That does not mean a divorce from OpenAI. A revised partnership struck in October 2025 gives Microsoft roughly a 20–30% economic stake in OpenAI, long‑term access to its IP and models through 2032 and an enormous commitment that OpenAI will buy about 250 billion dollars of Azure cloud capacity over time, making it one of Microsoft’s largest anchor customers. The pact was intentionally designed to let both sides pursue independent opportunities, so Microsoft can build its own frontier models while still treating OpenAI as its primary external model partner – a hedge that spreads technical risk, locks in cloud revenue and keeps the company central to the next wave of AI, no matter which lab is ahead at any given moment.

A new arrangement with OpenAI: less dependence, more freedom

The new agreement rewrites the existing relationship. Microsoft retains a significant minority stake in OpenAI and access to its models, but is no longer the "mandatory" exclusive provider of computing power. Three points are crucial:

• Microsoft $MSFT has licensing rights to use OpenAI's models and products until 2032, including those that eventually arise after general AI is achieved.

• OpenAI has committed to take a huge bundle of Azure services in the coming years, guaranteeing Microsoft long-term cloud revenue.

• Both parties have explicitly freed up their hands to develop their own, competing models - so Microsoft is no longer just "packaging" OpenAI's technologies into its products, but can compete for the cutting edge of AI itself.

This is what Suleyman identifies as a key prerequisite: making sure the partnership with OpenAI remains, but also having the contractually confirmed right to go its own way if it is strategically advantageous.

First custom models: speech transcription, voice and images

The most obvious proof that this is not just about marketing is the newly introduced speech-to-text model. Microsoft claims that its system achieves a significantly lower error rate than the previously commonly used solutions from OpenAI and other competitors, while running at lower computing power requirements. This is important for anyone looking to transcribe calls in bulk - from customer service lines to media houses.

Alongside this, the company has also released its own voice and image production models for widespread commercial use. This makes it clear that it doesn't want to depend solely on external partners in these segments. From the point of view of users of Microsoft 365, Azure and other products, this means that it is no longer just OpenAI running "under the hood", but increasingly pure Microsoft technologies.

Huge infrastructure investments: those who want their own models must have their own "power plant"

In order to compete realistically with the biggest players, Microsoft needs not only researchers, but more importantly computing power and energy. That's why the company is significantly increasing its investment in data centres and GPUs on which the big models are trained. In fiscal year 2026 alone, more than $120 billion is expected to go into AI infrastructure.

Meanwhile, Microsoft boss Satya Nadella has long said that the race for AI is not just about the number of chips, but also their efficient use and the cost of energy. That's why Microsoft is also investing in optimising its models to do more work with less computing resources. This doesn't just apply to training, but more importantly to day-to-day operations - the cheaper a single model query is, the easier it is for AI to pay for itself in real-world applications.

The partnership continues, but the AI strategy is expanding

Despite all the changes, Microsoft is publicly presenting the partnership with OpenAI as strong and long-term. OpenAI's products continue to run on the Azure cloud and Microsoft is deeply integrated into their offerings. The difference is that the bosses in Redmond no longer want the company's entire AI story to stand on OpenAI.

Rather, the new direction is "multi-model": the OpenAI models, Microsoft's own models and third-party models can run side by side in a single service. The company is going to make decisions based on which system is best suited to the task at hand - sometimes it will be high-end, general-purpose models, sometimes it will be specialized, cheaper, faster variants.

AI across products: Copilot as a gateway

At the same time, Microsoft continues to push AI across its entire portfolio. Copilot in Office, Windows, and other applications is becoming a major face of the new strategy. It's gradually transforming from the original "assistant" to a tool that can proactively perform tasks and work across services - for example, find needed documents, prepare analytics, or automatically resolve simple customer requests.

So far, it's a minority of users who are paying extra for Copilot, but this is where Microsoft sees huge room for growth. Every extra percentage point in adoption among the hundreds of millions of Microsoft 365 users means billions in additional revenue per year. The combination of proprietary models, a strong cloud, and deep integration into everyday tools gives the company a chance to be not just an "AI component vendor" but the main interface through which businesses and individuals interact with AI.

Instead the stock changes hands at roughly 2.7× trailing revenue and about 18× free cash flow, multiples that sit at the low end of the peer group. Management is guiding for more than 2.1 billion dollars of revenue and around a 25% adjusted EBITDA margin by 2028, which implies a mid‑20s compound growth rate from here if they hit the targets.

Top points of the analysis

• Revenues have grown at least 20% a year for six years in a row, reaching $1.3 billion in 2025 and jumping by nearly half in the past two years.

• The company has its own database of about 245 million consumer profiles in the U.S., giving it an advantage in targeting accuracy and cost over competitors that depend on external data.

• Gross margins are around 60-70%, adjusted EBITDA margin is over 20% and free cash flow has a margin of around 14%, effectively turning revenue growth into cash.

• Net income is still slightly negative, but operating profit has turned positive for the first time and the loss has narrowed significantly from hundreds of millions to tens of millions of dollars in just a few years.

• The balance sheet is healthy: the company has net cash of over $400 million, virtually no long-term debt, and financial ratios suggest low risk of payment problems.

• The stock trades at about 2.7 times annual sales, well below the average of similar tech firms of about 5.9 times, and fair value models point to long-term upside room for the price

What Zeta Global does and what it does for a living

Zeta Global $ZETA is not an advertising network or agency. It's a company that gives big businesses the brains behind their marketing - the Zeta Marketing Platform. This platform collects and stores data about customers and their behavior, builds unified profiles from it, designs segments and campaigns, and controls who sees what when on email, web, app, or other digital channels. The goal is not to buy more advertising, but to use existing budget to maximize sales and loyalty.

Zeta earns primarily on platform subscriptions and campaign processing fees. Most of its contracts are multi-year enterprise contracts that run in the hundreds of thousands to millions of dollars a year. It adds services - help with implementation, setting up data flows, creating segments or integrating into existing systems. As a result, revenues are relatively predictable and are based in significant part on recurring revenue.

The backbone of the business is approximately 450 large customers who spend in excess of $100,000 annually. For these so-called scaled customers, the average annual revenue is close to $1.9 million, growing by almost a fifth year-on-year. This means that Zeta isn't just growing by adding new names to its roster, it's growing by going deeper with existing clients - into more countries, product lines and marketing channels.

Data empire: 245 million profiles and deterministic identity

What sets Zeta apart from most of its competitors is its data foundation. Over the years, the company has built its own database of some 245 million consumer profiles in the US for which it has consent to process data. These are not anonymous aggregate numbers, but individual identities that combine email addresses, mobile identifiers, campaign response history, purchases and other signals into one coherent picture.

The key is that Zeta connects these profiles deterministically. This means that when a new signal appears in the system, such as activity from a mobile app or email, it can match it to a specific profile with a high degree of confidence. Unlike probabilistic methods, which only guess if two different signals belong to the same person, Zeta works with unambiguous links. As a result, matching accuracy is reported to be above ninety-five percent, whereas probabilistic models tend to be around eighty.

This has two major advantages. The first is the quality of targeting - when Zeta says it is sending a message to a particular segment, it does so with less "noise" than competitors that depend on third parties. The second is cost - once you build a database, the cost of each additional unit of data is relatively low, whereas buying external data packages tends to get more expensive with volume. This gives Zeta a scale advantage: with each new client and each additional signal, its data empire becomes both more valuable and relatively cheaper per unit.

Athena AI: an AI agent that translates data into decisions

Above this data base is a layer of artificial intelligence. Zeta doesn't sell it as a standalone AI "toy" but as part of a marketer's regular job. The most visible product is Athena - an AI agent that the company has deployed into the platform as a conversational interface.

Athena can answer questions in everyday language like, "Which segments responded best to the last campaign?" "What message generated the highest return for customers in a particular region?" "Which customer groups are likely to leave soon and how to reach them?" The agent can suggest new segments based on the data, automatically generate campaign variations for different groups, and adjust budgets between channels based on actual consumer behavior.

For clients, this means they don't need an army of data analysts to leverage a comprehensive data foundation. For Zeta, it means its platform is harder to replace - when one part of marketing learns to rely on Athena's help, this way of working often spreads to other departments. This increases the value of the platform, and with it, the average revenue per customer.

Growth and margins in numbers

When we line up Zeta's key numbers over the past few years, it looks like a textbook example of a growth technology company reaching profitability. In 2022, the company made roughly $591 million. A year later, revenue had grown to $729 million, a growth rate of over twenty percent. In 2024 came a jump to 1.006 billion, a 38 percent growth, and in 2025, revenues moved to 1.304 billion, another nearly thirty percent increase.

Gross profit grew even faster over the same period. From about $324 million in 2022 through $403 million in 2023, it rose to $550 million in 2024 and climbed to $791 million in 2025. Thus, gross margin today is roughly 61% and, after adjusting for one-time effects, is in the 65-70% range, which is typical for high-quality software with a significant data component.

The operating result has improved step by step. From an operating loss of nearly $259 million in 2022, through losses of $165 million in 2023 and $60 million in 2024, Zeta has moved to an operating profit of around $5 million in 2025. Net profit is still slightly negative - a loss of around $31.5 million - but that's a major shift by the standards of a company that was burning through hundreds of millions not so long ago.

When you look at cash flow, the picture is even better. Adjusted EBITDA margin is about 20.5%, and the company has managed to improve adjusted EBITDA margin year-over-year for nineteen consecutive quarters. Free cash flow grew by over seventy percent in Q3 2024, free cash flow margin was around fourteen percent and we see similar levels in the full year numbers. This means that Zeta is getting to the stage where every new dollar of sales is bringing in more cash than before.

Balance sheet, cash and share buybacks

The financial side of Zeta is relatively conservative today. The company has net cash of over $400 million, virtually no long-term debt, and debt-to-asset or equity ratios near zero. An Altman Z-score of around 3.3 suggests that Zeta is in a healthy zone in terms of bankruptcy risk or acute financial problems.

Management adds active capital work. Zeta has announced a $200 million share buyback program, which at a market capitalization of about $3.5 billion represents a little over twelve percent of the company. Combined with rising free cash flow, this means that even if the pace of revenue growth slows slightly, earnings per share can continue to grow thanks to higher margins and a declining number of shares outstanding.

Management

Zeta is led by co-founder David A. Steinberg. His approach can be summed up in two words: consistency and scaling. Just as he built the company's data layer incrementally, he is building investor confidence. Zeta has had more than fifteen consecutive quarters in which it has beaten its own revenue and earnings guidance and then raised the outlook. That said, management doesn't want to look good with one big announcement, but prefers to repeatedly deliver slightly better results than promised.

At the same time, the team has learned to manage costs so that growth is not bought out by endless losses. The transition from operating loss to operating profit, rising EBITDA margins and discipline in debt show that Zeta is not "scorched earth for growth at any cost". Instead, it is moving into the category of companies that are combining double-digit revenue growth with a gradual but clear improvement in profitability.

Zeta Plan 2028: growth and profitability targets

At the investor day, Zeta presented a roadmap to 2028 that summarises where it wants to take the company. Revenue targets are to exceed $2.1 billion, up from $1.3 billion, which implies annual growth of around twenty percent. Adjusted EBITDA margin is expected to move to 25%, a few percentage points higher than today, and adjusted EBITDA is expected to exceed $525 million in absolute terms. Free cash flow is expected to grow above 340 million as a result, and the free cash flow margin should reach sixteen percent.

The path to these numbers leads in three main directions. The first is deepening relationships with existing clients and increasing revenue per customer. The second is expanding the portfolio of features and leveraging AI to enable Zeta to charge higher prices for better value. The third is geographic expansion - particularly in Europe and other regions where marketing budgets are growing and digital is gaining momentum, but the market is not yet fully saturated.

Competition: Adobe, Salesforce and the big tech players

Zeta is in a segment where it is up against both specialist firms and tech giants. On the marketing cloud side, the most prominent are Adobe Experience Cloud $ADBE and Salesforce Marketing Cloud $CRM. Both systems have the advantage of size, breadth of features and an entrenched position in the IT architecture of large companies. Zeta defines itself against them by offering depth in several key areas - a unified view of the customer, deterministic identity matching, and a proprietary data base that competitors often replace with a mix of client data and external sources.

Alongside this, there is pressure from big technology players such as Google $GOOG, Amazon $AMZN and Microsoft $MSFT. These combine advertising systems, cloud services and AI and can absorb some of the demand for marketing technology in their own ecosystems. Zeta is therefore betting that many companies will want a partner that is not tied to one advertising channel, but can collect and activate data across the entire digital environment.

Valuation: cheap or fairly priced?

Looking at basic valuation metrics, Zeta today trades at roughly 2.7 times revenue, 18.8 times free cash flow and more than four times book value. For a company that is growing sales at a rate approaching thirty percent, has a gross margin of over sixty percent, an adjusted EBITDA margin of over twenty percent, and a double-digit free cash flow margin, these are not inflated multiples.

Comparisons with other marketing technology players, who often trade around five times revenue, suggest that the market is not fully factoring growth and margins into Zeta's price. Fair value models that combine discounted free cash flow and peer comparisons often come out around $20 per share, sometimes higher. This implies a near-term potential of about a quarter to a third higher than today's price if current growth rates and margins can be maintained.

But at the same time, the market is discounting Zeta for risks that are real: data regulation, competition from big players, concentration of large clients, and the fact that book net income is negative so far. All of this explains why multiples aren't higher - and why investors who believe in the Zeta 2028 story feel there may be room for overvaluation.

Risks that cannot be ignored

The first risk is the regulatory environment. Zeta is building much of its advantage on the use of personal data. Laws like GDPR and California's CCPA have shown that regulation can fundamentally limit what marketers can do with data. Further tightening, whether in the US or elsewhere, could limit Zeta's access to data or increase the cost of protecting and managing it.

The second risk comes from the marketplace. Adobe, Salesforce, and the big tech players have more resources, stronger brands, and broad portfolios that can in some cases displace Zeta or at least slow its growth. Zeta may have specialisation and proprietary data, but it must constantly innovate to maintain this advantage.

The third risk is concentration. Because Zeta works with a relatively small number of large customers, the loss of one or two key contracts can take a significant hit to revenues. In an environment where large companies sometimes consolidate their suppliers into a smaller number, this is not a purely theoretical risk.

A fourth risk is the integration complexity of acquisitions, particularly Marigold. Merging platforms, databases and teams in a way that does not affect clients through outages or reduced quality is a complex task. Any major setback could damage Zeta's reputation as a reliable partner.

Investment scenarios

Baseline scenario - delivering Zeta's 2028 plan at a slight discount

In the base case, Zeta succeeds in delivering most of what the Zeta 2028 plan promises. Revenue grows at a rate of around twenty percent per annum, reaching around $2.1 to $2.3 billion around 2028, as indicated by both management's own targets and the estimates of some analyst houses. Adjusted EBITDA margin is gradually moving from today's approximately twenty percent to the targeted twenty-five percent, free cash flow is approaching or exceeding the $340 million per year mark, and free cash flow margin is around sixteen percent.

In such a situation, the market would likely stop valuing Zeta as a "suspiciously cheap" growth title and would be willing to pay roughly four to six times each dollar of revenue, more than the current 2.7 times, but still at some discount to the biggest names in the industry. If such a revaluation were to happen alongside earnings per share growth boosted by buybacks, an investor coming in today could see a combination of decent earnings per share growth and reasonable rerating multiples over a three- to five-year horizon.

Growth scenario - Zeta as the "hidden" leader of AI marketing

In a growth scenario, more will be accomplished than what is locked into a conservative plan. The monetization of Athena and other AI features will accelerate platform adoption, Zeta's share of the marketing cloud market will roughly double as some analyses predict, and revenues will reach or exceed the high end of estimates around 2028. If revenues move more towards $2.3 billion and free cash flow margins reach the target sixteen percent, we are talking about free cash flow of around $340 million per year, as reported by several external models.

In an environment where Zeta has built a reputation as one of the leaders in AI marketing, it would not be unusual for the market to value such free cash flow at a multiple of twenty to twenty-four, which, with free cash flow in that range, supports a future enterprise value of roughly $6.8 billion to $8.2 billion. Against today's market capitalization of about 3.5 billion, this would imply a potential well above fifty percent over a multi-year horizon, but at the cost of higher execution intensity and sensitivity to market developments.

The cautious scenario - slowing growth and a sustained discount

In the more cautious scenario, Zeta's growth slows faster than the plan envisages. Revenues will grow not around twenty but rather around fifteen percent per year, and the company will still improve margins but will not quite reach the targeted twenty-five percent adjusted EBITDA margin or sixteen percent free cash flow margin. The regulatory environment will tighten and competition from the big players will siphon off some of the growth where Zeta has less differential value.

In such a scenario, the market would hardly be willing to move the price to earnings multiple well above three times. The stock could then trade somewhere in the 2.5 to 3.5 times earnings range over the long term, giving investors a return based on incremental earnings per share growth rather than a large revaluation multiple. With a strong balance sheet and free cash flow, it would still be a relatively safe growth title rather than a speculative one, but expectations would need to be proportionately more sober.

What to take away from the article

Zeta Global is a growth company that is no longer in the promise phase. It has a multi-year history of double-digit revenue growth, gross margins of over sixty percent, improving operating margins, and double-digit free cash flow margins. Under the surface, it's built on a data infrastructure of 245 million consumers and AI layers that translate that data into actionable marketing decisions. Management has a history of conservative communication and repeatedly beating its own targets, and has a plan in front of it that can take the company to the level of a free cash flow generator in the hundreds of millions of dollars within a few years.

The current valuation is partly consistent with this and partly not. A revenue multiple of around 2.7 is at the bottom, rather than the top, of the range for such growth and margins, but it also carries a discount for the real risks - regulation, competition, client concentration, and the still unfinished path to stable profitability under accounting standards. If you believe that Zeta can deliver on most of the Zeta 2028 roadmap and that its data and AI advantage will stand up to the big players, the current price may represent an interesting growth position with the potential for revaluation. But if you're nervous about the combination of data regulation and the power of Big Tech, it makes sense to keep an eye on Zeta, but keep it more as a minor weight or as a candidate for a later entry when the path to target numbers becomes even clearer.

TSMC has just announced that its second Japanese fab in Kumamoto will start production on the most advanced 3nm process in 2028.

Originally, this fab was supposed to produce chips on significantly less advanced 6 to 12nm processes.

Japan’s multibillion-dollar bet

Total investment in both Japanese fabs is expected to exceed $20 billion. The second fab alone could cost about $17 billion, although TSMC has not officially confirmed that number.

The Japanese government has approved subsidies of up to $4.62 billion for the second fab.

The entire project is being led by TSMC’s subsidiary JASM, which was established in 2021 with the support of Sony $SONY and was later joined by DENSO $DNZOY and Toyota $TM.

The planned monthly capacity of the second fab is 15,000 twelve-inch wafers on the 3nm process, with both fabs together expected to reach a capacity of 100,000 wafers per month.

The numbers

TSMC’s annual revenue $TSM for 2025 reached $122.3 billion, representing a year-on-year increase of 38.5%. For comparison: in 2024 it was $88.3 billion and in 2023 "only" $70.6 billion. So revenues nearly doubled in just two years.

Gross margin in Q4 2025 climbed to 62.3% and operating margin reached 54%. EPS recorded a year-on-year increase of 46.4%.

Global expansion at scale

Japan is only one piece of the puzzle. In Arizona, TSMC is investing a total of $165 billion to build five new fabs and two packaging plants, making it the largest foreign direct investment in U.S. history. The first Arizona fab began mass production on the 4nm process at the start of 2025 and supplies key clients such as Apple $AAPL and Nvidia $NVDA.

The second fab is slated to begin high-volume production in the second half of 2027, and the third is already under construction. The company is also planning fabs in Germany and continuing expansion in Taiwan.

Capital expenditures (CapEx) for 2026 are set at $52 to $56 billion, an increase of 27% to 37% compared with last year’s roughly $41 billion. Approximately 70% to 80% of that will be directed to the most advanced 2nm and 1.6nm processes.

How the stock is doing

Forward P/E is approximately 23, which isn’t excessive for a company with such growth and market dominance. TSMC holds roughly a 70% share of the global foundry market.

Do you have $TSM in your portfolio, or are you betting more on its customers like $NVDA or $AMD?

Here's why these companies are leading the market in 2026

The year 2026 has brought some major surprises to the stock markets. Instead of the AI tech giants that dominated returns in 2025, oil companies benefiting from the conflict in the Middle East and the closure of the Strait of Hormuz, as well as semiconductor equipment manufacturers benefiting from a massive cycle of investment in chip manufacturing capacity, topped the performance charts.

The size of these companies also plays an important role. All of the companies in today's survey have a market capitalisation of over $100 billion, meaning that these are not speculative bets on small growth companies, but established giants that can generate growth even at the huge scale of their businesses. It is the combination of strong share price growth this year and massive market capitalization that makes these companies interesting to say the least.

ExxonMobil $XOM

ExxonMobil is one of the absolute winners of 2026 in the US stock market. Shares are up approximately 33% since the beginning of the year, an extraordinary result for a company with a market capitalization of over $650 billion. This rise is mainly due to the dramatic increase in oil prices as a result of the conflict in the Middle East and the closure of the Strait of Hormuz by Iran in early March 2026. The price of Brent crude oil climbed to $126 per barrel in March, the highest level since 2022.

ExxonMobil is one of the biggest beneficiaries of this development. The company has an extensive portfolio of producing assets in the United States, particularly in the Permian Basin, which is one of the world's most productive oil regions. Unlike some of its competitors, ExxonMobil is not directly dependent on the transportation of oil through the Strait of Hormuz, which ensures operational continuity even in times of peak stress. At the same time, the company has completed the key Golden Pass LNG project, which strengthens its position in the global LNG market.

Fundamental strength and dividend stability

From a fundamental perspective, ExxonMobil operates with a P/E ratio of around 24, a dividend yield of approximately 2.4%, and continues a streak of 43 consecutive years of dividend increases.

The company generates strong operating cash flow and maintains a disciplined approach to capital spending. Analysts at Citigroup $C raised their target price to $150 in March (current price per share for $XOM is $160), highlighting a structural shift in institutional investor exposure to the oil sector. Interestingly, ExxonMobil's forward P/E even outperformed Nvidia's forward P/E in April.

Chevron $CVX

As the second-largest U.S. oil producer, Chevron has appreciated approximately 30% year-to-date this year, with the company's market capitalization rising to nearly $400 billion. Shares hit an all-time high of $214.71 apiece in March. As with ExxonMobil, the oil shock caused by the closure of the Strait of Hormuz remains the main catalyst for growth.

Chevron has significantly higher exposure to the price of oil compared to some of its competitors due to the structure of its portfolio. The company's primary focus is on the upstream segment, i.e. oil and gas production, which means that rising commodity prices translate directly and significantly into its profit margins. An important milestone in recent months has been the completion of the acquisition of Hess, which expanded Chevron's upstream portfolio to include high-quality assets in Guyana.

Valuation and risk

Chevron's current P/E ratio is around 30, which is well above the historical average and reflects investor expectations of a continued favorable oil price environment.

The dividend yield is approximately 3.5% with the company having increased the dividend by 4.1% this year. Analysts at DBS reiterated a Buy recommendation, but it should be noted that the consensus analyst target price is around $186, which is below the current share price. This suggests that the market has already largely priced in a positive scenario, and a potential drop in oil prices could lead to a quick correction.

Applied Materials $AMAT

In the case of Applied Materials , it is not an energy company, but a company that is a key supplier of manufacturing equipment to the semiconductor industry. The stock is up about 36% since the beginning of the year and has a market capitalization of over $275 billion. Applied Materials is the largest manufacturer of semiconductor wafer fabrication equipment in the world and holds the leading market share in the deposition segment, which is the process of depositing new materials onto silicon wafers.

Several structural factors are behind this year's growth. The shift to advanced Gate-All-Around transistor architectures, the development of backside power delivery technologies, and growing demand for High Bandwidth Memory (HBM) for AI applications are creating record order books. The company reported record sales in both its Logic and DRAM segments in the first quarter of fiscal 2026, and management announced a 15% increase in the quarterly dividend.

Competitive advantage in the AI chip era

Applied Materials' key competitive advantage is the breadth of its product portfolio. Unlike more specialized competitors such as Lam Research $LRCX or KLA $KLAC, Applied Materials covers virtually the entire manufacturing chain from etch to deposition to metrology and inspection. This integrated approach reduces dependence on a single process cycle and allows the company to better protect margins.

Intel $INTC

Intel is undoubtedly the most surprising name in this ranking. The stock is up about 37% since the beginning of the year and more than 160% over the past 12 months. The company, which most of the market wrote off as a lagging giant two years ago, is transforming itself into one of the most watched turnaround stories in the entire semiconductor industry under new CEO Lip-Bu Tan . Market capitalization has climbed above $250 billion.

The dramatic turnaround is primarily due to the successful completion of the program, with the company starting volume production on the 18A process in October 2025. This production node is reportedly about a year ahead of TSMC's rival 2nm process $TSM in implementing its key PowerVia technology for backside power delivery. In March 2026, the company began shipping its first commercial 18A-based products, specifically Core Ultra Series 3 (Panther Lake) processors for notebooks with integrated AI capabilities.

Strategic investments and risks

Intel has received massive strategic support: the US government holds approximately 10% stake through the conversion of an investment from the CHIPS Act program, Nvidia has invested $5 billion and SoftBank $2 billion.

These investments confirm the technological direction of the company, but also show the huge capital intensity of the entire turnaround. The firm still operates with negative free cash flow and GAAP gross margins remain around 30%, well below those of a healthy semiconductor company. CEO Tan calls 2026 an execution year, noting that he expects the real growth inflection point to be in 2027. A key test will be winning the first large external customer for the 14A process, which is scheduled for production in 2027.

TotalEnergies $TTE

European energy leader that has taken advantage of market chaos

TotalEnergies is the only European company in today's selection and one of the biggest winners of this year's oil shock. Shares on the New York Stock Exchange are up more than 42% since the start of the year and the market capitalisation is around $200 billion. As an integrated energy company, the company is benefiting not only from higher oil prices, but also from the extremely favorable refining margins and business opportunities that the current crisis is creating.

According to the Financial Times, TotalEnergies was able to execute one of the largest deals in the history of oil markets in March 2026. Its traders bought around 70 cargoes of crude from the Emirates and Oman, more than double the February figure, and took advantage of the dislocation in the Dubai crude benchmark market, where the price climbed from around $70 to $170 a barrel. Total profits from this trading operation reportedly exceeded $1 billion.

Diversification as a long-term advantage

Unlike pure US oil companies, TotalEnergies has a significantly more diversified business model. The company operates not only in oil and gas production, but also in renewable energy, LNG and electricity.

It recently announced a $2.2 billion partnership with Abu Dhabi Masdar focused on renewables in nine Asian countries and signed a 12-year nuclear power supply agreement with EDF. A P/E ratio of around 16 and a dividend yield of over 4% make TotalEnergies one of the most attractively valued oil majors in the market. Analysts at JPMorgan $JPM and Berenberg raised their price targets on the company in April.

Comparison to the broader market

The performance of these five companies significantly outperforms the broader market indices. While the S&P 500 is up around 4% year-to-date, all five companies in today's review are achieving double-digit appreciation.

The energy sector as a whole is one of the strongest sectors in 2026 thanks to geopolitical momentum, while the semiconductor sector is benefiting from a continued wave of investment in AI infrastructure and manufacturing capabilities.

Interestingly, this year's winners represent a rather unusual combination. Oil companies and semiconductor equipment manufacturers are not usually in the same group. This reflects the specific nature of 2026, when two independent but concurrent catalysts are shaping the markets: the energy crisis in the Middle East and the structural investment cycle in the semiconductor industry.

A strategic view

From an investor perspective, it is crucial to distinguish between cyclical and structural sources of growth for individual companies. ExxonMobil, Chevron and TotalEnergies have benefited primarily from the geopolitically driven oil shock, which is inherently temporary. If the Strait of Hormuz were to reopen or a diplomatic resolution to the conflict were to occur, oil prices could fall rapidly and with them the valuations of oil companies.

On the other hand, Applied Materials and Intel are benefiting from structural trends in the semiconductor industry that are longer-term in nature.

For energy companies, it is therefore important to monitor geopolitical developments and the real impacts on physical oil flows. Analysts warn that oil inventories are declining rapidly and if the Strait of Hormuz remains closed until mid-April, the world is in for a much harsher oil shock. For semiconductor companies, on the other hand, the pace of investment in new production capacity and the ability of companies like Intel to actually execute their plans remains a key factor.

What to watch next

• Developments in the negotiations between the US and Iran over the reopening of the Strait of Hormuz and its impact on oil prices.

• Q1 2026 earnings season: ExxonMobil and Chevron (April 24) and Intel (April 23).

• Intel's progress in securing external customers for the 14A process in the second half of the year.

• Developments in demand for HBM memory and advanced manufacturing processes that directly impact Applied Materials' order books.

• OPEC+ decisions on further production increases and market reaction to the potential release of strategic petroleum reserves.

Conclusion

The top 5 companies with the highest growth and market capitalization over $100 billion show how fundamentally the composition of market leaders can change in response to geopolitical and structural events. While just a few months ago the rankings were dominated exclusively by AI technology firms, today oil giants profiting from the energy crisis and semiconductor equipment manufacturers building infrastructure for the next generation of chips are coming to the fore.

At the same time, strong growth this year does not automatically make for an attractive entry point. For some of these stocks, the positive scenarios are already largely priced in, and a possible reversal in geopolitical developments or a slowdown in the semiconductor investment cycle could lead to a quick correction. The key to success therefore remains distinguishing between temporary catalysts and truly structural changes in individual business models.

For David Ellison, the head of Paramount Skydance and son of Oracle co-founder Larry Ellison, this solves the fundamental problem with the whole deal: a company with a market capitalisation of around $12bn is buying a business valued at more than $110bn including debt, in a package that includes an obligation to pay Netflix $2.8bn in severance for the cancellation of a previous deal. Without a massive injection of outside capital, it simply wouldn't have worked.

How the Gulf money is put together

According to leaks so far, the structure is supposed to look like this: the Saudi Public Investment Fund (PIF) will supply about $10 billion, the Qatar Investment Authority and the Abu Dhabi entity L'imad Holding the remaining roughly $14 billion. This money will be added to the equity of Ellison and RedBird Capital and debt financing from banks and private-equity partners such as Apollo Global Management.

Crucially, the Arab funds are to be formally non-principal. As structured, they enter with no voting rights, no board seats and no direct control over the combined firm. This structure has two functions:

• reassure regulators that this is not about "selling CNN to the Saudis" or about direct control of the media house

• and to allow the funds to share in the proceeds of a potentially very profitable consolidation of global media brands

Paramount therefore argues that the transaction should not come under the sharp scrutiny of the Committee on Foreign Investment(CFIUS) because foreign capital does not receive corporate control.

The political reaction: money yes, influence no - really?

Seven Democratic senators, led by Cory Booker, have already sent a letter to the FCC chairman expressing "deep concern" that billions of dollars from PIF, QIA and Abu Dhabi investors are helping to fund the takeover of Warner Bros. Discovery $WBD, including such brands as CNN and HBO.

Their argument is twofold:

• Even without formal voting rights, there is room for some power - through joint projects, content financing, future investments

• and the combination of capital from the Middle East region and from China (for example, through Tencent for other media assets) creates complex, sometimes competitive relationships that deserve more than a "superficial review"

The Senators therefore call for a thorough examination of whether such structured participation increases the risk that foreign governments may be indirectly involved in influencing editorial decisions and business priorities in sensitive areas - for example, just for CNN's news coverage or content licensing for markets where they have political interests.

As a result, even if the legal structure formally envisages 'money without a voice', the regulatory and political level may not agree. The FCC, DOJ, and potentially CFIUS will be under pressure to look at the scheme not only from an antitrust perspective, but also from a national security perspective.

Regulatory hurdles and timeline

The DOJ has already begun issuing subpoenas as part of its antitrust review of the transaction. That's standard for a deal of this size, but it suggests that everything won't just run on a fast track. Antitrust experts don't expect a ban yet - the combination of Paramount and WBD will indeed create a very strong player, but the market is still fragmented (Netflix $NFLX, Disney $DIS, Amazon $AMZN, Comcast $CMCSA, others).

Still, it's to be expected:

• A long approval process

• possible conditions (sale of some assets, commitments to continue open licensing of content, etc.)

• the involvement of European and Canadian regulators, who have already begun informally gathering input

Time is running out: the deal provides for a "ticking fee" - if the deal does not close by September 30, 2026, WBD shareholders will start collecting $0.25 per share for each additional quarter of waiting, which means a total of about $650 million per quarter extra for the buying party. Paramount thus faces a double whammy: convincing regulators and getting it done in time so the deal doesn't start eating into its own costs.

Why Paramount is doing this in the first place - and what the GCC investors are doing about it

From Paramount's perspective, the motivation is clear: to acquire Warner Bros. movie studios, the HBO and Max brands, CNN news and other assets in a single transaction and create a media group that can match Netflix or Disney in size and catalog. In an environment where streaming services are under margin pressure and where the price of talent and sports rights is rising, consolidating the library and brands makes strategic sense.

For Gulf funds, it is in turn part of a broader strategy to diversify the economy beyond oil and buy the global brands they bring:

• long-term cash flow from content, licensing and rights

• the opportunity to develop their own film and entertainment projects at home (studios, festivals, production)

• a reputation as a player who co-determines what the world watches

Just because they don't formally have voting rights doesn't mean they won't have a voice at the table when it comes to major co-productions, filming in the region or strategic investments.

Sector Technology – Software is currently taking a solid hit and most well-known names have dropped by tens of percent:

👉 And it's not just these big players — for example Duolingo ($DUOL) is about -45% year-to-date and is at the lowest levels since 2023.

👉 What's behind this?

- higher interest rates → pressure on growth stock valuations

- slowing growth at some companies

- but mainly FEAR OF AI 🤖

👉 The market is afraid that:

- AI could make software cheaper and reduce pricing power

- companies will lose part of their competitive edge

- some products will be partially or completely replaced

👉 Result? Investors are now often "selling everything," because it's unclear who will be the winners and who will be the losers.

👉 What is important to realize:

Most of these companies still have:

- strong cash flow 💰

- high margins 📊

- dominant market positions 🌍

Just look at $ADBE — where revenues are clearly rising! Still, the company is down by about 30% year-to-date.

📌 Question for investors: Is this the start of a bigger problem, or a typical correction that has historically created the best buying opportunities?

💬 Personally, I see these situations as moments when some quality companies reach more attractive valuations. Even so, I consider AI a huge risk for these companies and I tend to avoid most of them. For me, for example, DUOLINGO is misunderstood, as is UBER.

👀 Do you follow any of these companies? If so, which one and most importantly why that one? 👇

$CRM $UBER $INTU $FICO $PAYX $NOW $ADP

To me, this is one of the drawbacks of investing in ETFs: we're "forced" to buy overpriced companies, like SpaceX.

Personally, I'm also bothered by Tesla's large weighting in the S&P 500, but SpaceX's valuation is truly out of this world.

Crucially, the new commitment sits on top of a 2.9 billion dollar pledge from 2024, taking Microsoft’s announced investment in Japan above 13 billion dollars in a few years. Partnerships with Sakura Internet and SoftBank are meant to ensure that the added Azure capacity is physically located in Japan and architected to meet domestic rules on data residency and security, giving local enterprises and public agencies comfort that AI workloads can run at scale without leaving the country.

Why Microsoft $MSFT is focusing on Japan

Japan is a large and rich market, but one that has lagged behind the United States or Western Europe in some areas of digitization and cloud. This creates a strange combination - a lot of room for catching up, relatively high corporate and government purchasing power, and a strong emphasis on data security and sovereignty. By building local AI infrastructure, Microsoft is buying access to these future budgets. For the Japanese government, in turn, having a global technology partner that is willing to invest in-country and tailor services to local regulatory and security requirements is attractive.

The security dimension is also important. The package includes enhanced information sharing with the National Cyber Security Authority and joint projects aimed at protecting critical infrastructure. In the context of tensions in the region (China, North Korea, Russia), this is a signal that Microsoft wants to be seen as a strategic partner of the state, not just a private contractor. This opens the door to segments where contracts are long-term and barriers to entry for competitors are high.

Local partners: what role do Sakura Internet, SoftBank and the "big five" IT houses play

Microsoft will not build everything on its own. It announced that it will partner with datacenter provider Sakura Internet and SoftBank $SFTBY to increase Azure's computing capacity in Japan. It was a strong signal for Sakura, which is listed on the Tokyo Stock Exchange - its share price immediately jumped by around 20% as the market is pricing in the fact that some of the investment and new contracts will physically manifest themselves in its data centres and infrastructure.

SoftBank adds another layer. As a telecom operator and technology investor, it can connect its network, cloud capabilities and portfolio of AI firms with Microsoft's ecosystem. This is particularly interesting in the "we provide the complete solution" play - from connectivity, to infrastructure, to applications and services for end customers. From Microsoft's perspective, this provides a local "leg" that knows the market, the regulation and has its own customer base.

Another part of the programme targets human capital. Microsoft wants to train up to 1 million engineers and developers in cloud, AI and cybersecurity over several years in partnership with major Japanese IT firms - Fujitsu, Hitachi, NEC, NTT Data and SoftBank. This has a twofold effect: first, it will strengthen the overall skills of the country's workforce, and second, it will naturally "glue" this generation of IT professionals to Microsoft's tools and platforms. Those who learn on Azure, Visual Studio and Copilotech are less motivated to move to competing ecosystems.

What's in it for investors: Microsoft and select Japanese players

For Microsoft shareholders, the $10 billion move is a relatively small item given the size of the company, but it makes sense strategically. It helps the company:

• Strengthen the position of Azure and AI services in a region where it has not yet been as dominant as in the U.S.

• position itself as the preferred partner for the digitization of Japanese corporations (automotive, industrial, finance) and government

• and diversify growth outside the US market at a time when competition in AI among the "Big Three" (Microsoft, Google $GOOG, Amazon $AMZN) is intensifying

The risk is execution - whether Microsoft will be able to fill the new capacity quickly with real contracts and projects, and whether market conditions (regulation, security requirements) will inhibit the adoption of AI applications by more conservative Japanese companies. From a valuation perspective, however, the investment fits into a broader trend: Microsoft has been systematically building regional AI hubs around the world, and the Japanese package is one of the more visible steps in that mosaic.

On the Japanese side, three groups of companies in particular are worth watching.

• Sakura Internet, as a local datacenter partner, can benefit from higher demand for capacity, technology upgrades and reputational effect. However, the significant rise in the exchange rate also means that market expectations are high and any delay or weaker contract reality may lead to a correction.

• SoftBank, which has built its story on investments in technology and AI in addition to telecom services, can leverage partnerships in its investor communications and in specific digitization projects for corporate clients.

• Fujitsu, Hitachi, NEC and NTT Data will be at the forefront of integrating AI solutions and training - if Microsoft's program really gets the wave of projects rolling in the corporate and public sector, these companies may see increased demand for consulting and implementation services.

What other questions investors should ask

For the investor looking beyond the announcement itself, there are a few things to keep an eye on:

• What proportion of the $10 billion invested goes into physical infrastructure (datacenters, servers, networks) and what proportion goes into the "soft" components (training, security projects, software)

• how quickly specific large AI contracts will start to appear in segments such as car companies, banks, industrial conglomerates or government

• whether competitors (Google Cloud, AWS) will come up with similarly large programs, or whether Microsoft will manage to temporarily get a head start

Sometimes it's worth being patient with high-potential growth tech stocks, even though they come with increased volatility. These kinds of names often reward disciplined investors who wait for the right entry instead of trying to catch the exact bottom and top.

Over the past six months I had $BITF (Bitfarms) on my watchlist with a planned entry at 2.00 USD. After careful monitoring and recent price action, I opened a new position yesterday at 1.88 USD with a weight of 1% of the portfolio. My target price is set at 4.50 USD.

$BITF is among the largest publicly traded companies focused on Bitcoin mining. It operates large, energy-efficient mining farms in North America with an emphasis on sustainable energy sources.

Why I view $BITF as an interesting medium- to long-term investment:

Attractive valuation after the drop: The current price offers an entry significantly below my original target and reflects a reasonable valuation considering the mining capacity and plans for further growth.

Exposure to Bitcoin's long-term growth: As a pure-play miner, $BITF directly benefits from growing Bitcoin adoption, demand for network security, and potential price appreciation, without the risks typical for early crypto projects.

Operational efficiency and expansion: The company optimizes its hardware, secures cheap energy contracts, and increases its share of global hash rate, which positions it well even for the post-halving period.

High upside potential: Successful expansion and potential corporate catalysts could lead to a significant re-rating of the stock over a 12–24 month horizon.

Main downsides and risks:

High dependence on Bitcoin's price: A significant drop in $BTC would immediately impact mining economics and the stock price.

Energy costs and regulation: Rising electricity prices or regulatory changes (especially in key regions) can negatively affect margins.

Strong competition in the industry: Larger players with stronger balance sheets may gain an edge in efficiency and scaling.

Risk of the broader crypto market: Negative sentiment or regulatory interventions can cause severe declines across the sector.

I'm also watching other interesting names in this segment of tech infrastructure:

$IREN (Iris Energy) – planned entry 32 USD

$CIFR (Cipher Mining) – planned entry 11 USD

$NBIS (Nebius Group) – planned entry 76 USD

$CRWV (CoreWeave) – planned entry 52 USD

Overall, I view $BITF as an attractive way to gain leveraged exposure to Bitcoin's long-term growth and related technological infrastructure. At current prices, a small, well-managed position makes sense to me.

What do you think? Does $BITF make sense at these levels for you, or are you waiting for lower prices? What's your take on the other mentioned stocks?

Tickers mentioned in the post: $BITF $IREN $CIFR $NBIS $CRWV $BTCUSD

You can find the English version of this post on my eToro profile. If you'd like to follow me there or copy my USD portfolio, I'd appreciate it!

2026 confirms that gold is not just a relic of the past or a speculative asset for "old-fashioned" investors. The price of gold broke the psychological USD 5,000 per troy ounce mark in January this year and was in an extremely volatile range during the first quarter. It is currently trading around USD 4 700 per ounce, representing an appreciation of around 8 to 9% since the beginning of the year. Over the past 12 months, gold has delivered a total return to investors in excess of 55%.

Behind this development is a combination of several structural factors that have gradually strengthened over the past few years, reaching a critical point in 2026.

The position of the major banks

Central banks around the world continue to accumulate gold. They purchased more than 1,000 tonnes in 2025, the third consecutive year with such a high volume. Although the pace has slowed slightly in early 2026 (5 tonnes in January versus an average of 27 tonnes per month in 2025), the demand base is broadening. Traditional buyers such as China and Poland are joined by Malaysia and South Korea, which have increased their gold reserves for the first time in many years.

Geopolitical instability, particularly the conflict over Iran and the closure of the Strait of Hormuz, together with the ongoing war in Ukraine, is creating an environment in which institutions are looking for assets outside the reach of financial sanctions. Another important factor is the weakening confidence in the US dollar as a reserve currency. Rising US government debt, uncertainty around the future direction of the Federal Reserve and the overall fragmentation of the global financial system are leading to a structural shift of capital towards gold.

Major investment banks such as J.P. Morgan $JPM, Goldman Sachs $GS and Wells Fargo $WFC have set their gold price targets for the end of 2026 in the range of USD 5,400 to 6,300 per ounce.

Access to ETFs

For retail investors, however, there is no such thing as gold. How you gain exposure to this precious metal can fundamentally affect your portfolio's return, volatility and overall risk profile. Physical ETFs tracking the price of gold behave differently than funds investing in mining companies. The following four ETFs represent four fundamentally different approaches to the same asset.

SPDR Gold Shares $GLD

SPDR Gold Shares is the oldest and largest physically-backed gold ETF in the U.S. The fund was launched in November 2004 and has since become the de facto standard for investors seeking direct exposure to the price of gold. Each share of the fund represents a fractional ownership of physical gold that is stored in secure vaulted storage. The fund does not invest in futures contracts, shares of miners or other derivatives.

With assets under management in excess of $155 billion (as of March 31, 2026), GLD is by far the most liquid gold-focused ETF in the world. For institutional investors and large traders, this means extremely tight bid-ask spreads and an established options market that allows for sophisticated hedging strategies. The fund's expense ratio of 0.40% per annum is higher than some competitors, but the liquidity premium the fund offers often offsets this disadvantage for large positions.

Performance and market position

Since the beginning of 2026, $GLD has appreciated approximately 8%. Over the past 12 months, then, the total return exceeds 54%. The fund has moved within a wide range during the first quarter, reflecting the extreme volatility of gold prices during this period. GLD is ideal for investors who want clean and transparent access to gold without the operational risk associated with mining companies. It is also the preferred choice for short to medium term trading and hedging positions.

iShares Gold Trust $IAU

BlackRock 'siShares Gold Trust is the second largest physically backed gold ETF on the market. The fund was launched in January 2005 and is virtually identical in structure to GLD. Each share represents a stake in physical gold stored in vaults managed by JPMorgan Chase $JPM. From an investor's perspective, it is an alternative approach to gold that differs from GLD primarily in the cost structure and size of each share.

The main differentiator of IAU is the lower expense ratio of 0.25% per annum. The 15 basis point difference from GLD may seem small, but this cost differential accumulates over a long term holding period. For an investor who holds a position in gold for five or more years, the lower expense ratio of IAU can provide a measurably better total return. Assets under management of the fund are approximately $71 billion, making IAU the second largest gold ETF in the world and a very liquid instrument.

IAU's performance is virtually identical to GLD because both funds track the same underlying asset. YTD appreciation is approximately 8.1%, with a total return over 12 months in excess of 53%. IAU is an ideal choice for cost-conscious investors who plan to hold a position for the long term and do not need maximum depth in the options market. The lower price per share (IAU trades at approximately one-fifth of GLD's price) also makes it easier for smaller investors to gradually build a position.

VanEck Gold Miners ETF $GDX

The VanEck Gold Miners ETF represents a fundamentally different approach to gold exposure. Instead of holding the physical metal, it invests in shares of companies that mine and process gold. The fund was launched in May 2006 and tracks the MarketVector Global Gold Miners Index. The portfolio currently holds 57 titles, dominated by the world's largest gold mining companies.

The largest positions include Agnico Eagle Mines $AEM (around 12% of the portfolio), Newmont (11%), Barrick Mining $B (7.5%), Franco-Nevada $FNV (around 5%) and AngloGold Ashanti $AU (around 5%). The fund is sector concentrated exclusively in the basic materials segment, with Canada dominating geographically with around 55%, followed by the US and Australia. The fund's expense ratio is 0.51% per annum and AUM is approximately $29 billion.

A key feature of GDX is its higher beta to the gold price. When the price of gold rises, miners' shares typically rise faster as higher commodity prices are more reflected in margins and profits when mining costs are relatively fixed. But this effect works both ways. During the March gold selloff, miners' shares depreciated much faster than the metal itself. The fund's five-year monthly beta against the S&P 500 is about 0.66, but volatility against gold is significantly higher.

The fund's YTD appreciation is around 10%, but over the past 12 months, miners' stocks have delivered a total return in excess of 107%. In 2025, the fund posted over 144% appreciation in a single year. GDX is an option for investors who believe in a continuation of the upward cycle in gold and want to gain increased exposure while accepting higher risk.

WisdomTree Efficient Gold Plus Gold Miners $GDMN

WisdomTree Efficient Gold Plus Gold Miners Strategy Fund is the youngest and most aggressive fund in today's selection. It is an actively managed ETF launched in December 2021 that uses the principle of so-called return stacking. For every $100 invested, the fund allocates approximately $90 to a basket of global gold mining stocks (weighted by market capitalization) and another $90 to gold through futures contracts. This leaves approximately USD 10 as cash collateral. This results in a total exposure of USD 180 for every dollar invested, which is equivalent to approximately 1.8 times leverage.

This structure allows investors to gain simultaneous exposure to both physical gold and miners in a single instrument without having to sell other positions in the portfolio. WisdomTree presents the fund as an alternative to traditional leveraged ETFs on miners because its leverage comes not from daily resets (as with 2x or 3x leveraged funds) but from the structural layering of the two exposures.

The results are accordingly. In 2025, the fund attributed a whopping 239%. Since the beginning of 2026 (through the end of February), it is up about 42%. But this return comes with corresponding volatility. The fund's annual price range ranges from approximately $30 to $147, illustrating the extreme amplitude of the movements. The fund's AUM is approximately $231 million, significantly less than the other funds in today's comparison, and the expense ratio is 0.45% annualized.

GDMN is clearly designed for experienced investors with a high tolerance for volatility. In a strong gold growth environment, this fund significantly outperforms other gold funds. However, in the event of a correction, losses can quickly reach tens of percent. Moreover, the low liquidity (average volume of around 87,000 shares per day) and the limited size of the fund bring additional risks for larger positions.

Comparison of the four ETFs

Strategic view

Choosing the right gold ETF is not a question of which fund is objectively better, but of what role gold should play in a given portfolio. For a conservative investor who wants gold as a long-term hedge against systemic risk, GLD and IAU are the logical choices. IAU offers lower costs with an almost identical structure, while GLD excels in liquidity and a broader options market.

GDX makes sense for investors who believe in the continuation of the commodity supercycle and want to leverage the operating leverage of mining companies. In an environment where the gold price is rising and mining costs remain relatively stable, shares of miners can generate significantly higher returns than the metal itself. However, they also carry operational risk, political risk (many mines are located in volatile regions) and cyclical risk associated with the industry's investment cycle.

GDMN is designed for investors who are betting on gold's continued growth and who want to maximize their exposure without having to use traditional daily leveraged funds. The capital efficiency of this fund allows it to preserve other positions in the portfolio, which is an interesting feature from a portfolio construction perspective. However, it should be clearly understood that the 1.8x leverage works both ways and during a correction the fund can lose value significantly faster than the underlying assets.

What to watch next

• Fed rate developments and potential change in leadership: the nomination of Kevin Warsh to replace Jerome Powell as head of the Federal Reserve adds further uncertainty to the market. Any move towards more aggressive monetary easing would be positive for gold.

• Geopolitical developments: the conflict over Iran, the closure of the Strait of Hormuz and the ongoing war in Ukraine remain key factors for demand for safe-haven assets. Any easing of tensions could trigger a short-term correction in gold prices.

• Central bank purchases: the broadening base of buyers (Malaysia, South Korea, other emerging economies) signals a structural change in access to reserve assets. If this trend is confirmed, gold may receive further long-term support.

• Inflation Expectations and Real Yields: If the energy shock caused by the closure of the Strait of Hormuz translates into higher inflation, real US Treasury yields may remain negative or low, which has historically been strongly supportive of the gold price.

Summary

The current global gold market environment is creating conditions that have not been seen in this combination for decades. Structural central bank demand, geopolitical fragmentation, questions around the future of the dollar system, and record levels of government debt are all combining to support the long-term case for gold as an asset class. At the same time, however, it is important to remember that after a period of very strong growth, a correction may come, which will be significantly different for different types of exposure.

Each of the above offers a different risk-return profile, a different cost structure and a different sensitivity to the various factors affecting the gold price, so each ETF is suitable for someone else and neither may fit into some portfolios.

In January 2026 the balance sheet was rewired to match those ambitions, with roughly 960 million dollars of fresh capital lifting the cash pile to about 1.5 billion dollars and giving management room to chase a minimum 375 million dollars of revenue this year – more than seven times 2025 – without worrying about liquidity. The market, however, still largely files the name under “small wireless gear vendor”, which creates a gap between label and reality that a patient investor can exploit if the company executes on turning its border contract, Blue UAS status and Palantir‑powered data stack into a repeatable, high margin autonomous‑drone business.

Top points of the analysis

• Revenue jumped from about $7.2 million in 2024 to $50.7 million in 2025, up about 605 percent, with the fourth quarter of 2025 alone bringing in $30.1 million - about 60 percent of full-year revenue.

• The Autonomous Systems (OAS) division increased revenue from about $5.3 million to $49.7 million - an increase of about 840 percent in one year.

• Holds the only FAA type certification for a small UAS designed for autonomous safety and data collection

• Has the only FAA-approved drone system to fly beyond visual line-of-sight without a human operator on site

• Secured a government contract to deploy thousands of autonomous drones across state borders

• Signed a strategic partnership with Palantir

• The counter-UAS systems market is estimated to grow from roughly $6.6 billion in 2025 to more than $20 billion by 2030.

What Ondas does and what it makes money from

Ondas Inc $ONDS is now a holding company for two main parts: Ondas Autonomous Systems (OAS) and Ondas Networks. Ondas Networks was the original business - industrial wireless networks for railroads and critical infrastructure. But most of the investment story has shifted to the OAS division, which aims to deliver governments, militaries and critical infrastructure operators a complete autonomous defense system against drones and other threats.

The core of revenue today is a combination:

• Systems sales (drones-in-a-box, anti-drone systems, ground robotics, aircraft protection equipment).

• services and support (installation, training, maintenance, upgrades)

• Software and data services - control software, integration with AI and analytics platforms (Palantir $PLTR) and, over time, mission data output

For the customer, Ondas is not a "single drone supplier" but an integrator to deliver a whole suite of technologies, certifications, processes and data matched to actually reduce the risk of a drone attack to an acceptable level at a particular airport, power plant or military base.

Why is this company different than the typical drone title

There are plenty of companies on the market that can make a drone, radar or jammer. Ondas is different in a few hard-to-copy ways:

• The company says it is the only player to have full FAA type certification for the Optimus small autonomous system, designed for continuous patrol and data collection without a pilot

• is the only system in the U.S. to be FAA-cleared to fly beyond visual line of sight(BVLOS) without a physical operator on site, a major advantage for civil airspace

• won a government contract to deploy thousands of autonomous drones along the nation's border, an infrastructure, not just a test project

• increased autonomous systems sales by 840 percent and company-wide sales by more than 600 percent in one year

• in 2025-2026, going from "one company" to a conglomerate of nine acquisitions that form a complete multi-domain defense stack, from a drone in the sky to a robot on the ground.

For a long time, Ondas has been looked at as a cheap "micro-cap wireless provider," which explains why the stock was still trading below one dollar in 2025 before rising above $11 within a year.

Products and ecosystem - six layers of a defensive stack

To make sense of what Ondas is building, it needs to be viewed not as individual products, but as six layers to address the entire drone threat lifecycle - from detection to analysis.

1) Guardian: Optimus - a drone in a box

Optimus is a fully autonomous system that looks like a larger container with a charging station and a backroom for the drone. Installing it on an object means:

• the drone takes off on its own on schedule

• flies along a defined route or according to events

• collects imagery and other data

• lands, charges and repeats the cycle, without a pilot on site

This system is no longer just on paper - it's deployed in the UAE and with customers in the Middle East, Europe and elsewhere.

Two strong differentiators:

• FAA type certification - four years of approval and testing

• BVLOS permit without operator - additional years of operational data collection

This is a regulatory lead measured in years rather than months.

2) Raider to Raider: Iron Drone

Raiderhttps://www.youtube.com/embed/m4it_Z6HXLk?rel=1

The Iron Drone Raider is a fully autonomous anti-drone system. Once the detection system confirms an enemy drone, the Raider:

• automatically takes off

• ascends to the target

• launches a reusable net that physically intercepts the drone

• and then safely lands

Speed is key - human decision-making is often too slow with high-speed drones. Raider is designed to reduce the time from detection to neutralization to a minimum. In November 2025, Ondas won two consecutive contracts of $8.2 million each from a major European security agency, and each involved a different airport - no longer a pilot project, but the start of an infrastructure.

3) Cyber Layer: Sentrycs

It is not always appropriate to "take down" a drone physically. Over a crowd, expensive or critical technology, a silent intervention is better. Sentrycs, an Israeli firm that Ondas is buying for about $225 million, offers a "Cyber-over-RF" approach - the system takes over radio control of the drone, redirects it, and perches it safely.

The benefits:

• No interference with other systems (unlike crude jammers)

• no debris and collateral damage

• data retention in the drone - it can be traced back to where it came from and who was controlling it

Sentrycs has already been deployed in live operations, including protecting airspace at the World Economic Forum in Davos in 2026 - a reputationally extremely strong reference.

4) Ground Robotics: Roboteam and Apeiro

Ondas doesn't stop at the air. Roboteam, an Israeli manufacturer of tactical ground robots, supplies machines for explosive ordnance disposal, reconnaissance and urban operations that are used by militaries and security forces in more than 30 countries. Apeiro Motion complements other ground-based systems and tethered drones powered from the ground.

This allows Ondas to offer a response not only in the air but also on the ground - for example, to send a robot to secure a downed or captured drone, or to reconnoiter the area from where the threat came from.

5) Gateway to the U.S. military: Mistral

Mistral, whose acquisition for about $175 million was announced in March 2026, is a prime contractor for the U.S. military and special forces. It manages over $1 billion in existing contracts, mainly in weapons and unmanned platforms.

Crucially, Ondas also buys so-called IDIQ contracts with Mistral - framework contracts through which the US military can repeatedly buy without a full tender. While it takes years for a new player to get into such a mode, this move gives Ondas a shortcut to the world's largest defense budget.

6) Brain: Palantir

The final layer is data. The Palantir Foundry is meant to be the nervous system that connects stratospheric systems, Optimus drones, anti-drone systems, Sentrycs, and ground robots into one operational view.

Simply put: data from radars, RF sensors, cameras, drones and robots converge into Palantir, which helps:

• classify threats

• design responses

• evaluate incidents and learn the system for next time

For Ondas, this means a twofold advantage - technological (an AI layer it would take years to build on its own) and distribution (Palantir has relationships with virtually every major defence ministry and intelligence agency in NATO).

What airport deployment looks like in practice

Imagine an international airport that has had to shut down several times because of hobby drones costing a few hundred dollars - one stopped hour means millions of dollars in losses. At such an airport, Ondas installs Optimus stations along the perimeter that autonomously take off at regular intervals and patrol the runways and surrounding area.

At 2am, the system detects a drone entering the restricted area. The AI evaluates its trajectory and signal as suspicious - at that moment:

• Sentrycs will attempt to take control of the drone and guide it to a safe landing in the defined zone.

• if the takeover fails, the Iron Drone Raider takes off, climbs to the target and catches the drone in its net

• Roboteam robot goes to the landing site and secures the machine for analysis

The entire incident is logged into Palantir to evaluate the drone's behavior, possible origin and improve the detection model for the next time. The airport doesn't have to stop traffic, no one has to shoot, and the security team has a report on their desk in the morning, not just a "what the hell happened last night" question.

Why it's hard to replicate

Trench 1: The control wall

It took four years to get FAA certification for the Optimus System. BVLOS exemptions required a multi-year process of submitting operational data to prove the safety record of the system on a large scale. No competitor in the U.S. currently has equivalent certifications.

Why is this so important? Because government contracts (the primary market for this product) often require certified systems. Without FAA certifications, a competitor cannot bid for the same contracts Ondas is winning today. They would have to start the regulatory process from scratch.

Estimated time to replicate these certifications: at least three to five years, assuming the competitor already has a comparable product and starts the process today. Most won't.

AeroVironment (AVAV), the closest publicly traded competitor, has a deep relationship with DoD and a long history with tactical drones. But AeroVironment's products were designed for military operators, not for fully autonomous, unmanned commercial and security applications. Their certification architecture is different. Their BVLOS approval is different. They build in the same space, but from a different starting point.

Moat 2: System of Systems Architecture

One company can be outperformed. System of systems players are fundamentally harder to replace.

When the government buys an Ondas stack, it doesn't buy a drone. They are buying a complete autonomous security infrastructure: aerial surveillance, kinetic interception, cyber interception, ground response, AI-driven command and control. Each component reinforces the others. Replacing one component requires the exchange of interfaces, data protocols, training, maintenance relationships, and regulatory approvals.

This is a switching cost that is measured not in dollars but in years of program disruption.

Red Cat Holdings (RCAT), the most aggressive small-company competitor in this area, focuses primarily on ISR drones and has strong tactical drone capabilities. But Red Cat doesn't have a capability against UAS. It doesn't have a ground-based robotics stack. It doesn't have a Sentrycs cyber layer. It builds one dimension to the problem. Ondas is building a complete stack.

Moat 3: Combat Proven Operational Record

Iron Drone Raider has been deployed in Israel, UAE, Europe and several other areas. Sentrycs has operated in active defense environments. Roboteam platforms are used by military forces in 30 countries.

In defense acquisitions, operational track record is the currency. When the government evaluates a system against drone settlements, it is not just buying technology, it is also managing risk. A system with a documented operational history in GPS-denied environments, adverse weather, and complex airspace has a credibility advantage that cannot be produced in a laboratory demonstration.

Management

The transformation of Ondas is being led by Eric Brock, who combines the roles of CEO and Chairman of the Board.

Under his leadership, the company:

• made nine acquisitions in just a few quarters - Sentrycs, Roboteam, BIRD Aerosystems, Apeiro, Mistral and others

• Completely pivoted the portfolio from "wireless networking and commercial drones" to a multi-domain defense stack

• Raised nearly $1 billion in capital in January 2026, bringing cash on hand to about $1.55 billion

Brock openly says 2025 was a watershed year - a transition from a portfolio of technologies to a "scaled operating platform" with a growing backlog and global demand for integrated autonomous systems. But this also brings a downside - massive dilution: the number of shares jumped from tens to hundreds of millions in just a few years.

So the key question for investors is whether this capital allocation will look in retrospect like a smart bet on a new defence standard, or an overpriced roll-up.

The numbers: growth, losses, cash and balance sheet

Revenue and margin growth

• 2022: sales of about $2.1 million

• 2023: 15.7 million (more than six times)

• 2024: drop to 7.2 million (trimming old parts, rebuilding)

• 2025: 50.7 million, a growth of about 605 percent

The OAS division jumps from 5.3 million to 49.7 million, about 840 percent. Gross margin from five percent to about 40 percent - a major signal that the size of contracts is already starting to tip the fixed cost-to-revenue ratio.

Losses and cash burn

Operating costs rise from about 35 million in 2024 to 78.5 million in 2025, so the operating loss is about 58 million, with a net loss of $137 million. Adjusted EBITDA is around minus 32 million, roughly minus 9-10 million in Q4 2025 - so the company is still in a cash-funded growth phase.

Management is targeting product profitability around Q3 2026, segment profitability in OAS sometime in 2027, and company-wide profitability after that - these are targets, not certainties.

Cash and debt

Ondas has $594m in cash at the end of 2025, and has raised around $960m in new capital in January 2026, for a total of around $1.55bn. Debt is negligible, with a debt-to-assets ratio of a tenth of a percent, and an Altman Z-score above 4.

In other words, the risk of "running out of money" is currently low, the real risk is "under what conditions will capital be able to appreciate".

Market and competition

The counter-UAS market is set to grow from roughly $6.64 billion in 2025 to $20.31 billion in 2030, a rate of around 25 percent per year, according to MarketsandMarkets. The drivers are:

• Experience from Ukraine and other conflicts where low-cost drones have fundamentally changed tactics

• incidents at airports like Gatwick, Frankfurt, Schiphol, where hobby drones have managed to shut down operations

• pressure to protect critical infrastructure and large events

Ondas stands between here:

• Big weapons companies (Lockheed, Northrop, RTX) - they have the resources and relationships but are slower and often solve the "2035 problem", not the "2026 problem"

• specialized firms like D-Fend Solutions, Dedrone - strong point solutions (cyber, detection) but lack system-of-systems

• Smaller drone companies (e.g. Red Cat) that are strong in ISR and tactical drones but do not have a cyber layer, ground robotics or full integration

Ondas is trying to lean on that:

• has a regulatory edge (FAA certification and BVLOS)

• has an air-to-ground-to-cyber combination

• has references from real deployments (airports, Davos, Middle East)

• and through the Mistral can gain passage into US contracts

Valuation and what the market sees in it today

According to the data, with a market capitalization of around $4.9 billion and revenues of $50.7 million in 2025, Ondas is trading at roughly

• price to sales around 97 times

• price to book over 11 times

• negative P/E (loss-making company)

That in itself looks brutally expensive.

Context:

• If the company actually delivers 375 million in sales in 2026, the P/S would fall somewhere near 13 times - still expensive, but already comparable to other fast-growing defense-tech names

• Fair value of around $4.9 from a conservative screener based on historical numbers and punishes the firm for extreme volatility and losses, hence the low "valuation level"

• Reality is somewhere in between: the market is paying for ambition and stack today, but hasn't yet fully priced in a 10-year "defense infrastructure" scenario - for that, Ondas must first tap into hundreds of millions in stable revenue and better margins

Investment scenarios

Baseline scenario

• Revenues reach $170-200 million in 2026, integration of acquisitions is slower than the $375 million target would suggest.

• Gross margin stays around 40 percent, operating expenses grow a bit slower, losses shrink relatively, but the company is still in the red.

• Cash burn in the tens of millions a year is manageable due to cash, the market is gradually moving the multiples down (to 8-12x earnings, for example), so the return to shareholders is more a combination of modest price and valuation growth plus potential "rerating" if a clear path to profit is shown.

Growth scenario

• The integration of Sentrycs, Roboteam, BIRD and Mistral is running smoothly, revenues approach $300-375 million, backlog grows faster than reported revenues and passes $150-200 million during 2026.

• Gross margin moves above 40 percent, operating losses decline, EBITDA approaches zero in 2027.

• The market is willing to pay 15-20x expected revenues as Ondas is seen as a benchmark player in counter-UAS, similar to Palantir in defensive analytics - in which case the return to shareholders may be in the multiples, but it is a high execution scenario.

A cautious scenario

• Integration is harder than management expects - different cultures, systems, contract delays, Mistral brings in fewer contracts than thought.

• Revenues only grow to 100-150 million, margins stagnate or fall, market loses patience, revenue multiple falls into the 4-8x range.

• For the investor, this means that much of the "premium" valuation evaporates and the return is more about patience than a quick rerating.

Growth catalysts

• The accelerating wave of drone incidents - airports, infrastructure, borders - is making counter-UAS a priority for governments, not a "nice to have".

• Increasing NATO and allied budgets, shifting money from conventional platforms to asymmetric defense and counter-drone systems.

• US grants and programs to protect major events (World Cup, World Cup, national celebrations) create a specific domestic demand where Ondas has a head start thanks to the FAA.

• Partnering with Palantir increases Ondas' visibility with customers Palantir already works with - at the point where projects need autonomous drones and robotics, Ondas is a natural candidate.

Risks

• Extreme dilution - if the story stalls and the company nevertheless continues to fund growth through acquisition, existing shareholders continue to bear the cost of a series of issues without a corresponding increase in value per share.

• Ambitious targets - the jump from 50.7 million to 375 million in revenue in two years is extremely aggressive even in the context of roll-up strategies.

• Integration risk - bringing together nine companies that each had their own processes, culture and technology is difficult in such a short space of time; any major "stumble" will impact margins and reputation.

• Competition from large arms companies - their entry or acquisition of a similar stack can change the balance of power within a few years, especially in the US.

• Regulatory uncertainty - BVLOS permits are an advantage today, but a change in the regulatory framework (e.g., the FAA's new Part 108) may rewrite the rules of the playing field.

What the investor will take away

• Ondas today is more of a small "drone defense Palantir" than the former wireless network for industry - it composes a complete stack for autonomous defense against drone and ground threats and has a fundamentally unique combination of regulation, credentials and capital.

• The numbers are extreme on both sides: revenues are up hundreds of percent, margins are up, but losses and dilution are high, making Ondas a typical high-risk, high-reward bet.

• If you're looking for a second Palantir in a different corner of the defense market, Ondas may be a candidate - but more likely for a smaller, risk-conscious position with a longer horizon, where the path is not expected to be straight.

In practice, the pressure will fall heaviest on companies that depend on high value finished imports rather than on those that already have deep US manufacturing footprints. Multinationals that ship patented drugs from Europe or Asia, speciality steel and aluminium producers using offshore mills, and import‑heavy distributors with thin margins are at the front of the line, while integrated US based players with pricing power may find themselves in a stronger negotiating position if rivals are suddenly forced to rework supply chains.

1) Pharma: 100% on paper, zero for those who comply

The administration formally imposes a 100% tariff on imports of patented drugs and active ingredients, but at the same time opens a wide path to a 0% tariff.

Companies can avoid the duties if:

• they sign an agreement with the government to build new manufacturing capacity in the U.S. within 4-6 months; and

• agree to the Most-Favored-Nation (MFN) pricing scheme for selected segments (e.g. Medicaid, cash payments).

Other exceptions apply to countries that have trade agreements with the U.S. - EU, Japan, South Korea, Switzerland, UK - where base rates are significantly lower (15% or less) and typically also conditioned on MFN agreements. Generics and biosimilars remain out of scope for the time being, the duty does not apply to them and the situation is not due to be reviewed for another year.

Where is the real impact:

• Companies that already have an MFN and on-shoring agreement (typically Pfizer $PFE, Eli Lilly $LLY, AstraZeneca $AZN) will have a 0% tariff while fixing price and manufacturing commitments in the US.

• Manufacturers who dither will risk rates of 20-100% and will be under intense pressure from investors and the government to get on board.

The most exposed big players:

• Global big-pharma with a large share of patented drugs manufactured outside the US - Johnson & Johnson $JNJ, Novartis $NVS, Roche, Sanofi $SNY.

• Contract manufacturers and active ingredient suppliers in Europe and Asia, who will be forced to decide whether to invest in US pipelines or accept that US supply will be less competitive.

For investors, this means: negotiating noise in the short term rather than an immediate bump in P&L, but real production reshuffling and price pressures in the US in the medium term. Those with US production and MFN agreements will be privileged in the domestic market.

2) Steel, aluminum, copper: 50% stays, but duty is calculated differently - cargo will go up

Formal tariffs on steel, aluminum and copper remain at 50%, but the way they are calculated is changing: duties will now be based on the US spot price, not the value declared by the importer. The aim is to prevent under-invoicing and "optimization" of duties. The administration itself admits that it expects a higher volume of duties actually collected.

At the same time, the rules for products containing metals are being clarified:

• products with less than ~15% steel/aluminium/copper will be exempted from metal duties (only normal duties apply)

• products "substantially" made of metals will now typically fall under the 25% band on derivatives, while pure metal products remain at 50%.

Sectors and companies most affected:

• Construction and Infrastructure: producers of structural steel, pipe, sheet metal - U.S. Steel, Nucor $NUE, Steel Dynamics - may benefit from more expensive imports and greater domestic market protection.

• Automotive and white goods: Ford, General Motors, Stellantis, Whirlpool, Electrolux; higher import costs for steel and aluminum will put pressure on margins unless everything can be reflected in final prices.

• Electrical and engineering: Caterpillar, Deere, Siemens Energy (for equipment for the US market), cable and distribution equipment manufacturers, where copper is a significant part of inputs.

In practice, this means: it is more difficult for importers to circumvent tariffs, and for many products the effective duty moves closer to the nominal rate. Domestic metal producers and some construction firms may be relative winners, while industrial metal buyers will feel the cost pressure.

3) Construction and infrastructure projects: more expensive materials, uncertain contracts

A fixed 50% tariff on basic metal products, 25% on derivatives and 15% on some industrial equipment means that:

• large infrastructure projects (bridges, railways, energy) will have to rely more on domestic suppliers

• imports of some more complex equipment will become more expensive if metal content exceeds key thresholds.

Impact on companies:

• Domestic suppliers of building and structural components - e.g. Martin Marietta $MLM, Vulcan Materials $VMC (more aggregates, but often bundled with steel), regional steel mills - benefit from a more competitive pricing position relative to imports.

• Large EPC and construction giants - Fluor, Bechtel (non-traded), Jacobs $J or KBR - will need to focus more on optimizing supply chain and contract terms so that tariffs don't destroy margins on fix-price contracts.

What matters to investors is who has long-term contracts with the ability to pass on costs to the client, and who is locked into fixed-price contracts and will "swallow" the duties on their margin.

4) Retail and logistics: duty drawbacks, planning uncertainty

In parallel with the new tariffs, the government is launching a scheme to recover some $160 billion in unjustified duties collected from the already abolished IEEPA tariffs. Some 25,000 importers, including big players such as Costco and FedEx, are waiting for refunds that could improve their cash flow, but the system is rolling out gradually and payments can take up to 45 days from the approval of a claim.

Impact by company type:

• Large retail importers (Costco $COST, Walmart $WMT, Target $TGT, Home Depot $HD) - they will get a one-time boost from duty refunds, but at the same time, the new structured tariffs on metals and possibly other categories complicate their cost and pricing planning.

• Logistics and courier companies (FedEx $FDX, UPS $UPS) - refunds will help improve the balance sheet in the short term, but the overall complexity of customs schemes increases administrative costs and the risk of declaration errors.

5) Geopolitics and next steps: China, new investigations and the threat of further tariffs

The current wave of changes comes along with two new Section 301 investigations that target "structural overcapacity" and other practices in roughly 60 trading partners. The result may be new country-specific tariffs similar to those that previously fell under IEEPA, only now under a different legal heading.

China has launched its own investigation into US practices, and there is a real risk of further escalation of reciprocal tariffs ahead of the Trump-Si meeting in May. Pro:

• Exporters to China (e.g. manufacturers of manufactured goods, agricultural commodities - Deere, Caterpillar, agribusinesses like Archer Daniels Midland), this means an additional layer of uncertainty,

• global chains across automotive, electronics or engineering, this increases pressure to diversify production outside China and the US (Mexico, South East Asia, Europe).

What does this mean for investors

• For pharma, who has MFN and on-shoring agreements is more important than the nominal 100% rate itself. Companies with U.S. manufacturing and agreements will be relative winners; late responders may come under fire in the short term.

• For metals and industrials, expect more of an effective load, even if rates don't formally change. Domestic metals producers and some construction firms will be more protected, but industrial buyers will take another cost hit.

• For retail and logistics, duty drawbacks bring short-term relief, but the overall tariff regime is becoming more complex and less predictable, increasing the risk to long-term planning.

Personally, I definitely find selling a winning position more difficult, because I'm never sure it's the right time and I keep thinking I might hold a little longer since the uptrend may not be over. Paradoxically, a losing position bothers me less, because if the fundamentals haven't changed, I see it as a buying opportunity and expect to keep the shares in my portfolio for a long time.

Putting together five consecutive months of rising volumes matters in a market as brutal as China’s EV space. It suggests that recent pricing, refreshed trims and marketing tweaks are stabilising demand, even if Tesla is a long way from the dominance it once enjoyed and must now fight for every incremental sale against domestic brands that keep launching cheaper, feature rich models.

Sales rose to 85,670 vehicles after a seasonal lull

Figures from the China Passenger Car Association show that $TSLA sales include Model 3 and Model Y vehicles destined for both the domestic and export markets. The March figures also reflect a rebound after a seasonal downturn around Chinese New Year, when production traditionally drops due to a two-week break.

For the entire first quarter of 2026, Tesla sales in China grew 23.5% year-over-year, a significant acceleration from just 1.9% growth in the previous quarter. The growth was boosted by a recovery in demand in Europe, where Tesla exports a significant portion of its vehicles from its Shanghai plant.

BYD posted a 20.5% drop despite the recovery

Tesla's main Chinese competitor, BYD $BY6.F, is facing challenges. BYD sold 300,222 electric vehicles in March, up 57.85% from February but down 20.45% year-on-year. This is the seventh consecutive month with year-over-year sales declines.

China's EV market remains in a price war, with manufacturers fighting for market share through discounts, putting pressure on profit margins. BYD disclosed last week that its 2025 net profit fell 19% due to a prolonged domestic price war.

Competition in the Chinese market is gaining momentum

Tesla's share of the Chinese EV market has fallen to 8% from 10% last year in 2024. Competition from domestic brands has intensified significantly, with Xiaomi's YU7 SUV dethroning Tesla's Model Y as the best-selling car in China in January and Geely's Xingyuan becoming the best-selling model in February.

Xiaomi's SU7 sedan, which starts at around $33,000, competes directly with Tesla's Model 3 and outsold it in sales in China in just a few months of 2025. The YU7 SUV launched in mid-2025 for less than $50,000 targets the Model Y.

European market provides relief for Shanghai plant

Registration data points to improving demand across several European markets. In France, Tesla registrations rose 203% year-on-year to 9,569 vehicles in March, just below the December 2023 record of 9,572 units.

Norway saw a 178% increase to 6,150 vehicles, while Sweden and Denmark saw growth of 144% and 96% to 1,447 and 1,784 units, respectively. In other markets, registrations rose 72% in the Netherlands to 1,819 vehicles and 25% in Spain to 2,477, according to RAI and ANFAC data.

The strategic pivot away from EVs continues

Tesla is expanding its focus beyond EVs. The company is positioning solar power, humanoid robots and autonomous robotaxis as future growth engines. In China, Tesla achieved a breakthrough in early 2026 when it received a full-fledged local FSD training permit, allowing it to compete more effectively with local tech rivals like Xiaomi and Huawei.

After a challenging 2025, which saw its first major annual decline in shipments, Tesla is now betting on a trio of catalysts: the mass Model 2, the commercialization of the humanoid robot Optimus, and the regulatory rollout of fully autonomous driving.

While Chinese-built car sales and European demand are showing signs of recovery, Tesla's near-term performance will continue to be closely tied to its upcoming delivery dates and its ability to navigate increasing competition across key markets.

For income‑focused investors, the appeal is a combination of oligopoly economics and a shareholder friendly capital return policy rather than headline growth. The dividend yield sits in the 2.3–2.6% range with a history of high single digit annual growth, and total shareholder yield climbs above 4% when buybacks are included, all backed by a balance sheet and asset base that AI cannot disrupt and that should make incremental efficiency gains from automation over time. The crux is whether the current multiple already fully prices this durability and cash return profile, or whether the market is still treating UNP too much like a cyclical volume story and too little like a long term, infrastructure‑backed dividend compounder.

Top points of the analysis

• Union Pacific posted 2025 revenues of $24.9 billion (+2% YoY), net income of $7.1 billion (+6% YoY) and diluted EPS of $11.98 (+8% YoY), marking a best-ever year from an earnings perspective.

• The dividend story is strong: an annual dividend of $5.52 per share, a yield of around 2.5-2.6%, a payout ratio of 46-48% of earnings, 18 consecutive years of dividend growth, and roughly 10% annual dividend growth over the past decade.

• In 2025, the company returned $5.9 billion to shareholders - a combination of dividends and buybacks, implying a 4.3% total "shareholder yield" (dividend + share buybacks).

• The capital plan of $3.3-3.4 billion per year goes not only to maintenance, but also to intermodal terminal expansion, locomotive upgrades and AI-driven projects (predictive maintenance, energy management).

• Management expects mid-single-digit earnings growth for 2026.

What has changed: from PSR to PSR 2.0 with AI

Union Pacific $UNP, like other Class I railroads in the U.S., has shifted over the past decade to precision scheduled railroading (PSR ) - a strategy that minimizes cars standing still, minimizes transloads, increases locomotive utilization, and reduces staffing to improve OR and profitability. The first wave of PSR was accompanied by negative effects: disgruntled customers complaining about reliability and pressure from regulators over the perception of a "squeeze" system.

The second wave - PSR 2.0 - is different in that it integrates AI, predictive maintenance and data analytics. In Q2 2025, Union Pacific car velocity (miles per car per day) increased 10% YoY to 221 miles per day, fuel efficiency improved 2% to 1,058 gallons per 1,000 ton-mile at an average fuel price of $2.42/gallon, directly improving OR. These technologies allow PSR to do PSR "finer" - increasing efficiency while keeping service levels at a level that is acceptable to customers.

For the investor, this means: much of the "fruit" in PSR has already been picked in recent years, but data-driven optimizations (AI, predictive maintenance, energy management) can be done for a long time to come. Every percentage improvement in OR has a huge impact on EPS, because the fixed costs of the railway are high and any savings go directly into profits.

What has to work out

• Continued improvement in OR towards the "upper 50%" without degrading service quality enough to bring regulators into play with heavy-handed interventions.

• Capex on AI and infrastructure must deliver real savings and higher capacity to grow intermodal volumes (high margin segment).

• The merger agenda (regulatory process to transcontinental plans) must not take too much management attention away from operations.

• Product mixes (more intermodal, less coal) must support the margin profile, not erode it.

How does this become money for shareholders

1) Stable, growing dividend

Union Pacific is not a typical high-yield dividend title like utilities or REITs, but rather a stable, growing dividend. An annual dividend of around $5.52 per share means a yield of ~2.5% at a price of around $220 - which is average in a higher rate environment, but the key point is this:

• payout ratio is roughly 45-48% of earnings (covered by earnings)

• the dividend has grown for the last 18 years in a row

• the 10-year dividend CAGR is close to 10% per annum

This means the investor is getting a combination:

• today's yield of 2.5%

• potential 7-10% annual dividend growth in the years ahead

Thus, over a 10-year horizon, the dividend yield to cost basis can exceed 5-6% if dividend growth holds and EPS continues to grow at mid-single to high-single levels.

2) Buyback as the second leg of the return

In 2025, the company returned $5.9 billion to shareholders in dividends and buybacks - that's a 4.3% shareholder yield. With a capex of $3.3 billion and a stable FCF, the firm can afford to continue at a similar pace, barring an extreme recession or extremely capital-intensive regulation.

Per-share buybacks like UNP, which has relatively stable cash flow and no massive "holes" in its balance sheet, works as an ongoing EPS improvement and price support in periods of weaker sentiment. If the stock is trading around or slightly below fair value, the buyback increases value for the remaining shareholders.

3) A bet on US growth

Union Pacific loads a huge portion of U.S. industry onto its rails: containers from the Pacific, grain from the Midwest, oil and chemicals from energy regions, automotive components, coal, and increasingly intermodal (train + truck combination) transportation. Trends such as:

• Manufacturingreturning to the U.S.

• growth of intermodal terminals

• investment in infrastructure(US Infrastructure Act)

are directly increasing volumes.

For shareholders, this means that even in a mild recession, UNP is covered by a diversified sector mix, and conversely, in periods of growth, it has leverage over high fixed costs - as tracks and locomotives become better utilized, earnings grow faster than revenues.

Figures that support the growth thesis

Key figures FY2024-FY2025

• 2025 revenues: $24.9bn (+2% YoY).

• Net profit 2025: USD 7.1bn (+6% YoY).

• Diluted EPS 2025: USD 11.98 (+8% YoY), adjusted EPS USD 11.66 (vs. USD 11.11 in 2024).

• Q4 2025 EPS: USD 3.11 reported, USD 2.86 adjusted (vs. USD 2.96 in Q4 2024).

• Net income Q2 2025: USD 1.9 billion, car velocity +10%, fuel efficiency improved by 2%.

• Cash return to shareholders 2025: USD 5.9bn (+25% YoY).

• Dividend: USD 5.52/year, yield 2.5-2.6%, payout ratio ~46-48%.

• Dividend growth: 18 years of consecutive increases, 10-year CAGR around 10%.

• Shareholder yield (dividend + buyback): 4.3%.

Dividend and sustainability

UNP is a textbook example of a "future dividend aristocrat": long history of dividend growth, payout ratio around 45-50%, stable free cash flow and disciplined policies. The company is not yet officially in the Dividend Aristocrats index (it needs 25 years of continuous growth), but with 18 years under its belt, it is well on its way.

Dividend sustainability stands at:

• Oligopoly position (realistically only BNSF, CSX, NSC as competitors)

• infrastructure that is not economically feasible to duplicate

• strong balance sheet profile (no extreme debt)

• healthy payout ratio and continued EPS growth

If EPS grows at a long-term mid-single digit rate (5-7%), dividends can grow similarly or even slightly faster if the company maintains a payout ratio around 50%.

Valuation - what's included and what's not

UNP's current valuation (at a price somewhere around $220) corresponds to:

• trailing P/E of ~18-19× on EPS of $11.98

• forward P/E on expected 2026 EPS of $12.5-13.0 somewhere around 17-18×, depending on the specific estimate

This is slightly above the long term average for the rail sector, but in line with the premium that UNP has been commanding over the long term:

• the best network positioning in the western U.S.

• higher efficiency and margins vs. some competitors

• solid track record of capital allocation

Analyst consensus (MarketBeat, StockAnalysis) shows:

• 26 analysts, average rating of "Moderate Buy"

• Median price target $250-260, range $220-290

Implications:

• Today's price offers mid-single digit annualized upside (EPS growth + small re-rating) plus a 2.5% dividend, for a return of 8-10% annually

• The upside scenario (stronger volume growth, further OR improvement, re-rating to P/E of 20x) gives a price potential of USD 270-280 on a 2-3 year horizon

So UNP is not a "deep value", but neither is it an "overinflated growth". It is a quality core holding with a premium valuation that makes sense if an investor wants a stable compounder.

Macro and market

Railroads are an "economic bellwether" - they respond to industrial production, exports/imports, the commodity cycle and consumer demand. This is a two-way street:

• in a recession, volumes fall, revenue carloads fall, OR can get worse

• in an expansion, line utilization picks up, fixed costs are spread over higher volumes, and profits grow faster than revenues

Current macro (2026) is a mix:

• Intermodal growth (e-commerce, retail)

• Coal stabilization (long-term downward trend, but still relevant)

• Volatility in industry (chemicals, metals) due to tariffs and geopolitics

UNP comes out of this as a relative winner because:

• it has diversified customers in many industries

• benefits from infrastructure projects in the US

• can become more expensive due to weak competition, even if volumes are stagnant

The mega merger of Union Pacific and Norfolk Southern: what's at stake and why

Union Pacific (UP) and Norfolk Southern (NS) have agreed to merge in the summer of 2025 to create the first truly transcontinental railroad in the US, with more than 50,000 miles of track in 43 states and connections to about 100 ports. Formally, this is an end-to-end connection, as the UP dominates the western US, while the NS has a strong presence on the East Coast and in the industrial regions of the Midwest.

The companies submitted an application for approval to the regulator Surface Transportation Board (STB) in December 2025. The rigorous regulatory process, which includes additional document requirements and competitive impact analyses, is still ongoing as of spring 2026. The regulatory framework is significantly tightened following previous industry consolidations. UP therefore has to prove that the merger is in the public interest, i.e. that it will not only lead to higher profitability but also to a better service for customers and stronger competition with road transport.

What the company promises: synergies, volume growth and a more efficient network

The management of both companies communicates mainly three strategic pillars, which are volume growth, network synergies and improved customer service. In its regulatory filings, UP claims that the combined firm will enable it to transform approximately 10,000 existing sessions from interline mode, the transfer of goods between two railroads, to single line service within a single carrier. This is expected to significantly speed up services and reduce the risk of delays. In addition, management identifies the potential to create approximately 84,000 new sessions at the county level, where truck traffic now dominates and where it may make economic sense to switch to rail.

In terms of growth, UP is presenting an estimate to the regulator of approximately a 12 percent increase in ridership. Approximately three-quarters of this growth is expected to come from actual freight shifting from highways to railroads, rather than customers being pulled from other railroads. According to the company, this should generate additional revenues of about $4.2 billion. It would also better leverage the so-called Midwest Watershed Market, the area around the Mississippi that is now fragmented among multiple carriers. At the same time, the companies are counting on classic economies of scale. These include better utilization of locomotives and cars, consolidation of dispatch centers, investment in automated inspection portals, and greater transparency of car movements across the route.

What specific benefits the merger creates

Synergies are not just theoretical. The geographical overlaps and new connecting routes have clearly defined practical use cases. The analyses highlight two key connections in particular. The first is a direct line from Kansas City to Springfield, Illinois, which will improve the flow of goods between the Midwest and Upper Midwest. The second is the so-called Meridian Speedway corridor through the Dallas Fort Worth area. This corridor is key for transportation from Southern California to cities in the Southeast. This partially bypasses chronically congested hubs such as Chicago and New Orleans, which is a significant argument for speed and reliability.

For customers, in an ideal scenario, this means simpler contracts, less transshipment, shorter transit times and greater predictability. It is these parameters that railways have often failed to offer in recent years compared to trucking. UP and NS also frame the merger as a chance to reboot the industry's reputation. Indeed, in the past, railroads have focused heavily on revenue maximization and drastic capacity reductions. This led to service disruptions and volume outflows. If the merged company does indeed pull significant freight volumes off the highways, it could be a structural shift in favour of rail transport, even in terms of environmental impact.

Where is the investment opportunity

For UP and NS shareholders, the potential opportunity consists of two components. The first component is direct synergies, the second component is a potential re-rating of valuation if the market believes in sustainable growth in volumes and margins. Analysts note that if the merger is successfully approved and reasonable terms do not force significant asset divestitures, the combination could have significant scope for appreciation. This is due to greater efficiency, network exclusivity on certain corridors and a stronger negotiating position vis-à-vis large industrial customers.

The second level of opportunity is more situational and relates to working with probabilities during the approval process. Extensive regulatory oversight and potential requirements to divest portions of the grid may create increased volatility in UNP and NSC stock prices. Thus, investors who are able to make informed guesses about regulatory scenarios may benefit from short-term mismatches between perceived and actual risk. This may relate, for example, to relative trades between UNP and NSC shares or positions in rival railways that could gain access to divested lines.

Conversely, where do the main risks lie?

The main risk remains regulation. The Surface Transportation Board and other authorities have repeatedly indicated that they will judge large mergers in an industry with high barriers to entry extremely strictly. In March 2026, the regulator requested additional information on the impact on competition and on future expansion. This prolongs the process and increases the risk of stricter conditions that could significantly worsen the economics of the transaction. These could include expanded reciprocal transhipment, mandatory access for competitors to certain corridors or forced divestments of key lines.

Reputational and operational risks are another category. Customer groups have already called on regulators to either block the merger or impose very strict conditions on it, fearing the growing pricing power of the merged railway and a possible deterioration in service quality in less attractive regions. Internally, the integration of two large and culturally different companies risks leading to short-term service disruptions, strikes or problems with the implementation of new systems. Historical experience with rail mergers shows that synergies planned on paper can easily turn into several years of operational chaos if integration is underestimated.

Investment scenarios

Optimistic scenario

EPS grows 7-9% per year due to:

• a combination of volume growth, better OR and buyback

• mid-single digit revenue growth

• stable margins

Dividend grows 8-10% annually, P/E expands to 20x due to investor preference for quality infrastructure "bond proxies". Share price in 3-5 years moves into the $260-280 range with a total annual return of 10-12% including dividends.

Realistic scenario

EPS grows 5-7% annually, revenue stagnates or grows at a low rate, OR improves modestly, dividends grow 7-8% annually, P/E remains around 17-18×. Share price in 3-5 years is somewhere between $240-260, total return 8-10% per year (EPS growth + dividend + slight re-rating).

Pessimistic scenario

Regulation and cyclical downturn shrinks volumes, EPS stagnates or falls to ~$11-12, OR deteriorates, P/E compresses to 14-15×. Share price falls into the $160-180 range, dividend grows more slowly or stagnates. Annual return over 3-5 years is 0-3%, with short-term drawdown potential of 20-25%.

What the investor should take away

• Union Pacific is an oligopolistic rail business with high barriers to entry, stable free cash flow and disciplined capital allocation.

• 2025 confirmed the strength of the model: revenue of $24.9 billion, net income of $7.1 billion, EPS of $11.98, improved OR and massive cash return to shareholders.

• The dividend story is robust: 18 years of growth, yield 2.5-2.6%, payout ratio ~45-48%, combined shareholder yield 4+%.

• Valuation is premium but not extreme - P/E of 17-18x reflects quality and oligopoly position; for the long-term investor, UNP offers 8-10% annual compound with relatively low "drama".

• The key is to monitor OR, volumes, regulation and capital discipline; if these four pillars remain solid, Union Pacific has all the ingredients to become the "dividend aristocrat" of the next decade.

Revenue declines in established, stable companies operate differently than in technology startups. While it may be a transitional phase of growth investing for young companies, for established giants with historically stable revenues, negative growth often signals a structural problem. Indeed, it is this type of company that investors value primarily for its predictability and ability to generate consistent cash flow. Once this characteristic wears off, reassessment of valuation tends to be quick and painful.

2026 provides an interesting example of this phenomenon. Several global leaders in their industries are facing revenue declines that are not caused by one-off factors but by a combination of macroeconomic changes, shifting consumer preferences and geopolitical tensions. While some companies are able to use the crisis to restructure and return to growth, others are struggling with fundamental changes in their markets. Can the following companies reverse this trend?

LVMH $MC.PA

LVMH Moët Hennessy Louis Vuitton represents the largest player in the global luxury goods market. The company owns a portfolio including fashion brands such as Louis Vuitton, Dior and Fendi, jewelry including Tiffany & Co., alcoholic beverages with Moët & Chandon and other segments from cosmetics to watches. It is this diversification that has historically been seen as a competitive advantage, allowing the company to withstand fluctuations in each category. However, in the last year, it has become clear that when overall sentiment towards luxury declines, diversification does not help.

A picture of the decline in numbers

Over the last twelve months, LVMH's sales have fallen by around 4.6%. While this is not a dramatic fall, it is a significant signal of a change in trend for a company of this type. The luxury sector has historically grown at a steady pace due to a combination of product price inflation and expansion into new geographic markets. Stagnation or decline in revenue is therefore read by the market as a structural problem, not as transient volatility.

The greatest pressure comes from the fashion accessories and leather goods segment, which accounts for the bulk of total revenues. It is here that the cooling of demand is most pronounced. Chinese consumers, who have driven the growth of the entire luxury sector in recent years, have significantly reduced their purchases. The combination of economic uncertainty, regulatory crackdowns on the wealthier classes and a change in sentiment towards ostentatious consumption is creating an environment where even consumers with sufficient capital are putting off expensive purchases.

China as a major problem

The Chinese market has been a major source of growth for LVMH over the last decade. The expansion of the middle class and the growing number of millionaires in China have created an ideal environment for luxury brands. However, the situation has changed significantly in the last two years. Economic growth has slowed, the property market is in crisis and consumer confidence is falling.

The regulatory environment is also playing a significant role. As part of its shared prosperity strategy, the Chinese government has tightened controls on ostentatious consumption and public flaunting of wealth. This puts pressure on wealthy Chinese to curb visible spending on luxury goods. The combination of these factors has led to a segment that used to grow at double-digit rates now stagnating or even declining.

The company is trying to compensate for the weakness of the Chinese market by growing in other regions, especially the US and Europe. Here, however, we run into a valuation problem. The US and European luxury markets are much more mature and do not have the growth potential of the Chinese market. Unless Chinese demand recovers, LVMH faces a structural slowdown in growth with no obvious catalyst for a turnaround.

Valuation pressure

LVMH shares have historically traded at a premium to the broader market due to steady growth and strong pricing power. However, once revenue growth moved into negative numbers, the market began to question the legitimacy of this premium. Forward P/E ratios are hovering around values that are still above the long-term average for the consumer sector, creating pressure for further valuation correction as sales decline.

The key question is whether this is a cyclical decline related to the macroeconomic environment or the beginning of a structural change in consumer behavior. If younger generations in China and other markets are indeed shifting their preferences away from ostentatious luxury consumption towards other forms of status expression, this could pose a long-term problem for LVMH.

What to watch next

• The evolution of consumer sentiment in China and any signs of a recovery in demand in the region

• How the company manages to maintain margins in an environment of declining sales. If management starts to push for discounting or if average transaction prices fall, it could signal a deeper pricing power problem

• The firm's ability to activate growth in other regions. The U.S. market remains relatively resilient, but its size and growth potential are not enough to offset the fallout from China. The European market faces its own challenges, including higher interest rates and economic uncertainty

Unilever $UL

Unilever is one of the world's largest consumer goods companies with a portfolio of brands covering food, beverages, home care and personal care. Among the best known are Dove, Axe, Knorr, Hellmann's and Ben & Jerry's. Unlike LVMH, this is not a luxury segment but a mass consumer market where competition is extremely intense and growth has historically been slower but more stable.

Restructuring chaos

Over the past twelve months Unilever has seen sales fall by around 2 per cent. This figure in itself does not look bad, but in the context of several years of efforts to transform the business it is a significant disappointment. In recent years, the company has sold a number of brands it considered unviable and focused on categories with higher growth potential. The result so far is not convincing.

Management is trying to shift the portfolio towards the premium end of the market and towards higher value-added categories. In practice, this means divesting traditional food brands and strengthening segments such as skin care and health nutrition. However, this shift comes at a time when consumers in many markets are facing inflationary pressures and returning to cheaper alternatives. As a result, premium categories are experiencing a decline in demand at the very time when Unilever is betting on them.

Pressure on margins

One of the company's biggest challenges is maintaining margins in an environment of rising raw material, energy and distribution costs. Unilever has historically achieved stable operating margins due to its strong bargaining position with retail chains. In recent years, however, this dynamic has been changing. Retail giants such as Walmart $WMTTesco $TSCDY are strengthening their own private labels to compete with Unilever products at lower prices.

The company tries to offset the pressure on margins by raising prices, but this leads to a decline in sales volumes. This effect has been particularly noticeable in developed markets, where consumers are more sensitive to price changes than in developing economies. The result is a situation where, although a firm keeps the nominal value of sales relatively stable, the actual volumes sold fall.

Geography and structural changes

Geographical exposure also plays an important role in the current developments. Unilever $UL has historically benefited from growth in emerging markets, where a growing middle class has increased demand for branded consumer goods. However, this trend has slowed in recent years. Economic problems in some key markets such as India and Brazil have reduced the pace of growth, while developed markets have stagnated.

Another factor is the shift in consumer preferences towards local and smaller brands. This trend is particularly evident among younger generations who prefer a more authentic and less commercial image. Large multinationals such as Unilever face competition not only from other giants, but also from more agile smaller players who can respond more quickly to changing trends.

Dividend under pressure

For many investors, Unilever is primarily a dividend stock with a long history of payouts. However, falling sales and pressure on margins are creating questions about the sustainability of the dividend policy. The company has no plans to cut the dividend for now, but if fundamentals do not improve, management may have to reassess the priority between paying shareholders and investing in growth.

Unilever's market capitalization has been under pressure over the past year precisely because of uncertainty about future growth. The stock has lost some of the premium that investors have traditionally accorded to stable consumer companies. The forward P/E ratio has declined. According to the Fair Price Index on Bulios, $UL stock is currently already below its fair value.

What to watch next

• How quickly management can complete the portfolio restructuring and whether the new brand mix will actually translate into renewed growth. The market will be watching the evolution of organic sales growth by region and category. If it turns out that the premium segments on which the company is betting remain under pressure, this may lead to a further adjustment of the strategy

• The company's ability to maintain margins. If it is forced to continue with price increases without adequate product differentiation, it risks a further decline in market share. Conversely, successful innovation or brand strengthening could restore investor confidence

Sony Group $SONY

Sony represents a different case than the previous two firms. It is a technology and entertainment conglomerate with a diversified portfolio that includes electronics, the PlayStation gaming segment, movie and music studios, and financial services. It is this diversification that has helped the firm in recent years to balance weaker segments with stronger ones. But in the past year, it has become clear that when multiple divisions face problems at the same time, overall sales decline.

Electronics under pressure

Sony' s traditional electronics business has been facing structural challenges for years. The segment comprising classic TVs, audio devices and cameras is facing intense competition from Chinese manufacturers offering similar quality at lower prices. Sony is trying to maintain its premium position by focusing on the high-end segment, but this market is more limited and sensitive to economic fluctuations.

Over the last twelve months, sales from the electronics segment have fallen by around 5%. This is a more pronounced decline than the overall company, as other segments such as music and the gaming business have remained relatively stable. The problem is not only competition, but also overall weaker demand for consumer electronics in key markets.

PlayStation and the gaming segment

The gaming division has historically represented one of Sony's strongest pillars. The PlayStation 5 has entered a mature lifecycle phase after initial supply issues, which typically means stable hardware and software sales. However, the last year has seen a decline in both console and software sales, which is a worrying sign for investors.

There are several reasons for this. First of all, the console market has reached saturation point. Most people already own a console and purchases of new units are mainly limited to new customers or replacements of older devices. At the same time, the gaming industry as a whole is facing a slowdown after the covid boom. During the pandemic, there was an extreme increase in time spent playing, which led to high sales.

Another factor is the change in the consumption pattern of games. More and more gamers are switching to free-to-play titles or subscription services, which reduces revenue from the sale of full $70 games. Sony does own the PlayStation Plus subscription service, but its monetisation does not reach the levels of traditional software sales.

Film and entertainment studio

Sony's film division is going through a period of uncertainty. While some films have achieved commercial success, the overall box office remains under pressure. Audiences are going to cinemas less and less and moving more to streaming platforms. Sony doesn't have its own massive streaming service like Disney+ $DIS or Netflix $NFLX, which puts the company in a complicated position.

The company licenses some of its content to other platforms, which brings in steady revenue but also means it doesn't have full control over the monetization of its movies. This model may be advantageous in the short term, but in the long term it risks the company losing direct contact with its audience and becoming dependent on third-party decisions.

Valuation

Sony's stock has historically traded at a discount to U.S. tech giants, in part because it is a Japanese firm with a complex structure. The decline in sales has exacerbated this discount. Investors are unsure how to value a company with such a diversified portfolio, where some segments are growing and others are declining.

But the question remains whether current valuations reflect the risks associated with the continued decline of key segments or whether the market is overly penalizing short-term fluctuations. However, shares were still trading at absolute highs in the second half of last year. Since reaching them, however, a sharp sell-off has been triggered and has accelerated this year. Shares of $SONY are currently trading 30% below ATH. The Fair Price Index on Bulios already shows that they are below their intrinsic value based on DCF and relative valuation.

What to watch next

• Gaming segment developments. If PlayStation 6 delivers strong growth or if the company can successfully monetize live-service games, it could reverse the overall narrative. Conversely, further decline in this division would signal a deeper structural problem.

• Whether the company can maintain its premium position and margins (in the electronics segment) in an environment of price competition. Innovations in automotive sensors or expansion of the professional equipment portfolio could provide new sources of growth.

• The film division remains uncertain territory, where success depends on individual films and trends in media consumption. Strategic partnerships or potential acquisitions could shift the company's position in this segment.

Summary comparison of key metrics

Strategy

All three companies share a common problem. Their historically stable revenues began to decline at a time when the market no longer tolerated stagnation in premium-valued titles. After years of low rates and optimism, investors are now demanding concrete evidence of growth, not just stories about future potential.

LVMH $MC.PA faces geopolitical and macroeconomic risks associated with the Chinese market. Unilever $UL is struggling with internal restructuring challenges and changing consumer preferences. Sony $SONY is trying to manage a complex conglomerate in an environment where different segments are going through different phases of the cycle.

Earnings are still attractive, but the sustainability of payouts depends on the ability of companies to stabilize earnings and restore growth. If the decline in earnings continues for an extended period, management could be forced to cut dividends or limit share buybacks.

What to watch next

The key factor for all three companies will be developments over the next two to three quarters. If sales stabilize or even begin to grow slightly, the market may interpret this as the end of the downturn and the beginning of a new phase. Conversely, a further decline in revenues could lead to a more significant valuation correction.

For LVMH, the sentiment of Chinese consumers and any signs of economic recovery in the region will be crucial. Unilever needs to demonstrate that the portfolio restructuring is indeed delivering the expected results in terms of higher margins and renewed growth. Sony needs to show that the gaming segment remains a strong profit driver and that the company can monetise its content effectively.

The market will also be watching to see how each company's management responds to the current situation. Aggressive cost-cutting may boost profitability in the short term but threaten competitiveness in the long term. Conversely, investing for growth in a time of declining revenues requires courage and a belief that the fundamentals of the company are still strong.

Blackstone invests USD 2 billion, debt to be led by Bank of America

Blackstone Inc.'s $BX will provide an equity investment of about $2 billion, while Bank of America Corp. $BAC will lead an additional $14 billion in debt financing. This financing was originally planned as a construction loan, but is now expected to be structured as a bond issue.

The financing is expected to close this month. Related Digital, a subsidiary of New York-based development giant Related Cos. confirmed that the financing will be completed shortly.

Saline Township project generates over 1 gigawatt of capacity

Related Digital, Oracle $ORCL and OpenAI announced plans in October 2025 to develop a data center campus with more than one gigawatt of capacity in Saline Township, Michigan. The multibillion-dollar investment in the project is part of a partnership between OpenAI and Oracle to deliver an additional 4.5 gigawatts of capacity to the Stargate project.

The project consists of three single-story, 550,000-square-foot buildings on 250 acres. The investment is expected to create more than 2,500 union construction positions and a total of more than 450 on-site jobs, plus 1,500 positions countywide.

Oracle has already raised $58 billion in AI infrastructure funding

The financing follows other massive debt packages put together by banks for Oracle data centers: a $38 billion debt deal to build facilities in Texas and Wisconsin and $18 billion for a site in New Mexico . In total, Oracle has raised approximately $58 billion in debt financing for its data center projects.

The Texas and Wisconsin package would be the largest debt package associated with AI infrastructure. The $38 billion debt offering consists of $23 billion and $15 billion of term loans led by JPMorgan, while Oracle is preparing another $18 billion of other debt financing for a data center in New Mexico led by SMBC, MUFG, BNP Paribas and Goldman Sachs.

Stargate project expands across US with $500bn investment

Stargate was launched by OpenAI in January 2025 in partnership with SoftBank, Oracle and Abu Dhabi's MGX with an initial investment of US$100 billion, which will increase to US$500 billion over the next four years. In addition to the campus in Abilene, Texas, Stargate campuses are being developed across the US in Shackelford County, Texas, Doña Ana County, New Mexico, and Port Washington, Wisconsin.

The project brings Stargate to more than 8 gigawatts of planned capacity and more than $450 billion of investment over the next three years. Oracle has a contract to provide OpenAI cloud services worth approximately $300 billion over five years.

Related Digital has a $45 billion development pipeline with more than 5 gigawatts of near-term capacity across the U.S. and Canada . Oracle's stock has fallen in recent months and the cost of securing debt against default has risen, leading some potential investors to question, but spokespeople for Oracle, Bank of America and Related Digital confirmed that the financing is progressing as planned.

March 2026 was a volatile and ultimately negative month for global markets, driven by macroeconomic uncertainty, geopolitical tensions around Iran, profit-taking in the technology sector, and renewed concerns about interest rates. My USD portfolio – composed of individual stocks, ETFs and cryptocurrencies – returned -2.50% in March, outperforming the S&P 500, which fell -5.09%.

Year-to-date (YTD, Jan 1 – Mar 31, 2026) my portfolio stands at -7.71%, while $^GSPC 500 recorded -4.63%.

I’d rather not talk about my CZK portfolio; after several strong months and years it’s at -3.22% M/M and -4% YTD.

Closed positions:

$NVO (Novo Nordisk) – risky investment closed with a loss of -49%

$NIO (NIO Inc.) – profit target reached, gain +30%

$EIX (Edison International) – profit target reached, gain +28%

$DELL (Dell Technologies) – profit target reached, gain +35%

$CVX (Chevron Corporation) – closed 4 positions with an average gain of +39%

Newly opened positions:

$BABA (Alibaba Group) – entry at 132 USD

$BA (Boeing) – entry at 210 USD

$META (Meta Platforms) – entry at 544 USD

$RR.L (Rolls-Royce Holdings) – entry at 1172 GBP

I also continued regular Wednesday purchases of cryptocurrencies ($BTC, $ETH, $ADA) as part of my DCA strategy.

All proceeds from holding and selling assets are either reinvested or held as free cash waiting for the right opportunity.

The portfolio remains well diversified with an approximate allocation of 80% equities / 3% ETFs / 17% crypto and a relatively low correlation to the S&P 500.

Why the current portfolio composition makes sense (medium to long term):

Broad diversification across quality stocks, ETFs and crypto provides resilience even in tougher periods

Exposure to structural trends (AI, digital advertising, cloud, blockchain) supports the potential for higher performance

Discipline in risk management – controlled position sizing, a cash reserve and regular rebalancing

Long-term potential thanks to companies with strong free cash flow and assets with significant network effects

Main risks for the coming weeks and months:

Macroeconomic uncertainty and central bank actions may trigger further volatility

Risk of an oil shock in case of escalation or non-resolution of the conflict with Iran

Regulatory pressure on big tech and cryptocurrencies (especially in Europe and Asia)

Sector concentration – potential slowdown among AI/tech leaders

High volatility in cryptocurrencies, which can amplify portfolio moves in both directions

Overall, March 2026 confirmed that the portfolio is (hopefully) well positioned for a medium- to long-term horizon. Active allocation and the crypto component should provide relative resilience to the risks mentioned and potentially higher performance compared to the S&P 500.

How do you view the current portfolio allocation? Would you add more crypto, or would you rather stay more in traditional stocks?

You can find the English version of this post on my eToro profile. If you'd like to follow me there or possibly copy my USD portfolio, I’d be happy!

LVMH $MC.PA is experiencing the worst start to a year in its history. The French luxury conglomerate's shares fell 28% in the first quarter of 2026, a worse result than during the 2008-2009 financial crisis, the Covid-19 pandemic in 2020 and the bursting of the dot-com bubble in 2001, according to a Bloomberg analysis.

The war in the Middle East has shaken the luxury market

The reason for the dramatic decline is the impact of the war in the Middle East, which has clouded the global economic outlook and exacerbated problems with demand for luxury goods. Attacks between the US, Israel and Iran have forced airlines to cancel thousands of flights.

According to a CNBC report, the Middle East averaged a mid-to-high single-digit percentage of luxury brand sales, but the region was the fastest growing luxury market last year, growing between 6 and 8 percent, while global growth was zero. The region now accounts for about 6 percent of global luxury sales.

Bernard Arnault's wealth plunged by $55.9 billion

LVMH's stock collapse means CEO Bernard Arnault 's net worth fell by $55.9 billion in the first quarter alone, according to the Bloomberg Billionaires Index, bringing his total fortune to about $152 billion. This is the second largest loss among the world's 500 richest people after Oracle founder Larry Ellison, according to a FashionNetwork analysis.

During the first quarter, the Arnault family's stake in LVMH crossed the symbolic 50 percent mark. According to John Plassard, director of investment strategy at Cité Gestion, "LVMH has become more than a luxury stock, it is now a barometer of global confidence".

The luxury sector has lost $100 billion in market value

The war conflict has wiped roughly US$100 billion of market capitalisation from luxury sector companies, with LVMH and Hermès each losing more than US$40 billion in value. According to a CNBC report, shares of major luxury companies have fallen 15 percent or more since the war began.

• Richemont:

Zurich shares down about 20 percent in first three months

• Hermès:

Lost almost a quarter of its value over the same period

• Kering and others:

Kering shares fell by 5%, Brunello Cucinelli by 4.6% and Burberry by 4.3%

Tourism and travel under pressure

The decline also reflects the disruption to travel and tourism on which sales of the most expensive items are so dependent. According to Luxuri analysis, around 60 per cent of luxury spending in the UAE comes from tourists, with Dubai's appeal resting on security, tax benefits and political stability.

Bernstein analyst Luca Solca told CNBC, "If people don't get back to normal and we have more problems with oil and gas supplies from the Gulf, then the likelihood of a global recession could increase, and that would certainly dampen discrete sectors like luxury".

The outlook remains uncertain

LVMH is due to release first-quarter earnings later this month. Its core fashion and leather goods division is likely to have seen a 0.65% rise in organic sales during the period, according to analysts' preliminary estimates. The division includes the biggest brand Louis Vuitton as well as Christian Dior Couture.

According to a UBS research report, luxury investor sentiment is "the most bearish in years", while "heightened geopolitical uncertainty is likely to weigh on near-term earnings and delay a long-awaited turnaround in fundamentals".

The company is a Norwegian software group built around a family of web browsers for desktop and mobile, complemented by its own advertising stack, content and gaming services plus a growing layer of AI and Web3 features. Its strategy is to use the browser as the default gateway to the internet for roughly 300 million monthly active users, then monetise that attention through search and ad revenue while differentiating with integrated AI tools, built in crypto wallets and a dedicated gaming browser aimed at users who spend a disproportionate amount of time and money online.

Top points of analysis

• The stock trades around $14, with an estimated fair value coming out at roughly $20.5 - a potential of nearly 50 percent without factoring in the dividend.

• Revenues in 2025 were roughly $617 million, plus 28 percent year-over-year, after previous growth of 21 and 20 percent in 2024 and 2023.

• Net income for 2025 rose to about $109 million, earnings per share of $1.21, a net margin of about 18 percent and an operating margin of just under 15 percent.

• Free cash flow in 2025 was over $115 million, growing over 40 percent year-over-year, with very low capital expenditures.

• The balance sheet is extremely conservative: debt-to-equity and asset ratios around one percent, Altman Z-score of 7.5, high liquidity, and cash of over $155 million.

• Valuation multiples are still moderate - price to earnings around 11-12 times, price to sales 2 times and price to free cash flow around 11.6 times.

• The company pays a dividend of $0.39 per share, at today's price about a five percent dividend yield, and has also authorized a buyback program of up to $300 million for the next two years.

The business and how Opera makes money

Opera $OPRA is a technology company in the Internet content and services industry, best known for its web browsers - from basic browsers for everyday users to mobile variants to specialized products for gamers and users focused on artificial intelligence. The basic principle is simple: bring as many users as possible into its ecosystem and then monetize their attention through advertising, search engine partnerships and add-on services, gradually moving the company from a "simple browser" to a smaller ad-fintech platform.

Revenue today consists primarily of two main components:

• advertising revenue - targeted browser ads, often tied to e-commerce and other high-margin segments

• Search revenue - a share of revenue from search engines for traffic brought in and user queries

According to the latest data, advertising already accounts for roughly two-thirds of total revenue, while search covers the remaining third - and advertising is growing much faster, e.g., growing at more than 40 percent year-over-year in Q2 2025, while search grew by the lower teens. The third, smaller but fast-growing component is financial services like MiniPay, which is already seeing its first meaningful revenue and future potential.

Products and sources of growth

Opera is building its growth on a combination of several product lines and the ability to select specific groups of users that it can monetise more effectively than a 'big' universal browser.

Major browsers

• Opera One and Opera Mini - for mainstream users, with an emphasis on speed, data savings (particularly important in Africa and other emerging regions) and built-in features such as ad blocking, VPN or integrated wallet.

• Opera Air - for users focused on a clean, lightweight experience and mobile usage, targeted especially at emerging markets where speed and low data usage are key.

Opera GX gaming browser

Opera GX is a dedicated browser for gamers, now used by tens of millions of users per month (over 30 million at last count). It offers features such as CPU, memory and data usage limitations, integrated game news, menus and deep appearance customization. This target audience is attractive to advertisers - they are often younger, tech-savvy users with a higher propensity to spend in games and technology - and allows Opera to sell advertising in this segment at higher prices.

AI browsers and the Aria assistant

Opera is integrating artificial intelligence across its browsers. In 2025, it launched new AI-focused browsers Opera Air and Opera Neon, which build on the integration of large third-party language models (such as $GOOG's Gemini) and its own "orchestration" - a concept where the browser is the hub through which the user interacts with various AI services.

Aria's AI assistant, available in several products, serves as a tool for searching, summarizing content, writing text, and other tasks directly in the browser - for Opera, it's a tool to increase user engagement and prepare for future charging for more advanced features via subscription in premium versions of the browser.

Fintech - MiniPay and wallets

MiniPay is a non-custodial wallet built on the Celo blockchain, integrated into Opera Mini while functioning as a standalone app. It allows you to send money between users virtually instantly and with low fees, and just via a phone number - targeting especially Africa and other regions where traditional banking services are not so easily available. MiniPay already has more than 10 million activated wallets , and the company openly says that these services are already profitable, although it is reinvesting much of the revenue into growth and partnerships.

The fintech component is important because it gives the firm additional leverage beyond pure advertising - transactional revenue from payments, cross-border transfers and potentially other financial products can make up a significant portion of revenue over time and better diversify the cyclicality of the advertising market.

Management and strategy

The company is headed by CEO Lin Song, who is leading Opera through its current phase of transition from an "underappreciated browser" to a broader platform combining browsers, AI and financial services. Management has a track record of delivering and raising guidance - for example, it has raised full-year revenue and adjusted EBITDA guidance several times during 2025, indicating relatively conservative target-setting.

Key strategic points that management reiterates:

• Product segmentation - preferring multiple specialized browsers for specific groups (gamers, AI users, data-constrained users in Africa) rather than one universal product

• Shift to higher monetization - focus on higher average value users, more e-commerce advertising and affiliate programs where conversion rates are higher

• Development of AI as a "control layer" - browser as a hub through which the user interacts with various AI services, with the possibility of premium pricing for advanced features in the future

• Extending fintech services - MiniPay as a tool to bring payments and financial services to users directly in the browser, building on the demographic growth of Africa and other regions

The way management handles capital shows a combination of growth and return on capital: alongside investment in products and marketing, dividends and a new buyback program are worth up to $300 million for the next two years, a number significant against a market capitalization of around $1.2 billion.

Where the company is growing the most

Opera's growth rests on three pillars:

1. Advertising and e-commerce - the fastest component of revenue, driven mainly by e-commerce partners, where user intent to buy is high, and thus the price of advertising is higher.

2. Gaming and AI browsers - Opera GX for gamers and AI-based browsers such as Air and Neon increase user engagement and deliver premium model capabilities.

3. Fintech MiniPay - a fast-growing wallet that has already surpassed 10 million activated accounts and opens up space for transactional revenue in underbanked areas.

Geographically, Opera has a very strong presence in Africa, where its data-saving browsers (Opera Mini) hold a mobile market share of around 7-8 percent and bring in hundreds of millions of users per month - the company has long invested in local partnerships and 'free data for browser use' campaigns. In addition, it is targeting users in Southeast Asia and Latin America, where smartphone and Internet adoption is still in a growth phase.

Competition and market position

Opera operates in an environment dominated by three giants - Chrome, Safari and Edge - with global market shares of tens of percent. Opera's share is around 2 percent, but that doesn't necessarily indicate a weak position: the company has chosen niche segments where it can offer something extra.

Opera compensates for its smaller share with higher monetization per user - thanks to segmentation (gamers, AI users, e-commerce users) and smart ad targeting, it can extract a higher average revenue from each active user than the average "free" service. It also benefits from being an early entrant and having a strong brand in many regions, such as Africa.

Valuation and fair price

At today's levels, Opera is trading at approximately:

• a price-to-earnings ratio of around 11.8 times

• price to earnings ratio of 2.0 times

• price to book value 1.23 times

• price to free cash flow of around 11.6 times

With sales of about $617 million and net income of about $109 million, this implies that the market is willing to pay about $11-12 for every dollar of earnings and two units of firm value per unit of annual sales. Fair value with a conservative setup (sales growth of about 15-20 percent, net margin gradually in the 15-18 percent range, and a stable dividend) comes out to about $20-21 per share.

Given that the company has virtually zero net debt, a very strong balance sheet and already pays a five percent dividend, the current valuation looks more like a cheaper entry into a growing ad-fintech platform than an overpriced growth title.

Investment scenarios

Base case scenario

• Revenues grow in the 15-20 percent per annum range, margins remain near current levels or improve slightly.

• Earnings per share grow in the higher single digits to lower double-digit percentages annually.

• Valuation approaches fair value around $20, dividend yield stays around 4-5 percent, buyback program improves earnings per share growth.

Positive scenario

• AI browsers and the gaming segment will add more than expected today; MiniPay and financial services will ramp up significantly, delivering higher ARPU and a new, less cyclical revenue component.

• Revenues grow closer to 25-30 percent annually, net margins move towards 20 percent, and the market is willing to pay 15-17 times earnings.

• The stock could move well above today's fair value estimate in a few years if Opera is confirmed as a stable player in browsers, AI and fintech.

A more cautious scenario

• Growth slows to single digits - competition from large browsers, weaker advertising market or slower monetization of MiniPay and AI products.

• Earnings per share stagnates or grows very slowly; valuation falls to 8-10 times earnings.

• The yield for the investor is then mainly the dividend and any buybacks, while the share price does not move much in the long term.

What may surprise Opera in the future

• Tighter monetization of the gamer community and AI products - higher engagement, premium features and subscriptions.

• Faster growth of MiniPay and financial services, especially in Africa and other emerging regions.

• Greater and more aggressive use of buyback programs if share price remains low relative to earnings and cash.

• Greater attention from analysts and funds, possibly inclusion in broader indices, which in itself may bring rerating multiples.

Risks - what can spoil the scenario

• Strong competition from large browsers, which will steal share even in segments where Opera is strong today.

• Changes in the advertising market, cookie regulation and tracking that will reduce the effectiveness of targeting and the size of advertising budgets.

• Lack of adoption of premium AI features and financial services - users are sticking with free versions with no willingness to pay.

• Currency and political risk in emerging markets where a large portion of the user base is.

What to take away from the analysis

• Opera is not just a small browser next to the big names, but a profitable growth company with a clean balance sheet, strong cash flow and a mix of businesses - advertising, AI and fintech.

• The current valuation of around twelve times earnings and twice sales doesn't look excessive given the growth and profitability, rather the opposite - especially when combined with the five percent dividend.

• The key to success will be in the ability to continue to increase revenue per user, develop MiniPay and premium AI browsers, and maintain discipline in capital - i.e. a combination of growth, dividends and buybacks.

Would you consider reducing your position in $OXY after a 50% gain, or would you still hold?

There are thousands of companies in the stock market, but only a handful of them can sustain a long-term return on equity above 20% while earning a buy recommendation from most analyst houses. ROE (Return on Equity) measures how efficiently a company turns shareholder capital into net income. In practice, a value above 20% means that for every dollar of equity invested, the firm generates more than 20 cents of net profit per year, well above the broad market average.

In 2026, this indicator becomes even more important. As higher interest rates increase the cost of debt financing, investors are increasingly looking for companies that can operate efficiently with equity capital and are not reliant on cheap debt. At the same time, the analyst consensus "buy" becomes an important filter because it signals that experts see room for further upside in a given stock even at current valuations.

Today's selection includes seven companies from very different sectors. From tech giants to financial infrastructure to the gaming industry. The common denominator is precisely the combination of high capital efficiency and positive analyst consensus. Let's take a closer look at each of them.

Amazon $AMZN

Amazon isn't just the world's largest e-commerce platform. In recent years, the company has increasingly profiled itself as a technology conglomerate whose most profitable segments lie outside of traditional retail. The cloud division of AWS generates around 60% of the group's operating profit and is growing at a rate of around 20% a year. Added to this is the advertising business, which is approaching $70 billion in annual revenues and growing at over 20% year-on-year.

Amazon's current ROE is around 22%, a historically strong level. As recently as 2022, the company was in negative territory due to massive investments in logistics during the pandemic. A return above 20% reflects a shift from a phase of aggressive investment to a phase of monetizing the ecosystem it has built. Meanwhile, the firm is planning capital expenditures in excess of $100 billion this year, primarily focused on AI infrastructure and data centers.

The analyst consensus is Strong Buy with an average target price of around $285, which at the current price of around $208 represents room for upside of approximately 37%. Regulatory pressures, including the FTC antitrust proceeding scheduled for October 2026, and the question of returns on massive AI investments remain a risk.

Walmart $WMT

Walmart tends to be seen as a defensive retail giant, but its current ROE of around 23% points to a company that can generate above-average returns even with low margins. The key is a combination of extreme sales volume, fast inventory turnover and a strong negotiating position with suppliers. The company reported sales in excess of $713 billion this year, with year-over-year growth of about 5.8%.

Walmart's earnings structure has been changing in recent quarters. The e-commerce segment is growing at a rate of over 27% annually, the Walmart Connect advertising division has added over 37%, and the Walmart+ membership program is strengthening customer loyalty. These segments have significantly higher margins than traditional sales, which helps maintain high ROEs even in an environment of consumer spending pressure.

The analyst consensus is Strong Buy with a target price of around $131. However, Walmart trades at a P/E ratio of over 45x, which is more than double Nvidia's $NVDA valuation. At such a tight valuation, the stock could experience increased volatility if new segment growth slows or if tariff policy increases the cost of imports.

Tencent $TCEHY

Tencent is China's largest technology company, operating an ecosystem that includes social networks (WeChat with 1.42 billion active users), game studios, cloud services, fintech and digital advertising. In 2025, the company reported record annual revenue of RMB 751.8 billion with 14% year-on-year growth. Gross profit rose 21% to RMB 422.6 billion, reflecting a shift in revenue mix towards high-margin segments.

The gaming segment remains the revenue driver. Domestic gaming revenue grew 18% year-on-year and international gaming revenue surpassed the $10 billion mark for the first time. The company invested approximately RMB 18 billion in AI in 2025 and plans to double this amount in 2026. AI is already contributing to improved advertising targeting and higher player engagement.

Tencent's ROE is around 21%, and the firm operates with minimal debt (Debt-to-Equity of 0.32) and holds nearly RMB500 billion in cash on deposit. The analyst consensus is Buy with an average target price of around $96, representing an upside of over 50%. The main risk remains the regulatory environment in China and geopolitical tensions between the US and China, which may affect foreign investor sentiment.

Intuit $INTU

Intuit is the operator of one of the strongest software ecosystems in the US economy. Its business combines accounting software QuickBooks, tax platform TurboTax, financial portal Credit Karma and marketing tool Mailchimp. The platform generates revenues of over $18 billion annually with year-over-year growth of around 15%. Key drivers are extremely high customer retention and growing average revenue per user through cross-product sales.

Intuit's ROE is 23.5% as the company benefits from a capital-light software model. The platform now generates approximately 77% of revenue from recurring subscription services. In addition, Intuit is aggressively integrating artificial intelligence into its products, which adds value to users and strengthens the traction of the ecosystem.

Shares of Intuit are down approximately 34% this year, a result of a broader sell-off in software titles in 2026. The analyst consensus remains Strong Buy with a median target price of $800, which at the current price of around $432 represents upside potential of up to 84%. The market is concerned about the impact of AI on traditional software models, but Intuit is among the companies actively using AI to its advantage, not as a threat.

Interactive Brokers $IBKR

Interactive Brokers is a global broker that provides access to more than 150 exchanges in 34 countries. Founded in 1977, the firm has built on technological automation of trading since the beginning. As a result, it achieves operating margins of over 76%, well above the brokerage sector average. In 2025, the firm saw client accounts grow 30% year-over-year to nearly 290,000 accounts in the fourth quarter, and options trading volumes grew 26%.

Interactive Brokers' ROE is approximately 20.4%, which is outstanding in a sector where the average is around 10%. The firm's five-year average ROE exceeds 20%. This efficiency is driven by an automated operating model that requires minimal human intervention and allows the firm to scale trading volume without a proportional increase in costs.

The analyst consensus is Strong Buy with an average target price of around $69. The firm is benefiting from growing retail investor activity, expansion into new markets and increasing interest in options trading. The risk is the dependence on market volatility, which directly affects trading volumes and thus commission income.

Cadence Design Systems $CDNS

Cadence Design Systems is one of the least known firms in the semiconductor value chain, even though no modern chip could have been created without its products. The company develops integrated circuit design and verification software that is used by virtually every major semiconductor manufacturer in the world. Annual sales exceed $5.3 billion with year-on-year growth of around 15% and gross margins of an impressive 86%.

Cadence's ROE is around 21.8%. The company is benefiting from the increasing complexity of chip design driven by demand for AI accelerators. The more complex the chip, the more licenses and compute time customers need, which directly increases the average order value for Cadence. In addition, the company is expanding into systems analysis, focusing on thermal simulation and electromagnetic analysis of datacenters.

The analyst consensus is Buy with a target price of around $410. The firm's market capitalization is around $75 billion. The P/E ratio of 66x reflects the premium valuation the market accords to companies with virtually irreplaceable products in the critical supply chain. The risk is sensitivity to investment cycles in the semiconductor industry and competitive pressure from Synopsys $SNPS, a major rival in this duopoly market.

NetEase $NTES

NetEase is China's second largest gaming company and one of the world's most important game developers. The company operates a portfolio including both massively popular titles for the Chinese market and internationally successful games. It reported annual revenue of RMB 112.6 billion in 2025, with year-on-year growth of just under 7%. Profit grew by nearly 14% due to a better product mix and cost control.

NetEase's ROE is 22.4%, which is remarkable for a gaming company. This efficiency is due to a combination of high margins on its own game titles, minimal debt (Debt-to-Equity of just 0.04) and disciplined capital allocation. The company holds more cash in its accounts than its total debt, which provides it with a significant financial cushion.

At a P/E ratio of under 15x, NetEase is by far the cheapest company on today's list. The analyst consensus is Buy with an average price target of around $157, which represents upside potential of approximately 30%. Additionally, the company is benefiting from Apple's recent $AAPL reduction in App Store fees in China, which should positively impact game developers' margins. The risk is the regulatory environment in China and the dependence on the success of new game titles.

Overview of key data

D/E = Debt-to-Equity ratio

Strategic view

Looking at all seven companies, several commonalities stand out.

1. The high ROEs of these companies are not the result of excessive debt, but reflect true operating efficiencies and strong competitive positions. Most of them operate with conservative debt-to-equity levels.

2. Firms differ significantly in their valuations. At one end of the spectrum, we find NetEase with a P/E below 16x and Tencent below 20x. At the other end is Cadence with a P/E of over 68x and Walmart over 45x. This dispersion reflects the market's differing expectations for future growth. Investors pay a premium for companies where they see a structural competitive advantage and a predictable growth story. But these are also the ones that carry more risk.

3. Most of these companies have direct or indirect exposure to the AI trend. Amazon is investing tens of billions in AI infrastructure through AWS, Tencent is doubling its AI spending, Intuit is integrating AI into its products, Cadence is benefiting from the growing complexity of AI chips, and Interactive Brokers is using the technology to automate trading. This AI exposure, while not speculative, is already embedded in real revenue and operational savings.

What to watch next

• Fed interest rate developments and their impact on valuations of growth companies, particularly for Intuit and Cadence

• Amazon's Q1 2026 results (due April 23), which will give clues about AI ROI

• Regulatory developments in China that may impact the valuations of both Tencent and NetEase

• The pace of trading volume growth at Interactive Brokers amid a potential decline in market volatility

• Walmart's ability to sustain growth in new segments (advertising, e-commerce) while maintaining margin discipline

• The evolution of investment cycles in the semiconductor industry, key to the sustainability of Cadence's growth

How to approach these companies

The combination of ROEs above 20% and strong analyst support provides a fairly exclusive filter that separates firms with real capital efficiency from those whose growth has been driven primarily by cheap debt and favorable sentiment. All 7 firms in today's selection managed to translate their competitive position into above-average returns on equity while maintaining the confidence of analyst houses.

At the same time, meeting these two criteria alone is not an automatic reason to buy. Valuations vary significantly from company to company and some are trading at historically tight multiples. Therefore, it is key to watch whether the fundamental thesis for each of these companies will be borne out in the coming quarters, especially in the context of the ongoing AI investment cycle, geopolitical tensions and a volatile interest rate environment.

At the same time, the “Win Now” turnaround program is deliberately putting extra pressure on near term earnings. Management is pruning lower quality product, resetting inventories and rebalancing channels away from an over concentrated DTC push, moves that create several points of revenue headwind and add severance and restructuring charges today, but are meant to leave a cleaner, more sport centric portfolio and healthier margin structure from fiscal 2027 onward.

How Q3 2026 turned out

• Revenue: $11.279 billion, virtually unchanged from $11.269 billion a year ago, down about 3% on a currency-adjusted basis.

• Nike brand revenue: approximately $11.0 billion, roughly +1%, with growth in North America offsetting declines in Europe and China.

• Wholesale: $6.5 billion, +5% (+1% on a currency-adjusted basis), the main driver of growth.

• Nike Direct: $4.5 billion, -4% (currency-adjusted -7%), digital -9%, owned stores -5%.

• Converse: $264 million, -35%, down in all regions.

Gross margin declined 130 basis points from 41.5% to 40.2%. The main reason for this is higher tariffs in North America, which increase unit costs of imports. Selling and administrative expenses increased 2% to $4.0 billion:

• "Demand creation" (advertising + sports marketing) was around $1.1 billion, roughly in line with last year - higher sports marketing spend and currency effects offset lower brand marketing.

• Operating overhead was up about 3% to $2.9 billion due to severance costs and foreign exchange effects, partially offset by lower other administrative expenses.

Operating profit fell from $844 million to $650 million (-23%). Net income fell from $794 million to $520 million (-35%), and diluted EPS fell from $0.54 to $0.35 (-35%). The effective tax rate jumped from 5.9% to 20% as last year's period included a one-time non-cash tax benefit that boosted after-tax earnings.

Inventories were approximately $7.5 billion, down 1% from a year ago - lower volume and a different product mix partially offset higher unit costs due to tariffs. Cash and short-term investments fell to $8.1 billion, about $2.3 billion less than a year ago, as operating cash flow was insufficient for a combination of dividends, debt repayments, capex and share repurchases.

• Full presentation with results.

Management Commentary

CEO Elliott Hill described the quarter as a period of "significant steps to improve the health and quality of the business". He stressed that the pace of improvement varies across the portfolio, but the areas that Nike has prioritised are already showing early signs of recovery. At the same time, he acknowledged that the work is not done, but he said the direction is clear and teams are moving forward with a focus on speed and discipline.

CFO Matthew Friend said third-quarter results were in line with internal expectations. However, he also pointed out that the "Win Now" actions will impact results for the rest of the calendar year - so the company expects the clean-up phase to be reflected in the numbers in future quarters. Still, he says the moves are intended to set Nike up for long-term profitable growth, even if they do worsen margins and earnings per share in the short term.

Why the stock fell after the results

The stock closed around $52.8 during the trading day, but fell about 9% to about $48.2 in after-hours trading after the results were released. There are several reasons:

• Earnings per share of -35% - EPS of $0.35 is noticeably weaker than last year, and the market is concerned that it will take longer to return to previous levels, especially since management itself says that "Win Now" will continue to hurt the numbers in the short term.

• Stagnant sales and weak Nike Direct - sales are flat and falling in real terms, with the channel that Nike has long bet on suffering the most: direct sales and digital. Investors see that growth is now being driven by wholesale, which has lower margins and less control over the brand.

• Pressure on margins due to tariffs - gross margins are down 130 basis points and the company says the main cause is high tariffs in North America. This isn't a simple problem that can be solved in one action - it's a structural pressure that can drag on for longer.

• Unclear turning point - management talks about discipline and long-term growth, but explicitly warns that restructuring will weigh on results for the rest of the year. The market doesn't see a specific point at which EPS and margins will start to clearly improve.

Combined with the trend of recent years (declining annual sales and earnings), this is leading investors to reassess valuations - Nike $NKE is still a strong brand, but no longer with a profile of steady earnings growth.

Long-term results

For the most recent full fiscal year (ending May 31, 2025), Nike earned $46.3 billion, down nearly 10% from $51.4 billion in the prior year. The year before, revenues were about $51.2 billion, and in 2022 they'll be about $46.7 billion, so the company is hovering around the same level rather than growing appreciably after a pandemic period.

Gross profit fell to $19.8 billion in 2025 from $22.9 billion in 2024. Meanwhile, operating expenses held high (around $16.1 billion, down only slightly from a year ago), so operating profit fell from $6.3 billion to $3.7 billion (-41%). Net income fell from $5.7 billion to $3.2 billion, and earnings per share from roughly $3.76 to $2.17.

The number of shares is down slightly due to share buybacks (about 1.61 billion diluted shares in 2022 and about 1.49 billion in 2025), but the drop in EPS is still significant. EBIT is down from about $6.3 billion to $3.9 billion, EBITDA from $7.2 billion to about $3.7 billion. So Nike has been generating solid profits over the long term, but the trend of recent years has been down, not up.

Shareholders

Insiders (management and directors) hold approximately 1.5% of Nike's stock. Institutional investors own around 83-85% of the stock and free float, which is typical of large US blue-chips.

The largest owners include:

• Vanguard Group with a stake of approximately 9.8%.

• BlackRock with approximately 7.7%.

• State Street around 5%

• Capital World Investors approximately 4%

This means that the share price is heavily influenced by the sentiment of the large funds. When they collectively decide that the growth profile and margins are weaker, there is a rapid repricing - just like after the current quarter.

News and action over the past quarter

• "Win Now" program - Nike continues a package of measures aimed at rapidly improving efficiency: layoffs, simplification of organizational structure, greater focus on key categories and channels. In the short term, this increases costs (severance), but should reduce the fixed cost base.

• Shift in sales structure - additional focus on wholesale partners in North America to clear inventory faster and strengthen share in key segments. This is a partial departure from the earlier "Direct-to-Consumer at all costs" strategy.

• Inventory Control - The company continues to work on reducing inventory in problem categories and better production planning so it doesn't have to rely as much on discounts.

• Marketing mix - Nike is shifting spending towards sports marketing (athletes, leagues, events) while pure brand advertising is more moderate. The goal is to strengthen the core brand in sports, not just lifestyle image.

• Dividend - The company paid about $609 million in dividends to shareholders in the quarter, up about 3% from a year ago, and continues to maintain a 20+ year streak of annual dividend increases.

What makes the Delta win more important than the raw fleet number is the strategic signal. Delta, which has been rolling out free Delta Sync Wi‑Fi across roughly 1,000 jets using Viasat and Hughes systems, is now adopting a multi vendor strategy where Amazon supplies both hardware and cloud links into AWS and digital services for 163 million SkyMiles members. In‑flight connectivity is becoming another front in the competition between hyperscale clouds, and with Kuiper Amazon is showing it is prepared to go after Starlink’s future high value airline revenue, not just challenge SpaceX in launch through Blue Origin.

What exactly Amazon has worked out with Delta

The deal calls for installing Leo terminals on Delta's new $DAL machines and launching the service in 2028 on flights within the continental US, with gradual expansion as the satellite constellation grows. According to internal materials referenced by WSJ and CNBC commentary, Leo is to offer speeds roughly 3-5X faster than Delta's current solution, with Wi-Fi to remain part of the "free Wi-Fi" for SkyMiles members - i.e., no direct paywall for passengers.

Delta now covers roughly 1,200 planes with satellite connectivity from Viasat and Hughes Network Systems, and boasts that 163 million passengers have already used its free Wi-Fi. So switching part of its fleet to Amazon Leo $AMZN means not only a technology upgrade, but also a realignment of relationships across the supply chain - from satellite operators to ground infrastructure to integration with cabin systems. For Amazon, it is also a reference contract where it can show other airlines that it can deliver end-to-end solutions: hardware, connectivity and cloud backend.

In addition, the contract is Amazon's second major win in the airline vertical in a short period of time. Back in 2025, JetBlue chose Project Kuiper (today's Leo) for roughly a quarter of its fleet, with the goal of deploying faster, free Wi-Fi by 2027. In combining with Delta, Amazon is thus acquiring two brands that pioneered free Wi-Fi in the U.S., building on their "stream anywhere" marketing - this time not over land lines, but via its own constellation of low-orbiting satellites.

How Amazon is building Leo and why it's a thorn in Starlink's side

Leo/Kuiper is one of Amazon's most ambitious side projects: the company has pledged at least $10 billion in capital to build a global satellite network to connect homes, businesses and carriers. As of April 2025, Amazon had put 214 satellites into orbit, and it plans to roughly double the pace in the next 12 months with more than 20 planned launches. In total, it has around 100 launches contracted with Blue Origin, United Launch Alliance and also SpaceX (Falcon 9), with an aggregate contract value of several billion dollars.

Yet today Starlink is far ahead technologically and infrastructurally: as of 2019, it has launched over 10,000 satellites and has become the largest satellite operator in the world, benefiting from its own reusable Falcon 9 rocket and vertical integration of manufacturing. This has allowed it to aggressively rollout services - from home internet to aviation, where it already has contracts with airlines such as United Airlines, Alaska Airlines, Hawaiian Airlines and, more recently, Southwest. Amazon is entering the field relatively late, but with a different package of advantages: money, cloud, a retail ecosystem and the ability to cross-sell services.

Amazon's weakness so far is the pace of deployment of the constellation. In January, the company asked the US FCC for a two-year extension to its original deadline (it was supposed to launch half of the 3,200 planned satellites by July 2026), prompting sharp criticism from SpaceX and regulators led by Commissioner Brendan Carr. Amazon has responded by saying it is doing "the best it can within what it can control" and that commercial launch of services is only "a few months" away - in small regions at first, with gradual expansion as the constellation grows.

Why Starlink has reason to be nervous about this for the first time

Starlink has so far built a position in aviation as almost the default solution for airlines looking to move Wi-Fi from a slow and paid service to a fast and often free one. Hawaiian, Alaska, United, and Southwest have gradually announced Starlink integration, creating a winner-takes-most effect: those who want to say "fastest Wi-Fi in the air" in marketing terms are reaching for Musk's network. The arrival of Amazon with Delta and JetBlue breaks this narrative - airlines now have two strong brands to pit against each other in tenders and price negotiations.

For Starlink, this threatens to erode future margins in a segment that at the same time has attractive unit economics: a commercial airline brings in thousands of paying passengers a day, with operating costs per megabit falling with each additional satellite. Adding Amazon to the game pushes up prices and bundling - Delta can combine satellite connectivity with AWS services (passenger analytics, content personalization, real-time operational data), something Starlink has no way to replicate.

Moreover, in terms of the strategic map, this is another front where Musk and Bezos clash indirectly: SpaceX vs. Blue Origin in launchers, Starlink vs. Leo in constellations, Tesla vs. AWS partners in automotive cloud solutions. Every big Delta-type contract now also takes on a symbolic "who the industry chose" dimension. For investors, this reinforces the thesis that satellite internet is no longer a one-horse race and that in verticals like aviation or trucking, Amazon and LEO can realistically bite off Starlink's piece of the pie.

What's in it for Amazon: more than just data in the air

Amazon isn't just gaining new revenue from connectivity with this move. Leo is another distribution channel for its entire suite of services - from Prime Video streaming to AWS cloud services to retail marketing. Delta talks about "the next era of connected travel and onboard digital experiences," which in practice means integrating personalized offers, entertainment and loyalty programs directly into the Wi-Fi portal. Who controls the onboard data feed can influence what services passengers see and what data goes back to the cloud.

Plus, it's a long-term locked-in customer. Installing terminals, certification for different aircraft types, integration into avionics and service contracts create high switching costs - an airline doesn't change its satellite Wi-Fi provider every year like a mobile operator. This gives Amazon a recurring B2B revenue stream with good usage insight (telemetry) and the potential to cross-sell other services (analytics, advertising platforms, cloud-based traffic solutions).

Another benefit is reputational capital in the regulatory and aerospace sectors. Contracts with major airlines help Amazon argue to regulators (such as the FCC) in favor of greater flexibility in constellation deployment timelines: it can argue that delays are not only technical problems, but also hinder innovation in key sectors such as aviation. From Blue Origin's perspective, moreover, each additional Kuiper/Leo satellite means an incentive to scale up New Glenn launchers more quickly - reducing Amazon's dependence on SpaceX to take out its own Starlink competition.

Anyone looking at the stock after that kind of run has to hold two ideas in their head at once. On one side sits a dominant, cash generative advertising machine that still earns tech leading margins and funds everything else; on the other is an AI and hardware investment program on a scale that recalls the original metaverse push, layered on top of more than 80 billion dollars in cumulative losses at Reality Labs that prove Meta is willing to go all in on long term bets. Whether this capex avalanche becomes the next great leg of the equity story or another expensive detour is exactly to otázka, kterou tenhle článek rozebírá číslo po číslu.

Top points of the analysis

• Q4 2025: revenue $59.9B (+24% YoY), operating income $24.7B (+6%), operating margin 41%, net income $22.7B (+9%), EPS $8.88 (+11% YoY) - all above expectations.

• 2026 outlook: Q1 2026 revenue $53.5-56.5B (above consensus), CapEx guidance $115-135B (vs. ~$72B in 2025, +73%), primarily into AI datacenters, GPUs and custom chips.

• Reality Labs: cumulative operating losses of over USD 80bn over the last 5 years, billions in losses again in 2025; Meta starts "pullback from metaverse", job cuts and more focus on AI products instead of pure VR/AR.

• Analysts' Predictions: 40+ analysts see Meta as a Strong Buy, then average 12-month target is ~$820-860, median around $850.

• Key disagreement: can Meta monetize AI (personalized assistants, recommendation systems, creator tools) so that higher CapEx yields higher future earnings, or is it repeating the Reality Labs pattern of burning FCF on an uncertain long-term outcome.

What has changed: from metaverse to "personal superintelligence"

Zuckerberg has practically reframed the company's story: from a "metaverse company" back to an AI-driven social & ads firm that additionally builds personal superintelligence for billions of people.

The reality is hybrid:

• The family of apps (Facebook, Instagram, WhatsApp, Messenger) is still the core of the business - accounting for $58.9 billion of the $59.9 billion in revenue in Q4 2025.

• Reality Labs had ~$1B in Q4 revenue (-12% YoY) but cumulative losses of over $80B, and Meta $META is reducing pace of investment, restructuring teams, laying off ~700 people as part of broader restructuring.

• AI is now officially "priority #1": Meta is building Meta Superintelligence Labs, buying GPUs in bulk, designing custom chips, building new datacenters, and pivoting product development as well (AI assistant in apps, AI for creators, AI search in social feed).

Zuckerberg explicitly described this as a shift from "metaverse first" to "AI first, metaverse later". This has two implications for investors:

• Positive: the AI story is much better understood by the market than the metaverse. AI helps the core ad business - better targeting, better performance, more engagement, more spend.

• Negative: CapEx numbers are massive, FCF is down and Real Labs is still burning billions.

How AI becomes money (and not just a second metaverse)

1) AI as the "engine" of the ad business

The most important layer is AI in advertising (the advertising arm). Already in 2023-2025, Meta has repeatedly said that improvements in AI recommendation systems will directly lift:

• the number of ad impressions

• relevance to users

• return on ad spend (ROAS) for advertisers

The Q4 2025 results confirm this: Family of Apps revenue +25% YoY, in an environment where digital advertising is growing less overall. Meta is getting a bigger share of digital ad budgets thanks to AI, and that's a net gain.

2) AI products for users and creators

Meta is introducing AI:

• As an assistant in Messenger, WhatsApp and Instagram

• as a content creation tool (generating images, text, video)

• as "personal superintelligence" - perzonalized answers, recommendations, search

In the short term, the monetisation of these features is low, but strategically there are two layers of value:

• They increase engagement (more time in apps → more ads).

• They open up space for new forms of advertising (branded AI experiences) and tools for creators/advertisers to charge a premium.

So AI is not just a "cool feature" but a direct multiplier for core business.

3) AI infrastructure and cloud monetization?

It's clear from the data above that Meta is primarily building AI for itself, not for external customers. Unlike Amazon or Microsoft, it doesn't have a traditional cloud, however:

• may eventually make some of its AI capabilities available externally (e.g. Llama-based APIs, B2B AI tools)

• already plays a role in open-source AI(Llama models) that increases its influence

CapEx shock: 115-135bn in 2026 - what this means for FCF and valuation

The CapEx plan for 2026 is the biggest number in all of "Big Tech" - $115-135 billion in one year, vs ~$72 billion in 2025. Most will go to:

• AI datacenters

• GPUs (Nvidia, custom chips)

• network infrastructure expansion

• partly to Reality Labs

Free cash flow in 2025 has already fallen 16% to $43.6 billion, even as revenue grows 22%. In 2026, we can expect:

• Further decline in FCF due to CapEx

• Strong EPS numbers (thanks to accounting - CapEx goes into depreciation gradually)

• FCF yield that may temporarily appear low (for cash-oriented investors)

This is why part of the market fears an "FCF squeeze": if Meta invests massively for several years in a row, the market may not be willing to pay P/FCF multiples at today's levels, even if the P/E looks reasonable.

Valuations and expectations

• Forward P/E: >25x on expected EPS (according to most models).

• Analyst consensus: Strong Buy, average 12m target ~$820-860, median $850, bullish scenarios up to $935-1,144.

• JPMorgan target: $825 (overweight), argument - revenue growth >20% even in 2026, better cost visibility.

• FCF 2025: USD 43.6bn, FCF margin ~30%, expected decline in 2026 due to CapEx.

Valuation - is this a falling knife or an AI overvaluation

Forward P/E around 25x on the company with:

• 20+% revenue growth

• 40+% operating margin

• 30+% FCF margins (when CapEx normalizes)

• extremely strong core business

Not overpriced per se - it's premium, but in light of AI and the competency position it makes sense.

The question is different: what's already priced into that price?

• AI ads effect: improving revenue growth to 20+% - we already see this in the numbers, so some of the AI value is "realized".

• Personal superintelligence, AI assistants, B2B AI: this is more of a long-term option that is not yet in the numbers.

For the investor who is "catching the knife" today, sensitivity is key:

• If Meta delivers 20% revenue growth and maintains a 40% margin, a P/E of 25× can be stable or even expand.

• If CapEx pushes FCF margin down over the long term and AI products don't deliver new monetization, P/FCF compression can drag the stock down even with decent EPS numbers.

Macro and competition

Meta is struggling on several fronts:

• Advertising: competition from Google $GOOG, TikTok, Snap $SNAP. AI creates differentiation - better performing ads vs. others. So far it looks like Meta is gaining share, not losing.

• AI: competes with Microsoft $MSFT + OpenAI, Google, Amazon $AMZN, but on a different playing field - not in enterprise AI, but in "consumer + social" AI. Here it has the advantage of data, scale and distribution.

• Metaverse/VR: Reality Labs vs. Apple $AAPL (Vision Pro), Sony $SONY, Valve. Here Meta is "deep in the red" so far with no clear timeline on profitability.

Macro risk comes primarily from the advertising cycle: if a recession hits and advertisers cut budgets, sales will slow, and AI alone won't compensate.

Why Meta may be undervalued

1) Mismatch between CapEx cycle and investor horizon

The CapEx shock ($115-135 billion in 2026) happens in one to two years, but the AI infrastructure that Meta is thereby building will generate revenue for perhaps five to ten years. Most DCF models and retail investors tend to:

• Penalize the short-term downside of FCF

• while conservatively modeling the long-term impact of AI on revenues and margins

The result is a "time mismatch" - costs are fully visible, but revenues are only partially priced by the market so far. If Meta can show that AI infrastructure in the next 2-3 years:

• Holds 20% revenue growth

• and sustainably raises ARPU.

the current discount may disappear faster than models predict.

2) The added value of AI in the "private graph" that public markets can't price

Most of the AI narrative is built on LLM and the public internet. Meta has something different: the private social graph of billions of people, their behaviors, relationships, groups, interests. The AI that sits on top of that dataset has:

• a different type of "moat" than the models trained on the web

• a unique ability to personalize feeds, recommendations, social search

This translates in numbers as:

• Higher engagement (more time in apps)

• better ad performance

• less sensitivity of advertisers to the price per ad (because ROI is higher)

The market often values AI by "model parameters" and GPU spend, but doesn't underestimate the uniqueness of the data set - and this is where Meta can win over companies that have great models but less unique data. That's a layer of value that a standard P/E or P/S multiplier can't capture.

3) The Hidden Value of Reality Labs Restructuring

Reality Labs today looks like a pure capital burn - $80+ billion in accumulated losses, paltry revenues. But Meta:

• Already actively cutting back at Reality Labs (layoffs, reorganization)

• and is shifting some of the team and technology towards AI and "mixed reality" use-cases that may be more monetizable.

Implicit in the price today is the "Reality Labs = perpetual loss" scenario. Once Meta:

• Clearly limits the annual cash burn

• or shows the first product linkage between RL and AI that makes business sense (e.g. AI-driven productivity, gaming, B2B collaboration)

there is room for positive re-rating as the market re-evaluates how much of the capitalization RL "eats".

4) The unappreciated effect of buybacks combined with high growth

Even after CapEx, Meta generates tens of billions of FCF per year and has a long history of aggressive equity buybacks. If:

• EPS grows at a double-digit rate

• and the share count is declining in real terms every year

Standard valuation multiples (P/E) often understate the true growth in value per share. Investors look at headline revenue and EPS growth, but are less attuned to the effect that every future dollar of profitability is attributable to fewer shares. This creates situations where a company looks "fairly valued" on a P/E basis, but from a per-share compounding perspective is cheaper than it appears.

5) Psychology after the "Zuck comeback"

Meta has had a "comeback story" - after falling below $100 in 2022, it has become one of the biggest winners of the AI rally. Psychologically, this creates two effects:

• a portion of investors still have the old Metaverse "faceplant" in their heads and are extremely skeptical of any big CapEx announcement

• another part is "late to the party" and waiting for the perfect entry

This may keep sentiment and multiples lower than would be consistent with pure fundamentals until Meta $META delivers several consecutive quarters where:

• Revenue growth

• Margins

• and FCF

clearly show that AI CapEx was an investment, not another metaverse gamble.

Risks

1) CapEx overshoot

115-135 billion USD in 2026 is a brutal number. If:

• Adequate revenue/EPS growth doesn't materialize, or AI products fail to monetize.

Could repeat Reality Labs pattern - years of FCF pressure with no clear ROI. This would lead to valuation compression even if the core business remains strong.

2) Reality Labs "black hole"

80+ billion in cumulative losses, yet no clear path to significant revenue. Unless Meta makes a clearer capital allocation for RL (e.g. annual loss cap, milestone-based investing), RL may continue to drag down FCF and margin.

3) Regulation

Antitrust, privacy, content moderation. Meta is permanently under political and regulatory pressure in the US and EU. Any interference (data restrictions for AI training, fines, restrictions on targeting) can reduce the effectiveness of ads and limit growth.

4) AI competition

If AI capabilities become a commodity (everyone has similarly good recommenders and assistants), the Meta vs. other differentiation will fall. In that case, AI investments would be a necessary hygiene investment, but not an edge that would justify higher valuations.

5) Post-run sentiment

Meta has more than tripled since 2022. Even small negatives (e.g. weaker guidance, signals that AI products are not delivering expected returns) can lead to sharp corrections as a portion of investors are "hot money" looking for AI stories.

Investment scenarios

Optimistic scenario

• Revenues grow 20-22% annually 2026-2028.

• Operating margins hold at 38-42% even with high CapEx.

• AI products (assistants, tools for creators) start to deliver a new revenue stream.

• Reality Labs partially restructures, losses are capped.

• P/E remains 25-28x on EPS approaching e.g. $45-50 in 2028.

→ Price >$1k over 3-4 years, 12-15% annual return.

Realistic scenario

• Revenue grows 15-18%, margin gradually declines to 35-38% due to investments, but FCF stabilizes after 2026.

• AI is a strong tailwind for ads, but new AI products are more of an add-on.

• Reality Labs remains loss making but in a sustainable range.

• P/E compresses slightly to 20-23x on EPS of $35-40.

→ Price around $800-920, annual return of 8-11%.

Pessimistic scenario

• CapEx becomes unsustainable - FCF falls, market starts looking at P/FCF.

• Revenue slows below 12-14% due to weaker advertising, AI effect wears off.

• Regulatory/antitrust restrictions hit Meta's data advantage.

• P/E compression to 15-18x, EPS $30-35, sentiment is tipping.

→ Price $450-630, potential -10 to -35% from today's level in one of the worst case scenarios.

What the investor should take away

• Meta has an extremely strong core today - advertising with 40+% margin, 20+% revenue growth and one of the strongest AI stacks in the consumer world.

• AI is not "new business" but a turbo for existing business - giving Meta more share of digital ads, more engagement, more spend.

• CapEx of 115-135bn in 2026 is two-sided: could create a huge moat, or an FCF trap - key will be how quickly AI monetization shows up beyond pure ad improvement.

• Reality Labs remains a black hole, but the firm is also clearly "flipping" priority to AI - investors should watch whether RL remains a controlled experiment or rampant capital burn.

• Valuation is not cheap, but not meaningless either - forward P/E >25x reflects real growth and AI optionality; upside is interesting, but volatility will be high.

As investors, we are once again in "headline trading" mode: one leaked report about a possible end to the war, unconfirmed and without details, is enough to trigger a positive reaction in stocks, even though the fundamentals — a closed Strait of Hormuz, uncertain supplies, strained reserves, and the risk of persistently higher energy prices — remain unchanged. This is exactly the kind of moment when it's worth asking whether we're responding to a real shift in reality, or to politically convenient "spin" of the story that tomorrow's headline may rewrite within a few hours.

The Fed at a crossroads

On 30 March 2026, US Federal Reserve Chairman Jerome Powell spoke to nearly 400 students at Harvard University as part of an introductory macroeconomics course. His address was one of his last public speeches before his second term as Fed chairman expires on 15 May 2026. Markets followed his words with extreme attention, not only because of the content but also because of the context in which the speech was delivered.

The Fed is currently in an unprecedented situation. The key interest rate remains in the 3.50% to 3.75% range, where the central bank moved it after three quarter-percent cuts at the end of 2025. The oil shock caused by the conflict in the Middle East, which has entered its fifth week, has driven the price of Brent crude above $112 a barrel and brought new inflationary pressures. The core personal consumption expenditures indicator, excluding food and energy, is holding at 3%, still well above the Fed's 2% target. Adding to all this is the question of who will be running the central bank in six weeks.

Powell's speech at Harvard

Rates are in the right place

In a discussion moderated by Professor David Moss, Powell repeatedly emphasized that he believes the current monetary policy settings are appropriate for a wait-and-see approach. The central bank, he said, is in a position where it can monitor developments without having to react immediately by changing rates. The key argument was precisely the temporary nature of the oil shock. Powell recalled that energy shocks have historically come and gone quickly, while the effects of monetary tightening have been much delayed. If the Fed were to raise rates now, the effect would come at a time when the oil shock would likely have passed and the economy would be unnecessarily burdened.

This wait-and-see setup is confirmed by market data. According to CME FedWatch, the markets are currently assigning a 95.3% probability that rates will remain unchanged at the April FOMC meeting. The probability of a rate hike is just 4.7% and markets are not currently pricing in a cut at all. The dot plot from the March meeting still points to one cut during 2026 and another in 2027, but most analysts expect that under current conditions it will not occur until the second half of the year at the earliest.

Powell's speech at Harvard

Inflation, tariffs and the oil shock

Powell candidly named the three main inflationary layers facing the US economy.

• The first is tariffs on imports, which he estimates add about 0.5 to 1 percentage point to inflation.

• The second layer is the oil shock itself, which directly affects energy costs and feeds through to the prices of goods and services.

• The third layer is structural factors in the labour market, where labour shortages persist in some sectors and wages in services remain above a level compatible with the 2% inflation target.

In its March projections, the Fed revised its estimate of PCE (Personal Consumption Expenditures) inflation for 2026 to 2.7% in both headline and core inflation components. This represents a significant shift from the December projections, when the central bank expected headline inflation at 2.4% and core inflation at 2.5%. Meanwhile, Powell said that without the impact of tariffs, inflation would be closer to the 2% target, suggesting that trade policy remains one of the main factors preventing a return to price stability.

Private credit market under scrutiny

Powell's comments on the private credit market also attracted market attention. In recent months, nervousness has increased in this segment after the failed merger of Blue Owl Capital sparked a wave of requests to buy back fund shares. The situation has also been complicated by concerns that artificial intelligence could disrupt the business models of software companies whose bonds make up a significant portion of private credit fund portfolios. A direct example of this was the sell-off in $BLK shares, which fell by as much as 7% in a single day. Our analyst team has produced a comprehensive video about it on the Bulios channel.

Powell called the situation a correction, not a systemic risk. He stressed that the Fed is monitoring the links to the banking system and looking for potential channels of a possible emerging problem, but has not found them yet. Still, he acknowledged that regulators are paying close attention to this fast-growing segment. An important milestone will be June 30, 2026, when private credit funds and investment companies will publish semi-annual reports and have to revalue their positions to fair market value, which may bring the first real transparency about the extent of losses in the sector.

Artificial intelligence and the labour market

In a section of the discussion focused on the future of work, Powell advised students to invest time in mastering artificial intelligence tools. He identified large language models as tools that significantly increase productivity, and admitted that he himself actively uses them. But he also acknowledged that graduates are entering one of the most difficult job markets in recent years, with both technological changes and adjustments in immigration policy contributing to weaker job creation.

The question of Powell remaining in office

Transition of power: Powell versus Warsh

President Donald Trump has nominated Kevin Warsh to be the new Fed chairman on January 30, 2026. Warsh, a former Fed governor from 2006 to 2011, is known for his historically hawkish stance on inflation, but has recently shifted his position, arguing in favor of lowering rates that productive gains from AI technology allow monetary policy to be eased without the risk of rekindling inflation.

The formal nomination was submitted to the Senate on March 4. But the confirmation process has become complicated. Senator Thom Tillis has announced that he will block any nominee for Fed chairman until the investigation into the $2.5 billion renovation of the central bank's headquarters is resolved. Meanwhile, US Attorney Jeanine Pirro subpoenaed Powell for questioning in connection with the renovation, but the court quashed her subpoena and she appealed the decision. The Senate Banking Committee has scheduled a hearing on Warsh's nomination for the week of April 13.

If the Senate does not confirm Warshe by May 15, when Powell's term as chairman expires, Powell could technically remain in office as acting chairman by virtue of his membership on the Board of Governors, which runs through February 2028. That situation would create an unprecedented period of uncertainty about the leadership of the world's most important central bank.

Jerome Powell, Kevin Warsh

What a change in Fed leadership would mean for markets

Historically, every change in the Fed's leadership brings a repricing of assets across the yield curve.

The yield curve is a chart that shows what interest rates (yields) government bonds have depending on their maturity. On one side are short-term bonds, such as 3 months or 2 years, and on the other side are long-term bonds, such as 10 or 30 years. This chart shows investors how the market perceives future developments in the economy and interest rates).

The transition from Alan Greenspan to Ben Bernanke in 2006 heralded a fundamental change in the approach to crisis management. The transition from Bernanke to Janet Yellen was key to the exit from quantitative easing. And the arrival of Powell in 2018 brought a period of quantitative tightening that generated significant market volatility.

Warsh's arrival could bring a combination of lower short-term rates and more aggressive Fed balance sheet reduction. Analysts at JP Morgan $JPM expect Warsh to push through a series of rate cuts once he takes office, with the first quarter-percentage-point cut coming as early as June 2026. At the same time, however, the Fed could shift from a passive approach to actively selling securities from a portfolio that currently stands at about $6.6 trillion. This approach, referred to as QT-for-cuts, would create conflicting pressures on different parts of the yield curve.

Market context and reaction to Powell's speech

Source: FOMC projections (March 2026), Trading Economics, EIA

The S&P 500 Index fell 0.39% on Monday, March 30 and was approximately 9% below its all-time high. The technology sector led the decline with a drop of over 1%, while financials and value stocks posted modest gains. The VIX volatility index crossed the 30 mark, signaling increased investor nervousness. Oil remains a key factor. Brent crude oil rose over 55% in March, the steepest monthly increase in the benchmark's history.

The bond market reacted to Powell's speech with a drop of around 10 basis points in US 10-year Treasury yields. The markets interpreted his words as confirmation that a rate hike was out of the question for the remainder of his term. With the probability of a rate hike at just 2.2% at the April meeting, the most likely scenario is that Powell leaves office with rates exactly where they are today.

A strategic view

For investors, the current situation represents a classic case of tension between mandates. The Fed faces both downside risk on the labor market side, which would play in favor of keeping rates low, and upside risk on the inflation side, which in turn could require a more restrictive approach. Powell has openly named this dilemma and acknowledged that the central bank is actively operating with these inputs.

From an asset allocation perspective, the Fed's wait-and-see stance creates an environment that favors defensive sectors and shorter-dated bonds. The energy sector is the only segment of the S&P 500 in positive territory for March with a gain of over 9%. Conversely, growth technology stocks remain under pressure as higher inflation expectations reduce the present value of future cash flows.

The political dimension of the Fed's leadership transition introduces a type of risk to markets that cannot be quantified by standard models. Should fears of an erosion of central bank independence be confirmed, historical precedents from the 1970s show that the consequences could be far-reaching. At that time, political pressure from President Nixon on then Fed Chairman Arthur Burns contributed to keeping rates too low, allowing inflation to spiral out of control. The subsequent tightening under Paul Volcker required rates to rise to 20%, plunging the economy into recession.

What to watch next

• Kevin Warsh Hearing (week of April 13): The key will be what stance Warsh takes on Fed independence and whether he will reaffirm his recent dovish rhetoric on rate cuts. Senators' questions will suggest how complicated the path to confirmation will be.

• March inflation data (CPI and PCE): the first hard data to capture the effect of the oil shock. The key will be whether higher energy costs start to spill over into core inflation or remain confined to headline.

• April FOMC meeting (April 29): penultimate meeting under Powell's leadership. Markets expect rates to be left unchanged; tone of communication and any dissents will be key.

• Oil price developments and the situation in the Strait of Hormuz: Analysts warn that a crossing of the $120 per barrel mark for Brent could trigger a recession. Conversely, any progress in diplomatic negotiations could bring a rapid decline in prices.

• Private credit funds' semi-annual reports (30 June): the date when the true extent of losses will become apparent. Revaluation to fair value may trigger another wave of redemption requests.

The Fed's stance

Powell's speech at Harvard yesterday confirmed that the Fed is sticking to a wait-and-see strategy in the final weeks of his tenure. The central bank has decided to overlook the oil shock as a temporary phenomenon and not to resort to further tightening even though inflation remains well above target. This decision is slightly positive for equity markets in terms of funding costs, but it also signals that the Fed sees enough risks on the growth side to tolerate higher inflation rather than risk a contraction in the economy.

The question of a change in Fed leadership adds an element to the equation that markets are only beginning to fully appreciate. Warsh's ascension could bring a paradoxical combination of lower short-term rates and higher long-term yields, which would fundamentally change the environment for asset allocation.

Investors should keep a close eye not only on macroeconomic data in the coming weeks, but also on the political process surrounding the confirmation of a new Fed chairman, as this may be as important to market sentiment as the inflation and employment numbers themselves.

But at the same time, it's not just about the logo on the pitch. The partnership is designed to bring tangible benefits to cardholders, strengthen fans' attachment to the Amex card and demonstrate whether the company can turn an expensive sports contract into measurable revenue growth and customer loyalty.

What exactly is Amex gaining

The deal gives Amex $AXP cardholders priority access to advance ticket sales, special on-site experiences and other benefits at select NFL events in the U.S. and abroad. This includes the Super Bowl, NFL Draft, international league games and other "marquee" events where Amex plans branded zones, priority entries and promotions.

As a first swallow, Amex is opening early-access to pre-sale for the Los Angeles Rams - San Francisco 49ers game in Melbourne in September 2026, the first ever NFL regular season game in Australia. The partnership fits into Amex's broader expansion in sports: the company now has over 50 contracts with leagues, teams, arenas and major events around the world and is now the official payment partner of MetLife Stadium, Mercedes-Benz Stadium, the New York Giants, Jets and Atlanta Falcons.

How much money is involved and why the NFL was on the hook

Noofficial terms have been announced, but according to the Sports Business Journal, the deal is a seven-year contract worth roughly $910 million to $950 million - or roughly $130+ million per year. That's roughly 2.5 times what Visa, which had been an NFL partner since 1995 and decided not to renew the contract after 30 years, was paying for the same category.

For the NFL, it's a net boost in sponsorship revenue, plus the ability to separately monetize other financial categories like retail banking or peer-to-peer payments that were previously bundled with Visa $V. For Amex, it's an expensive but strategic entry into the biggest sports business in the U.S., where fan loyalty often translates into payment card choices.

What American Express hopes to get out of this

Amex has long built its brand on exclusivity - "membership" is supposed to bring access to things others can't. The NFL deal extends that story to the sport with the highest viewership in the US.

The company can promise three main effects from this:

• image enhancement and acquisition channel: fans who switch from another card to access pre-sales and league-related rewards because of the NFL Extra Points Amex card benefit

• Higher spending: Amex cardholders will be incentivized to pay with Amex at stadiums and around NFL events (merch, travel, hospitality), which raises transaction volume and fees for Amex

• Deeper data on premium customer behavior that overlaps with NFL audiences (corporate hospitality, high-spend segment), which can be leveraged in other products and partnerships

On the product side, Amex is set to launch the NFL Extra Points American Express credit card issued by Comenity Capital Bank in the US this year to reward NFL-related purchases - from tickets to merch to travel. This creates a direct channel for the company to monetise its fan base outside of game time itself.

Possible scenarios

1) Partnerships as a growth engine

The NFL still has the highest viewership in the US and the league is aggressively expanding abroad - London, Frankfurt and now Melbourne. If Amex can systematically use presales and experiences as a magnet for new clients, the NFL could become one of the most important acquisition channels for premium and co-branded cards, especially in the US and select markets. In such a scenario, an annual investment of around $130 million can pay off in the form of higher fees, interest income and client loyalty that competitors (including Visa) cannot replicate.

2) Brand "must-have" but financially only neutral

The more moderate option is that the NFL deal will boost brand perception, but the benefit in numbers will be more spread out - hard to separate from the entire Amex sports portfolio, which already includes dozens of leagues and events. The partnership, then, fulfills a primarily marketing role: Amex is "visible wherever there is something going on", but the return on an individual contract cannot be accurately calculated. In such a scenario, the bottom line is that the NFL simply must have an Amex - because if it weren't there, the space would be occupied by competitors.

3) An overshot bet on a crowded advertising market

The pessimistic scenario assumes that sports marketing in the United States is already so saturated that the next big contract brings dilution rather than enhancement of the effect. The NFL is full of partners, fans are overwhelmed by promotions, and a portion of the audience additionally watches games in environments where Amex benefits are not relevant (international streaming, secondary ticketing). In addition, if the league's growth abroad does not meet expectations, a seven-year, nearly $1 billion contract could retroactively appear to be an expensive ticket into a space where Amex is hard to differentiate from the competition.

What this means for the league and for Visa

For the NFL, the arrival of Amex is a clear demonstration of strength: after thirty years, it was able to replace Visa with a new partner for roughly 2.5 times the annual fee and still retain the ability to sell other financial categories separately. The league also fits into Amex's portfolio as a global brand - which is important for international matches and building the NFL as a global sports brand.

Visa loses a long-standing "hero" contract but gains the freedom to redirect marketing money elsewhere - to the Olympics, football, digital payments and fintech, where it may be more strategically positioned. So it is essential for investors to watch whether Amex's NFL partnership will bring a visible acceleration of acquisition and spendy on key cards, or remain primarily a symbol that Amex belongs to the clubs of brands that can afford the most expensive sports rights in the world.