Melius cuts Microsoft stock price target on Copilot concerns By Investing.com

Melius cut its MSFT price target to $400 from $430 while the shares trade at $385.32 (−21% YTD), citing Copilot reorganization, weak IP sharing with OpenAI, higher-than-expected R&D and increased internal Azure capacity use that could constrain customer adoption. Offsetting headwinds, Microsoft reported Q2 revenue up 15% with Azure growth of 38%, Evercore reiterated an Outperform with a $580 target, and the company declared a $0.91 quarterly dividend payable June 11, 2026 (record May 21). Legal/IP risk with OpenAI and capacity/cost pressures leave the outlook uncertain despite strong cloud growth; the developments are material to MSFT (likely individual-stock moves ~1–3%) while near-term tech sentiment was also boosted by a five-day delay in strikes on Iranian infrastructure.

Analysis

The immediate structural takeaway is a bifurcation between compute-hardware winners and cloud incumbents forced into higher capital intensity. GPU-driven suppliers (infrastructure OEMs and chip vendors that capture marginal rack-level spend) get a multi-quarter cadence of above-trend order flow as customers and hyperscalers re-architect for model training and inference, while a cloud provider that internalizes much of that consumption faces a double hit: higher COGS today and deferred monetization of AI features. Expect this to show up as sequential margin compression and elevated capex/maintenance spending for at least the next 2 quarters, with normalization only if pricing or tiering for AI workloads becomes broadly accepted by enterprise buyers over 6–12 months.

The legal/IP channel is a binary medium-term catalyst: a favorable resolution that clarifies model-IP rights could unlock reuse of third-party models and capacity relief within 6–18 months; an adverse outcome increases verticalization and self-provisioning by hyperscalers, accelerating hardware spend. Geopolitical de-escalation is a lower-conviction, near-term volatility dampener rather than a structural driver — it reduces one-way risk premia but does not alter the capacity economics that will dominate earnings revisions.

Second-order consequences matter: enterprises delaying seat consolidation will slow SaaS churn and compress ARPU upside for AI bundles, creating a window where independent software vendors with lighter inference footprints can outcompete larger vendors on price. Conversely, firms selling data-center orchestration, software-defined rack solutions, and custom interconnects (the “ops” layer) will see disproportionate demand growth as companies choose on-prem or co-lo placement to avoid public-cloud capacity squeezes.

Key Decisions for Investors

• Long NVDA (3–9 month): buy a 3–6 month call spread to express upward GPU demand with defined risk. Target 2.5x upside if NVDA captures incremental data‑center spend; cap premium outlay to 1–2% of fund. Reduce or take profits on first 25% gain; stop-loss at 12% of premium.
• Pair trade (3–9 month): long NVDA / short MSFT equal notional to play hardware vs cloud-capacity squeeze. Size at 1–2% net exposure with target 20–35% relative return if capacity tightness persists; hard stop if pair moves >15% against position over any 30-day window.
• Short spatialomics names BRKR and TXG (3–6 month): buy puts or short on weakness size-constrained at 0.5–1% each. Rationale: demand migration risk from AI models is underappreciated; target 3:1 reward:risk on continued re-rating, stop-loss at 20% adverse move.
• Tactical long EVR (6–12 month): add small overweight to benefit from sustained analyst positivity and M&A/IB flow; target modest capital appreciation (15–25%) as sector research earnings revise. Keep position at 0.5–1% and reassess on next quarterly cadence.
• Risk management: avoid naked long MSFT exposure until legal/IP clarity (6–18 months). If unwilling to cut equity, hedge with 6–12 month put spreads (cost-limited) sized to offset 30–50% of current position’s downside exposure.