HSBC sees more upside in Intel stock on server CPU opportunity

HSBC raised its Intel price target to $200 (highest on Wall Street) while keeping a Buy rating, citing a stronger server CPU outlook and newly including Intel’s foundry business in its valuation. Analyst Frank Lee lifted 2026 server CPU shipment growth to 25% (from 20% YoY) and increased 2026 data center & AI revenue to $24.1B (+4% vs consensus), while raising 2027 server CPU growth to 30% and DCAI revenue to $33.0B (+20% vs consensus). HSBC’s sensitivity work suggests broader EMIB adoption could lift Intel’s 2028 EPS by 23% versus the base case, with EPS estimates still ~21% (2026) and ~55% (2027) above consensus.

Analysis

The market should treat this less as a one-day target reset and more as a debate about whether Intel can re-rate from a troubled hardware vendor into a capacity-constrained supply-chain asset. Near term, the real earnings lever is still server CPU mix: if shipment growth is accelerating, fixed-cost absorption can move margins faster than revenue, which is why the stock can work even before foundry dollars show up. The foundry piece is mostly a valuation bridge today; design wins in 2H26-2027 matter because they would reduce the "proof gap" between narrative and cash flow.

Second-order effects are more interesting than the headline. TSM faces not just potential unit share pressure, but a bargaining/packaging story: if Intel can credibly offer alternative advanced packaging, it gives hyperscalers and AI chip buyers leverage on pricing, lead times, and dual-sourcing. That said, this is not an immediate moat break for TSM; the more realistic near-term effect is multiple compression risk if investors start pricing in a second source of advanced packaging capacity.

Contrarian view: the street may be underestimating how much capex and execution risk is embedded in a foundry turnaround, and overestimating how quickly design commitments translate into margin-accretive revenue. If Intel’s 2026-27 server gains are real but foundry margins stay low, the equity story is still only partially de-risked. Falsifiers: any slippage in 2H26 customer design commitments, a downgrade in server CPU growth, or evidence that packaging wins come with inferior economics versus TSM's alternatives.

Key Decisions for Investors

• Tactically long INTC on pullbacks for a 3-6 month horizon, but size it as a sentiment trade rather than a fully de-risked fundamental re-rating; use failure to show 2H26 design wins or weaker server CPU shipment momentum as the stop.
• Pair trade: long INTC / short TSM into the next round of foundry/packaging commentary if you want to express packaging-share optionality; the thesis works only if investors start assigning meaningful value to non-TSMC supply, and it breaks if TSM retains pricing power and utilization remains tight.
• Buy INTC LEAPS or a call spread only if you can wait for 2026-27 catalysts; the risk/reward is tied to eventual margin leverage and foundry credibility, not this quarter's print.
• Watch NVDA as a beneficiary of incremental packaging capacity, but do not chase it unless there is evidence Intel’s packaging scales without quality or yield penalties; otherwise, the leverage stays with TSM and the AI supply chain remains constrained.
• Alert item: if sell-side 2027 EPS estimates for INTC stop moving higher over the next 1-2 quarters, fade the re-rating — the stock would then be trading on narrative instead of independently verifiable cash generation.