It’s a single layer of carbon atoms packed into a hexagonal honeycomb that is 200x stronger than steel, lighter than paper, and more conductive than copper. It could supercharge EV batteries, double the lifespan of structural concrete, and eliminate rust almost entirely.

But there was always a catch. You could either produce high-purity graphene in tiny, expensive batches, or you could make low-grade carbon soot by the ton. This is what the industry calls the Graphene Paradox.

For years, every company in the space was stuck on one side of that dilemma.

Until now.

1. The Hyperion Method

HydroGraph Clean Power Inc. (CSE: HG or OTCQB: HGRAF) invented a revolutionary method of producing high-purity graphene at an industrial scale.

Traditional graphene production starts with graphite through a method known as liquid-phase exfoliation — essentially, it’s when graphite gets stripped apart layer by layer using harsh acids, massive water consumption, and high amounts of energy. This method is inconsistent, impure, and accompanied by toxic industrial waste. It’s the side of the paradox most commercial producers have been on — producing tons graphene at low quality.

HydroGraph’s patented Hyperion reactor system works from a completely new perspective. Instead of tearing graphite layers down into graphene, Hyperion builds graphene up.

The process: feed acetylene and oxygen (two cheap, abundant gases) into a compact modular reactor, then trigger a controlled explosion. In less than a second, the gases convert into 99.8%+ pure graphene.

HydroGraph’s Hyperion method has all the purity of a lab and all the output of commercial producers — at only 70% of the cost per ton.

2. Hyperion’s Incredible Unit Economics

This breakthrough means nothing if the numbers don’t hold up. HydroGraph’s do.

A single Hyperion reactor costs less than $150,000 and only takes two to three months to build. Once running, that one reactor can over 10 metric tons of high-purity graphene per year — converting pennies worth of gas into millions of dollars of high-purity graphene with nearly an 80% profit margin per batch.

And because each reactor is modular and identical, HydroGraph doesn’t need to raise billions for a megafactory when they want to increase their production capacity. They can scale to consumer demand as needed by just building a modular addition; which is reminiscent of SaaS-like agility usually not found in the materials world.

3. The Asymmetry

Here’s where the story shifts from an interesting technology to a generational opportunity.

Graphene is a performance multiplier that touches everything physical. Since HydroGraph can produce it at purity and scale through their revolutionary Hyperion technology, their total addressable market isn’t just one sector. It’s the entire global physical economy — every bridge, skyscraper, building, battery, and foundation.

In 2017, Nvidia was trading at $4.50 per share. It had just released its now-famed GeForce GTX 1080 Ti GPU, a foundational innovation in the GPU world. At the time, Nvidia was a niche gaming company. What the market missed was that Nvidia had built the engine today’s AI boom (or bubble) is running on. Since it traded at $4.50 a share, it’s risen 4600% — a 46-bagger. This is because Nvidia’s GPU turned out to be the essential infrastructure for multiple massive markets at once.

Today, HydroGraph is trading at the same price, with the same type of advantage over its competitors, subject to the same market blindness. In nine years, imagine how a graphene-based world would value HydroGraph.

HydroGraph’s upside is only contained by the size of the global physical economy — this is the very definition of asymmetry.

Trading around ~$4.50, HydroGraph is priced like an R&D-stage speculative biotech firm. But the Hyperion system is built, the purity benchmarks are proven, and the commercial transition is underway. The gap between where this stock sits today and where the technology points is wide — and that’s exactly the kind of asymmetry I look for.

Yes, it certainly trades at a premium compared to today’s financials. A $2B market cap based on only $60K of FY2025 revenue results in an astronomical 10,000x P/S ratio. serves to illustrate the market’s demand for near-perfect future execution. Any missteps, share dilution, or growing pains could send HydroGraph’s share price cascading downward. For value investors, it’s certainly far outreaching its calculated intrinsic value — but as a growth investor, HydroGraph is a tempting ticket with virtually unlimited reward.

“Given a 10% chance of a 100x payoff, you should take that every time.” — Jeff Bezos

Growth investing isn’t about avoiding risk; it’s about finding asymmetry.

But if HydroGraph scales the way their total addressable market and unit economics suggest they can, the upside here is measured in multiples instead of percentages.

Ten years from now, high-purity graphene will likely be the backbone of the physical economy. The companies that can produce it — reliably, cleanly, and at scale — will be the ones that matter.

Right now, HydroGraph is one of the only companies on Earth that can check all three boxes. Neither its competitors nor the market have caught up yet.

But out of the SaaSpocalypse emerges a company that won’t just survive the landscape—it’ll rule it through its dominance and superiority in the world of advertising.

After returning a famous 700% in 2024 and another 100%+ in 2025, AppLovin ($APP) has spent the first half of 2026 mired in a mildly downward trend.

Caught in a storm of profit-taking, unfounded short-seller noise, a highly competitive ad-tech realm, and generalized SaaS anxiety, shares have slid from an all-time high of $745 down into the low-$480s1. Even with a post-Q1 earnings rebound pushing it back to around $484, the stock remains down roughly 25% year-to-date.

This 25% discount (as well as 34% to ATHs and 45% to consensus analyst price targets) is truly a gift from the market. This is a business growing revenue at 59% year-over-year (YoY), printing an 84% EBITDA margin (!!!), and boasting a 72% Free Cash Flow margin— all in the 95th percentile among its S&P 500 peers.

1. The Financial Profile: An Elite Class of One

To understand the absurdity of $APP’s valuation, look closely at its structural profitability. In its fresh Q1 2026 earnings report, AppLovin dropped a financial bomb on the street:

• Revenue: $1.84 billion (up 59% YoY).

• GAAP Net Income: $1.21 billion (up a staggering 109% YoY).

• Adjusted EBITDA: $1.56 billion, translating to an jaw-dropping 84% EBITDA margin for the quarter.

Let’s contextualize this. NVIDIA, operating at the peak of a historic semiconductors monopoly, operates with EBITDA margins hovering around 65%. Visa and Mastercard—the gold standards of asset-light businesses—sit around 68% to 70%. Simply put, no business is as effective as AppLovin in converting revenue to cash.

2. The AXON 2.0 Engine: The AI Advantage

The market frequently mischaracterizes AppLovin as just another mobile gaming ad network. This is a fundamental misunderstanding of their structural moat. AppLovin is an AI advertising company whose initial sandbox happens to be mobile gaming.

At the heart of the business is AXON 2.0, a proprietary machine learning engine. AXON executes millions of ad auctions per second. It doesn’t look at oft-lagging historical demographic data; instead, it processes real-time contextual signals to predict the near-exact Return on Ad Spend (ROAS) an advertiser will get for a specific placement.

The advantage of AXON is extreme. Whereas Meta and Taboola face stagnating ROAS metrics due to low engagement and lagging data, AppLovin is able to convert more of the advertising spend of a customer into actual results.

A comparison (table made with Claude):

2b. Why AXON Is Safe From Replication

The bears argue that rival platforms or startups will replicate AXON’s code — as they see it as just another SaaS wrapper for AI. They can’t, because AXON’s power isn’t just the math—it is the data.

AppLovin owns the mediation layer between advertising/cloud and local usage, meaning it controls the real-time bidding infrastructure for the vast majority of global mobile apps.

1. More advertisers spend on AppLovin because it delivers the highest ROAS.

2. Higher spend grants AXON more real-time auction data.

3. More data trains the model to become exponentially smarter at predicting user actions.

4. Smarter targeting increases ROAS, attracting even more advertisers.

If a competitor builds a copycat algorithm tomorrow, they have zero data to feed it. They would be forced to underwrite auctions blindly, losing millions of dollars in ineffective ad spend while AppLovin’s AXON engine prices risk and conversion metrics to the millimeter.

3. David & Goliath Comparison

A key sentiment dragging AppLovin down in early 2026 was the fear that Meta, with its industry-standard Advantage+ AI advertising suite, would march into AppLovin’s territory and crush them. This narrative is completely false. In direct, head-to-head performance comparisons within mobile environments, Meta’s suite is consistently worse-performing than AppLovin’s AXON, turning fewer eyeballs into paying consumers.

Meta’s AI models are built to optimize for user engagement inside a closed ecosystem (i.e. Instagram, Facebook, Threads). Meta tracks consumer demographic and behavior data across its famous privacy-last policies.

Meta’s biggest vulnerability is its reliance on tracking you across the web to build a user profile. If Apple or Google change their privacy settings, Meta’s ability to collect data is harmed. AppLovin doesn't have this problem because it doesn’t care who you are; it cares what you are doing.

By pulling data straight from its mediation network, AXON targets users based on actions inside apps. Regulators are clamping down on traditional ad-trackers, but because AppLovin is baked directly into the app infrastructure itself, its data collection is naturally built to survive the privacy crackdown (though AppLovin is the subject of an ongoing SEC investigation into their own data collection practices2.)

Meta: uses data on who you are online to improve user targeting.

AppLovin: uses data on what you do in-app to improve user targeting.

4. The Explosive Angle: Entering E-Commerce

If AppLovin were only planning to be just a mobile gaming ad network, it would be a highly profitable cash cow — like it is now. But the true explosive opportunity lies in its massive, ongoing product expansion: The Self-Serve AXON E-Commerce Advertising Platform.

For the past couple years, AppLovin’s e-commerce tools were restricted to a tight group of beta testers. Despite the restrictions, roughly 57% of qualified leads (highly engaged viewers) converted directly into live, paying consumers, with a customer acquisition cost (CAC) breakeven timeline of an astonishing 30 days.

This is jaw-dropping. According to metrics from EightX, Meta’s advertising CAC breakeven time is well over six months for most users and its qualified lead to consumer pipeline is well under 20%. — vastly underperforming AXON.

In June 2026, AppLovin is scheduled to launch its self-serve AXON e-commerce platform into the public sphere. This opens the floodgates to millions of independent web and e-commerce brands worldwide.

By applying AXON 2.0's advantage to retail, AppLovin is positioning itself to become the performance ad engine for the entire open web, allowing Shopify merchants and direct-to-consumer (DTC) brands to bypass the expensive, saturated ad channels of Google and Meta.

This is the asymmetric angle. The core mobile advertising business is delivering industry-leading fundamentals, yet at current levels the stock trades at a discount of over $200 per share relative to consensus FY2026 analyst targets — presenting investors a moment in time to buy the possible future gold standard in e-commerce.

Let’s review this setup:

Investors have the chance to buy a company guiding for over 50% revenue growth in the upcoming quarter, maintaining an 84% EBITDA margin, and sitting on a brand-new e-commerce catalyst—all for only a forward earnings multiple of 30.

Subscribe now

You’d expect a company like this to trade at an all-time high. Instead, it’s currently sitting at its lowest valuation in years.

Enter Adyen ($ADYEY in US markets). Powered by their ironclad position within the mega-cap tech realm, superiority of their streamlined payment infrastructure, and explosive growth in new SaaS business banking products, Adyen is a high-growth fintech trading at the valuation of a stagnating legacy IT company.

Adyen’s technology simplifies how large businesses accept and manage money. Instead of forcing companies to use a messy patchwork of separate credit gateways, local banks, and third-party systems, Adyen replaces the entire chain with a single, unified software platform. Through this one system, global enterprise giants (like Uber, Spotify, and Microsoft) can seamlessly process online, mobile, and in-store payments anywhere in the world while using real-time data to prevent fraud, boost checkout success rates, and offer embedded business banking services (high-margin SaaS offerings).

1. Simplicity Advantage

Payment processing is the nervous system of global commerce. In an era of currency volatility, fraud, and distressed trade relations, multinational enterprises are desperate for simplicity. The old model of international expansion required stitching together a dozen regional acquirers, gateways, and risk management tools. It was a bloated mess.

Adyen’s advantage is in recognizing that the future belongs to a single, unified global platform. They don't just process payments; they act as the source for an enterprise's financial flows. When a customer taps a card at a convenience store register in London, that transaction flows through the exact same backend as an online mobile purchase in Tokyo.

If legacy infrastructure is a complicated zig-zag through international terminals, conversions, and banks, Adyen’s is a smooth, straight line from Point A to Point B.

2. Big Tech Exposure

And here is the strategic angle that the market is entirely not realizing: Adyen is the shadow beneficiary of mega-cap tech growth. As its primary source of revenue lies in taking small commissions for both the amount of the payment and completing the payment, a higher frequency of larger payments is directly how it’s possible for Adyen to make $662M of revenue in Q1 alone (a 16% QoQ increase).

When you buy Adyen, you are essentially buying exposure to the continued dominance of the world’s largest tech monopolies.It’s the quiet financial engine running underneath the tech giants — trading far below their astronomical valuations.

As these mega-caps continue to scale at rapid rates, Adyen automatically scales alongside them. If Uber launches in a new country, Adyen captures the volume. If Microsoft’s cloud revenue surges globally, Adyen takes its toll on every dollar. You get direct exposure to the relentless compounding of Big Tech, but without having to pay the high premiums of the AI boom.

3. Adyen and AI

Theory says that as AI agents become more sophisticated, they will begin executing transactions on behalf of users (e.g., an AI assistant booking your flights, optimizing your SaaS subscriptions, or reordering your groceries).

Traditional payment gateways are designed for humans typing in CVV codes to ensure authenticity (that’s the security code on your credit or debit card!); they will break down under the nature of AI transactions.

In 2025, Adyen rolled out Adyen Uplift. Instead of relying on static, human-centered checkout protocols that block legitimate buyers (i.e. CVV codes!) or lets fraud slip through, Adyen's AI models analyze the data on every transaction in milliseconds for authenticity.

Adyen’s dynamic, intelligence-first layer is purpose-built to recognize and authorize agent-led commerce safely — they’re literally building the financial rails for the AI economy. Adyen remains one of very few enterprise payment processing ecosystems prepared to adapt to an agentic AI world.

4. Undervaluation Thesis

Currently trading around €943 per share1, Adyen sits at a market cap of roughly €30 billion. This implies a trailing Price-to-Earnings (P/E) multiple of around 28x based on their FY 2025 earnings per share (EPS) of €33.72.

Let’s analyze that. Management is guiding for 20% to 22% net revenue growth in 2026 and targeting EBITDA margins expanding past 55% by 2028. They have virtually zero debt and operate with CapEx at only 5% of revenue — meaning Adyen is an absolute cash cow behemoth trading at an absurdly low valuation of 28x trailing earnings.

Paying only 28x earnings for a 20%+ compounder with 50%+ margins—one that is tethered directly to mega-cap tech growth and the dawn of AI agentic commerce—is a severe mispricing.

It’s over 50% below consensus analyst price targets at the time of writing with over twenty-six “Buy” ratings from independent analysts.

Thanks for reading!

This isn’t financial advice. This is investment research aimed to provide you with deep, high-quality knowledge — but always analyze every trade and take great care when investing.

21 May 2026 — APEX | Deacon Brantley

While the market is asleep at the wheel, the neobank Dave Inc. ($DAVE) is quietly building a behemoth. Powered by the efficiency of their CashAI underwriting engine, new product launches in subprime credit and BNPL, and explosive user growth as consumer distress increases, we’re watching a hyper-growth compounder trade at highly modest prices — an opportunity too good to pass up.

1. The $18 CAC Loop

The fatal flaw of most consumer-focused fintechs is customer acquisition inefficiencies. Companies like SoFi ($SOFI) or Chime ($CHYM) spend hundreds of dollars per customer, relying on complex long-term strategies to recoup the initial costs that often take years until breakeven.

This model changes wholly when the product itself is the hook. By leveraging too-good-to-be-true cash advances of up to $500 (the ExtraCash product), Dave is able to spend very little on direct marketing and still maintain an explosively growing user base.

• CAC Efficiency: New transacting members are acquired at a flat $18.

• Gross Profit Payback: The initial $18 CAC to acquire a new user is offset within only three months — meaning that after the initial period, any activity by the customer results in pure revenue for the company.

• Extreme Scalability: In Q1 2026 alone, this highly efficient onboarding mechanism brought in 695,000 new members, expanding the Monthly Active Users (MAU) footprint by 18% YoY to 3 million.

Dave’s highly scalable, efficient model means that the business can grow at a fast rate without consistent and heavy capital injections; and that what capital would traditionally be directed at marketing and acquisition can instead strengthen Dave’s whopping 72% gross profit margin (as of Q1 2026) — which lies above the 90th percentile for the fintech sector. Dave is effectively a money printer functioning as a high-margin software company, not a distressed loan services lender.

2. Macroeconomic Tailwinds

In a time where inflation is rising, wages are stagnating, and the price of goods are steadily increasing, consumers are steadily more distressed. But here’s the alpha: macroeconomic distress is Dave’s customer acquisition funnel.

As traditional banking institutions tighten their credit standards to protect their own balance sheets, they lock out millions of subprime and underbanked consumers. Because of this, Dave’s total addressable market has expanded to an estimated 185 million Americans who are actively looking for alternatives to the slate of predatory options such as high-interest loans, credit cards, or overdraft fees. That’s where Dave’s ExtraCash comes in.

When a consumer needs an extra $100 to fill their gas tank (or these days, $120), traditional options slice into their remaining liquidity. Dave’s no-fee, no-hassle, no-strings-attached ExtraCash offering steps in as an essential utility rather than a luxury product. This contrarian resilience is why Dave experiences robust growth even when the broader consumer environment looks highly pressured — making it a highly alluring investment in this time of consumer distress.

• Counterintuitive Growth: As macroeconomic conditions worsen, more and more consumers become locked out of traditional credit institutions — expanding Dave’s possible customer base.

• ExtraCash Advantage: Instead of charging high fees, Dave uses a proprietary AI engine to discern the level of risk to take on for the consumer (explained below) while maximizing user satisfaction.

3. The CashAI Edge

The biggest bear case leveled against any alternative lender targeting the subprime market is simple: credit risk. Bears assume that aggressive volume growth in this segment inevitably leads to a cascade of defaults and debt write-offs.

Dave has thoroughly shattered this narrative through CashAI, its proprietary machine-learning underwriting engine.

Instead of relying on outdated, lagging FICO scores, CashAI evaluates real-time banking data, transaction habits and direct deposit behavior to determine true creditworthiness. The data from their blowout Q1 2026 earnings report proves that this engine isn’t just functional—it is an absolute competitive beast:

• Explosive Volume vs. Tightening Risk: ExtraCash origination volume surged 37% YoY to a staggering $2.1 billion in Q1 2026.

• Record-Low Delinquencies: Despite that massive influx of volume, Dave’s 28-day past due payment rate dropped to a record Q1 low of 1.69%. Delinquencies are lowering as the volume of subprime credit users increase — the antithesis to traditional economic thought.

4. A Super Bank Ecosystem

Dave is successfully executing a classic gradual expansion playbook. They hook the customer with ExtraCash, transition them into a primary banking user via the Dave Debit Card (which processed over $534 million in spend during Q1!), and are now layering on high-margin credit products with Dave Flex and their new BNPL offering.

Dave Flex is designed as a direct, responsible alternative to predatory subprime credit cards. Backed entirely by CashAI underwriting, it requires no credit checks, carries no compound interest, and charges no late fees — a revolutionarily appealing product in the subprime world

By cross-selling this product into a devoted user base of nearly 3 million members, Dave is poised to supercharge its Average Revenue Per User (ARPU), which has already jumped an impressive 24% YoY to $212 in Q1. Now contrast that with only an $18 CAC!

5. The Undervaluation Thesis

Let’s look at the math. At the updated full-year guidance midpoint of $16.50 EPS, the stock is currently (May 19) trading at roughly 14.5x forward earnings. For context, you are buying a company growing its top-line at nearly 30%, expanding its GAAP net income by triple digits, and maintaining software-equivalent 44% Adjusted EBITDA margins for a multiple typically reserved for stagnant, legacy regional banks.

Furthermore, management is aggressively capitalizing on this atrocious mispricing. By utilizing excess cash alongside a strategic $200 million convertible note execution, they deployed $194.9 million in Q1 alone to retire 7.0% of the entire outstanding share float — a signal of confidence in the future of Dave.

When a hyper-growth compounder aggressively shrinks its active share count whilst raising earnings guidance, it creates a bomb of increased EPS (EPS = earnings/share count, so reduced share count = higher EPS) With Wall Street consensus price targets clustered between $310 and $340 and current prices hovering around $230, Dave Inc. represents a rare, highly asymmetric alpha opportunity hiding in plain sight.

The market is treating it like a sluggish subprime lender, but the underlying numbers show a high-margin consumer technology juggernaut firing on all cylinders.

This isn’t financial advice! Take great care when trading and evaluate every decision.

19 May 2026 APEX | Deacon Brantley

Subscribe now

Welcome to APEX! I’m so glad you’re here. If you’re new, our mission is straightforward. We bring you the raw, deep, and insightful investing alpha you need to make sense of where the markets are actually going. No scams, no fraud, no shady discord -- just quality DD and research.

Today, we are tackling one of the biggest elephants in the tech room: Microsoft. For the past few years, the mainstream narrative has been entirely unanimous: Microsoft is an unstoppable behemoth effortlessly printing money as it leads us into a new era. But when you lift the hood, a very different story emerges.

TL;DR: Microsoft Is Running Out Of Time

Underneath the AI hype lies an overvalued legacy giant subsidizing its own disruption while its core software moats evaporate in real-time.

• The OpenAI Predator Shift: OpenAI is no longer an exclusive asset—it’s an active competitor attacking the Office moat. Microsoft lost its cloud exclusivity (OpenAI now partners with AWS, Oracle, and Google) while paying margin-crushing royalties for non-exclusive tech.

• The Copilot Money Sink: Copilot has failed to achieve mass adoption, stalling at a 3.3% penetration rate among M365 users. Its enterprise market share cratered from 18.8% to 11.5% over the past year, consistently ranking dead last in user popularity polls.

• The Death of “Per-Seat” SaaS: Microsoft’s business model relies on charging per human user (seats). As autonomous AI agents replace human workflows, enterprise clients will radically downsize their overall Office 365 license counts.

• Hardware Ecosystem Decay: Xbox console revenue is in a historic multi-year freefall, and the Surface line remains stagnant and overpriced. Lacking a mobile ecosystem, Microsoft is starting from nothing in the AI-first wearable market.

• The $190B Capital Trap: Microsoft is burning an eye-watering $190 billion in CapEx for 2026 (a 61% spike). Wall Street is pricing Microsoft as an agile, high-margin software enterprise, but it is mutating into a sluggish capital-intensive infrastructure utility.

____

1. The Failure of Microsoft’s AI Strategy

What was once Microsoft’s greatest asset—its exclusive partnership with OpenAI—has become a structural risk and an existential threat.

• The Partner-to-Predator Change: The 2026 landscape has exposed the friction. OpenAI is no longer a research lab; it is an enterprise software company. With the launch of OpenAI’s native Search and Workplace tools, Sam Altman is directly attacking the Office moat. Microsoft isn’t just renting a brain; they are subsidizing the R&D of the very entity designed to replace them with each passing quarter.

• The Middleman Tax: Unlike Google (Gemini) or Meta (Llama), Microsoft is not vertically integrated. Every time a user interacts with Copilot, Microsoft is paying a margin-crushing royalty to OpenAI. They have essentially turned Azure—once a high-margin cloud play—into a discounted utility provider for their own partner turned predator.

• Loss of Exclusivity: OpenAI is no longer exclusive to Azure as of April 2026. It has just signed massive computing deals with Oracle, AWS, and Google Cloud — Azure’s direct and most dangerous competition.

• Oh, Copilot: These stats explain why Copilot is not only a losing bet — it’s a money sink cratering Microsoft’s credibility.

  • Fewer than 40% of employees with access actually use it daily.

  • Only 3.3% of eligible users have signed up for paid Copilot, which is a far cry from the 30% goal Microsoft needed to hit to ensure profitability.

  • Copilot’s enterprise market share has dipped 7.5% over the last year while CapEx to keep Copilot running has increased 61%.

  • Copilot consistently ranks last in User’s Choice popularity polls, often receiving less than 10% of the vote, trailing behind Claude (57%) and Gemini (34%).

2. Microsoft Office: Death By Agent

Many of Microsoft’s revenue streams (i.e. Office) are built on the per-seat SaaS model. But, the AI agents Microsoft is peddling are fundamentally anti-seat.

• The Math: If a Copilot agent allows one employee to do the work of four, an enterprise doesn’t need five Office 365 licenses — it only needs one.

• Office Erosion: As AI agents (i.e. Claude Cowork) continue to get more capable, individuals will rely more on the agent’s capability to engineer organization and productivity rather than their actual direct use of Office products, creating a problematic middleman layer in between Office and the consumer.

• Google Stack: As Gemini continues to improve, its direct access into the free Google One ecosystem is a direct threat to Copilot and the paid Office ecosystem. If Gemini were to develop a cheaper, cloud-based agent capable of better in-browser work, Google would outcompete Microsoft, as it would possess both the intelligence and the productivity suite, rather than just the suite.

3. The Hardware Identity Crisis

Microsoft’s hardware track record is a graveyard of us-too products — attempts to emulate the commercial success of better-thought-out products. From the Zune to the Windows Phone, and now to the stagnating Surface and Xbox lines, Microsoft has failed to build a dynamic hardware ecosystem that can compete with Apple or even the new wave of AI-first hardware startups.

• Xbox in Freefall: With hardware revenue cratering, Microsoft is abandoning the console wars to become a third-party publisher. This comes after nearly five years of 30-40% YoY revenue loss in Xbox sales and the significant loss of market share to the more agile and user-friendly Steam and PlayStation. In addition, the recent decision to publish flagship Microsoft-produced games such as Forza on PlayStation consoles have eroded the value of an Xbox, as it no longer becomes necessary to own an Xbox to play Microsoft games.

• Surface Stagnation: The Surface has failed to capture the enterprise audience, battling extraordinarily high costs (often $1,800+ for premium setups), which often run 1.5 to 2x higher than the Surface’s enterprise-level competitors and provide less technical allure than the high performance of Apple’s M-series.

• The HoloLens & Wearables Miss: After nearly 20 years, Microsoft still hasn’t figured out how to build a pocketable device. In a new world where AI-first wearables (rings, glasses, pendants) are the new frontier, Microsoft is starting from zero. The $5B HoloLens wearable project, once hailed as the future of the company, has largely been mothballed for general purchase and exists only in niche defense and military contracts.

4. History Doesn’t Repeat, But It Often Rhymes

In the 2000s, Microsoft sat on its Windows laurels while the world shifted to mobile. Today, Microsoft is sitting on its SaaS laurels while the world shifts to autonomous agents.

• Institutional Hubris: Just as Steve Ballmer famously laughed at the iPhone for lacking a keyboard, Satya Nadella’s Microsoft initially underestimated the speed at which foundational models would become widely adopted or coveted. By the time they realized the AI boom was here to stay, they were so far behind they had to buy a $13B get out of jail free card with OpenAI to maintain relevance.

• The Legacy Anchor: Microsoft’s biggest weakness is its greatest success. They are so tethered to protecting Office and Windows that they cannot pivot to the AI-first future with full commitment. They are defending a 20th-century fortress against 21st-century guerrilla warfare.

• The Final Act: In the mobile era, Windows lost relevance and market share to MacOS and Android, even being completely locked out of the mobile market. In the AI era, Azure is now at risk of becoming a utility for someone else’s intelligence. They are repeating the exact same pattern: arrive late, spend billions to catch up, and settle for being the infrastructure for their more agile competitors.

Photo by Mike Blake || Reuters

5. A Valuation Trap

Microsoft is currently locked into the brutal CapEx arms race, projecting a mind-boggling $190 billion in expenditures for 2026 alone, a 61% increase from 2025.

• Capacity vs. Cash: Microsoft is aggressively buying up GPUs, land, and multi-decade power contracts to power data centers that keep getting delayed.

• Poor Revenue Thesis: Microsoft is in a unique position among its peers in the AI race. It doesn’t produce the GPUs or memory chips the AI boom runs on or even the actual intelligence itself, making Microsoft have to outcompete and outspend on both fronts. Its only opportunity to offset those two massive costs is in the software layer, which is currently being eroded by its competitors as I write this article. Therefore, it must make more revenue from the software sector alone than it spends on both hardware and AI.

• Loss of Market Share: Early FY2026 data indicates Copilot’s enterprise market share has dipped from ~18% to ~11.5% as companies realize that generic AI assistants are often less useful than specialized, industry-specific AI tools from smaller startups or powerful AI agents.

• Overvaluation: Microsoft is valued as a $3T high-growth software company. But underneath lie flat-to-negative growth in hardware, incredibly high CapEx, and the erosion of Microsoft’s strongest revenue streams. We are witnessing the possible fall of a giant — and Microsoft is racing against the clock to maintain relevance in an AI world.

Leave a comment