Stock Market Today, Sept. 25: Microsoft Stock Jumps 4% After Revamping Copilot With Code Generation and Agentic AI Tools

Microsoft closed at $516.17 after unveiling expanded Copilot capabilities, with trading volume of 37.9 million shares, nearly 15% above its three-month average. Investors are focused on whether AI adoption and cloud demand can justify more than $115 billion of planned 2026 capex, primarily for data centers and AI infrastructure. At 21x cash from operations versus a five-year average of 25x, the stock is viewed as reasonably valued if its AI investments generate returns.

Analysis

The investable question is not feature breadth but whether Copilot evolves from bundled retention tool into a measurable per-seat and per-workload monetization engine. Persistent agents and coding workflows can raise switching costs in Microsoft 365, GitHub and Azure simultaneously, but they also increase inference intensity; near-term gross-margin dilution is plausible before revenue recognition catches up. The next 1-3 months should therefore be driven less by product reception than by management disclosure around paid-seat conversion, Azure AI workload growth and incremental depreciation.

Microsoft’s infrastructure build creates a second-order read-through for NVDA, but the stronger relative beneficiary may be Oracle if enterprise customers seek lower-cost capacity or multi-cloud deployments while Azure capacity remains constrained. Conversely, a broad agent layer could weaken standalone SaaS vendors whose value proposition is workflow automation rather than proprietary data, particularly if Microsoft prices Copilot aggressively inside existing enterprise agreements. The market is likely assigning too much value to demonstrations until procurement budgets show that agents displace labor or consolidate software spend.

At the current valuation framing, upside requires proof that AI revenue scales faster than capital intensity; multiple expansion is unlikely if cash conversion deteriorates despite strong bookings. A downside surprise would be a deceleration in Azure growth, a higher capex outlook without associated remaining-performance-obligation acceleration, or evidence that Copilot usage is largely included-seat consumption rather than paid incremental adoption. Over 6-18 months, the key structural risk is that enterprise AI economics migrate toward model and compute suppliers, leaving application vendors to absorb inference costs in order to defend seat pricing.

Key Decisions for Investors

• Maintain a modest long MSFT only into the next earnings print; add on evidence of accelerating paid Copilot penetration and Azure growth reacceleration, not product announcements. Target 8-12% upside over 3-6 months if monetization offsets capex concerns; exit/reduce if capex guidance rises while cloud growth or operating-margin guidance weakens.
• Express relative AI-infrastructure demand through long NVDA / short MSFT in equal dollar risk for the next 1-3 months if Microsoft’s capex ramp is confirmed but Copilot pricing and paid-user metrics remain opaque. The pair benefits if compute spend accrues faster to suppliers than to Microsoft’s software P&L; cover on clear Copilot ARR disclosure or material Azure margin resilience.
• Watch ORCL for a long entry only after confirmation that cloud infrastructure backlog converts to revenue without a further deterioration in funding or capex intensity. Oracle is a plausible capacity-constrained Azure spillover beneficiary, but its balance-sheet and execution sensitivity make it a higher-beta confirmation trade rather than an anticipatory position.
• Avoid treating the one-day move as a standalone catalyst. Set an earnings alert around three falsifiers: Azure growth below consensus, no quantified paid Copilot adoption, or operating-margin pressure exceeding depreciation expectations; any two would support trimming MSFT and increasing the NVDA/MSFT relative trade.