|
TL;DR: Everyone's debating Microsoft's AI lead. Almost nobody's talking about February's announcement: Microsoft may have solved quantum computing. If they're right, AI is just the warmup. Here's what value investors need to know, and why each layer of analysis tells a different story.
The Story Everyone Missed
In February 2025, while the market obsessed over Copilot revenue and Azure AI growth, Microsoft quietly announced something that could be far more significant.
Majorana 1. The world's first quantum chip powered by topological qubits.
If you haven't heard of it, you're not alone. The announcement sparked intense scientific controversy, which meant the business press moved on quickly. But DARPA (the agency that funded the internet, GPS, and stealth technology) selected Microsoft as one of only two companies for their utility-scale quantum computing program.
When DARPA bets on you, pay attention.
Today I want to do something different. Instead of just running our valuation models and giving you a verdict, I want to walk through each layer of analysis and explain what it reveals. The numbers matter. But understanding why each methodology exists matters more.
What Microsoft Claims to Have Built
Let me translate the physics into investment terms.
The Problem with Current Quantum Computers
Today's quantum computers are fragile. Their qubits (quantum bits) constantly break down, requiring massive error correction. IBM and Google are building quantum computers with thousands of qubits, but most of that computing power goes to fixing errors, not solving problems. It's like hiring 100 employees but spending 90 of them just managing the other 10.
Microsoft's Different Approach
Instead of building fragile qubits and error-correcting them, Microsoft spent 20 years developing qubits that are stable by design. They call them "topological qubits," built on exotic particles called Majoranas that don't exist in nature. The engineering is complex. The investment thesis is simple: if stable qubits work, Microsoft skips a decade of error-correction work everyone else is doing.
The Breakthrough
Majorana 1 is an 8-qubit chip with a clear engineering path to 1 million qubits. For context: all the world's current computers operating together can't do what a one-million-qubit quantum computer will be able to do. Microsoft's timeline is "years, not decades."
Why Physicists Are Skeptical
Here's where it gets interesting for investors.
Microsoft's claims face significant scientific pushback.
Professor Sergey Frolov at the University of Pittsburgh called the project "fraudulent," arguing the data doesn't prove functional topological qubits. Nature published an editorial note on Microsoft's paper requesting further verification (unusual for a corporate announcement). Science News reported physicists are "mostly unconvinced." And a previous Microsoft-affiliated Majorana paper was retracted in 2021.
This isn't a slam dunk. The scientific community has real doubts.
But here's the investment insight.
Controversial breakthroughs are discounted. If the market were certain Majorana 1 worked, Microsoft would trade at a higher multiple. The controversy creates a situation where you might be getting trillion-dollar optionality for free. The skepticism is doing your discounting for you.
Layer 1: The Quality Firewall
Before we value any company, we need to know if it's safe to value at all. This is what Layer 1 does. It answers one question: can this company survive long enough for our analysis to matter?
Three tests. Pass or fail.
Piotroski F-Score: 6/9 (Neutral)
Joseph Piotroski, an accounting professor at Stanford, designed this test in 2000 to catch "value traps" (companies that look cheap but keep getting cheaper). It checks nine things: Is profitability improving? Is leverage falling? Is the company diluting shareholders? Each question gets one point.
Microsoft scores 6 out of 9. Not a red flag, not a green flag. The company is profitable, but a few efficiency metrics are flat. For a mature tech giant, this is fine.
Altman Z-Score: 4.46 (Safe Zone)
Edward Altman created this formula in 1968 to predict bankruptcy. It combines five ratios (working capital, retained earnings, EBIT, market cap, and revenue) into a single score. Below 1.8 means danger. Above 3.0 means safe. Microsoft sits at 4.46, comfortably in the safe zone.
Why does this matter? Because a company can look cheap right before it goes bankrupt. The Z-Score is a sanity check. Microsoft passes easily.
Beneish M-Score: No Flags
Messod Beneish designed this test to catch earnings manipulation. Companies that manage earnings aggressively tend to blow up later. The M-Score looks at revenue growth, margin changes, and accrual patterns to flag suspicious accounting.
Microsoft shows no manipulation signals. The numbers are what they appear to be.
Layer 1 Verdict: PASS. Microsoft is financially healthy, not cooking the books, and not going bankrupt. We can proceed to valuation.
Layer 2: What Is Microsoft Worth?
This is where most investors start and stop. They pick one valuation model, run it, and call it a day. That's a mistake.
Every valuation model makes assumptions. Some assume no growth. Some assume infinite growth. Some trust book value. Some trust cash flows. Using multiple models reveals where they agree (high conviction) and where they disagree (uncertainty you need to understand).
We ran seven models. Here's what they say.
Graham Number: $145
Benjamin Graham, the father of value investing, wanted a simple formula that any investor could calculate by hand. His "Graham Number" multiplies earnings per share by book value per share, takes the square root, and multiplies by 22.5. It gives you a conservative ceiling price.
For Microsoft, that ceiling is $145. At $450, you're paying 310% above Graham's maximum. Graham would not touch this stock.
Why it matters: The Graham Number represents the value investor's floor. It asks: if everything goes wrong, what's the minimum this company is worth based on assets and earnings right now? Microsoft fails this test badly.
Damodaran DCF: $143 (Base Case)
Aswath Damodaran teaches valuation at NYU Stern and publishes all his models online. His DCF (discounted cash flow) approach projects future free cash flows, then discounts them back to today at the company's cost of capital.
We used his methodology with Microsoft's actual numbers: 12% revenue growth fading to 4% terminal growth, 35% operating margins, 9% cost of capital. The result: $143 per share.
Why it matters: A DCF is the most intellectually rigorous valuation approach. It forces you to state your assumptions explicitly. Microsoft would need to grow faster, for longer, at higher margins than historical patterns suggest, just to justify $143. The current price of $450 implies growth assumptions that are hard to defend.
Lynch Fair Value: $284
Peter Lynch, the legendary Fidelity fund manager, used a simpler test. He believed a fairly valued company should have a P/E ratio equal to its earnings growth rate (this is called a PEG ratio of 1). If Microsoft grows earnings at 15%, it should trade at 15x earnings.
Using this framework, Microsoft's fair value is $284. Still 40% below the current price, but the most generous model in our set.
Why it matters: Lynch's approach favors growth. If any model would justify Microsoft's price, it would be this one. It doesn't.
Owner Earnings DCF: Can't Compute
Warren Buffett prefers "owner earnings" to accounting earnings. Owner earnings subtract the cash a company must reinvest just to maintain its competitive position. For some companies, reported earnings overstate true profitability because so much gets reinvested.
For Microsoft, our data doesn't cleanly separate maintenance spending from growth spending, so this model returned incomplete results.
Earnings Power Value (Greenwald): $130
Bruce Greenwald at Columbia teaches a method called Earnings Power Value. The idea: ignore growth entirely. Ask what the company would be worth if it never grew again but maintained current profitability forever.
Microsoft's EPV is $130. Even assuming zero decline and perpetual current profits, the stock is overvalued.
Why it matters: This is the most conservative DCF variant. It says: forget the growth story. Just look at what exists today. And what exists today doesn't justify $450.
Dividend Discount Model: $103
If you only care about cash returned to shareholders through dividends, Microsoft is worth $103. This model assumes you're buying a stream of dividend payments and values that stream at a required return.
Why it matters: Microsoft is not primarily a dividend stock, so this model is less applicable. But it anchors the low end: if growth stalls and Microsoft becomes a dividend utility, this is where the price gravitates.
Layer 2 Verdict: OVERVALUED. Zero out of seven models support the current price. The most generous (Lynch) implies 40% downside. The most conservative (DDM) implies 75% downside. The composite fair value across all models is $183.
Layer 3: Quality and Moat
Here's where Microsoft shines.
Valuation models tell you what to pay. Quality models tell you what you're buying. A cheap stock can still be a bad investment if the business deteriorates. An expensive stock can still work out if the quality is extraordinary.
Greenblatt Magic Formula: 100/100
Joel Greenblatt's "Magic Formula" ranks companies on two dimensions: how good the business is (return on invested capital) and how cheap the stock is (earnings yield). Companies that score high on both tend to outperform.
Microsoft's quality is so high that it scores 100/100 on the quality dimension. The issue is the price dimension dragging down the combined score.
Why it matters: The Magic Formula confirms what we intuitively know: Microsoft is an exceptional business. But Greenblatt would tell you to wait for a better price.
ROIC: 50.5%
Return on invested capital measures how much profit a company generates for every dollar it invests in the business. Microsoft generates 50 cents of profit for every dollar of capital. This is extraordinary (the median company earns 10-15 cents).
Why it matters: High ROIC is the quantitative signature of a moat. It means competitors can't easily replicate what Microsoft does. The business earns excess returns because it has some advantage (brand, switching costs, network effects, patents) that protects it.
Operating Margin: 44.7%
Microsoft keeps 45 cents of every revenue dollar as operating profit. The sector median is 18%. This margin advantage is why the company is worth so much.
Moat Durability: 8/10 (Wide)
We assess moat durability by looking at margin stability, customer retention, pricing power, and competitive positioning over 5-10 years. Microsoft rates 8 out of 10, meaning its advantages are likely to persist.
Balance Sheet: AAA
Microsoft is one of two companies in America with a AAA credit rating (the other is Johnson & Johnson). It has $80 billion in cash and a net debt to EBITDA ratio of 0.19x. This company could weather almost any economic scenario.
Layer 3 Verdict: EXCEPTIONAL. Microsoft is objectively one of the highest-quality businesses in the world. The moat is wide, the margins are thick, the balance sheet is a fortress. The problem isn't the business. The problem is the price.
Layer 4: The Qualitative Thesis
Layers 1-3 use numbers. Layer 4 asks: what has to be true for this investment to work? What are the bull cases? What are the bear cases? And how do we weigh the quantum angle?
The Bull Case: Seven Pillars
Azure AI Dominance. The OpenAI partnership plus enterprise distribution channels gives Microsoft a lead in AI monetization.
M365 Switching Costs. 400 million commercial seats with 90%+ retention creates recurring revenue that almost never churns.
Fortress Balance Sheet. $80B cash, AAA rating, 0.19x net debt to EBITDA. Microsoft can weather anything and buy anything.
Capital Return Discipline. $70B+ returned annually through dividends and buybacks. Management isn't empire building.
Cross-Sell Network Effects. Azure plus M365 plus Dynamics plus LinkedIn plus GitHub creates an ecosystem where each product reinforces the others.
Gaming Optionality. Activision acquisition plus Game Pass creates Xbox as a platform, not just a console.
Quantum Computing. Majorana 1 topological architecture could leapfrog competitors by 5-10 years. This is the new pillar.
The Bear Case: Five Risks
Valuation Stretch. Zero out of seven models support the price. This is the critical issue.
AI Commoditization. Open-source models (Llama, Mistral) and competition from Google and Amazon could erode Azure AI margins.
Antitrust Overhang. DOJ and EU scrutiny on cloud bundling, Activision integration, and potential AI regulation.
Growth Deceleration. Microsoft's revenue growth (12.4%) is at the sector median, not above it. The market is pricing in acceleration that may not come.
Quantum Vaporware. The scientific skepticism is real. The 2021 retraction happened. Majorana could be another dead end.
The Taleb Addendum: A Contrarian View
Here's where I have to be intellectually honest with you.
I've been framing Microsoft's quantum play as "asymmetric optionality," and it is. But Nassim Taleb, who literally wrote the book on optionality (Antifragile), would probably reject buying Microsoft at $450 as the wrong way to access this bet.
The Taleb Critique
His barbell strategy says: put 85-90% in ultra-safe assets, and 10-15% in extremely speculative bets with massive upside. The key is to keep them separate.
Microsoft at $450 violates this principle. You're not buying "safe plus speculative lottery ticket." You're buying:
A richly valued mega-cap (fragile to multiple compression), with embedded optionality you cannot separate.
If the quantum play excites you, Taleb would argue you should buy the optionality directly.
The Alternative Structure
Allocate 90% to T-Bills, cash, or short-duration bonds. Allocate 10% to a quantum pure-play basket (IonQ, Rigetti, D-Wave).
Or even simpler: Microsoft January 2028 deep out-of-the-money calls. Pay $5-10 for $600 strike options. You lose the premium if quantum doesn't materialize. You make 10x or more if it does.
The point: You can get quantum optionality without paying 30x earnings for Microsoft's cash flows.
Why I'm still saying HOLD (for existing positions):
Most readers already own Microsoft. The quantum angle is a reason not to sell. But Taleb is right that it's the wrong entry point for new money. If you're sitting in cash, consider his barbell instead.
Next issue, we'll analyze the quantum pure-plays directly: IonQ, Rigetti, and D-Wave. What does convexity look like when the entire company IS the bet?
Verdict
HOLD (existing positions) / AVOID (new money at $450) | Confidence: 65%
The valuation models say SELL. Every single one of them. The quality metrics say this is one of the best businesses in the world. The quantum optionality is real but unproven, and the controversy means it's probably not priced in.
The honest take:
If you own Microsoft, the quantum angle is a reason to hold rather than sell. The optionality exists. But don't kid yourself about the valuation. At $450, every model shows significant overvaluation.
If you're waiting to buy, set an alert for around $280 (Lynch fair value). At that price, you get quality plus optionality at a reasonable entry. That's the investment. At $450, it's speculation on continued multiple expansion.
If you believe strongly in quantum computing specifically, Taleb's barbell makes more sense. 90% safe assets plus 10% quantum basket gives you direct exposure without the Microsoft baggage.
We show our work. You make the call.
This analysis was generated by the Tenets engine. Every calculation is auditable. Every assumption is stated. We build conviction, not just answers.
Subscribe to get one deep-dive analysis like this every Sunday.
Next Issue: The Quantum Basket. IonQ, Rigetti, and D-Wave through the engine. What happens when traditional valuation models break completely?
Questions? Reply to this email. We read everything.
