Macquarie cuts Alibaba stock price target on AI investment costs By Investing.com

Macquarie trimmed its price target on Alibaba to $175.90 (from $176.20) while maintaining an Outperform rating; the U.S.-listed shares trade at $127.68 and are down ~26% over six months. Alibaba reported Q3 revenue and profitability below expectations and faces higher near-term investment into AI (Qwen) and cloud/training costs, prompting Susquehanna to cut its target to $170 (from $190) while Morgan Stanley kept a $180 target and expects long-term cloud margins near 20%. Strategic positives include the launch of the 5nm XuanTie C950 AI chip (>3x faster than its predecessor) and narrowing Qwen losses, which partly offset near-term profitability headwinds.

Analysis

The recent activity highlights an asymmetric bifurcation between incumbents that can monetize AI across cloud and at-edge hardware and those that are still funding product-market fit. For firms that control both silicon and service distribution, the second-order effect is a longer, stickier revenue stream — not just one-off chip sales — which should lift server ASPs and recurring software revenue over a 12–36 month window even if near-term gross margin is pressured by R&D.

Conversely, vertically integrating hyperscalers pursuing in-house ISA designs and AI stacks compress the addressable market for third-party accelerators and IP over the medium term; I estimate this could shave 10–20% off external ASIC TAM in specific China-centric enterprise verticals within 2–4 years. That shift also reroutes capex to in-house fabs/packaging and away from traditional outsourced suppliers, creating winners among systems integrators and niche packaging vendors while hurting some FPGA/accelerator specialists.

Key catalysts to watch are iterative benchmark releases, region-specific certification/regulatory headlines, and quarterly cloud metrics (token consumption, price-per-inference). Near-term reversal risks include failed integration proofs, customer churn to alternative AI stacks, or an acceleration of competitor wins that relegate incumbents to price competition; these events can play out over days (news-driven), quarters (guidance/earnings), and years (chip adoption curves).

The consensus is focused on headline AI progress, underweighting the profit-versus-investment tradeoff and the multi-year erosion of third-party silicon demand in localized markets. That creates both tactical opportunities to play implementation wins and strategic shorts where spending is outpacing monetization with limited near-term margin visibility.

Key Decisions for Investors

• Long INTC equity or 6–9 month call spread (buy 1x 9-mo ATM call, sell 1x 9-mo +20% call) — thesis: capture device-to-cloud integration premium; target 30–40% upside if adoption signals accelerate; stop-loss at -12% on position value to limit execution risk.
• Pair trade: Long GOOGL 6–12 month calls (targeting cloud monetization) / Short BABA 6–12 month puts or equity (ratio 1:1) — plays US cloud monetization vs China margin drag; target asymmetric R/R ~2.5:1 if China regulatory/capex headwinds persist; unwind if token/usage metrics diverge positively for BABA for two consecutive quarters.
• Short BABA outright with a 6–12 month horizon via puts sized to target 20–30% downside — rationale: continued investment cadence can depress near-term FCF; hedge with a small long-dated call (12–18 month) to cap tail risk if successful monetization accelerates (breakeven ~50% move in either direction).
• Buy Morgan Stanley (MS) exposure (6–12 months) via equity or call spread as a trade on information asymmetry and potential re-rating of coverage winners — modest sized position to offset liquidity in shorts; target 15–25% upside if sector rotation favors research-led flows, stop-loss -10%.