundational models improve. If it turns out that the integration layer Palantir sells can be easily replicated, the entire argument for a sustainable lead would be undermined. This is at the heart of the bears’ arguments, including the well-known criticism that the company is a house of cards built on a narrative about artificial intelligence applications.

What the Market Is Valuing Correctly—and What It May Not

The market appears to be correctly valuing the exceptional growth rate and actual profitability. Unlike many growth companies, Palantir generates both accounting profit and strong cash flow, and its high revenue retention rate confirms that customers continue to spend more after implementation. These qualities justify a premium valuation compared to standard software.

What is debatable is the extent of this premium. With a price-to-revenue ratio in the tens, the price reflects many years of flawless growth with high margins. Any setback—whether a slowdown in commercial growth, the cancellation of government contracts, or intensified competition—could lead to a sharp compression of the multiple, as demonstrated by this year’s decline in the stock price despite accelerating fundamentals.

As for the Golden Dome itself, it’s important to be precise. The figure of $185 billion is the budget for the entire program, not the value of Palantir’s contract. The company’s actual share is uncertain, and estimates vary; one of the more optimistic projections suggests that Palantir could secure around $18 billion in government contracts between 2026 and 2028—more than the current consensus. Furthermore, the program is not expected to be completed until around 2035, and a formal major contract for Palantir has not yet been confirmed. The argument that the market is not factoring this opportunity in is therefore only half true. The market does not fully reflect it in the price precisely because it is uncertain, far in the future, and difficult to quantify—not necessarily because it is being overlooked.

Numbers

• Revenue in Q1 2026: $1.63 billion, 85 percent year-over-year growth, the highest since the company’s IPO.

• U.S. revenue: $1.28 billion, up 104 percent, of which commercial revenue was $595 million (up 133 percent) and government revenue was $687 million (up 84 percent).

• Adjusted operating margin: 60 percent; adjusted free cash flow: $925 million, with a margin of 57 percent.

• Reported net income: approximately $870 million, about four times the previous year’s figure.

• Rule 40: score of 145 percent.

• Revenue retention rate from existing customers: 150 percent.

• Unreported contractual obligations: around $4.45 billion, up from $1.9 billion a year ago.

• Full-year outlook for 2026: revenue of $7.65 to $7.66 billion (growth of about 71 percent), U.S. commercial segment over $3.22 billion (growth of about 120 percent).

• Revenue for 2025: $4.48 billion, 56 percent growth, net income of $1.63 billion.

• Number of commercial customers: 1,007 over the past twelve months, a 31 percent increase.

• Cash and short-term investments: approximately $8 billion, net cash position, virtually debt-free.

Valuation

Palantir’s valuation is at the heart of the entire debate surrounding the stock. At a price of around $128 and a market capitalization of around $308 billion, the stock is trading at a price-to-revenue ratio in the tens and a price-to-earnings ratio exceeding 140, making it one of the most expensive large companies on the U.S. market. Even looking ahead to next year, the multiple remains extreme.

What is factored into such a price? Essentially, a scenario in which the company grows at a rapid pace for many years in a row, maintains margins above 50 percent, and gradually gains a significant share of the massive defense and commercial markets. In other words, today’s price doesn’t just assume success—it assumes near-flawless, long-term success.

This creates an uncomfortable asymmetry. For another upward rerating, the company needs not only to meet but to significantly exceed already high expectations—for example, by securing a specific, large contract within the Golden Dome project or by maintaining 100% growth in the commercial segment in the coming quarters. Conversely, it takes relatively little to trigger a decline in valuation—namely, any sign of a slowdown—because at such a high multiple, the market has no built-in buffer. This year’s decline in the stock price, despite accelerating growth, is a harbinger of just how sensitive the price is to shifts in market sentiment. Compared to companies like Baidu, where cash forms the basis of its valuation, or Alphabet, where the multiple remains reasonable despite a record-high price, Palantir is at the opposite end of the spectrum—a stock where fundamentals must catch up to the price, not the other way around.

Risks

The biggest risk is the valuation itself. At the multiples at which the stock is trading, the room for disappointment is minimal, and any slowdown in growth or compression of margins could lead to a sharp decline, even if the business remains healthy. This risk is independent of the company’s quality and stems purely from the price.

Dependence on government and political contracts is another risk. A large portion of revenue comes from U.S. government contracts, which may include termination clauses at the customer’s convenience and are sensitive to political and budgetary changes. The company’s ties to the current administration are both an advantage and a vulnerability.

International and regulatory risks are particularly evident in Europe, where there are concerns about strategic dependence on software controlled from the United States, and where, according to reports, the company lost a lawsuit in Switzerland regarding government contracts. This could hinder expansion into markets outside the U.S.

Competitive risk stems from the argument that barriers to entry are falling as underlying models improve. If the integration layer that Palantir sells were to lose its uniqueness, the entire argument for a sustainable competitive advantage would be weakened. This also includes criticism that the company’s value is based primarily on its narrative.

The risk of stock dilution remains relevant because the company has historically generously rewarded employees with stock, which increases the number of shares and dilutes earnings per share. This risk has lessened with the onset of profitability, but it has not disappeared. Financial risk, on the other hand, is low; the company holds net cash and has no significant debt.

Investment Scenarios

Optimistic Scenario

For this scenario to materialize, the U.S. commercial segment would need to maintain growth of around 100 percent in the coming quarters, the company would need to secure a specific, large contract under the Golden Dome program or other major defense programs, and margins would need to remain above 50 percent. In that case, the narrative that value is shifting toward the application layer would be confirmed, and the market would remain willing to pay a high multiple. Revenue could grow several-fold over the next few years, and the stock would return to the upper end of the analyst range—around $255 or higher—if the defense opportunity began to materialize with concrete figures.

Realistic Scenario

The most likely outcome is continued strong growth, which, however, will gradually slow from extreme levels as the customer base expands. The commercial segment continues to grow rapidly, but the pace normalizes from over 100 percent; the government business segment adds steadily; and Golden Dome contributes gradually and to a lesser extent than the program’s overall budget suggests. Margins remain high, and the company meets or slightly exceeds its outlook. In this scenario, the stock trades within a wide and volatile range around the analyst target price of roughly $183 to $194, with the tension between strong fundamentals and a high valuation persisting and the price reacting sensitively to each quarterly report.

Pessimistic Scenario

A negative outcome would involve a slowdown in commercial growth, disappointment regarding the pace or scope of defense contracts, intensified competition from cloud providers and model developers, and, potentially, the cancellation or reduction of government contracts. Given the extreme valuation, even a slight disappointment would be enough to cause a sharp drop in the P/E ratio. The stock could correct back to the lower end of its current range—well below $100—as suggested by the range of analyst estimates, whose lower bounds hover between $70 and $90. Unlike weaker companies, however, Palantir remains profitable and debt-free even in this scenario, so this is not an existential risk but rather a painful revaluation.

What to Watch Next

• Monitor the growth rate of the U.S. commercial segment. The key will be whether it maintains growth above 100 percent in the coming quarters, as that is currently the main driver and the primary basis for the valuation.

• Monitor specific developments regarding Golden Dome and other defense programs. The decisive factor will be whether a formal, quantifiable contract for Palantir materializes—not just the overall program budget.

• Monitor margins and share dilution. Maintaining an adjusted operating margin around sixty percent and controlling stock-based compensation will determine how quickly earnings per share will grow.

• Monitor international and regulatory developments, particularly in Europe, where concerns about strategic dependence on U.S. software and legal disputes could hinder expansion.

• Monitor the stock price’s reaction to earnings. With such a high valuation, it’s important to determine whether even excellent results will be enough to sustain the price, or whether the market will begin to demand an ever-higher bar.

Key Takeaways from the Article

Palantir is currently a company with exceptional fundamentals and an extremely high price tag. It is growing faster than ever before, is truly profitable, generates strong cash flow, and has a structural advantage in the defense sector thanks to the Maven system. This makes it one of the few growth companies that no longer needs to prove its economic viability.

The central argument regarding the $185 billion contract—which the market has not factored in—should be viewed with a grain of salt. This figure represents the budget for the entire Golden Dome program, not a contract for Palantir; Palantir’s actual share is uncertain, spread out over many years, and not yet formally confirmed. The opportunity is real and potentially very valuable, but it is an option for the future, not a guaranteed revenue stream, and that is precisely why the market does not fully reflect it in the price.

The main source of tension surrounding the stock lies not in the quality of the business, but in its price. At the multiples at which Palantir is trading, the price already factors in many years of nearly flawless growth, so the room for disappointment is minimal and sensitivity to market sentiment is enormous. For an investor who believes in the long-term thesis of a shift in value toward the application layer of artificial intelligence and is willing to tolerate high volatility, this could be a quality long-term position. Conversely, an investor sensitive to valuation should exercise caution, because even an outstanding company may not be a good investment at any price.

For shareholders, rising hardware prices could weigh on demand if buyers hold off on upgrades, but they may also help protect margins if the increases offset the higher input costs. The extent of the impact will depend on how consumers respond and whether additional products are repriced.

For shareholders, a government investigation into pricing practices adds regulatory and reputational uncertainty. The scope, legal basis, and potential consequences of the probe have not been disclosed.

Separately, EU regulators have said that Amazon and Microsoft cloud computing services should fall under the bloc's stricter technology rules. If that determination stands, it would subject their cloud offerings to additional compliance requirements in Europe.

$CEZ.PR was once a golden dividend machine, regularly paying out tens of billions into the state treasury and into the pockets of small investors. Now ČEZ stands at a crossroads: Andrej Babiš’s government has launched a process that will ultimately transform the energy giant beyond recognition. One part will pass into the hands of the state, while the other will end up in the hands of private investors. And in the meantime, roughly 160,000 small shareholders are waiting to see what the state will offer them.

The Historic Shareholders’ Meeting That Changed the Game

The June 2026 general meeting lasted over ten hours. But when it ended, the decision had been made: shareholders approved, with more than 90 percent of the votes, the spin-off of ČEZ’s non-generation division into a new subsidiary with the working name DSZS. This new entity will include ČEZ Prodej, ČEZ Distribuce, GasNet, ČEZ ESCO, and commercial activities—essentially everything that comes into contact with customers and regulated distribution.

The generation division—power plants, nuclear units, and coal-fired power sources—will remain within the original ČEZ and will be nationalized.

The new company is expected to be formally established within a matter of weeks, with the selected businesses transferring to it by the first quarter of 2027. The state will retain a 51 percent stake in the subsidiary and offer the remaining 49 percent to private investors —through an IPO, institutional funds, or a share swap with ČEZ’s current minority shareholders.

“The book value of ČEZ Distribuce’s assets amounts to 110 billion crowns, GasNet’s to another 11 billion, ČEZ Prodej’s to 9 billion, and ČEZ ESCO’s to 15 billion. However, the total valuation of the subsidiary will depend on market value, not book value.”

Pavel Cyrani, Vice Chairman of the Board of Directors of ČEZ

The key question: How much will the state pay for it?

This is the crux of the whole story. Nationalizing ČEZ’s generation division will not come cheap—the state holds approximately 70 percent of the shares through the Ministry of Finance, while the rest belong to private shareholders. In order to squeeze out minority shareholders under the Business Corporations Act, it must exceed the 90 percent threshold.

This means buying up roughly 20 percent of the shares voluntarily—and forcing at least some of the remaining shareholders to sell at a so-called fair price.

At the current market price of around 1,200 crowns per share, ČEZ’s market capitalization stands at approximately 660 billion crowns. The state would pay around 190 billion for the 30 percent stake held by private shareholders. However, analysts point out that this won’t be possible without a premium— key players such as Pavel Tykač and investor Michal Šnobr, who together with their groups hold over 10 percent of the shares, have a strong negotiating position. Without their consent, the state will not be able to acquire a 90 percent stake.

A comparison with France suggests what to expect: during the nationalization of EDF in 2023, the French government offered a premium of roughly 50 percent above the market price. Applied to ČEZ, this would mean a buyout price of around 1,800 crowns per share.

Why the Dividend Is Falling—and What It Signals

At the same annual general meeting, shareholders approved a dividend of 42 korunas per share—five korunas less than last year, and significantly less than the record 145 korunas in 2022. In total, the company will distribute roughly 22.6 billion crowns, of which the state, as the majority shareholder, will receive about 16 billion.

The declining dividend is no coincidence. ČEZ announced an investment program totaling 400 billion crowns for the years 2026 through 2030. The funds are earmarked for new energy sources, grid modernization, gas-fired power plants, and preparations for nuclear power plant construction. The more the company invests, the less it can pay out to shareholders.

At the current share price, the dividend yield is around 3.3 percent—not exactly a dazzling figure for investors who have traditionally viewed ČEZ as a reliable dividend stock.

“The market had already priced in a dividend of 42 crowns, so its approval did not trigger any movement in the stock price. Given the current market price, the dividend yield is about 3.3 percent, which is not particularly impressive.”

Tomáš Pfeiler, portfolio manager at Cyrrus

What Happens to the Money—and Who Bears the Costs

According to Deputy Prime Minister Karel Havlíček, the buyout of private shareholders should be financed from ČEZ’s own resources, not from the state budget. Part of the funding is to come from the sale of a 49 percent stake in the new subsidiary DSZS.

It’s an elegant arrangement on paper: the company will finance its own nationalization. But analysts are skeptical. The cost of buying out minority shareholders is estimated at a quarter of a trillion crowns, taking into account the necessary premium. At the same time, ČEZ faces massive investment demands. Managing both at once will be a key test for management.

The entire process is expected to take one and a half to two years. Havlíček has promised that the government will complete it by the end of its term.

What does all this mean for ČEZ shareholders?

The situation is unusual. ČEZ remains the most liquid stock on the Prague Stock Exchange and the backbone of the PX index. At the same time, the moment is approaching when the generation division will likely disappear from the stock exchange entirely —institutional investors, who by law can hold only publicly traded shares, will have to sell.

Retail shareholders—numbering around 160,000—face a simple question: sell now, or wait for the government’s buyout offer in the hope of a premium? The answer depends on the price the government ultimately offers. And no one knows that yet—Havlíček is deliberately keeping it under wraps so as not to influence the market.

ČEZ, as the Czech public has known it for the past 30 years, will cease to exist. Whether this will be good or bad news for those who hold the company in their portfolios will depend on the offer the government ultimately makes.

For shareholders, the resolution removes one legal uncertainty, though the absence of disclosed terms makes it difficult to assess the financial impact. Regulatory and legal pressure on large platforms over child safety remains a broader industry theme, and the case adds to the list of litigation Google has faced in this area.

The size of the customer commitments suggests that buyers are locking in supply well in advance, which the company said points to sustained appetite for memory chips. For shareholders, beating estimates alongside a large deal pipeline provides reassurance that demand trends remain favorable, even as the broader semiconductor market has faced periods of uncertainty.

For shareholders, the suit adds to the legal exposure the company faces. The financial impact would depend on how the litigation develops and whether any fault is attributed to the vehicle or its systems.

For shareholders, the concern is less about any single departure and more about what a sustained flow of senior AI staff toward a well-funded competitor could mean for Google's ability to stay at the forefront of AI development, a segment that management has identified as central to its long-term growth.

For investors, inclusion in the Dow can matter because index funds and products that track it must hold the stock, which can affect demand. The report highlights a broader challenge for index compilers in keeping traditional benchmarks relevant as tech-driven companies come to dominate market activity.

For shareholders, a government inquiry into pricing practices adds a layer of regulatory and reputational uncertainty. The outcome and any potential consequences would depend on how the probe develops.

For shareholders, NTSB investigations into crashes can lead to broader findings about vehicle systems and, in some cases, prompt regulatory action. The outcome and timeline of the investigation have not been disclosed.

For Amazon shareholders, Zoox's transition to a paid ride service is a key milestone in determining whether the unit can eventually justify the investment behind it. The timeline and specific markets for the expansion have not been disclosed.

For shareholders, the allegation introduces reputational and potential legal risk at a time when Alibaba is investing heavily in its own AI capabilities. The details of the accusation, including its scope and any formal proceedings, have not been disclosed in the available reporting.

For shareholders, the potential loss of customer data at a third-party facility could expose the company to reputational and legal risks, depending on the scope of the damage and what data was affected. Details on the extent of Google's exposure have not been disclosed.

Netflix is one of those stocks retail investors buy and often get burned on, because the thesis like "everyone has Netflix, it can’t fall" doesn’t hold. Netflix is currently trading around $72. Is this an opportunity or just a cheap-looking trap?

When you see a quality company drop 45% from its peak, the first instinct is obvious. Netflix has 325 million subscribers, nearly a 30% operating margin, and cash flow that grows every year. On paper, it’s every investor’s dream.

But what’s on paper can be misleading.

Before you jump into a position, you need to ask one question and answer it honestly. "Is this a real decline of a great company to an attractive price, or does the market see something the average investor is overlooking?"

Let’s break it down!

Why did the stock fall? It’s not just one reason

Over the past year Netflix fell from its 52-week high above $134 to roughly $74 today. That’s a 45% drop. Meanwhile the company itself reports record results.

A combination of factors piled up at once. The co-founder and the person who built Netflix from DVDs to global streaming dominance, Reed Hastings, is leaving the board. At the same time the mega-acquisition of Warner Bros. Discovery collapsed, then Netflix lost the bid for Roku, and the market now lives in uncertainty: what will they buy next and for how much, if they buy at all. :)

Add the macro context. The Fed is keeping rates higher for longer, a strong dollar is pressuring international revenues, and Meta $META has just launched Instagram for TV — a direct battle for screen time.

And on top of that: Netflix stopped reporting quarterly subscriber counts. One of the key metrics the market used to gauge growth momentum.

The business is truly exceptional

It’s important to be honest both ways. Netflix is not an overhyped startup without fundamentals. It’s one of the highest-quality media businesses in the world.

Revenue in 2025 reached $45.2 billion, year-over-year growth nearly 16%. Operating margin 29.5%, net income $11 billion and free cash flow $9.5 billion — that’s the number that matters to investors.

Yet as recently as 2022 Netflix was chronically cash-negative. It was even nicknamed "Debtflix" on Twitter — a company that borrows for content and never pays it off.

The competitive advantage has four layers that are hard to replicate. The data flywheel — Netflix knows 325 million people better than anyone else. It knows what you watch, when you close the app, what brings you back. It translates that knowledge into more efficient content investment than competitors who are burning cash blindly.

Original content that no one else has. No other streamer produces global hits with that consistency.

Pricing power. Netflix has repeatedly raised prices and subscribers remain. In April 2026 it did so again. That’s a strong signal — a company with a lower-quality product couldn’t have pulled that off.

And then there’s the ad business, which was virtually zero two years ago and now generates over $1.5 billion annually. Management targets doubling to $3 billion in 2026.

Red flags that can't be ignored

And now the part that made me decide not to rush into buying.

Insiders are selling aggressively. Reed Hastings sold shares worth over $37 billion. CFO Spencer Neumann sold roughly $5.4 billion. Over the past 90 days total net insider sales exceeded $61 billion. Insiders know the company better than anyone outside. They don’t sell this aggressively if they think the stock is cheap.

M&A chaos without strategy. Netflix unsuccessfully pursued Warner Bros. Discovery for about $72 billion. Then it lost the battle for Roku. And now the market is speculating about Lionsgate. The problem isn’t that Netflix wants to grow via acquisitions — the problem is investors don’t see a clear logic. Every report of a potential deal sends the shares down because investors fear management will overpay for the wrong target.

End of subscriber reporting. The market lost a key, simple way to measure whether the growth story is working. Netflix replaced subscriber numbers with ads and pricing power — these are legitimate metrics, but harder to read in real time.

Q2 outlook disappointed. Management admitted that part of the costs related to the aborted WBD acquisition will be pushed into 2026.

Valuation and fundamental analysis

Here’s the core of the matter, and I’ll say it straight — the numbers aren’t as convincing as they might seem at first glance.

DCF analysis shows three scenarios:

Pessimistic scenario — FCF grows 8% annually and the discount rate is 11%, giving a fair value around $40 per share. At $74 you’d be paying about an 85% premium over fair value.

Base case — 12% annual FCF growth and a 10% discount rate gives a fair value around $51. Still overvalued by roughly 45%.

Optimistic scenario — FCF growth 16%, discount rate below 10%, fair value approaches $73. So practically at today’s price.

In other words, today you’re paying as if everything will follow the optimistic scenario — no margin of safety, no room for error.

The FCF growth implied by the current market cap is roughly 3%. That’s not value in the classic sense. For a growing business with 12–14% annual revenue growth that’s acceptable, but it’s a bet that the growth story will continue without major stumbles.

My investment thesis

Netflix is exactly the kind of stock that tempts me and at the same time holds me back.

I really like the business. Margins are rising, the advertising segment is accelerating, subscribers stick around despite repeated price increases. And a 45% drop from the peak combined with a $25 billion buyback program suggests that management also thinks the stock is undervalued.

But insider sales of $61 billion over 90 days is something you can’t simply move past with a calm heart. Hastings spent his entire professional life building Netflix. If he sees the right moment for a massive exit now rather than a year ago at a higher price, you need to ask why.

Netflix had a clear identity: organic growth, original content, data. Every speculation about a big acquisition undermines that identity and adds a risky element.

My view is simple — I won’t buy before I see Q2 results that confirm Q1 guidance wasn’t a one-off fluctuation. And I need to see a clear signal that M&A appetite has calmed and capital is going back into buybacks. The entry zone where I’d be willing to start building a position is somewhere between $62 and $65.

Until then I’m watching, not rushing.

For now I’m putting Netflix on my watchlist, but I’m curious about you — did you buy shares of $NFLX or will you be watching?

And then there’s the peculiarity that has the company in the spotlight right now. A company burdened with over $8 billion in debt has decided to return roughly half a billion dollars to shareholders—not in cash, however, but in the form of newly issued 9% preferred shares. So instead of reducing its debt, it is taking on another fixed obligation. The question for an investor isn’t “Is this cheap?”—because relative to cash flow, it’s extremely cheap—but rather “Is this debt manageable, and what does this strange dividend actually signal?”

Key Points of the Analysis

• Revenue for 2025 totaled $4.44 billion and declined slightly year-over-year (by 0.33%). This is a mature, slowly shrinking telecommunications business with a gross margin of around 68%.

• The company has been operating at a loss on an accounting basis for a long time. The net loss for 2025 reached $611 million, mainly due to high interest expenses, depreciation, and foreign exchange effects. However, profitability at the operating cash flow level is healthy.

• Free cash flow rose by 42% to $306 million, with an FCF margin of 15%. Cash flow, not accounting profit, is the relevant metric here.

• Debt is the dominant issue: consolidated net debt is around $7.6 billion, and net leverage is roughly 4.5 times adjusted OIBDA. However, the subsidiary Liberty Puerto Rico is dangerously indebted (net leverage around 8x, covenant-based leverage as high as 14x).

• At first glance, the valuation is absurdly low: a price-to-free-cash-flow ratio of just 1.77 and a price-to-revenue ratio of 0.27. This is a level at which either companies on the verge of bankruptcy or deeply undervalued opportunities trade.

• Special dividend: For every 10 common shares, shareholders received one 9% preferred share with a liquidation value of $25, totaling approximately $500 million. It is paid in securities, not cash.

• Main risk: debt refinancing, particularly in Puerto Rico. If the company fails to meet maturities and interest payments amid weak operating performance, the value of the shares could plummet faster than that of a typical company due to leverage.

Company Overview

We’re talking about Liberty Latin America $LILA (also traded under the tickers LILAK and LILAB), a telecommunications operator active in the Caribbean and Latin America. The company is registered in Bermuda, managed from Denver, employs around 10,000 people, and is led by Balan Nair. It belongs to a broader constellation of companies associated with investor John Malone, which is an important context: these companies tend to be heavily indebted, managed with an emphasis on cash flow and tax efficiency, and often undergo spin-offs, mergers, and financial engineering.

At its core, the business is simple and stable. The company provides households and businesses with broadband internet, video, landline and mobile phone services, and—through its network of subsidiaries—subsea data links between regions. Revenue comes primarily from recurring subscriptions, with broadband accounting for roughly 80% of the average customer’s bill. For a financially savvy layperson, the best analogy is this: it is a collection of regional cable and mobile operators, acquired on debt, whose stable cash flows are intended to pay down that debt and gradually build value for shareholders. The whole story therefore stands or falls on whether operating cash flow can service the debt burden.

Segments and Where Cash Is Generated

The company is divided into several regional units, which vary in terms of financial health and debt levels. To understand the risk, it is helpful to examine them through the lens of each group’s debt:

• C&W (Liberty Caribbean, Panama, Liberty Networks): the largest unit with debt of around $4.9 billion and covenant leverage of 3.5x, making it the relatively healthiest part of the business. It also includes the submarine network and B2B services.

• Liberty Puerto Rico (LPR): the problem child. Debt of around $2.9 billion, net leverage close to 8x, and covenant leverage as high as 14x. Furthermore, revenue here is declining due to difficulties with the mobile network migration. This is the epicenter of refinancing risk.

• Liberty Costa Rica: a small but sound unit with debt of around $500 million and leverage of just 1.8x.

• VTR (Chile): Faces strong competition in the fixed-line market from players like Claro and Entel, but is benefiting from 5G and the growing B2B segment.

Historically, Puerto Rico was the core source of cash, but today it is also the greatest source of uncertainty. The Caribbean is undergoing recovery following Hurricane Melissa, which cost the segment approximately $27 million in adjusted OIBDA in the fourth quarter of 2025 and is estimated to have an impact on free cash flow of roughly $100 million in 2026. Management is therefore guiding investors to expect that the stronger half of 2026 will come in the second half of the year, with Jamaica’s full recovery.

Market, Competition, and Strategy

The company’s market is telecommunications in the Caribbean and Latin America—an environment characterized by slow growth, intense competition, and currency risk. Consolidated revenue is largely stagnating or declining slightly, as the decline in older products (landlines, roaming, video) is offsetting growth in broadband and mobile data. Structurally, the company is benefiting from growing demand for data and the convergence of fixed and mobile services; conversely, it is being hurt by the region’s macroeconomic weakness, inflation, and currency depreciation, particularly that of the Chilean peso.

Management’s strategy has three pillars. Operationally, the focus is on reducing costs and increasing margins, which is proving successful: adjusted OIBDA, net of capital expenditures, rose by 27% in 2025, and adjusted free cash flow by 29%, despite lower capital expenditures. On the financial front, the focus is on active debt management—specifically, refinancing and extending maturities (over 75% of the group’s debt is not due until 2031 or later). Strategically, there is talk of potentially monetizing assets—such as towers—following the example of European operators, which would allow the company to reduce debt without diluting shareholders. In addition, there are specific projects to expand capacity (Manta, Salvador) and a cloud partnership with AWS.

Financial Performance: Be Careful What You’re Tracking

Caution is needed here, as the company’s financial statements and its cash reality are at odds with each other. On paper, the company is a chronic loser. The net loss amounted to $611 million in 2025 and $657 million in 2024, while operating profit fluctuates wildly depending on one-time write-offs and exchange rate effects (in 2024, there was even an operating loss, while in 2025, operating profit was $720 million). Accounting profit is a virtually useless metric here.

The reason is simple: at the net income level, the company is dragged below zero mainly by interest expenses on its 8-billion debt, which historically have ranged from $500 million to $620 million annually, plus heavy depreciation charges on its capital-intensive telecommunications network. However, these are largely non-cash (depreciation) or structural (interest) items that say nothing about the operating performance of the business itself. On the contrary, the business is stable: a gross margin of around 68%, adjusted OIBDA in the range of 1.6 billion USD annually, and growing free cash flow. For this type of company, it is relevant to monitor OIBDA, free cash flow, and leverage trends, rather than the net income line.

Balance Sheet and Debt: The Real Heart of the Story

While the balance sheet is a side note for other companies, here it is the main story. Consolidated gross debt stood at approximately $8.4 billion at the end of 2025, compared with cash of around $800 million and an additional $900 million available in credit lines. Consolidated net leverage improved to roughly 4.3 to 4.5 times adjusted OIBDA, which is high for a telecom company but not unmanageable. The average interest rate after hedging was around 6.9%.

However, it is crucial not to look solely at the consolidated average, as it masks huge differences between the units. The healthy Costa Rica and the relatively solid C&W stand alongside Puerto Rico, whose covenant leverage of 14x is in the danger zone. Although the debt is largely long-term (75% maturing in 2031 and later) and the company is actively refinancing (Puerto Rico recently secured a new term loan of $200 million and a revolving line of credit of $140 million maturing by 2030), there is simply no certainty that it will be able to refinance on acceptable terms.

Other balance sheet metrics reflect this. Working capital is deeply negative (around -$1.8 billion), the current ratio is only 0.09, and the Altman Z-score is negative. Some of this is normal for telecom companies (negative working capital due to prepaid revenue and deferred payments), but the overall picture is clear: any risk associated with this stock is primarily financial, not operational. There is no risk that the company will stop generating cash. The risk is that debt service will swallow up that cash before it reaches shareholders.

Valuation: Cheap for a Good Reason, or Overlooked?

Based on standard multiples, the stock looks incredibly cheap. The price-to-free-cash-flow ratio is just 1.77, the price-to-revenue ratio is 0.27, and the price-to-book ratio is 1.73. There is no P/E ratio because the company is operating at a loss. A price-to-free-cash-flow ratio below 2 is a level the market typically assigns to companies it considers to be at real risk of bankruptcy, or to value plays that are deeply off the radar.

Here is the key to understanding the entire valuation: low equity multiples are a logical consequence of high leverage, not automatically a sign of undervaluation. When 90% of a company’s value is in debt, shareholders are left with a slim and highly volatile remainder. Essentially, the market is saying that it values the company as a whole (debt plus equity) at a reasonable multiple of roughly 5 to 6 times OIBDA, but because debt accounts for the overwhelming majority, only that small, inexpensive fraction is attributed to the equity portion. For the stock to prove “cheap,” the company must continue to reduce its leverage—whether through OIBDA growth, asset sales, or debt repayment—thereby shifting value from creditors to shareholders. For the stock to turn out to be a “trap,” all it takes is stagnant OIBDA amid refinancing pressure, and the stock’s value could plummet. Analysts’ price targets, in fact, fully illustrate this uncertainty: they range from roughly $5 among cautious banks to double that amount among optimists.

A Curious Dividend: Half a Billion in Preferred Shares

This is the heart of the whole story and the reason why the company is now attracting attention. In May 2026, it announced a special dividend, albeit in an unusual form. For every 10 common shares held, a shareholder received one newly issued 9% preferred share with an initial liquidation value of $25. This corresponds to a liquidation value of roughly $2.50 per common share and a total of approximately $500 million. The record date was June 1, 2026; the distribution took place on June 16; and as of June 17, the preferred shares have been trading separately under the ticker symbol LILPV.

What’s curious about this? Several things at once. First, the company isn’t paying out cash but is issuing a new security, so it isn’t enriching shareholders with money but rather changing the structure of their holdings. Second, and this is crucial, a company burdened with 8 billion in debt is voluntarily adding another fixed obligation through this move. A 9% coupon on $500 million represents a new cash outflow of around $45 million per year, which will also take priority over common shareholders. Third, the preferred shares are cumulative and perpetual, so the obligation is permanent and unpaid coupons accumulate.

How should this be understood? An optimistic interpretation suggests that management is so confident in future free cash flow that it can afford to add a fixed payment while simultaneously delivering value to shareholders. A 9% yield is attractive to income-seeking investors, and the separately tradable security gives them flexibility (they can sell it or hold it for income). The skeptical interpretation is that this is financial engineering that adds costs to a company that should rather be reducing its debt, and that indirectly benefits large shareholders affiliated with Malone. The truth likely lies somewhere in between: it is both a signal of confidence and an increase in balance sheet risk. For investors, it is crucial to understand that, following the distribution, the common stock is “lighter” by this amount (the price decreased by the value of the preferred stock around the ex-date) and that a new claim now exists that takes precedence over it.

What Can Move the Stock

Given the leveraged nature of the stock, the main catalysts are related to debt and cash, not revenue growth:

• Deleveraging. Any reduction in net debt—whether through OIBDA growth or debt repayment—shifts value from creditors to shareholders. For such a highly indebted company, this is the strongest driver of the stock price.

• Asset monetization. Selling towers or other infrastructure, following the European model, would allow for deleveraging without dilution. Management has repeatedly mentioned this as a possibility.

• Resolving the Puerto Rico issue. Sorting out refinancing and stabilizing revenue in the most indebted unit would remove the single biggest risk from the story.

• Hurricane recovery. A stronger second half of 2026 and a full recovery in Jamaica would confirm that cash flow is back on track.

What to watch: trends in consolidated and segment-level leverage (especially in Puerto Rico), adjusted free cash flow, progress on refinancing maturities, and any announcements regarding asset sales. Another positive signal is that value and distressed investors are opening positions rather than closing them.

Risks

• Debt refinancing. The single greatest risk. The Puerto Rican group faces the risk of being unable to meet maturities on acceptable terms, which could lead to restructuring at the subsidiary level. Signal: deteriorating covenant leverage and developments in its debt market.

• Leverage works both ways. A thin layer of equity atop massive debt means that even a slight deterioration in operations can cause the share price to plummet. Signal: a decline in OIBDA amid rising interest rates.

• Currency risk. Revenues in local currencies (particularly the Chilean peso) against debt largely denominated in dollars. Currency depreciation undermines the ability to service debt. Signal: exchange rate movements in the region.

• Structural revenue decline. A mature market, declining sales of older products, and fierce competition are keeping growth at zero. Signal: continued revenue decline in key units.

• New fixed-cost burden from preferred shares. A 9% coupon adds an annual cash outflow of approximately $45 million and carries seniority over common stock. Signal: any delay in coupon payments.

• Natural disasters. The Caribbean is vulnerable to hurricanes, as Hurricane Melissa demonstrated. Signal: seasonal damage and the pace of insurance claims.

Investment Scenarios

Pessimistic (12 to 24 months). Puerto Rico fails to refinance on reasonable terms and heads toward debt restructuring at the subsidiary level; OIBDA stagnates under pressure from competition and currency fluctuations; and the new preferred stock liability erodes cash. Due to leverage, the value of the common stock plummets to near zero or to a fraction of a percent of today’s price. This is a realistic scenario that every investor must take seriously, precisely because of the balance sheet.

Realistic (12 to 24 months). The company smoothly refinances its maturities; leverage remains around 4 to 4.5x and is gradually declining slightly thanks to growing cash flow; the Caribbean is recovering; and Puerto Rico is stabilizing, albeit slowly. The stock remains cheap relative to cash flow and trades in a range of roughly today’s price to double that price, depending on risk appetite and the pace of debt reduction. The yield on preferred stock adds to the income component.

Optimistic (18 to 36 months). The company will sell part of its infrastructure (towers) and significantly reduce debt; OIBDA will grow thanks to cost savings and B2B business; the Puerto Rico issue will be resolved; and the market will stop pricing the stock as distressed. The shift in value from creditors to shareholders will be amplified by leverage, and the stock could trade significantly above today’s price, closer to the targets set by optimistic analysts. Meanwhile, preferred shares will continue to pay their 9% dividend.

Key Takeaways from the Article

• This is the indebted telecommunications holding company $LILA, operating in the Caribbean and Latin America, whose stock represents a thin and highly volatile strip of equity above debt exceeding $8 billion.

• Its appeal is purely value-based and speculative: a price-to-free-cash-flow ratio below 2 and a price-to-revenue ratio of 0.27 mean that, if the company successfully deleverages, the value attributable to shareholders could increase significantly. Reported earnings are irrelevant here; OIBDA and free cash flow are what matter.

• The main weakness is debt, particularly the dangerously indebted subsidiary Liberty Puerto Rico, as well as the region’s currency risk and structural revenue stagnation. Leverage, however, works both ways.

• The special dividend is a double-edged sword: shareholders received approximately $2.50 per share in 9% preferred stock (totaling around $500 million), which signals confidence in cash flow but also represents a new fixed obligation.

• Value and distressed investors are circling the company (John Malone with about 7%, GCI Liberty, Oaktree), which is typical of Malone-style telecommunications plays built on leverage and financial engineering.

• Within the portfolio, this is a speculative, high-risk “special situation” position, not a defensive or income-generating holding. It belongs only in that part of the portfolio where the investor can tolerate the possibility of total loss in exchange for the potential for a multiple-fold return, should the debt restructuring succeed.

The pharmaceutical industry has one unpleasant characteristic: even the best drug will eventually lose its patent protection, and sales will collapse like a house of cards. AbbVie $ABBV went through this with Humira, once the world’s best-selling drug. And instead of recovering from that, it’s spending billions right away to ensure a similar scenario doesn’t happen again.

This time, it’s digging deep into its pockets. AbbVie announced the acquisition of biotech firm Apogee Therapeutics $APGE for $10.9 billion in cash. That’s roughly 240 billion crowns and also AbbVie’s largest acquisition since it swallowed up Botox maker Allergan for 63 billion in 2019.

A drug that only needs to be injected twice a year

The entire deal hinges on a single molecule. It’s called zumilokibart, an experimental antibody for atopic dermatitis—the most common form of eczema—and, in the future, for asthma as well.

What makes it interesting is not so much its effectiveness as its convenience. The current market leader, the drug Dupixent from Sanofi and Regeneron, is administered via injection every two weeks. Zumilokibart aims to allow patients to get by with just one injection every three to six months. In the second phase of trials, roughly two-thirds of patients achieved significant skin clearance after 16 weeks, with a dosage that can be counted on the fingers of one hand per year.

The market at stake is surprisingly underserved. Jeffrey Stewart, head of AbbVie’s commercial division, noted during a conference call that only about 8 percent of patients with atopic dermatitis are currently receiving treatment—the lowest rate among all immunological categories, yet the fastest-growing one. In other words: there is a clear market leader, but the overwhelming majority of patients still remain untreated.

"We believe the deal makes sense and that AbbVie is the ideal buyer to maximize zumilokibart’s potential."

Brian Abrahams, analyst at RBC Capital Markets

The Shadow of Humira, Which Will Never Completely Disappear

To understand why AbbVie is willing to pay nearly a 50 percent premium over Apogee’s market price, one must look at the company’s recent history.

For years, Humira was literally a cash cow. But when its patent protection expired and cheaper biosimilars flooded the market, sales began to plummet. According to available data, in the first quarter of this year alone, they fell by nearly 39 percent to $688 million. For a company that had built a large part of its business on a single drug, this was a test of survival.

And AbbVie passed it. Today, two successors are driving sales in place of Humira:

• Skyrizi —an antibody for psoriasis and inflammatory bowel disease, with sales of around $17.6 billion last year

• Rinvoq —a so-called JAK inhibitor for inflammatory diseases, with over $8 billion in sales last year

Together, this duo generated over $30 billion, and AbbVie’s entire immunology division grew by 14 percent last year. The company not only weathered Humira’s decline but emerged stronger than before. In this light, the acquisition of Apogee is a logical next step—AbbVie doesn’t want to wait until Skyrizi and Rinvoq eventually run out of steam.

A Bet on 2030 with an Open End

But this brings us to the most interesting—and at the same time, the riskiest—part. Zumilokibart is not yet a finished drug. The third, decisive phase of clinical trials is not scheduled to begin until the second half of this year, with results expected around 2028 and approval not anticipated until early 2030 at the earliest. AbbVie itself admits that the drug won’t start contributing to earnings per share until 2032.

In other words: the company is paying $11 billion in cash for a promise that won’t be confirmed for at least four years—and may not be confirmed at all.

Furthermore, some analysts point out a particular risk. The competing drug Dupixent attacks inflammation from two angles, while zumilokibart relies on only one. AbbVie’s management claims this isn’t a problem and that the efficacy data are strong. But if it ultimately turns out that the new drug is easier to administer but no more effective, it could end up as a niche option rather than the new market leader for which AbbVie is paying billions.

"Buying Apogee now makes sense. The drug has the potential to achieve mega-blockbuster status with peak annual sales exceeding $10 billion."

Robert Michael, CEO of AbbVie

The market has so far reacted mildly positively — AbbVie’s stock jumped by about 4 to 5 percent following the announcement and is trading around $234 today, placing the company among the pharmaceutical giants with a market capitalization of over $400 billion (about 8.8 trillion crowns). Apogee surged by about 50 percent to the offer price.

What the Numbers and Analysts Say

For investors, AbbVie is primarily attractive as a dividend stock. The company pays a quarterly dividend of $1.73 per share, which corresponds to a yield of around 3 percent, and is among the so-called “dividend aristocrats”—companies with a long history of regularly increasing their payouts. Meanwhile, the stock is trading near its annual high—over the past 52 weeks, it has ranged from roughly $182 to $245.

Key figures worth noting:

• Market capitalization: over $400 billion

• Dividend yield: around 3% (quarterly dividend of $1.73)

• Analysts’ average 12-monthprice target: roughly $254 (range from $184 to $328)

Zumilokibart itself, however, could be more than just a safety net. Following the release of Phase 2 data, analysts at Guggenheim Securities doubled their peak revenue estimate for the drug to $5.2 billion annually. While that’s a fraction of what Skyrizi and Rinvoq generate, it’s still a significant figure for a single drug.

"The market has long been searching for late-stage growth assets. This acquisition should therefore be well received by investors."

Chris Schott, J.P. Morgan analyst

The next test, however, won’t come from Apogee. It will be AbbVie’s second-quarter earnings report, which the company will release on July 24. Investors will be watching Skyrizi’s sales closely against the backdrop of a raised full-year outlook of around $21.6 billion. It is precisely this pair of Humira successors—not a drug that won’t go on sale for another four years—that is keeping the stock afloat right now.

Interestingly, this acquisition fits into a much broader trend. Major pharmaceutical companies across the industry are now aggressively buying up biotech firms with late-stage drugs, as patents for many of their key products are set to expire soon. Rather than patiently waiting for their own research—which takes years—they’re simply buying the next big hit. AbbVie is simply confirming that it is among the most active players in this game—following the acquisitions of ImmunoGen and Cerevel in 2023, Apogee is another piece of the puzzle. The question remains whether the $11 billion bet will pay off, or whether AbbVie has paid for an expensive option that will never turn to gold.

The End of an Era: Telecom Leaves the Dow Jones

When Verizon joined the Dow Jones Industrial Average in April 2004, it symbolized the expansion of the mobile era. Wireless networks were growing, smartphones were just emerging, and telecommunications looked like the sector of the future. Two and a half decades later, another symbol is emerging—this time pointing in the opposite direction.

S&P Dow Jones Indices announced on June 23, 2026, that Alphabet $GOOG will replace Verizon $VZ in the Dow Jones Industrial Average $^DJI. The change will take effect on June 29 before the market opens. After 22 years, the telecommunications giant is thus leaving the 30-member club of the world’s most closely watched stock index.

Since joining the index in 2004, Verizon’s stock has risen by only about 40% (excluding dividends)—while Alphabet has gained over 13,000 percent over the same period since its IPO in August of that year.

Why now—and why Alphabet?

The Dow Jones is a price-weighted index, which means that each stock’s influence is based on its absolute price, not its market capitalization. Verizon was trading around $47 per share and represented just 0.5 percent of the index’s value —a virtually negligible share. It had only a minimal impact on the Dow Jones’s day-to-day movements.

Alphabet, on the other hand, is currently trading around $346 per share and, upon inclusion, will become the sixth most influential component of the entire index. In explaining the change, S&P Dow Jones Indices stated that Alphabet better represents the communications services sector and will bring exposure to artificial intelligence, cloud infrastructure, and digital advertising to the index.

With the addition of Alphabet, the Dow Jones will, for the first time in history, become an index that includes all five of the largest tech giants simultaneously—alongside Alphabet, these are Nvidia $NVDA, Amazon $AMZN, Apple $AAPL, and Microsoft $MSFT. According to CNBC, Alphabet’s stock rose by approximately one percent in after-hours trading following the announcement.

"The importance of traditional telecoms has clearly diminished. Alphabet’s inclusion reflects the market’s recognition of its long-term growth potential and industry leadership."

Ayako Yoshioka, Senior Investment Strategist at Wealth Enhancement

What They Had to Do With the Index Price

Swapping two stocks with such different prices requires a mathematical adjustment. The Dow Jones is calculated as the sum of the prices of all thirty components divided by a special number—the so-called divisor. If we simply replaced Verizon with Alphabet without adjusting this number, the index would artificially rise without anything in the market actually changing in value.

S&P Dow Jones Indices will therefore adjust the composition before the market opens on June 29. The new value will be published on June 26 at the end of the trading day so that index funds can incorporate it into their portfolios in time.

And it is precisely these passive funds that are in for an interesting move: ETFs such as the SPDR Dow Jones Industrial Average ETF ($DIA) must buy Alphabet shares and sell Verizon shares. Given the size of these funds, the volume involved could visibly impact both stocks around June 29.

Verizon: 22 Years in the Index, but No Glory

Verizon joined the Dow in April 2004 as the successor to AT&T $T —at the time still with a reputation as a dynamic operator that had replaced the old guard. But the story didn’t unfold as planned.

Over the entire period it has been in the index, its stock (excluding reinvested dividends) has risen by only about 39.5 percent. By comparison, the broader market has appreciated several times more over the same period. Verizon was a dividend stock for conservative investors—with a yield of over 6 percent, but with negligible room for capital growth, as summarized in an analysis by The Motley Fool.

Paradoxically, in 2026 —the year Verizon completed its $20 billion acquisition of Frontier Communications and significantly expanded its fiber-optic network—it was removed from the index. The market remembers a decade of weak performance, not a recent strategic move.

“Verizon accounted for only half a percent of the index’s value. A component like that simply does not serve the purpose of a benchmark, which is to reflect the U.S. economy.”

Analytical commentary from S&P Dow Jones Indices on the reasons for the change

Alphabet Enters the Index: Sixth-Largest Weighting

Alphabet’s stock is up approximately 10 percent in 2026 compared to the start of the year— despite a turbulent period marked by regulatory pressures and legal disputes in both the U.S. and Europe. This marks the fourth consecutive year the company could end on a positive note.

In terms of business structure, Alphabet brings a level of diversification to the Dow that none of the existing technology members has offered. Google Search dominates approximately 90 percent of the global internet search market, YouTube is the second-most-visited platform in the world, the Google Cloud division is growing at double-digit rates, and the entire group is investing heavily in artificial intelligence through its Gemini series of models.

The inclusion in the Dow Jones is, to some extent, symbolic—Alphabet has long been part of the S&P 500 $^GSPC and the Nasdaq 100. But the Dow Jones has its own appeal. It is an index followed by millions of retail investors and cited in major news broadcasts as a barometer of the U.S. economy.

It wasn’t until 2022, when Alphabet carried out a 20-for-1 stock split and reduced the price from approximately $2,200 to about $110, that it became technically eligible for inclusion in the price-weighted Dow Jones. Before that, a single Alphabet share would have dominated the entire index and distorted its movements.

What This Says About Today’s Market

The change in the Dow Jones’ composition is not merely an administrative matter.It is a signal of where the center of gravity of the U.S. economy has shifted. As recently as 2004, a telecommunications operator was the natural representative of American business—infrastructure, millions of customers, stable cash flow.

In 2026, it is being replaced by a company whose most valuable assets are intangible: algorithms, data, artificial intelligence, and a global advertising platform. The Dow Jones is 130 years old and has undergone more than fifty changes during that time. Each of these changes was a snapshot of what was driving the American economy at that moment.

Situations like this show how sensitive defense companies can be to individual contracts. On the one hand, they’ve benefited from a strong long-term trend of rising military budgets; on the other hand, they can react sharply in the short term to any major order or change in government plans.

I’d be interested to hear how you view this situation.

Do you see the current drop in Rheinmetall as an interesting buying opportunity, or do you think the stock was already priced too high before this news? And do you follow any European defence companies at all, such as Rheinmetall, TKMS, Leonardo, or Saab?

Gross margin measures how much a company retains from every dollar of revenue after deducting the direct costs of producing goods or providing services. It is therefore one of the first—and best—indicators of profitability. While net income is affected by taxes, interest, one-time items, and investments in development, gross margin reflects the very essence of the business—namely, whether the company is selling something the market is willing to pay for.

Investor Warren Buffett has long pointed out that companies with consistently high gross margins tend to have what is known as an “economic moat.” The ability to charge prices significantly above costs usually stems from a strong brand, patent protection, network effects, or a dominant market position. Conversely, if a company’s margin declines over the long term, this is often the first warning sign that competitors are beginning to erode its advantage—even before this is reflected in revenue itself.

Visa $V

Visa is a textbook example of a company with an extremely high gross margin. The company does not operate banks or issue cards to end customers. Instead, it operates a global payment network through which transactions flow between merchants, banks, and cardholders in more than 200 countries. Once this network is established, the cost of processing each additional transaction is minimal, while fees from payment volume continue to grow. It is precisely this so-called network effect that explains why Visa’s gross margin has remained very high over the long term.

For the quarter ending in March 2026 , the company’s gross margin reached approximately 81%, a figure found among only a handful of companies in the real economy. Meanwhile, total net revenue in the second fiscal quarter of 2026 rose 17% year-over-year to $11.2 billion, driven by growth in payment volume, cross-border transactions, and the number of transactions processed. In addition, the company announced a new $20 billion share buyback program, demonstrating the strong cash flow its business is capable of generating.

Price and Valuation

Visa shares are currently trading around $327, which corresponds to a market capitalization of approximately $600 billion and a price-to-earnings (P/E) ratio of around 29. The dividend yield is rather symbolic, at around 0.8%, as Visa returns capital to shareholders primarily through share buybacks. Importantly, the high margin here is not a one-time phenomenon but a structural feature of the business. The main risk remains regulation of interbank fees and competition from alternative payment methods, rather than pressure on production costs themselves, which are minimal in the network model.

Coca-Cola $KO

The Coca-Cola Company, ticker $KO, is a global company that manufactures and sells concentrates and syrups and licenses the brand. In contrast, companies such as Coca-Cola Consolidated and Coca-Cola FEMSA are bottlers that physically produce, package, and distribute the beverages. While $KO’s gross margin is over 60%, that of bottlers is typically around 40% because they bear the costs of capital-intensive production and logistics.

This so-called “asset-light” model—that is, a model with low capital intensity—is at the core of $KO’s profitability. The company outsources the costly aspects of production and distribution to its partners and retains the most valuable part of the value chain for itself: the brand, the recipe, and marketing. The company’s gross margin in 2026 ranges between approximately 61% and 63%, reaching roughly 63% in the first quarter of 2026, according to the report—the highest level in the past eight quarters.

Cost Pressures and a Defensive Strategy

Even such a high margin, however, is not immune to cost pressures. The company itself acknowledged that rising aluminum prices, caused by geopolitical conflicts and import tariffs, have put pressure on margins in recent quarters. Coca-Cola is responding with pricing measures that have helped mitigate the impact. The stock trades around $80 with a market capitalization of approximately $345 billion, a P/E ratio of around 25, and a dividend yield of roughly 2.65%. Coca-Cola is also among the so-called “dividend aristocrats,” having increased its dividend for more than 60 consecutive years. For investors seeking a combination of stable margins and defensive characteristics, it is one of the most typical choices on the market.

Cisco $CSCO

Cisco is the world’s largest manufacturer of networking equipment, but its high gross margin now stems increasingly from software, subscriptions, and services, rather than solely from the sale of physical switches and routers. This shift in business model is crucial because software and service revenues typically have significantly higher margins than hardware. In recent quarters, the company’s overall gross margin on a GAAP basis has hovered around 64 to 65%, with the service segment reaching over 68%.

For fiscal year 2026, Cisco is targeting revenue of around $63 billion, and the company has repeatedly raised its outlook, thanks in part to strong demand for network infrastructure for artificial intelligence. Management refers to a so-called “network supercycle,” in which hyperscalers are ordering equipment to interconnect AI clusters. In just one quarter, orders for AI infrastructure from hyperscalers reached $2.1 billion. This helps maintain high margins, as more complex products with higher added value command better markups.

Valuation Following Sharp Growth

Cisco shares are trading around $120, near their all-time high, with a market capitalization of approximately $477 billion. However, following strong price growth, the P/E ratio has climbed to roughly 40, which is above average for a traditionally defensive hardware company. The dividend yield is approximately 1.4%. Cisco’s high margin is both realistic and structural, but the question remains to what extent the current enthusiasm surrounding AI infrastructure is already priced into the stock.

Merck $MRK

Merck is a pharmaceutical giant whose high gross margin stems from the patent protection of its drugs and vaccines. Once the company develops and approves a successful drug, the actual production of each additional dose is relatively inexpensive, while the selling price remains high due to exclusivity and clinical value. Merck’s flagship product is the cancer drug Keytruda, which generated over $30 billion in revenue in 2025 and ranks among the world’s best-selling drugs.

Merck’s gross margin on a GAAP basis for the full year 2025 hovered around 75%, and on a non-GAAP basis, it was as high as 80%. In the individual quarters of 2026, the margin fluctuated slightly due to write-offs, restructuring costs, and inventory revaluations following acquisitions, but it remains structurally very high. However, Keytruda is nearing the end of its primary patent protection, and Merck must fill the gap left by its sales with new drugs from its development pipeline.

Dividend and Valuation

Merck shares are trading around $114, with a market capitalization of approximately $281 billion and a P/E ratio of around 32—though this figure is distorted by one-time accounting items in recent quarters. The dividend yield is approximately 2.9%, and the company has increased its dividend for 16 consecutive years. For investors, this is a case where high margins reflect the true value of intellectual property; at the same time, however, it is important to monitor how the company manages to offset the impending decline in revenue from Keytruda.

American Express $AXP

We intentionally included American Express in this overview because it clearly illustrates the limitations of purely mechanical comparisons of gross margins across industries. Some sources report a gross margin of around 61 to 63% for American Express, which at first glance would place it among the companies on this list. The problem is that for a bank and card issuer, the term “gross margin” does not have the same meaning as it does for a manufacturing or software company.

American Express does not manufacture a physical product with a clear cost of goods sold. Its business is based on interest income, merchant fees, and membership fees, offset by interest expenses, rewards for cardholders, and, most importantly, provisions for credit losses.

What various websites refer to as “gross margin” is therefore more of an artificial construct that they calculate differently. It is no wonder, then, that their figures vary. One source cites 62.21%, another 62.76%, and yet another 61.03%. For financial institutions, it therefore makes much more sense to track metrics such as net interest margin, return on equity (ROE), or operating margin.

What the Numbers Say About the Company

If we look at more meaningful metrics, American Express appears very strong. Return on equity hovers around 34%, net margin around 14%, and operating margin around 20%.

In the first quarter of 2026, the company reported an 11% increase in revenue to $18.9 billion, and earnings per share rose by 18%. The stock is trading around $341 with a market capitalization of approximately $231 billion, a P/E ratio of around 21, and a dividend yield of roughly 1.1%. American Express is therefore a high-quality company, but as an example of a high gross margin, it serves more as a warning that the same metric can mean something entirely different across various industries.

Comparison

* The gross margin for American Express is methodologically problematic for a financial institution, and individual data sources differ in their calculations. The figure is provided for illustrative purposes only and is not directly comparable to other companies.

Strategic Perspective

A high gross margin alone is no guarantee of a good investment, but it is an excellent initial filter. It indicates that the company has something for which the market is willing to pay a significant premium, whether it be a global payment network (Visa), a brand and an asset-light model (Coca-Cola), a shift toward software and services (Cisco), or patent-protected drugs (Merck). What these companies have in common is that their margins stem from a structural advantage, not from a one-time price fluctuation.

At the same time, however, each of these companies faces its own specific risks. For Visa, it’s fee regulation; for Coca-Cola, it’s input cost pressures and shifting consumer preferences; for Cisco, it’s stretched valuations following a price rally and the cyclical nature of demand; and for Merck, it’s the looming patent cliff for Keytruda. A high margin therefore indicates that the business is strong, but it says nothing about whether the stock is currently cheap or expensive. However, the Fair Price Index on Bulios can help you with that .

A gross margin above 60% is a rare trait found in the real economy only among companies with a truly strong competitive position. For all four companies, this is a meaningful and easily comparable metric that reflects the structural advantage of their business models.

We included American Express as a counterexample to show that the same metric does not necessarily mean the same thing across different industries. For financial institutions, gross margin is methodologically questionable, and it makes more sense to track other metrics.

There are approximately 1,400 companies listed on U.S. stock exchanges that regularly pay dividends. Fewer than seventy of them have increased their dividends every year for at least 25 years—and they bear the honorary title of “Dividend Aristocrat.” Only forty companies worldwide have managed to do so for fifty consecutive years.

McDonald’s $MCD has just joined this exclusive club known as the Dividend King.

The numbers speak for themselves

The quarterly dividend currently stands at $1.86 per share, or $7.44 per year. At today’s share price of around $272, this corresponds to a yield of approximately 2.7%.

Anyone who held McDonald’s stock ten years ago was receiving a significantly lower dividend. According to Digrin data, the average annual dividend growth over the past three years has been approximately 8.8%. In other words, the dividend doubles every eight years. For investors who hold the stock for the long term, the effective return on the original purchase price thus increases year after year.

Key figures today:

• Annual dividend: $7.44 per share

• Yield: 2.7% (at a price of $272)

• Number of consecutive annual increases: 50

• Average dividend growth over 3 years: 8.8%

• Market capitalization: $193 billion

Why the Golden Arches Won’t Run Dry

McDonald’s isn’t just a restaurant chain. It’s a real estate and franchise machine that generates stable cash flow regardless of how consumers are faring.

Over 95% of its more than 40 ,000 locations are operated by independent franchisees. McDonald’s owns or leases the land, collects fees and royalties—and shifts operational risk to the franchisees. The result is a business with operating margins hovering around 46%, a level that many tech companies would envy.

The full-year 2025 results confirm this model in practice: global revenue for the entire system rose to $139 billion, comparable sales for 2025 increased by 3.1% year-over-year, and net income reached $8.56 billion, up 4.1% year-over-year. It is precisely the stability of these figures that guarantees shareholders will receive a dividend in the next quarter as well.

"McDonald's is one of the few companies where you can be relatively certain that the dividend will continue to grow in five, ten, and fifteen years—and that is a rare quality for income investors."

Dividend Strategy Analyst, Sure Dividend

Stock at a Yearly Low—Opportunity or Trap?

This is where things get complicated. MCD stock is trading near its 52-week low in June 2026—on June 23, it ranged between $270 and $274, while the 52-week high was $341.75. That’s a decline of roughly 20% from the peak.

McDonald’s suffered from a decline in foot traffic in the first half of 2025 —in the first quarter of 2025, comparable sales in the U.S. fell by 3.6%, with customers in the lower and middle income brackets feeling the impact the most. This was compounded by adverse currency effects and restructuring costs.

Since then, however, the situation has turned around. In the first quarter of 2026, global comparable sales rose by 3.8%, with a 3.9% increase in the U.S. and a 3.9% increase in international markets. The company also reaffirmed its full-year financial targets.

The average consensus among 34 analysts is a price target of $331 —which would represent more than 20% upside potential from current levels. Of course, no price target is a guarantee.

"Consumers today face uncertainty, but they can always count on McDonald’s—for consistent value that helps them stretch their family budget."

Chris Kempczinski, CEO of McDonald’s, Q1 2026 earnings call

McDonald’s > NEXT: A Bet on the Next Five Years

Just as its stock is losing ground, McDonald’s has unveiled a new strategic plan. At the June franchisee conference in Las Vegas, CEO Kempczinski unveiled an initiative called McDonald’s > NEXT — the first comprehensive global strategy since 2020, when the “Accelerating the Arches” plan was launched.

The plan is built on four pillars: a new restaurant design, better-tasting food and beverages, customer-driven innovation, and enhanced hospitality. Specific financial targets are set to be announced at the investor day in September 2026.

In terms of numbers, this currently translates to an ambition to reach 50,000 restaurants by the end of 2027 —approximately 10,000 more than today. China is the key market for expansion. In addition to physical expansion, McDonald’s is investing in order automation, currently testing the ARCHY system—designed for automated order processing—at five U.S. locations.

It is interesting to note the focus on chicken and beverages as the categories with the greatest potential. In the chicken segment, McDonald’s has long been losing customers to specialty chains—a fact acknowledged by Kempczinski himself, who noted in an internal memo that “traditional competitors are refining their menus, and a new wave of specialists is redefining quality in the areas of chicken, beef, and beverages.”

Half a century is a strong argument

McDonald’s is not a stock for speculators. It’s a story of compound interest applied to dividends —the sooner you start reinvesting your dividend payments back into the stock, the more significant the effect will be in twenty years.

Current valuation— a P/E ratio around 22, a yield over 2.7%, and the stock at its annual low—looks reasonable on paper. Risks include pressure on consumers amid high prices, growing competition, and higher expansion costs. Financial details of the new NEXT strategy are still lacking.

But the basic equation is simple: 50 years of continuous dividend growth doesn’t happen by accident. It results from a combination of a recession-resistant franchise model, management discipline, and tangible assets in the form of real estate scattered across the globe. That is value that won’t expire—no matter what people eat over the next fifty years.

I came across information that Japan has begun selling U.S. Treasuries in order to have funds to buy yen. I dug into it a bit because there are more layers to it than it appears at first glance.

Why is Japan intervening at all

This year alone the yen repeatedly tested the 160-per-dollar mark. That's the level Tokyo considers a red line. A weak yen by itself wouldn't be such a problem; the issue is its combination with expensive imported energy, because oil jumped above $113 per barrel after the Iran war. That made living costs in Japan rise sharply and the Ministry of Finance had to step into the market. By the end of May it had intervened a record roughly $73 billion, which was the largest operation of its kind.

Where they get the money

To finance those yen purchases Japan apparently sold U.S. Treasuries. A small note that media often skip: the BOJ acts only as the executor; the Ministry of Finance makes the decisions. The mechanics are exactly as described: sell bonds, obtain dollars, and buy yen with them. Hard data backs this up too — Japan's holdings of foreign securities fell by more than $75 billion at the end of May compared to April, which matches the scale of the intervention.

A quieter and possibly more important flow

More important than the one-off intervention, it seems to me, is a second, less visible movement. Japanese investors, insurers and pension funds sold nearly $30 billion net in U.S. Treasuries in the first quarter — the largest quarter in almost four years — and the pace is accelerating month by month.

The reason is elegant. The BOJ is limiting purchases of domestic Japanese bonds, so yields at home are rising and Japanese institutions can finally earn returns domestically without currency risk. Suddenly they don't have to go to the U.S. for yield.

Why shareholders should care

Japan is the largest foreign holder of U.S. debt, holding over $1.2 trillion. When it sells or simply buys less, bond prices fall and U.S. yields tend to rise. Estimates suggest pressure on the 10-year yield on the order of 20 to 50 basis points in the medium term. And higher long-term rates flow through to mortgages, corporate financing costs, and stock valuations.

A bigger risk, however, is the carry trade — those hundreds of billions to a trillion dollars borrowed cheaply in yen and poured into riskier assets around the world. If the yen suddenly strengthens sharply, whether due to intervention or rate hikes, investors have to rush to close these positions. That exact mechanism knocked markets down in August 2024.

I'm curious how you see it. Do you take the strengthening yen and Japanese sales as a real risk for equities, or as noise that nobody will care about in a few months?

When a company triples its revenue year-over-year, it’s usually cause for celebration. But with OpenAI, it’s more complicated. According to documents the company sent to its shareholders— first reported by The Information—the creator of ChatGPT generated $5.7 billion in revenue in the first quarter of 2026. At the same time , however , it burned through $3.7 billion in cash during the same period.

Both figures tripled year-over-year. And that symmetry is precisely where the problem lies.

Growth That Undermines Itself

Revenue of $5.7 billion in a single quarter is something almost every tech company in the world would envy. When we add the fact that they’re growing at three times the rate, we’re talking about one of the fastest-growing businesses in history. The catch is that the costs of achieving this growth are rising just as fast.

In the real world, that’s called a problem. Large companies usually gain what’s known as “operational leverage” over time—the more they sell, the cheaper it becomes to serve each additional customer. At OpenAI, this isn’t the case yet. What it sells is inference using state-of-the-art AI models, and that becomes more expensive—not cheaper—with every additional user.

The specific figures for the quarter are as follows:

• Revenue: $5.7 billion (roughly 125 billion Czech koruna)

• Cash burn: $3.7 billion

• Operating loss: $9.3 billion

• Net loss: $21.3 billion (largely due to a one-time accounting item of $12.4 billion)

If we break this down into simple math, as Barron’s analysts did, OpenAI loses roughly $1.22 for every dollar it earns. Revenue has tripled, but the loss ratio has remained the same. The company is growing and burning through cash at the same time—and it’s doing both on a record scale.

Why This Isn’t Like Building a Mobile Network

OpenAI’s defenders like to argue that this is a classic capital-intensive infrastructure project—expensive at the start, cheap afterward. But this parallel doesn’t hold up.

Take the construction of a mobile network. Building a cell tower costs a fortune, but once it’s up, it serves a million users for the same cost as a thousand. The cost per customer decreases as scale increases. It is precisely this economics that has made telecommunications and software such profitable businesses.

OpenAI’s GPU clusters don’t work that way. A cluster handling a million queries costs proportionally more than one that handles a thousand. Marginal cost does not decrease with scaling—it increases. And this is precisely what makes OpenAI a business that is structurally different from any trillion-dollar company that came before it.

According to data from the analytics firm Sacra, the cost of inference alone reached $8.4 billion in 2025 and is expected to climb to $14.1 billion this year. That’s a 68 percent increase.

“The implied multiple of 35 times future revenue is priced for a monopoly outcome that does not yet exist.”

Greg Jensen, co-head of the investment division at Bridgewater

A cushion that looks better than it actually is

One number sounds reassuring. At the end of the quarter, OpenAI had over $73 billion in cash and marketable securities, up from $40 billion at the end of December. At the current burn rate, that amounts to roughly a five-year reserve without the company having to raise a single additional dollar.

But that jump from $40 billion to $73 billion didn’t come from business operations. It is largely the result of a massive funding round closed in late March —a $122 billion round, the largest private tech funding round in history, backed by Amazon $AMZN, Nvidia $NVDA, and SoftBank $SFTBY. This round valued the company at $852 billion.

So the cushion is real. But it’s also freshly inflated from the outside, not earned. And that’s a pretty significant difference when monthly spending is counted in the billions.

A trillion-dollar IPO with an uncomfortable question lurking beneath the surface

These numbers aren’t coming at a random time. On June 8, OpenAI confirmed that it had confidentially filed for a U.S. initial public offering (IPO), which could take place as early as September and value the company at up to $1 trillion. That would make it the largest stock market debut in history. Goldman Sachs $GS and Morgan Stanley $MS are leading the IPO.

This is precisely where the quarterly numbers become ammunition for both sides. The bulls point to the revenue curve; the bears point to the chasm below it. Neither side is clearly right, and the S-1 prospectus, which the company will publish before the actual IPO, will be the first opportunity to see the audited financials.

There is no shortage of skeptics. HSBC $HSBC estimates that OpenAI may need over $207 billion in additional capital by 2030, even under an optimistic scenario. Furthermore, internal documents suggest that the company expects a loss of around $74 billion in 2028 alone, before turning a profit in 2030. Meanwhile, it has committed to computing capacity contracts worth up to $1.4 trillion over eight years.

"These are great companies. But a great company doesn’t automatically mean a great investment."

Jay Ritter, director of the IPO Initiative at the University of Florida

A comparison with the competition, however, is not flattering. While OpenAI is losing $1.22 for every dollar of revenue, rival Anthropic is expected to report a small operating profit for the second quarter, according to Barron’s, after doubling its revenue to an annualized $47 billion.

The whole bet can thus be summed up in a single sentence. Investors are supposed to buy a company for $1 trillion that is banking on inference costs falling before it runs out of money. A five-year cash reserve buys time. But it doesn’t buy certainty that the model’s economics will ultimately balance out.

The second layer of the story explains why the company is in the spotlight right now. It is a real estate fund focused on healthcare properties that pays a dividend of around 6% per year in monthly installments, which is unusually convenient for income investors. After rebounding from last year’s lows, the stock has returned to the upper end of its annual range, and the question is whether it still has room to grow. The thesis argues that they do, and it rests on three pillars: a stable, cash-backed dividend; a demographic tailwind in the form of an aging population; and the potential for a revaluation should interest rates fall.

Key Points of the Analysis

• Revenue for 2025 rose by 4.5% to $2.82 billion. Operating cash flow reached $1.25 billion, and the gross margin hovers around 58%. This is a stable, moderately growing business built on rental income.

• The reported net income for a real estate fund is misleading: net income for 2025 was only $71 million (EPS $0.10), which results in a seemingly alarming P/E ratio of 60. However, the relevant FFO ranges from approximately $1.72 to $1.75 per share, which corresponds to a multiple of around 11.

• The dividend is $1.22 per year and is paid monthly at $0.10167, with a yield of around 6%. The payout ratio is roughly 70% of FFO, which is safe.

• The balance sheet has an investment-grade rating (BBB+ and Baa1) and net debt of about 5.4 times EBITDA, which is a reasonable and conservative level for a real estate fund.

• In March 2026, the company spun off its senior living business (Janus Living) into a separate public company, thereby simplifying its portfolio and focusing on outpatient healthcare and laboratory properties.

• Same-store rent growth (cash NOI) reached 3.9%, and lease renewals in the outpatient care sector were concluded with a positive price increase of around 4.4%. Operationally, therefore, the company is growing.

• Upside potential relative to price: at a price of around $19, the stock is trading below fair value estimates (analyst targets range from roughly $20 to $29, and the fair value model indicates around $29). Room for a revaluation exists primarily through a decline in interest rates.

• Main risk: interest rates and refinancing. Real estate funds are sensitive to interest rates, and the company also has a $650 million bond issue maturing in mid-2026. Added to this is weakness in laboratory real estate.

Company Overview

We’re talking about Healthpeak Properties, Inc. $DOC, one of the largest U.S. real estate investment trusts (REITs) focused on healthcare real estate. The company is headquartered in Denver, employs around 400 people, and is led by Scott Brinker. In its current form, it was created following the merger of the original Healthpeak with Physicians Realty Trust in 2024, when it also adopted its current ticker symbol, DOC. It holds an investment-grade rating, making it one of the more creditworthy players in the sector.

The real estate investment trust (REIT) model is simple at its core. The company owns a portfolio of buildings, leases them to tenants in the healthcare and scientific sectors, and collects rent, most of which it then distributes to shareholders in the form of dividends. This is because real estate funds are required to pay out the majority of their taxable income, which is why they traditionally offer high dividends. Value for investors thus comes from two sources: regular rent converted into dividends and long-term appreciation of the properties. A distinctive feature of healthcare buildings is their defensive nature. Demand for medical care is not very cyclical and is structurally growing as the population ages, making these properties an exceptionally stable asset.

What the Portfolio Consists Of

Following the IPO of its senior living division, the company is focusing on two main pillars and one supplementary one:

• Outpatient Medical Properties: the core of the portfolio, strengthened by the merger with Physicians Realty. These are medical buildings and clinics, often located near hospitals, with long-term lease agreements and high occupancy rates. Here, the company is securing lease renewals with positive rent increases of around 4.4%, which is a healthy sign of pricing power.

• Laboratory and Life Science Properties (Lab / Life Science): buildings for biotechnology and pharmaceutical research. Following the pandemic-era boom, this segment is facing an oversupply and weaker demand, so it is currently a source of caution rather than growth. However, the company is investing in it opportunistically; for example, it purchased the Gateway campus in San Francisco for a fraction of its replacement cost.

• Senior Housing and CCRC: Historically a major part of the business, but now largely spun off through the Janus Living IPO. The company retains a stake, but the operational risk and capital intensity of this segment are being shifted off the main balance sheet.

The logic behind this arrangement is clear: focus on stable and defensive outpatient care as the core business, cautiously weather a weaker cycle in the laboratories, and divest the operationally demanding senior housing segment. The result should be a simpler and more predictable real estate fund.

The Market and Demographic Trends

The main long-term theme for the entire sector is demographics. The population in developed countries is aging; the number of people over 65 is growing, and with it, the need for healthcare, outpatient examinations, and related services. This creates structural demand for healthcare real estate that is only minimally dependent on the economic cycle. Unlike office or retail buildings, which are threatened by remote work and e-commerce, medical facilities are relatively immune to these trends. People still visit their doctors in person.

This positive factor is offset by two structural weaknesses. First, real estate funds, as an asset class, are extremely sensitive to interest rates. When rates rise, property values fall, debt refinancing becomes more expensive, and investors prefer safe bonds over dividend-paying stocks, causing real estate fund prices to drop. This is precisely why the past few years have been painful for the sector. Second, the laboratory real estate segment experienced an oversupply following the pandemic, so not all parts of the healthcare market are growing at the same pace.

Competition: Inexpensive and Generous Among the Expensive

The appeal of this stock is best highlighted when compared to its competitors, as healthcare real estate funds are currently divided into two camps. On one side are the expensive growth stars; on the other, the cheaper income plays—and this company clearly belongs to the latter group. That’s not necessarily a disadvantage; it’s just a different yield profile.

The benchmark for the first camp is Welltower $WELL, the world’s largest healthcare real estate fund with a market capitalization of around $150 billion, focused on operating senior living facilities. Welltower literally exploded in 2025. Its same-store rents in senior housing grew by over 21%, and total returns to shareholders exceeded 49%, ranking it among the best real estate funds of the year. However, the market values it at roughly 33 times FFO, with a dividend yield of just 1.5%. Welltower is therefore a bet on growth and capital appreciation, not on income. The second-largest player, Ventas ( $VTR), has a similar growth-oriented profile and is only slightly cheaper; it also focuses primarily on senior housing.

In contrast, $DOC trades at around 11 times FFO with a yield of about 6%, which is roughly one-third of Welltower’s multiple and four times its dividend yield. The reason for this difference is legitimate. The company has bet on outpatient healthcare and laboratory real estate, which are growing more slowly (same-store growth of around 3.9%) than the booming senior housing sector, and has also partially divested itself of the operationally demanding senior housing business through the Janus Living IPO. The market therefore rightly premiums growth, and where growth is lacking, it assigns a lower multiple and compensates with a higher dividend.

And that is precisely why the company is attractive. Investors choose their profile. Those who want growth and are willing to pay 33 times earnings and settle for a 1.5% yield will buy Welltower. On the other hand, those who want high, immediately paid-out income backed by tangible assets at a reasonable multiple—plus the potential for a revaluation if interest rates fall—will opt for this stock. Furthermore, compared to smaller and more indebted competitors like American Healthcare REIT, this company stands out with its investment-grade rating, a more conservative payout ratio, and more predictable rental income. It isn’t the cheapest or highest-yielding stock in the sector, but it is one of the few that combines a high yield with an investment-grade balance sheet.

Financial Performance: Why Accounting Profit Is Misleading

When it comes to real estate investment trusts, you need to set aside the standard income statement. At first glance, the company’s financial figures appear contradictory. Revenue is growing, but net income for 2025 fell to just $71 million, and individual items on the income statement fluctuate wildly due to reclassifications and one-time effects from the merger. The reason is that a real estate fund’s financial statements are burdened by massive property depreciation charges, which, however, do not represent actual cash expenses. A building is depreciated even as its market value rises.

That is why real estate funds are evaluated using FFO (funds from operations)—that is, earnings adjusted for these depreciation charges and one-time sales—and the even stricter AFFO, which also takes maintenance investments into account. And here, the picture is significantly healthier. FFO for 2026 is estimated to range from approximately $1.71 to $1.75 per share, operating cash flow exceeds $1.25 billion, and same-store rent is growing by 3.9%. The only downside is that FFO is expected to decline slightly in 2026 from last year’s roughly $1.84, mainly due to the divestiture of senior housing and higher interest rates. The multi-year growth streak will thus likely be interrupted this year before moderate growth resumes.

Balance Sheet, Debt, and Liquidity

The balance sheet is one of the company’s strengths, and for a real estate fund, this is a critical factor, as the entire business model is based on debt. Net debt of around 5.4 times EBITDA is reasonable and conservative for the sector, and the company holds an investment-grade rating from all three major agencies (BBB+ from S&P and Fitch, Baa1 from Moody’s), which it has earned in recent years precisely through asset sales and debt reduction. In addition, it holds over $1.1 billion in cash, so liquidity is not an issue.

The only immediate concern is refinancing. The company has a $650 million bond issue maturing in mid-2026, and like any real estate fund, it is generally exposed to the interest rates at which it can roll over its debt. Given its investment-grade rating and ample cash reserves, this is not an acute threat, but it is a factor that keeps the valuation under pressure as long as rates remain high. Simply put, the company is in good operational shape and its financial risk is under control, but the entire sector is waiting for relief from interest rates.

Valuation

Based on the right metrics, the stock looks reasonable, not extreme. At a price of around $19 and FFO of around $1.72, it is trading at roughly 11 times FFO, which is slightly below both the historical and sector averages (around 12 to 12.5 times). The price-to-operating-cash-flow ratio is around 14.7, and the price-to-revenue ratio is 4.7. The book P/E of 62 has no meaningful value for a real estate fund.

The argument that the stock has room to grow rests on two pillars. The first is a slight revaluation of the multiples back toward the historical average, which in itself provides some upside potential. The second and stronger pillar is sensitivity to interest rates. If rates fall and discount rates normalize, real estate funds historically experience significant revaluation, as property values rise and the attractiveness of competing bonds declines. This is precisely why fair value estimates are significantly above today’s price, ranging from an analyst consensus of around $21 to models pointing to $29 and higher. For the stock to prove undervalued, all it essentially takes is for rates to come down and for FFO to remain stable. For it to prove overvalued, rates would have to stay high for a long time, the laboratory segment would have to continue weakening, and FFO would have to fall more than expected.

Dividend: The Hero of the Story, but with Scars

The dividend is the main reason most investors hold this stock, and it deserves a closer look. The company pays out $1.22 annually, distributed in monthly installments of $0.10167, which, at a price of around $19, corresponds to a yield of about 6%—higher than the sector average of around 5%. The monthly frequency is convenient for income investors because it smooths out cash flow and makes it easy to reinvest.

In terms of sustainability, the dividend appears secure. The payout ratio is roughly 70% of FFO and around 76% of the stricter AFFO, which means the company earns the dividend comfortably and still has room for share buybacks (in the first quarter of 2026, it repurchased $100 million worth of shares). Cash coverage is therefore not an issue.

There is, however, a blemish that it’s only fair to acknowledge. This company is not a dividend aristocrat. It has cut its dividend twice over the past decade, most recently by 19% in 2021 due to the impact of the pandemic. This is an important point to consider. The dividend is reliable but not untouchable, and history shows that in a crisis, management will cut it if it deems it necessary. What would have to happen for the dividend to be at risk of a cut? A significant and sustained decline in FFO that would push the payout ratio to unsustainable levels, or a debt or refinancing shock that would force the company to hoard cash. With a current payout ratio of around 70% and an investment-grade rating, such a scenario is rather unlikely, but history teaches us not to take the dividend for granted.

What Could Drive the Stock

• A decline in interest rates. The strongest driver. A normalization of discount rates would lead to a revaluation of the entire sector—and this stock along with it.

• Demographic growth in demand. An aging population is driving long-term growth in demand for outpatient healthcare and, consequently, for medical facilities.

• Portfolio simplification. Following the IPO of its senior living division, the company is more transparent and less capital-intensive, which could raise the multiple the market assigns to it.

• Acquisitions below replacement cost. Opportunistic acquisitions, such as the Gateway campus, add value if the laboratory market recovers.

• Share buybacks. There is room for further share buybacks as long as the price remains below the estimated fair value.

What to watch: FFO trends and outlook, same-store rent growth, occupancy and demand in the laboratory segment, refinancing progress, and, of course, the trajectory of interest rates.

Risks

• Interest rates. Real estate funds are more sensitive to interest rates than almost any other sector. High rates keep valuations down and make refinancing more expensive. Key indicator to watch: the trend in long-term bond yields.

• Refinancing. A $650 million bond issue maturing in mid-2026 and a general reliance on debt rollovers. Key indicator: terms of new issuances and rating trends.

• Weakness in laboratory real estate. An oversupply and weaker demand following the pandemic boom. Key indicator: occupancy rates and rents in this segment.

• Decline in FFO in 2026. Interruption of the growth streak due to the divestiture of senior housing and higher interest rates. Signal: FFO outlook performance.

• History of dividend cuts. Two cuts over the past decade indicate that income is not untouchable. Signal: payout ratio above healthy levels.

• Focus on healthcare. Regulatory and reimbursement risks associated with the U.S. healthcare system also indirectly affect tenants. Signal: pressure on healthcare providers’ financing.

Investment Scenarios

Pessimistic (12 to 24 months). Interest rates will remain high for longer, refinancing will become more expensive, the laboratory segment will continue to weaken, and FFO will fall more than expected. The market remains bearish on real estate funds, and the stock stagnates or falls back toward the lower end of its annual range, around $15 to $17. The dividend is maintained but shows no growth, and a yield of around 6% is the only consolation.

Realistic (12 to 24 months). Rates gradually decline; FFO stabilizes after this year’s drop and returns to modest growth of around 3% annually; the outpatient segment grows steadily; and the portfolio becomes more transparent following streamlining. The stock will be revalued closer to its historical multiple and trade in the range of roughly $21 to $25, with a monthly dividend of around 6%. The combination of modest growth and income results in a decent total return.

Optimistic (18 to 36 months). Interest rates will fall significantly, discount rates will normalize, and the entire real estate investment trust (REIT) sector will be revalued upward. Demographic demand and the recovery of the laboratories will boost FFO; the market will assign the company a premium multiple, and the stock will head toward fair value estimates of around $29 or higher. Furthermore, the dividend will begin to grow again.

The range of possible outcomes here is significantly narrower and less binary than with leveraged or growth plays. The downside is protected by tangible assets and a 6% dividend; the upside is linked primarily to falling interest rates. This makes the stock a relatively defensive income play with reasonable—albeit not spectacular—appreciation potential.

Key Takeaways from the Article

• This is a real estate investment trust (REIT) focused on healthcare properties (NYSE: DOC), which primarily owns outpatient medical and laboratory buildings and pays a monthly dividend of around 6%.

• The book P/E of 60 is misleading for a real estate investment trust because it is dragged down by property depreciation. The relevant FFO yields a multiple of around 11, which is slightly below the historical average, opening up room for a revaluation.

• The annual dividend of $1.22 is covered by roughly 70% of FFO and appears secure, but the company has cut it twice over the past decade, so it is a reliable—but not untouchable—income stream.

• The main growth thesis rests on falling interest rates (to which the sector is extremely sensitive) and a demographic tailwind in the form of an aging population and growing demand for healthcare.

• Compared to its competitors, it is a cheap and generous player. While the growth-oriented Welltower trades at around 33 times FFO with a yield of just 1.5%, this company trades at around 11 times FFO with a 6% yield. It’s a choice of income and value over growth and a premium, plus it has an investment-grade balance sheet.

• The main weaknesses are interest rate sensitivity, the upcoming refinancing of a $650 million bond issue, a weaker laboratory segment, and an expected slight decline in FFO in 2026.

• Within a portfolio, it makes sense to view this stock as a relatively defensive income position tied to the interest rate cycle. It offers a stable monthly dividend backed by tangible assets, with reasonable upside potential if rates fall. It is neither a speculative nor a high-growth play, but rather a more stable stock built on cash flow.

The Dow Jones Industrial Average is the oldest and most closely watched U.S. stock index, but it operates on a principle that seems a bit archaic today. Unlike the S&P 500, where each company is weighted according to its market capitalization, the Dow is what’s known as a price-weighted index. This means that a specific company’s influence on the index’s movement is determined solely by the price of a single share, not by the company’s total value. A stock priced at $900 thus carries much greater weight in the index than a stock priced at $50, even if the cheaper one belongs to a significantly larger company.

If we rank the index components by market capitalization—that is, by the company’s actual size on the stock market—we get a completely different order than the one based on their weight in the index. And at the very bottom of this ranking are currently four companies whose market value is significantly lower than that of the rest of the index.

A small market capitalization within the Dow does not mean these are small companies. All four are global leaders in their respective industries with billions in revenue. Rather, it means that each of them has gone through a period in recent years that the market has viewed with greater caution than the rest of the index.

Travelers $TRV

Travelers is the second-largest underwriter of commercial property and liability insurance in the United States and the only pure-play insurance company in the Dow index. Despite its size and stable business, it is currently one of the two smallest members of the entire index, with a market capitalization of around $66 billion.

A defensive business that’s performing well this year

Unlike the other companies in this selection, Travelers did not end up at the bottom of the list due to problems. On the contrary. Its shares are trading near their all-time high of around $310 and have posted gains in the low single digits since the beginning of 2026—a solid result in a year as turbulent as this one. This is due to the defensive nature of the insurance business. Premiums are paid regardless of how the economy is performing, and in an environment of higher interest rates, insurance companies also earn more from investments in their insurance reserves. The stock’s low beta of around 0.5 confirms that it is a stock that trades significantly more steadily than the rest of the market.

In the first quarter of 2026, the company reported operating income of $1.7 billion and a return on equity (ROE, i.e., the ratio of net income to equity) of 19.7%. At the same time, it raised its quarterly dividend by 14% to $1.25 per share. All of this points to a company in good shape, not a laggard in the index.

Be Cautious When Interpreting a Low P/E Ratio

At first glance, the price-to-earnings (P/E) ratio of around 9 to 10 may catch the eye—a figure that is often associated with undervaluation for high-quality companies. The year 2025 was exceptionally favorable for Travelers in terms of underwriting, thanks in part to below-average losses from natural disasters, which pushed earnings per share to an unusually high level. The low P/E ratio thus partly reflects a one-time strong year, rather than a permanently cheap valuation. In the insurance sector, it is more advantageous to track the valuation relative to book value, where the stock trades at roughly double that level—which is no longer a bargain range.

Analysts are divided. While Piper Sandler and Roth Capital, for example, are raising their price targets above $340, Barclays $BCS, on the other hand, has downgraded the stock to “underweight,” arguing that insurance premiums are beginning to soften and the pace of growth is slowing.

Nike $NKE

Nike is the world’s largest sports footwear and apparel brand, but unlike Travelers, it has fallen among the Dow’s smallest members for the opposite reason. Its market capitalization of around $64 billion is the result of a long and painful decline in its stock price.

A Decline of Tens of Percent

Nike shares have lost roughly 32% since the beginning of 2026 and are trading around $44, near their annual low. This is a level the stock last reached many years ago (2014).

A combination of several factors is behind the decline. Revenue in fiscal year 2025 fell by nearly 10% to $46.3 billion, and net income plummeted by more than 40%. In addition, the company is facing a significant slowdown in demand in China, where revenue is expected to decline by about 20% year-over-year, as well as pressure from U.S. tariffs, which management estimates at roughly $1 billion.

Paradoxically, it is precisely this price decline that has pushed Nike’s dividend yield to an attractive roughly 3.63%, the highest among the four companies. Meanwhile, the price-to-earnings ratio remains relatively high at around 29, as the company’s earnings are currently under pressure. A high P/E ratio here does not mean an expensive stock in the traditional sense, but rather temporarily suppressed profitability.

Betting on a Turnaround Under New Leadership

The investment thesis for Nike today centers on a turnaround. CEO Elliott Hill is striving to return the brand to its roots—namely, sports and innovation—while streamlining its bloated distribution channels. The company has also announced a collaboration with Google’s artificial intelligence to create a new shopping experience. Key events will include the upcoming release of fourth-quarter fiscal results at the end of June and the World Cup, which is traditionally a major marketing opportunity for the brand. However, many analysts remain cautious and are lowering their price targets to around $50.

Sherwin-Williams $SHW

Sherwin-Williams is the world’s largest manufacturer of paints and coatings, and with a market capitalization of around $78 billion, it sits in the middle of our quartet. It is a classic, high-quality company with a long history and a strong position among professional painters, but it is being held back this year by cyclical market conditions.

Dependence on the Construction and Housing Markets

More than 80% of the company’s revenue comes from North America, and a large portion of its business is tied to construction, real estate sales, and renovations. This is the core of this year’s problem. Higher interest rates have cooled the U.S. housing market, reducing both home sales and the pace of new construction, which directly impacts paint sales volumes. The stock has lost about 2% since the start of the year and is trading around $320.

Despite the unfavorable environment, the company increased revenue by 6.8% to $5.67 billion in the first quarter of 2026, primarily thanks to the strength of its brand and its ability to command higher prices. This so-called “pricing power”—the ability to raise prices without losing customers—has long been the company’s main competitive advantage. However, the stock remains expensive, with a price-to-earnings ratio hovering around 30 and a dividend yield of just under 1%, the lowest among the four companies.

An Unsuccessful Attempt at a Major Acquisition

An interesting development this year was Sherwin-Williams’ attempt to expand through an acquisition. Together with Japan’s Nippon Paint, it submitted a bid to acquire its European competitor AkzoNobel at 73 euros per share. Under the agreement, Sherwin-Williams would have acquired the automotive, marine, protective, and powder coatings divisions. However, AkzoNobel’s board of directors repeatedly rejected the offer, preferring its own merger with Axalta. In early June, Sherwin-Williams withdrew its bid.

3M $MMM

The last and largest company in today’s selection is the industrial conglomerate 3M, with a market capitalization of approximately $83 billion. The company manufactures tens of thousands of products, ranging from adhesive tapes and respirators to abrasive materials and components for the automotive and electronics industries. For years, it served as a textbook example of a so-called “value trap”—a stock that appeared cheap but was dragged down by endless litigation.

The End of the Era of Lawsuits?

The main obstacles were two massive cases: disputes over military earplugs and lawsuits related to so-called “forever chemicals” (PFAS), or perfluorinated substances that are virtually non-biodegradable. Over the past two years, the company has managed to largely resolve both major cases through billion-dollar settlement agreements. The market welcomed this, and by 2025, the stock had risen in value by roughly one-third. However, the lawsuits have not completely disappeared.

3M still faces more than 15,000 individual lawsuits related to PFAS, and key “bellwether” trials—test cases whose outcomes will indicate how the litigation will proceed—are expected by the end of 2026. The company estimates the costs of these lawsuits at roughly $500 million per year.

Operating Revenue Under New Leadership

Under CEO William Brown, 3M has been focusing on streamlining processes and improving margins in recent quarters. In the first quarter of 2026, adjusted earnings per share rose by 14%, and the operating margin increased to 23.8%. At the same time, the company completed the spin-off of its healthcare business into a separate company, Solventum, in which it retained a minority stake. Following this spin-off, 3M significantly reduced its dividend; it currently pays $0.78 per quarter, which corresponds to a yield of around 1.9%.

The GAAP price-to-earnings ratio is high, at around 30, due to one-time items related to litigation, but the forward P/E ratio, calculated based on adjusted earnings, is significantly lower, at around 18. This better reflects the company’s actual operating profitability. The stock is trading around $160 and is roughly flat year-to-date.

A Comparison of the Four Smallest Stocks in the Dow Index

* For 3M, the accounting (GAAP) P/E and forward P/E based on adjusted earnings are shown.

Strategic Perspective

Although these four companies share the bottom of the Dow’s market capitalization rankings, from an investment perspective, they represent entirely different stories that should not be lumped together.

Travelers is a defensive, high-quality stock that is benefiting this year from higher rates and a calm disaster season. Investors here are not betting on a dramatic turnaround, but rather on stability and a growing dividend. The main risk is softening insurance premiums, a concern even more cautious analysts are highlighting.

Nike, on the other hand, is a pure bet on a turnaround. Those who believe that the new management can revitalize the brand and restore revenue growth can now buy the stock at a fraction of its former price, along with an above-average dividend yield. However, the risk is significant, as turnarounds for major consumer brands tend to be protracted, and China remains a major unknown.

Sherwin-Williams is a high-quality cyclical company whose fortunes are closely tied to the U.S. housing market. With a challenging valuation and a weak construction cycle, there is less room for error, but the company’s long-term pricing power remains intact.

3M is a bet on the completion of its turnaround. If the company manages to definitively settle its legal disputes and maintain margin improvements, a forward P/E ratio of around 18 times earnings seems reasonable. However, the PFAS lawsuits remain a major question mark that could easily swing the outcome either way.

The four smallest members of the Dow Jones Index together offer a colorful illustration of how different paths can lead a company to the bottom of the market capitalization rankings.

A company’s position in the Dow Jones Index does not, in and of itself, indicate the quality or attractiveness of its stock. The index’s price-weighted methodology means that companies at completely different stages of their development are listed side by side. Each of these four companies must therefore be evaluated individually, based on its own business model, valuation, and risks, rather than on its current position in a single specific ranking.

Record Numbers That Failed to Impress

Adobe $ADBE released its results for the second fiscal quarter of 2026 on June 11, and on paper, they look great. Revenue reached $6.62 billion — a record in the company’s history—with year-over-year growth of over 10%. Non-GAAP earnings per share came in at $5.96, once again beating analysts’ estimates. The company also raised its full-year 2026 revenue outlook to between $26.5 and $26.6 billion.

Nevertheless, the stock fell to a new seven-year low the day after the earnings report and is currently trading around $195. How is it possible that the market is punishing record results and an upward outlook with such a sharp decline?

The answer lies in one word: ARR.

What Is ARR and Why Does It Matter?

ARR, or Annual Recurring Revenue, is a key metric for Adobe. It measures how many subscribers pay regularly for Photoshop, Acrobat, Premiere Pro, or Creative Cloud creative suites. The higher the ARR, the more predictable and valuable the business.

Adobe told investors that ARR from individual subscribers will slow down in the second half of the year —and this is intentional. The company is redirecting web traffic from paid sign-up flows to free versions of its tools. Acrobat, Express, and the generative AI tool Firefly will be available without a payment gateway. Planned price increases for Creative Cloud have also been postponed.

Analysts immediately reflected this in their price targets: Citi lowered its target from $264 to $228, BMO Capital from $285 to $230, and Mizuho from $270 to $245. The consensus among 33 analysts now stands at an average target of $288 —still significantly above the current price, but with a predominance of “Hold” recommendations.

"The strategic shift toward acquiring freemium customers through Adobe and Firefly lowers our expectations for ARR growth from individual subscribers in the second half of the year."

Shantanu Narayen, CEO of Adobe

A gamble, or a well-thought-out long-term strategy?

Narayen deliberately defends this strategy by comparing it to one of his best moves from the past. In the 1990s, Adobe decided to make Acrobat Reader completely free. It was a risky bet—the company sacrificed direct revenue. The result: PDF became a global standard, and Adobe continues to benefit from it to this day. Narayen claims he’s doing the same thing now with Firefly and Express—building a massive user base that will eventually become paying customers.

The data supports this thesis. Monthly active users of Acrobat and Express have surpassed 850 million and are growing by 20% year-over-year. Freemium users of creative tools have grown from 50 million to 90 million in one year. Traffic to adobe.com has increased by more than 40% year-over-year. ARR from AI products—Firefly, Acrobat AI Assistant, and GenStudio—has exceeded $500 million and tripled compared to last year.

These are numbers to be taken seriously. The question is: how long will it take for them to convert to paid subscriptions?

Firefly vs. the World: The Battle for Creative AI

The second layer of the story is competitive pressure. Midjourney, Stable Diffusion, and tools from Google and OpenAI offer generative AI at a fraction of the price or for free. Adobe is responding by positioning Firefly as a “commercially safe” alternative —the models are trained exclusively on licensed content free of copyright disputes. For enterprise customers who need to avoid legal risks, this is a strong selling point.

According to Q3 2025 results, Adobe reached a total of 29 billion AI generations; Firefly Services grew 32% quarter-over-quarter; and customers’ proprietary models expanded by 68%. Integrations with Google $GOOG Image, Veo, Gemini Flash 2.5, and OpenAI show that Adobe isn’t just betting on its own models but is building a platform that ties everything together.

"Firefly is already the go-to production studio for a new generation of creators. The momentum proves that trust and security are just as important as innovation."

Shantanu Narayen, CEO of Adobe, Q3 2025 Earnings Call

The Outgoing CEO and the Open Question of Succession

Adding to the mystery of record revenue and falling stock prices, the earnings call delivered yet another surprise. Narayen announced in March that he is stepping down after 18 years as CEO—as soon as a successor is named. He will remain as chairman of the board. At the same time, Adobe announced a change in the CFO position: Dan Durn is leaving the company, and Steve Day has been named interim CFO.

Two key positions are up in the air at once: for investors who value management continuity, this adds an extra layer of uncertainty. Who will lead Adobe in the next, decisive phase—converting freemium users into paying subscribers?

At first glance, however, the valuation looks attractive. According to StockAnalysis, the stock is trading at a forward P/E of around 9.6—significantly below historical norms for software companies with this level of cash flow. For fiscal year 2025, Adobe reported revenue of $23.77 billion and net income of $7.13 billion.

The outgoing CEO’s big bet

Adobe stands at a crossroads of its own making. Management is consciously limiting short-term revenue in the name of a larger future user base. It’s a strategy that has worked in the past —the PDF Reader is proof of that. But back then, the competition wasn’t as fierce, and the pace of change wasn’t as rapid.

Narayen is staking his legacy on the idea that 90 million freemium users will become paying customers faster than the market expects. If he’s right, a stock price around $195 will look like an exceptional opportunity. If not, his successor will inherit a tougher situation than the numbers suggest.

For now, the market has opted for skepticism. The results of the next two quarters will show whether that skepticism is justified.

If you look at the Salesforce ($CRM) chart for 2026, you’ll get the feeling that something has seriously gone wrong. Shares of the leading customer relationship management platform have fallen by more than 43% since the start of the year, hitting a new 52-week low of around $149 in mid-June —the lowest level in recent years. Yet the company is reporting record revenue, improving margins, and aggressively investing in artificial intelligence.

How is this possible, and what does it say about the state of the software stock market?

Numbers That Don’t Match the Price

Let’s start where every analysis must begin: with the results.Salesforce closed fiscal year 2026 with record revenue of $41.5 billion, up 9.6% year-over-year, and profits rose by more than 20%. In the most recent reported quarter, the company beat analysts’ estimates across all tracked metrics —revenue reached $11.13 billion, up 13% year-over-year, and adjusted EPS of $3.88 beat the consensus by nearly 24%.

Operating margins—historically a weak point for Salesforce—reached record levels: non-GAAP margin at 34.8%, GAAP margin at 21.1%.

A key indicator of future revenue—current remaining performance obligation (cRPO), or contractually secured revenue due in the next 12 months — rose 12% year-over-year to $29.6 billion.

Agentforce: A Bet on Autonomous AI

The key story surrounding Salesforce today is Agentforce —a platform for deploying autonomous AI agents in enterprise environments. The results so far are impressive:

• Agentforce & Data 360 ARR exceeded $1.2 billion in Q1 FY2027 and is growing by more than 100% year-over-year

• More than 9,500 paid customer contracts for Agentforce

• The platform has processed over 3.2 trillion tokens

• Salesforce invested $5 billion in Anthropic, the developer of the Claude AI model

In June, the company is acquiring Fin, an AI platform for customer service formerly known as Intercom, for $3.6 billion. This includes the acquisitions of Informatica (data integration) and m3ter (usage-based pricing for AI products). Management has a clear vision: to transform Salesforce from a CRM tool into a comprehensive platform for enterprise AI.

By 2030, the company aims to exceed $60 billion in revenue, with an organic CAGR of over 10%.

"Salesforce is building what we call the Agent-driven Enterprise—an environment where AI agents work hand in hand with people. The company’s position in data and CRM gives Agentforce an advantage that competitors without a comparable data foundation will find difficult to replicate."

Brian White, analyst at Monness Crespi

Why Are Shares Falling When the Numbers Are Rising?

The answer lies on two levels. First, the market does not yet believe in the monetization of AI. Agentforce generates billions in ARR, but investors are asking whether AI agents will replace some of the functions that customers used to pay for. BofA maintains an “Underperform” rating with a price target of $160 precisely because of concerns that the shift to AI will slow new customer acquisition and limit opportunities for upselling.

Second, in Q1 FY2027, management issued annual guidance that fell slightly short of analysts’ expectations. While the company confirmed FY27 revenue in the range of $45.9 to $46.2 billion, the market had expected more. After the results were released, the stock briefly spiked, then continued to fall—and since June 1, it has fallen for 13 consecutive trading days without a break, an unprecedented streak for this company.

Add to that sector-wide pressure on software stock valuations and concerns about layoffs (Salesforce laid off employees in its AI division, MuleSoft, and Marketing Cloud), and you have the full picture.

What Analysts Are Saying

Despite the decline in sentiment, the consensus remains convincingly positive. Fifty-one analysts maintain a “Buy” rating, and the average 12-month price target is around $254—which would represent an appreciation of over 65% compared to the current price.

• Average target (54 analysts): $244

• Median target (64 analysts): $255

• Canaccord Genuity: Buy, target $225

• Bank of America: Underperform, target $160

Simply Wall St reports that its fair value model indicates a 57% undervaluation at current prices. Meanwhile, technical indicators signal significant oversold conditions (RSI in the oversold zone).

"The company, with a 21% share of the CRM market—three times that of Microsoft, its closest competitor—is trading today as if the market expects no growth. That is an overly pessimistic scenario."

Seeking Alpha Analytics Team

The Biggest Risk: Time

Salesforce has one of the strongest positions in enterprise software —a loyal customer base across the Fortune 500, unique data assets, and rapidly growing AI products. At the same time, it has alienated some investors with a massive, debt-financed $25 billion share buyback program, which boosted EPS but also increased financial leverage.

The key question is: when will the FY27 turnaround show up in the numbers convincingly enough to win over even the skeptics? Management promises acceleration in the second half of FY2027. If that doesn’t materialize, the current slump may be just the beginning.

Lately I've been increasingly enjoying following the robotics sector. People often talk mainly about humanoid robots, but I think the most interesting opportunities may lie elsewhere.

What has interested me most over the long term is the company Intuitive Surgical $ISRG.

At first glance it doesn't look cheap, but I believe investors are paying a premium primarily for an exceptionally high-quality business.

The company is built around the da Vinci system, and its strength isn't just in selling surgical robots. The key elements are recurring revenues from instruments, accessories and service, which give it a very stable and predictable business. That's why I see it more as a company that can grow steadily over the long term rather than a bet on a currently popular theme.

Among other companies, I'm also intrigued by $SYM , which focuses on warehouse and logistics automation. The market potential is huge, but it's also a significantly riskier investment than Intuitive Surgical. An interesting combination of robotics, automation and the semiconductor industry is offered by $TER.

I believe robotics won't be the story of a single winning company. Healthcare, logistics, manufacturing and industry will each have their own leaders. So far Intuitive Surgical has stood out to me the most in the healthcare sector. To me, robotics is not just a vision of the future but a highly profitable, high-quality business.

Which robotics company are you following? I'd appreciate any tips, because I feel this sector will become increasingly interesting in the coming years thanks to developing AI.

When it comes to Meta today, many investors are mainly focused on one thing:

Huge AI spending.

And honestly, it’s not surprising. Meta expects its capital expenditures in 2026 to be roughly 125 to 145 billion USD. That’s a staggering number and the market naturally asks whether Mark Zuckerberg is once again burning too much money on a vision that might not pay off.

But in my view, you can look at this completely differently.

What if Meta isn’t burning money, but buying the next decade of the internet?

Meta is more than just Facebook and Instagram

Many people still think of Meta as the company that owns Facebook and Instagram and makes money from ads.

That is true, but to me it’s an incomplete picture.

Today Meta is building something bigger:

An AI layer across its entire ecosystem.

Better content recommendations. Better ads. AI tools for businesses. WhatsApp business agents. Reels. Threads. Proprietary AI models. And perhaps the most intriguing — smart glasses.

This isn’t a single AI product.

This is AI embedded into apps that billions of people use every day.

And that’s why Jensen Huang from Nvidia’s remark that “nobody uses AI better than Meta” makes a lot of sense to me. Investors worry too much about capex and overlook how AI is improving Meta’s advertising business.

AI is already helping the core business

Meta’s biggest business is still advertising.

And AI is extremely powerful here.

Meta needs to know what content to show you, which ad to show you, at what time, in what format, and for what kind of behavior.

The better AI nails that, the longer you stay in the app. The longer you stay, the more ad inventory Meta has. The better ads perform, the more companies are willing to pay.

And this is no longer just theory.

In Q1 2026 Meta reported revenues of 56.31 billion USD, a year-over-year growth of 33%. Ad impressions grew by 19% and average ad prices rose by 12%.

So far it doesn’t look like AI is just an expensive toy.

It looks like AI is already improving the main engine that makes Meta money.

The biggest optionality? Glasses

To me this is the most underestimated part of the whole thesis.

Meta’s glasses are not interesting just because they have a camera.

They are interesting because they can be the first natural AI hardware.

Today, if you want to use AI, you have to pull out your phone, unlock it, open an app and type or speak something.

With glasses it’s different.

You look at something. You ask. You get an answer.

No phone. No friction. No taking a device out of your pocket.

Meta, together with Oakley, introduced the Oakley Meta HSTN as a new category of “Performance AI glasses.” The glasses include a built-in camera, open-ear speakers, IPX4 resistance, 3K video and the Meta AI assistant.

And that’s where the story, in my opinion, starts to get really interesting.

Imagine glasses that can help you with sports, travel, cooking, work, content creation, translation, navigation, or everyday questions during the day.

Not as a replacement for your phone tomorrow.

But as the first step toward shifting part of the interaction with technology from the hand directly to the eyes.

Content could be a huge driver

This is something many investors underestimate.

Meta doesn’t have to explain to people what to do with the glasses.

People already create Reels, Stories, TikToks, Shorts, vlogs, workouts, travel, food, cars, sports, everyday life.

Glasses just give them a new way to do it more naturally.

Hands free. A first-person view. A more authentic shot. Less posing. More moment.

This is exactly the kind of product that can start as a “cool gadget” and over time become an everyday part of life.

Like smartwatches.

At first many people didn’t get them. Now millions wear them daily.

The valuation doesn’t look crazy

And now comes the most interesting part.

Meta is not a cheap small company.

But looking at valuation, I don’t think it’s more expensive than the market.

StockAnalysis lists Meta’s trailing P/E at about 21x, forward P/E at about 17.6x and PEG ratio around 0.84.

For comparison, the S&P 500 has a FactSet forward 12-month P/E of roughly 20.1x, meaning the index trades at a higher forward earnings multiple than Meta.

And that, to me, is a very interesting mismatch.

Meta trades cheaper than the S&P 500 despite, in my view, having much stronger AI optionality than the average company in the index.

Of course, the market discounts Meta for huge capex, Reality Labs losses and regulatory risks.

But if AI spending proves to improve ads, engagement, WhatsApp monetization and at the same time opens a new category of smart glasses, today’s valuation might not be expensive at all.

Risks shouldn’t be ignored

So this shouldn’t read like pure hype — the risks are real.

Meta is spending enormous amounts. Reality Labs still generates big losses. Smart glasses may fail to achieve mass adoption. Regulations and privacy could be an issue. And if AI investments don’t pay off, the market can punish the stock.

This is not a risk-free investment.

But investing is often not about whether a company has risks.

It’s about whether the potential reward makes sense relative to those risks.

And for me, that equation still makes sense with Meta.

Conclusion

Meta, in my view, is one of the most interesting AI companies on the market, even though many people still treat it as just Facebook and Instagram.

It’s not a chip company like Nvidia. It’s not a cloud giant like Microsoft. It’s not a pure AI startup.

But Meta has something few others do:

billions of users, a massive advertising engine, Instagram, WhatsApp, Messenger, Facebook, Threads, proprietary AI models, data, distribution, cash flow, and possibly a new hardware interface via glasses.

To me, Meta is a company that can turn AI into real money in a very practical way.

And maybe that’s why the market still doesn’t fully appreciate it.

Because while the market worries about how much Meta spends on AI, I ask a different question:

What if Meta is literally buying a spot in the next era of the internet?

There are two layers to this story. The first is the advertising business, with a gross margin of over 90%, which is growing at a rate of around 70% per year and where operational leverage is only now beginning to take full effect. The second layer—and the one that fascinates the market—is data licensing: the archive of human-written conversations has become premium fuel for training artificial intelligence, and major AI labs are paying for access to it. The question for an investor isn’t “Is it expensive?”—based on trailing numbers, it certainly is—but “Is the earnings power growing faster than it can be reflected in the valuation multiple?”

Key Points of the Analysis

• Revenue for 2025 grew by 69% to $2.2 billion, with a gross margin of 91%. In the first quarter of 2026, revenue rose 69% to $663 million, of which advertising accounted for $625 million (+74%).

• In 2025, the company reported its first full-year net income of $530 million (net margin of 24%) after a loss of $484 million in 2024. For the first quarter of 2026 alone, net income was $204 million and diluted earnings per share were $1.01, roughly a 7-fold increase year-over-year.

• Daily active users (DAUq) rose by 17% to 126.8 million, and total ARPU increased by 44% to $5.23. Monetization is thus growing faster than the user base itself, which is the most valuable indicator for an advertising platform.

• Free cash flow for 2025 soared by 217% to $684 million, with an FCF margin of 34% and negligible capex of $6.7 million. This is an extremely “capital-light” model with an almost software-like economy.

• The balance sheet is rock-solid: low debt, a net cash position, a current ratio of 12.7, and an Altman Z-score of 68.8. Instead of paying a dividend, the company launched a $1 billion share buyback program.

• The second growth driver is AI data licensing: agreements with Google and others generate over $200 million annually, with the potential to grow by an order of magnitude upon renewal.

• Upside vs. price: at a price of around $175, the trailing P/E is roughly 48–72, but based on annualized earnings from the last quarter, the forward multiple falls to the range of around 35–40. Analyst targets in a bullish scenario range from $250 and up.

• Main risk: Dependence of logged-out users on Google Search. If AI summaries in search results consistently erode traffic, it will impact the fastest-growing audience segment—the risk is so real that it is the subject of a shareholder lawsuit.

Company Overview

We’re talking about Reddit, Inc. $RDDT, the operator of one of the world’s largest discussion platforms, organized into hundreds of thousands of communities covering virtually every conceivable topic. The company is headquartered in San Francisco, employs roughly 2,200 people, and is led by co-founder Steve Huffman. It went public as recently as 2024, so its public history is short, and its financials remain heavily impacted by IPO-related costs to this day.

Yet the business model rests on a simple but exceptionally strong foundation. Users create content and engage in discussions for free, and the company sells ads around that content—and a gross margin of over 90% reveals who bears the costs of “producing” the content: the community, not the company. In recent years, a second layer of monetization has been added to advertising in the form of licensing data for training language models. For the financially savvy layperson, the best analogy is probably this: it’s an advertising business with an economic model resembling that of software, to which high-margin data licensing has now been added. And it is precisely this combination of rapid growth, high margins, and a recent transition to profitability that makes the valuation so controversial.

How the Company Makes Money

Advertising accounts for the overwhelming majority of revenue—in the first quarter of 2026, it amounted to $625 million out of a total of $663 million, or roughly 94%. Advertising growth is driven by two factors simultaneously: more impressions (a growing audience and longer time spent on the platform) and higher ad prices thanks to better targeting and tools for advertisers. The most important metric here is ARPU, or average revenue per user, and its figures tell the whole story of untapped potential:

• Global ARPU: $5.23 (+44%)

• ARPU in the U.S.: $9.63 (+54%)

• International ARPU: just $2.02 (+51%)

This gap between U.S. and international figures is crucial. It means the company can grow significantly without acquiring a single new user—all it takes is for the international audience to start monetizing at levels closer to those in the U.S.

The second segment—still small in terms of volume but strategically of above-average importance—is data licensing (around $39 million per quarter). Advertising is thus the core of the business and the source of cash flow, the fastest-growing and most “optional” element is data licensing, and the driving force behind both remains audience growth—particularly abroad and among logged-out users who arrive via search.

Market, Competition, and Where the Company Is Vulnerable

The company is playing on two major playing fields simultaneously. The first is global digital advertising, where it remains a relatively small player alongside giants like Meta $META and Google $GOOG —which, paradoxically, means there is significant room to capture a portion of advertising budgets. The second, newer arena is AI data licensing, a market that was practically nonexistent a few years ago and now generates tens to hundreds of millions of dollars annually.

In advertising, the company competes for budgets with Meta, Google, Pinterest, and Snap. It is smaller than them, but it has one advantage that is difficult to replicate: authentic, community-curated, text-rich content with a high level of engagement. It is precisely the depth of discussions and comments that sets the platform apart from visually oriented networks—and at the same time makes its data valuable for AI. Here, comments aren’t just a metric, but paid input for training models and, at the same time, fuel for deeper advertising engagement.

But this vulnerability has a name: Google. It plays three roles at once in this story—it is a partner (paying for data), a customer, and a potential threat, because its AI summaries in search results could drive away traffic. And since search drives a significant portion of the fastest-growing audience (logged-out users), this role is the most sensitive aspect of the entire argument. Added to this is a more concrete threat: Meta is testing its own discussion app, which could strike directly at the core of the model.

Management and Strategy

The company is led by co-founder Steve Huffman, who has been associated with the platform since its inception and holds a class of shares with greater voting rights. This gives management strong control and a long-term focus, but at the same time reduces the influence of ordinary shareholders. As is typical for a public company, the management’s track record is short but operationally impressive—it has turned the business around from chronic losses to profitability while simultaneously opening up an entirely new revenue stream from data licensing.

The strategy rests on three pillars: accelerating the monetization of the existing audience, expanding internationally, and building data licensing as a second high-margin growth engine. However, one figure speaks volumes about the management’s discipline: operating costs in 2025 fell year-over-year, while revenue grew by 69%. It is precisely this operational leverage that distinguishes a truly scalable business from a typical internet company.

Financial Performance: A Breakthrough, Not Steady Growth

The growth profile is extraordinary. Revenue rose from $667 million (2022) to $804 million (2023), $1.3 billion (2024), and $2.2 billion (2025). Far more important than the growth itself, however, is the turning point in profitability, and here it pays to understand the accounting background.

From 2022 to 2024, the company reported losses, which peaked in 2024 with a net loss of $484 million. However, this loss was not operational—it was largely an accounting one, caused by a surge in costs related to stock-based compensation around the time of the IPO (operating costs jumped by 108% at that time). Anyone looking only at the “net income” line would see 2024 as a disaster; in reality, it was a one-time accounting effect of the IPO.

In 2025, the picture reversed: operating costs fell by 10%, revenue rose by 69%, and the company reported an operating profit of $442 million and net income of $530 million for the first time (EPS $2.84, diluted $2.62). And the pace is accelerating—in the first quarter of 2026 alone, net income reached $204 million and adjusted EBITDA reached $266 million, with a margin of 40%. The company’s profile is therefore neither stagnant nor mature; it is a company on the steep part of the operating leverage curve, where every additional dollar of revenue translates into profit at an above-average rate. That is precisely why “true earning power” is at the heart of this entire thesis: trailing figures simply cannot keep up with how quickly profitability is changing.

Cash Flow and Balance Sheet

The cash side of the business is even stronger than the income statement. Operating cash flow in 2025 grew by 211% to $691 million, and free cash flow by 217% to $684 million, representing an FCF margin of 34%. Since the platform does not require heavy capital expenditures (capex was only $6.7 million), virtually all operating cash is converted into free cash flow—an economic model more reminiscent of software than traditional media.

The only dilutive factor is stock-based compensation, which does not reduce cash but dilutes shareholders through the issuance of new shares. However, the trend is important: their volume is declining as a percentage of revenue (in the first quarter of 2026, approximately $68 million compared to $85 million a year earlier, despite significantly higher revenue). Dilution is therefore slowing down just as cash generation is accelerating.

And the balance sheet? It is one of the strongest aspects of the entire story. Debt is practically zero, the net position is cash-positive (net debt to EBITDA of -3.2), the current ratio is 12.7, and cash stood at over $950 million at the end of 2025. The Altman Z-score of 68.8 is so high that the risk of financial difficulties is practically zero from a balance sheet perspective. For investors, this has one key implication: all risk associated with this stock is operational and valuation-related, not financial. The company faces no threat of debt refinancing or pressure from creditors, and management has complete freedom to finance growth and share buybacks from its own resources.

Valuation: The Core of the Dispute

This is where everything is decided. The surface multiples look high—P/E in the range of roughly 48 (based on the latest data) to 72 (based on full-year 2025 figures), price-to-revenue around 13–16, and price-to-free cash flow over 46. For a value investor, these are all red flags.

However, the trailing P/E here measures a company that was still unprofitable not long ago and whose earnings have multiplied several times over in recent quarters. When the multiple is based on the annualized earnings power of the last quarter (diluted earnings of $1.01 per quarter, or roughly over $4 per year) and ongoing growth is factored in, the forward P/E falls into the range of roughly 35–40. That’s a world of difference: a stock that looks extremely expensive at 70 times earnings looks like a fast-growing growth stock with a justifiable premium at 35–40 times forward earnings—provided the growth continues.

For today’s price to be considered “cheap,” the company must maintain revenue growth in the double digits, continue to expand margins through operating leverage, and turn data licensing into a significant source of revenue. For it to prove “expensive,” all it would take is for advertising growth to slow to the pace of more mature competitors, for search traffic to decline, or for data contract renewals to fall short of expectations. Just how uncertain the answer is is best illustrated by the range of models: conservative fair value estimates range from around $85 to $150, while optimistic targets range from over $250 to more than $320. Such a wide range in itself suggests that the value of this stock depends almost entirely on the pace of future growth.

Share Buybacks Instead of Dividends

The company does not pay any dividends and has no plans to do so in the foreseeable future—which makes sense for a rapidly growing internet company. It returns value to shareholders in other ways: in February 2026, it approved a share buyback program worth up to $1 billion, which, at a price of around $175, corresponds to roughly 3% of all shares. The logic behind this is closely tied to dilution—the buyback primarily offsets the issuance of new shares as part of compensation packages, so its primary goal is to stabilize the number of shares rather than significantly reduce it.

For investors, this leads to three conclusions. Do not expect a dividend—the entire return must come from share price appreciation. The buyback is funded from the company’s own cash, not debt, so it is sound and sustainable. And the key metric to watch is net dilution—that is, whether buybacks and slowing stock-based compensation combined will keep the number of shares roughly stable. The latest data suggest that the fully diluted number of shares is holding relatively flat—a positive sign. It would be disappointing to see a renewed acceleration in stock-based compensation without corresponding buybacks, or a diversion of cash into overpriced acquisitions. So far, however, the company is acting with discipline and is accumulating cash rather than squandering it.

What Could Drive the Stock

There are several specific growth drivers:

• International monetization. The international ARPU of $2.02 is a fraction of the U.S. ARPU of $9.63. Even partial catch-up would mean a multiple increase in revenue without the need to acquire new users.

• Maturation of the advertising platform. Better targeting, performance-based campaigns, AI tools for advertisers, and integration with e-commerce, which expand the pool of advertisers.

• Data licensing for AI. Existing contracts are worth over $200 million annually, and their renewals could be significantly higher—though some analysts are skeptical about such a dramatic increase.

What to watch: the growth rate of DAUq (last quarter: +17%), ARPU trends (especially internationally), the trajectory of adjusted EBITDA margin (last quarter: 40%), and the specific terms of data contract renewals. The ratio of active to inactive users is also important—active users monetize better, while rapidly growing inactive user numbers are sensitive to changes in search trends. If these metrics hold up, the forward valuation is justifiable; if growth stalls, the high multiple will collapse quickly.

Risks

• Dependence on Google Search. Logged-out users, the fastest-growing segment, come primarily from search. AI summaries in search results could permanently reduce their traffic—a risk serious enough to have led to a shareholder lawsuit. Signal: a slowdown in the growth of logged-out users.

• Valuation risk. Based on trailing multiples, the stock is expensive, and the entire investment thesis rests on continued high growth. Any slowdown will depress the multiple disproportionately.

• Advertising concentration. Roughly 94% of revenue comes from cyclical advertising, which advertisers cut first during a recession.

• Competition. Meta is testing its own discussion app, and the company is competing for ad budgets with much larger players.

• Uncertainty surrounding data licensing. The future value of renewals is speculative, and part of the market expects moderate rather than exponential growth.

• Volatility and shareholder structure. A beta of 1.91 and behavior reminiscent of a meme stock mean sharp fluctuations; furthermore, the founder holds controlling voting rights.

Investment Scenarios

Pessimistic (12–24 months). AI in search will permanently erode traffic from logged-out users; DAUq and revenue growth will slow; data licensing will disappoint upon renewal; and the advertising market will cool. Margins will hold up, but the market will stop paying a growth premium. The forward multiple will be squeezed down to the level of mature social networks, and the stock will return to a range of roughly $80–120. The main threat here is permanent damage to the business model, not just a temporary slowdown.

Realistic (12–24 months). Advertising growth will slow but remain in the double digits; international monetization will improve; data licensing will contribute steadily; and operating leverage will continue to expand margins. The company will meet expectations, but no more. A forward P/E of around 35–40 will prove reasonable, and the price will range from roughly $180 to $250, driven mainly by earnings growth rather than an expansion of the multiple.

Optimistic (18–36 months). International ARPU will begin to significantly catch up with U.S. levels, data licensing revenue will multiply during renewals, AI in search will ultimately drive more traffic and visibility to sources, and operating leverage will push margins higher. The company will establish itself as a structurally profitable growth platform, and the market will assign it a premium multiple. The price will range from roughly $280–$350 and higher, in line with the most optimistic DCF models.

The range is therefore exceptionally wide—roughly $80–120 on the low end versus $180–250 in the middle and $280–350 on the high end. This makes the stock an extremely asymmetric growth bet, whose value depends almost exclusively on a single question: how long and how fast can the company grow before the forces of maturity catch up with it?

Key Takeaways from the Article

• It is one of the world’s largest discussion platforms (NYSE: RDDT), which generates revenue primarily from advertising around user-generated content and, more recently, from licensing data for AI training.

• Its appeal rests on three pillars: a gross margin exceeding 90%, revenue growth of around 69%, and a recent shift to profitability, where operating leverage is just beginning to take effect—which is why the trailing P/E ratio overestimates the company’s true, forward-looking earning power.

• The balance sheet is debt-free with a cash position, so all risk is operational and valuation-related, not financial. No dividend; capital is being returned via a $1 billion buyback.

• Key weaknesses: dependence on traffic from Google search (the subject of a shareholder lawsuit), concentration of revenue in cyclical advertising, high valuation based on trailing figures, and extreme volatility (beta of 1.91).

• Data licensing for AI is the most interesting growth driver—high-margin, recurring revenue of over $200 million annually with the potential for exponential growth, which did not exist three years ago.

• Within the portfolio, this is a high-growth, highly volatile bet with no income component: the potential for significant appreciation in exchange for sensitivity to the pace of growth and changes in search trends. This is neither a defensive nor an income-generating position—it is a concentrated bet that earnings power will grow faster than the market currently anticipates.

When Nvidia $NVDA invests in a company, the market takes notice. And when it also adds that company to its Nasdaq-100 index and Uber $UBER partners with it, it creates exactly the kind of story that retail investors go crazy for. We’re talking about the Dutch company Nebius Group $NBIS, a stock that has surged by hundreds of percent over the past year and has become one of the most closely watched names in the entire AI boom.

But behind all that excitement lies an unpleasant detail. While the price is breaking records, most Wall Street analysts expect the stock to fall. Let’s take a look at what Nebius actually does and why opinions on it differ so dramatically.

From the Ruins of Yandex to One of the Hottest AI Names

Nebius has one of the most unusual origins on the U.S. stock market. The company emerged from the breakup of Russian tech giant Yandex, which was forced to sell off its Russian assets following sanctions in 2022. What remained moved to Amsterdam, shed its geopolitical baggage, and transitioned into the role of a pure-play AI cloud provider.

Today, Nebius is building what’s known as “neocloud” infrastructure—data centers packed with graphics chips that companies use to train and run artificial intelligence. It’s the same business model as CoreWeave $CRWV, for example, and demand for it is exploding.

The numbers illustrate this starkly. A year ago, Nebius reported quarterly revenue of around $50 million. In the first quarter of 2026, that figure was $399 million, according to the company’s filings with the U.S. SEC. That represents year-over-year growth of around 680%, which is unprecedented in the sector.

"This isn’t a one-off spike. Demand for AI computing power is growing faster than most analysts’ models can keep up with."

An analyst tracking the non-cloud sector

Nvidia, Uber, and the Bet on “Full-Stack” AI

A key moment came in March, when Nvidia announced a direct investment of $2 billion (roughly 42 billion crowns) in Nebius. It’s not just about capital. Part of the agreement is a plan to deploy over 5 gigawatts of Nvidia computing systems by 2030—which, according to a Bloomberg report, is enough energy to power roughly 3.8 million households.

It’s important to understand the motivation here. Nvidia isn’t just selling chips to Nebius—it’s putting its own money into the company. When the world’s most valuable company risks capital alongside its customer, the market interprets this as a stamp of confidence. Moreover, Nvidia has made similar investments in other suppliers, so this is part of a broader strategy to kickstart the entire AI ecosystem.

The other side of the story is perhaps even more interesting. Nebius owns the Avride division, which develops autonomous vehicles and delivery robots. And it is Avride that has entered into a partnership with Uber, with both companies investing up to $375 million. Avride’s robots are already delivering food via Uber Eats in U.S. cities like Austin and Dallas, and a robotaxi service is set to launch on the Uber platform.

Taken together, this paints a picture of a company that aims to cover the entire artificial intelligence value chain: from data centers through the cloud to physical robots on the street. This is called a “full-stack” approach, and to investors, it sounds like a dream.

D-Day: Entry into the Nasdaq-100

Today’s date is no coincidence. On June 22, Nebius officially joins the prestigious Nasdaq-100 index, alongside other AI players including CoreWeave and Rocket Lab. For the company, this means an automatic influx of money from index funds, which are required to buy the stock, and, most importantly, greater prestige in the eyes of institutional investors.

And that’s not all. In recent weeks alone, Nebius has racked up a string of positive news:

• Inclusion in the Nasdaq-100 (June 22) and the resulting mandatory purchases by index funds.

• The completed acquisition of Eigen AI, a company focused on optimizing AI models, which is expected to strengthen its cloud platform.

• Expansion in the United Kingdom for approximately 1.7 billion pounds, including the construction of new data centers.

In short, one piece of good news after another. No wonder the stock price is soaring.

When a stock gains +500% in a year

The stock is currently trading around $286 and has gained over 500% in the last twelve months. Market capitalization has climbed to roughly $72 billion, which is about 1.5 trillion Czech korunas. For a company that was, until recently, an insignificant remnant of Russia’s Yandex, this is a staggering leap.

The catch that isn’t talked about much

But here’s where the story starts to unravel. Although everything looks rosy, analysts’ average 12-month price target hovers around $244, as shown by data from stockanalysis.com. In other words, the Wall Street consensus expects the stock to fall from today’s level—by more than ten percent.

There are several reasons for caution:

• Insane capital intensity. Nebius plans to invest (capex) $20 to $25 billion in 2026. That’s more than the company earns, and the whole plan hinges on its ability to build data centers on time, secure power for them, and fill them with orders.

• Profit quality. Analysts point to a high proportion of non-cash items in the financial statements, which makes it difficult to gauge true profitability, especially as the company simultaneously raises capital and issues convertible bonds.

• Competition from the giants. Amazon $AMZN, Microsoft $MSFT, and Alphabet $GOOGL also operate the same business—renting out computing power for AI—through their cloud services (AWS, Azure, Google Cloud). Nebius is competing with them for the same customers, but against companies with incomparably deeper pockets that also build their own chips.

And then there’s the simple question of valuation. With a price-to-earnings (P/E) ratio of around 105, Nebius is trading at levels that assume absolutely flawless execution of its future plans. Any hiccup—whether a construction delay, a permitting issue, or a robotaxi accident—could send the stock price tumbling. It’s no coincidence that firms like DA Davidson, Wolfe Research, and Seaport Research have initiated coverage of the stock with only a neutral rating, even though they view the business itself positively.

“The business is real, and the growth is real. But the re-rating from a purely speculative non-cloud stock to an institutional infrastructure stock has already largely taken place.”

Summary of the views of analysts at DA Davidson, Wolfe Research, and Seaport Research

What It’s All About

At its core, Nebius is a bet that the demand for computing power for artificial intelligence isn’t going away anytime soon, and that a small player from Amsterdam can carve out a decent slice of the pie even against the hyperscalers. Investments from Nvidia, a partnership with Uber, and its inclusion in the Nasdaq-100 all reinforce this thesis.

At the same time, however, the company is burning through a massive amount of capital, and its valuation leaves no room for error. Interestingly, just as euphoria reaches its peak—with the stock entering the elite index—the analyst consensus is cautiously standing on the sidelines. Neobius has thus become the perfect test case for the question the entire market is asking today: where does justified enthusiasm for AI end, and where does madness begin?

If someone had said in the summer of 2024 that Palantir’s stock would quadruple in a year, they would likely have been laughed at. Yet that’s exactly what happened. Palantir Technologies $PLTR climbed from about $21 in June 2024 to an all-time high of around $207 in November 2025 —and then the decline began. Today, the stock is trading around $128, which is 27% lower than where it started 2026.

Numbers That Leave Analysts Breathless

Palantir reported revenue of $1.63 billionin the first quarter of 2026 —up 71% year-over-year. Adjusted earnings per share reached $0.33, beating analyst estimates by more than 18%. Based on these results, management immediately raised its full-year outlook: the company now expects revenue of $7.65 to $7.66 billion in 2026, which would represent 71% growth compared to the previous year.

The U.S. business division—the one that attracts the most attention—is growing even faster. In the most recent reported quarter , U.S. commercial revenue rose 137% year-over-year. The company also reports over 950 customers and a “Rule of 40” score of 114—a figure that most established software companies can only envy.

The Problem of Valuation

This is where the story gets complicated. Even after the sell-off since the start of the year, Palantir is trading at approximately 144 times net income and roughly 75 times revenue. These are figures that are virtually unprecedented among established companies.

By comparison, $NVDA Nvidia, which is growing at a similar pace, trades at a fraction of that multiple. And RBC Capital, one of the most skeptical voices on Wall Street with a price target of $50, openly states that today’s price implies a decade of flawless execution without the slightest hesitation.

"Palantir’s current valuation assumes that the company will become the world’s largest software company—while maintaining margins that almost no company achieves today. That’s an extremely narrow path for investors."

Rishi Jaluria, analyst at RBC Capital Markets

At the other end of the spectrum are UBS, with a price target of $200, and Rosenblatt Securities, with a price target of $225 —both firms argue that Palantir is building a sustainable competitive advantage through its ontological platform, which is becoming, for customers, an operating system for AI-driven decision-making.

The range of analyst targets from $50 to $225 says it all about how divided the market is on Palantir.

From Intelligence to Business—and Now to the Military

Palantir was founded in 2003 as a contractor for U.S. intelligence agencies. The founders—including Peter Thiel and Alex Karp, who remains CEO today—built the Gotham and Foundry platforms to analyze massive datasets for the military and governments.

Today, the company generates roughly 54% of its revenue from government customers and 46% from the commercial sector. The government sector is not merely a source of recurring revenue—in August 2025, Palantir signed a 10-year contract with the U.S. military worth up to $10 billion, consolidating 75 existing contracts into a single agreement. Shortly thereafter came a billion-dollar deal with the British Ministry of Defense.

It is precisely this expansion into the European defense market that is a key topic to watch in 2026. Increased defense spending by NATO countries following Russia’s invasion of Ukraine is opening doors for Palantir to markets where, until recently, it had no presence at all.

AIP: A Bet That’s Paying Off So Far

The key product of the last two years is AIP—the Artificial Intelligence Platform. It is an environment that allows companies to deploy large language models on their own internal data —securely, without having to share sensitive information with external providers.

It is AIP that is driving the explosion in commercial customers. Palantir organizes so-called “boot camps,” where, over the course of a few days, it demonstrates concrete results on potential customers’ own data. According to the company’s reports, the conversion rate for these events is exceptionally high.

“AIP demonstrably compresses what would otherwise take months into days. We’re seeing customers sign multi-year contracts after just three days of boot camp—that’s something unparalleled in enterprise software.”

Alex Karp, CEO of Palantir, in a letter to shareholders, Q3 2025

So where does the truth lie?

Palantir is in a strange situation: operationally, it has never had better times, yet its stock has lost more than a quarter of its value this year. Part of the sell-off is due to the exit of some speculative capital, part to Michael Burry, who disclosed a short position in Palantir at the end of 2025— and part simply to gravity, which eventually catches up with every stock trading at extreme multiples.

The next important milestone is August 10, 2026, when the company will report its second-quarter results. If commercial growth in the U.S. remains in the triple digits and management raises its outlook again, the market will have a new reason to resume its rally. If the pace slows—even slightly—the bears will have exactly the data support they’ve been waiting for.

Dividend yield is one of the most closely watched metrics investors use to select stocks for income-oriented portfolios. The higher the yield, the faster capital returns in the form of cash, regardless of whether the stock price is currently rising or falling. Moreover, the current environment of higher interest rates plays into this approach, as many financial firms are able to benefit from more expensive funding, whether in the form of higher returns on invested reserves or through higher-cost lending.

The financial sector, in particular, has traditionally been a source of generous dividends. Insurance companies, banks, asset managers, and real estate funds typically generate stable cash flows and return a large portion of their profits to shareholders. However, a high yield alone says nothing about quality. Sometimes it reflects a strong and predictable business, while other times it serves as a warning sign that the market expects a future decline in profits or even a dividend cut. It is therefore always important to distinguish where the yield comes from and how it is backed.

Progressive $PGR

Progressive is one of the largest providers of auto liability and collision insurance in the United States, and with a market share of around 18.5%, it is one of the industry leaders. The company has long built a reputation as the most profitable major auto insurer on the market, primarily thanks to its advanced risk assessment. Progressive was among the first to begin making extensive use of telematics—that is, the collection of data on drivers’ actual behavior—and algorithmic models that allow for more accurate pricing of insurance policies for each client.

This technological edge is reflected in a key metric for insurance companies known as the combined ratio. This is the ratio of claims paid and operating expenses to premiums collected. A value below 100% means that the insurer is profitable on underwriting alone, even before accounting for investment returns. Progressive has consistently remained well below this threshold. For the full year 2025, the combined ratio reached 87.1%, and in the first quarter of 2026, it was as low as 86.4%.

Meanwhile,the company’s growth shows no signs of slowing down. For 2025, Progressive reported revenue of $87.6 billion, representing a year-over-year increase of 16%, and net income of $11.3 billion. The number of active policies increased by roughly 9% to nearly 40 million, with direct auto insurance sales growing the fastest.

A Dividend That Is Not What It Seems at First Glance

With Progressive, it’s absolutely essential to understand how the dividend is structured; otherwise, investors can easily be misled. The company pays a regular quarterly dividend of just $0.10 per share, or $0.40 per year. At the current price of around $199, this corresponds to a yield of only about 0.2%. That is certainly not a high dividend.

The vast majority of the yield comes from the so-called variable annual dividend, which Progressive declares once a year and whose amount is derived from its earnings and capital position. For 2025, the total payout reached approximately $13.90 per share, bringing the trailing twelve-month (TTM) yield to about 7%.

This yield reflects past performance and is not guaranteed. If earnings were to decline in any given year or if the company needed to strengthen its capital, the variable dividend could be significantly reduced or even eliminated. Meanwhile, the forward yield on the regular dividend remains at just the aforementioned 0.2%.

Why the Stock Is Falling This Year

Despite strong fundamentals, Progressive has been among the weaker stocks this year. Since the beginning of 2026, the stock has lost about 10%, and over the past twelve months, it has fallen by more than 20%, lagging behind a number of competitors. There are several reasons for this.

The market is concerned about a slowdown in premium growth now that competitors have caught up with premium rate hikes, and analysts therefore expect lower earnings per share for 2026 and 2027 than the record high of roughly $18 set in 2025. April’s data also contributed to the nervousness, as the combined ratio temporarily jumped to 90.2%, which, while still indicating underwriting profit, represents a significantly lower margin than in previous months. Trading at a price-to-earnings ratio of around 12 times expected earnings, Progressive is still trading at a premium relative to its cheaper competitors, and the market therefore reacts sensitively to any sign of weakness.

VICI Properties $VICI

VICI Properties is a real estate investment trust (REIT) that owns the land and buildings of the most famous American casinos and entertainment complexes, but does not operate them itself. Its tenants include giants such as Caesars Entertainment and MGM Resorts, which pay VICI long-term rent under so-called triple-net leases. Under this model, the tenant pays not only the rent but also taxes, insurance, and property maintenance, providing the landlord with a highly predictable and stable income stream with minimal operational concerns.

However, VICI’s portfolio has long since expanded beyond just casinos on the Las Vegas Strip. The company is gradually expanding its business into the so-called “experience economy,” which includes, for example, water parks, bowling alleys, golf courses, and youth sports facilities. This diversification reduces dependence on a single segment and gives the properties a status similar to that of critical infrastructure, which tenants are unlikely to abandon. Furthermore, VICI has reported 100% occupancy and 100% rent collection since its IPO, even during the challenging period of the pandemic.

Reliable and Growing Yield

From a dividend perspective, VICI is a model of stability. The company pays a quarterly dividend of $0.45 per share, or $1.80 annually, which, at a price of around $28, corresponds to a yield of roughly 6.8%. Unlike Progressive, this is a regular and predictable payout. Furthermore, VICI has increased its dividend eight times in a row since its IPO in 2018, with the dividend growth rate consistently outpacing inflation over the long term. The company targets a payout ratio of around 75% of its AFFO (adjusted funds from operations), which is the primary measure of cash earnings for real estate funds and a better indicator than traditional net income.

In 2025, VICI’s revenue reached $4.0 billion, and AFFO per share rose by more than 5% to $2.38, comfortably covering the dividend payout. The company maintains an investment-grade rating (BBB- from S&P and Baa3 from Moody’s) and is actively expanding its portfolio, most recently through the acquisition of seven casino properties as part of the so-called Golden Portfolio.

The main risk remains the concentration among a few large tenants in the gambling sector and the sensitivity of REIT prices to interest rate movements. If rates remain higher for longer, high bond yields compete with dividend-paying REITs and push their valuations down. You can find out more about the stock’s intrinsic value in the Fair Price Index on Bulios.

Blue Owl Capital $OWL

Blue Owl Capital is an alternative asset manager that has risen in recent years to become one of the most significant players in the private credit sector. This segment of the financial industry—where non-bank institutions provide loans to companies outside the traditional banking system—has been growing rapidly in recent years, and Blue Owl is one of its leading players. The firm operates three main platforms. Direct lending accounts for roughly 37% of assets under management, real assets such as real estate and infrastructure for 27%, and so-called GP strategic capital—that is, investments in stakes in other asset management firms—for another roughly 22%. You can find an overview of the company’s business here.

A key advantage of Blue Owl is its so-called “permanent capital,” which makes up the majority of assets under management and which clients cannot simply withdraw quickly. This ensures an exceptionally stable flow of management fees for the firm. As of the end of the first quarter of 2026, Blue Owl managed $314.9 billion in assets, representing a year-over-year increase of 15%. In addition, over the past twelve months, it raised $57 billion in new capital.

A Yield Over 9% and the Trap of a Misleading P/E Ratio

Blue Owl pays a quarterly dividend of $0.23 per share and has committed to a total payout of $0.92 for 2026. At the current price of around $9.50, this corresponds to a yield of roughly 9.65%. At first glance, the stock appears to have an extreme price-to-earnings (P/E) ratio of over 100, which would suggest an astronomical overvaluation.

However, this figure is highly distorted and cannot be taken at face value when it comes to asset managers. GAAP-compliant earnings for these companies are burdened by a number of non-cash items, such as amortization of intangible assets, which distort the real economic performance of the business. Investors therefore focus on other metrics, primarily so-called fee-related earnings and distributable earnings—that is, the cash that the company actually earns and can distribute to shareholders. Based on these adjusted earnings, Blue Owl is trading at roughly 11 times expected earnings, which is, in fact, a relatively low valuation. In the first quarter of 2026, fee-related earnings rose 14% year-over-year and distributable earnings rose 11%, so the company is certainly not losing momentum in its operations.

Why the Stock Has Fallen by Half

Despite solid operating results, Blue Owl ranks among the worst-performing stocks of the past year. From its high of around $21, the stock has fallen to roughly $9.50—a drop of about 64%—and has lost 36% in 2026 alone. The main reason is growing skepticism toward the entire private lending sector. Investors fear that after years of rapid growth, non-performing loans could begin to accumulate in the event of an economic slowdown. Added to this are signs of increased capital redemptions from BDC (business development company)-type retail funds, a segment in which Blue Owl also operates.

Conversely, some market participants argue that these concerns are exaggerated and that the capital-light model—based on fees from permanent capital—is more resilient than the market currently prices it. Although analysts’ average target price, following a series of downward revisions, hovers around $15 to $16—significantly above the current price—there is a wide range of opinions. Blue Owl is thus the riskiest, but also potentially the most profitable, stock among today’s trio. The high yield of over 9% largely compensates for the uncertainty surrounding the future of private lending.

Comparison: Three Different Paths to High Yield

Although all three stocks share an above-average dividend yield, their profiles differ fundamentally. Progressive is a high-quality growth insurance company whose high yield consists largely of a volatile variable component. VICI is a stable income REIT with a reliably growing dividend, and Blue Owl is a dynamically growing but riskier asset manager with the highest—yet least certain—yield.

The table clearly shows that the highest yield does not necessarily mean the best choice. With Progressive, the yield is actually very low when looking at the regular dividend, and the real appeal lies in the annual variable payout. VICI offers a compromise between yield and certainty. Blue Owl entices investors with the highest yield, but at the cost of the highest risk and the highest volatility.

Strategic Perspective

For an income-focused investor, it’s important to understand that each of these stocks plays a different role in the portfolio.

As a general rule, with high-dividend stocks in the financial sector, it’s always important to ask where the yield comes from and how it’s backed. A high dividend yield may be a reward for quality, but it may also be compensation for risk that the market has factored into the price. The difference between these two situations often determines whether the dividend is sustainable in the long term.

Three stocks, three different stories, and three distinct paths to high dividend yields. Progressive demonstrates that a high yield can consist largely of a volatile, variable component, and investors should therefore not be misled by a fluctuating figure. VICI Properties represents the opposite extreme: reliable and predictable income backed by long-term lease agreements and an eight-year history of dividend increases. Blue Owl, meanwhile, offers the highest yield, but at the cost of the highest risk and uncertainty surrounding the future of the private lending market.

The common lesson remains that the dividend yield itself is only the beginning of the analysis, not the end. The true value of a stock is determined by the quality of the business behind the payout and its ability to fund the dividend over the long term.

Accenture lowered its revenue growth outlook for 2026 from 3–5% to 3–4%. Earnings did beat estimates at $3.80 per share, but revenue and new bookings missed expectations, with new contracts down 3% year‑over‑year.

The stock has plunged nearly 40% over the past year.

The weakness is mainly due to a slowdown in the U.S. federal segment and more cautious corporate IT spending. Accenture is trying to reverse the trend with cybersecurity acquisitions. With $4.2 billion in acquisitions, it aims to lift margins where traditional consulting is stagnating.

Does anyone have $ACN in their portfolio? Will you be adding to this stock?

The reductions are part of an ongoing effort by management to streamline operations. For shareholders, the cuts signal that the company is pressing ahead with its cost-reduction agenda beyond its home market, extending the restructuring to international support functions.