MSFT
At $390, the market implies specific beliefs about Copilot adoption, OpenAI retention, and AI capex returns -- here is exactly what you are being asked to believe
At $390, we can decompose Microsoft's market price into exactly four layers: auditable balance sheet value, segment values anchored to comparable companies, growth premiums above those comps, and a speculative residual that represents three specific AI-related beliefs. This is not a valuation model -- it is a mirror showing what the market currently believes.
What The Price Implies
At $390, the market implies that Microsoft's existing businesses (cloud, productivity, personal computing) are worth approximately $361/share when anchored to comparable company multiples plus growth premiums. The remaining $22/share is a bet on three specific outcomes: that Copilot becomes the default enterprise AI platform (implied probability 35%), that the OpenAI partnership remains productive through 2030 (implied probability 40%), and that the massive AI infrastructure buildout generates adequate returns (implied probability 50%).
The Beliefs Embedded in the Price
- +70% of Fortune 500 have adopted M365 Copilot, record seat adds +160% YoY
- +E7 bundle ($99/user, May 2026) makes Copilot a default capability for top-tier enterprises
- +GitHub Copilot at 4.7M paid subscribers, 42% market share among AI coding tools
- +Accenture RCT: 8.69% increase in pull requests, 15% increase in PR merge rate
- −Only 3.3% penetration after 18 months (15M/450M seats)
- −Gartner: only 5% of pilot-completers moved to larger deployment
- −METR RCT: experienced developers 19% SLOWER with AI tools on mature codebases
- −Google includes Gemini free in Workspace ($12/user) vs $30 Copilot add-on
- −Developer trust in AI declining: only 29% trust AI, down 11pp (Stack Overflow 2025)
- +$250B incremental Azure commitment from OpenAI
- +Stateless API exclusivity until AGI; 20% revenue share through 2032
- +OpenAI pre-IPO docs flag Microsoft dependency as MAJOR business risk (mutual dependence)
- +$30B/year Azure consumption, growing 200%+ YoY
- −OpenAI signed $38B AWS deal + $300B Oracle Stargate
- −OpenAI developing custom Titan chip (TSMC 3nm, H2 2026)
- −Microsoft holds 27% equity but ZERO board seats, ZERO veto rights
- −Microsoft now lists OpenAI as a COMPETITOR in its 10-K (first time 2024)
- −OpenAI running ChatGPT on Google TPUs since June 2025
- +$625B contracted backlog (RPO) provides demand justification
- +Azure capacity-constrained: demand exceeds supply through mid-2026
- +Data centers have 10-15 year useful lives; Amazon warehouse analogy
- +Maia 200 custom chips: 30% better perf/$ vs third-party GPUs for inference
- −FCF conversion hit 16.4% quarterly in Q2 FY2026 -- worst in modern MSFT history
- −Capital intensity at 22.9% of revenue, resembling industrial companies
- −Sequoia: AI infrastructure requires ~$800B in product sales for adequate returns
- −1990s telco fiber parallel: $500B invested, 85% dark by 2005, $2T value destroyed
- −CFO reversed capex slowdown guidance -- FY26 growth rate will EXCEED FY25's 45%
At $390, Layers 1-3 (auditable + anchored + growth premium) account for $368/share, or 94% of the stock price. The speculative AI premium is $22/share (6%). Whether $390 is a good price depends entirely on whether you believe in these three AI outcomes. If you cannot evaluate these beliefs -- if you have no informed view on Copilot adoption, OpenAI retention, or AI capex ROI -- you are effectively taking a 50/50 bet on the biggest technology investment cycle in history. In that case, the index may be a better choice.
What Does $390 Buy You?
Price-Implied Expectations decomposition. Every dollar accounted for.
The Key Debates
The 14 questions that determine whether this stock is worth owning.
Bulls: Microsoft's bundling playbook (think E5 adoption trajectory) will eventually drive Copilot to 20%+ attachment. E7 at $99/user makes Copilot a default. 450M seat TAM with $360/year ARPU = $162B potential. Bears: Only 3.3% penetration after 18 months. Gartner says only 5% of pilots moved to full deployment. METR randomized controlled trial found experienced developers were 19% SLOWER with AI tools. Google offers Gemini free in Workspace, creating price pressure from below.
This is one of the most uncertain beliefs in the model. At 35% probability, it represents $8/share of the stock price. If you have no view on enterprise AI adoption, the implied 35% is a bet you may not want to take. Consider whether you have evidence that is not reflected in the current 3.3% penetration rate.
Microsoft Copilot is an AI assistant embedded in Office 365 that helps with writing, data analysis, presentations, and meetings. It costs $30/user/month on top of the existing Microsoft 365 subscription. The question is whether it becomes a must-have tool that most enterprises pay for, or a nice-to-have that only a fraction adopt.
- +M365 Copilot: 15M paid seats on 450M+ base = 3.3% penetration (Q2 FY2026 earnings)
- +Daily active users only 6M of 15M licensed -- 40% daily utilization (Recon Analytics survey)
- +E7 bundle at $99/user launching May 2026 bundles Copilot as default (Microsoft blog)
- +Gartner: only 5% of pilot-completers moved to larger deployment; 17% decided not to adopt (Gartner 2025)
- +METR RCT: 16 experienced OSS developers 19% SLOWER with AI tools on mature codebases (arXiv:2507.09089)
- +Accenture RCT: 450 developers saw 8.69% increase in pull requests with Copilot (GitHub blog)
- +Google includes Gemini free in Workspace at $12-18/user vs $30 Copilot add-on
- +Cursor: $2B ARR (Mar 2026), 60% enterprise revenue, Claude Code $2.5B ARR -- competitive pressure on GitHub Copilot
- +Stack Overflow 2025: only 29% of developers trust AI, down 11pp from 2024
What Would Change the Price
The highest-impact events, ranked by potential price impact.
The Beliefs Behind the Price
Each assumption embedded in the current price. Do you have an edge on any of them?
If AI capex generates adequate returns, incremental value is ~$95B. Scenario: FCF conversion recovers from 40% to 55-60% as capex normalizes. Cloud operating margins expand from 42.7% to 48-52% as Maia chips and scale economies kick in. This represents a 10-15pp improvement in FCF conversion across a $130B+ revenue base = $13-20B incremental annual FCF, capitalized at 25-30x = $325-600B incremental value. We use the lower end.
Market-implied probability that $100B+/year capex generates positive ROI: 50%. Evidence for: $625B contracted backlog, capacity constrained (demand > supply), data centers have 10-15 year lives, Amazon warehouse cycle analogy. Evidence against: FCF conversion hit 16.4% quarterly, capital intensity at 22.9% of revenue (industrial-level), Sequoia estimates $800B product revenue needed, telco fiber optic parallel ($500B wasted 1996-2002). Key difference from 1990s: funded by cash flow not debt.
If Copilot wins (becomes default enterprise AI), the incremental value is ~$171B. Scenario: 15-20% attachment on 450M seats = 67-90M paying users at blended $300/year after discounts = $20-27B incremental revenue at 15-20x multiple = $300-540B. We use $171B as the expected conditional value in this scenario.
Market-implied probability that Copilot becomes the default enterprise AI interface: 35%. Currently 3.3% penetration (15M/450M seats). Evidence for: 70% of Fortune 500 adopted, record seat adds +160% YoY, E7 bundle launches May 2026. Evidence against: only 5% of pilot-completers moved to full deployment (Gartner), 17% decided NOT to adopt, METR study showed 19% slowdown for experienced devs, developer trust in AI declining (29%, down 11pp).
IC anchored at 8.5x EV/Revenue based on cloud infrastructure comps. AWS implied ~10x EV/Rev (at 20% growth, 35% OI margin). Google Cloud implied ~4.5x (at 35% growth, 24% margin, subscale). Oracle OCI ~5x. Median ~7x, but adjusted to 8.5x reflecting Azure's margin superiority over GCP and scale advantages. This is the ANCHORED multiple -- any premium above this is Layer 3 growth premium.
IC growth premium $200B = $27/share. This is the market's belief that Azure will grow faster than the peer group for an extended period. Azure at 39% vs AWS at 20% and market average at ~25%. The premium represents the NPV of excess growth over what the 8.5x anchored multiple already captures. Azure AI contributing 16pp of growth, $625B RPO (110% YoY growth), capacity constraints suggesting demand exceeds supply. But also: $80B unfulfilled orders due to power constraints, neocloud competition, and Google Cloud accelerating to 48% growth.
IC revenue ~$128B annualized from H1 FY2026. Azure surpassed $75B in FY2025, growing 34-39%. Total IC includes Azure ($80B+), server products ($15B, declining), GitHub ($3B), enterprise services. Q2 FY2026: IC revenue $32.9B (+29% YoY). H1 FY2026 run rate ~$128B.
MPC anchored at 4.0x blended EV/Revenue. Gaming comps: EA ~5.3x, TTWO ~5.5x (publisher multiples). Windows OEM: ~3-4x (mature licensing). Search: Bing at ~3x (subscale vs Google Search at ~6x). Activision acquired at ~5x. Blended: ($16B*4x + $23.5B*5x + $16.5B*3x) / $56B = ~4.1x. Rounded to 4.0x.
MPC revenue ~$56B. Windows OEM & Devices (~$16B, +2%), Gaming (~$23.5B, +9%, 95% software/services post-Activision), Search & News (~$16.5B, +13% ex-TAC). H1 FY2026 run rate $56B. Flat overall: gaming growth offset by Windows OEM maturity. Q2 FY2026 MPC revenue -3% YoY due to tough Xbox hardware comps.
Net cash $54B as of Dec 31, 2025. Cash $24.3B + short-term investments $65.2B = $89.5B liquid; long-term debt $35.4B. Net = $54.0B. Debt-to-equity 0.12x (very conservative). WARNING: $28.9B IRS tax dispute outstanding (transfer pricing, 2004-2013).
If OpenAI partnership holds productively through 2030, incremental value is ~$140B. Includes: $250B committed Azure consumption generating cloud margin, 20% revenue share on OpenAI's growing revenue ($13B in CY2025, projected $40B+ by 2028), API exclusivity driving Azure AI adoption, 27% equity stake worth ~$135B at current OpenAI valuation. Value ABOVE what's already in Layer 2 cloud anchor.
Market-implied probability that OpenAI stays on Azure at substantial levels through 2030: 40%. Currently $30B/year Azure consumption, 45% of $625B backlog. Evidence for: $250B committed, stateless API exclusivity, 20% revenue share through 2032. Evidence against: $38B AWS deal, $300B Oracle Stargate, custom Titan chip planned, Microsoft now lists OpenAI as competitor, zero board seats on 27% equity.
P&BP anchored at 8.0x EV/Revenue based on enterprise SaaS comps. Salesforce ~7x (9% growth, 30% OI margin), Adobe ~9.5x (11% growth, 35% OI margin), SAP ~7.5x (10% growth, 30% margin), ServiceNow ~17.5x (25% growth, 28% margin). Median of mature SaaS ~8x. MSFT P&BP grows 16% with 61% margins -- best-in-class profitability justifies top-quartile multiple.
P&BP growth premium $100B = $13/share. The market prices P&BP above the enterprise SaaS comp-anchored value. The premium reflects: (1) Copilot driving ARPU expansion from $30/user add-on with 3.3% penetration today, E7 bundle to accelerate; (2) M365 seat growth of 6%/year on 450M+ base; (3) LinkedIn growing 9% with AI integration; (4) Dynamics 365 growing 19%. But also risks: Google includes Gemini free in Workspace, enterprise AI budget fatigue, and Copilot retention concerns (only 5% of pilot-completers moved to full deployment per Gartner).
P&BP revenue ~$134B annualized from H1 FY2026. M365 Commercial ($88B+, +14%), LinkedIn ($18B, +9%), Dynamics 365 (~$8B, +19%), M365 Consumer (~$8B, +11%). Q2 FY2026: P&BP revenue $34.1B (+16% YoY). Copilot revenue embedded in M365 Commercial cloud growth. E7 bundle ($99/user, May 2026) expected to accelerate Copilot attachment.
~7.44B diluted shares. Market cap ~$2.9T at ~$390/share. Declining via ~$20B/year buyback program ($60B authorization, $57.3B remaining). Share count declining ~1-2%/year.
Scenario Analysis
Pre-computed outcomes under different assumption sets.
Methodology
Price-Implied Expectations (PIE) framework based on Mauboussin & Rappaport's "Expectations Investing." Segments valued using comparable company multiples (Layer 2), with residual allocated to probability-weighted speculative businesses (Layer 4). Evidence sourced from SEC filings, earnings calls, and public reports.
PIE Model • v7.0 • Last updated: 3/26/2026