Swiss watch exports climb 9.2% in February on US, Japan demand

Swiss watch exports rose 9.2% YoY to CHF2.168bn in February, with wristwatches accounting for CHF2.075bn (+10% YoY). Major market movements: US +26.8% to CHF431.5m, Japan +23.7% to CHF155.9m, France +57.1% to CHF152.7m, UK +10% to CHF142.3m, Singapore +5.1% to CHF135.7m, while Hong Kong was the only major market to decline (-5.2% to CHF146.5m).

Analysis

Sustained, high-volume demand for advanced accelerators is reshaping capital allocation across the semiconductor supply chain: foundries, HBM memory vendors and advanced packaging providers see their pricing power and capacity lead times extend, which funnels incremental margin upstream faster than it translates to end customers. Expect lead-time elasticity to be the primary bottleneck over the next 3–9 months — capacity prioritization will show up as inflated ASPs for 5nm/4nm wafers and premium for CoWoS/EMIB slots, benefiting TSMC/ASML/SK Hynix more than fabless peers in the near term.

At the component level, elevated accelerator volumes create a multiplier effect for high-bandwidth memory and power delivery ecosystems: HBM supply tightness can add $50–150+ of bill-of-materials per top-end GPU module, squeezing competitors that lack vertical integration or long-term supply contracts. This dynamic widens the moat for incumbents who control packaging & memory relationships, making it harder for challengers to undercut on price without meaningful capital outlays or long-term purchase commitments.

Key reversals to watch are policy and architecture risk. Export-control changes or a sudden pivot by large OEMs toward custom in-house inference/training ASICs are 6–24 month tail-risks that would materially compress consensus growth multiple for accelerators. Near-term market risk centers on quarterly demand visibility and earnings beats — a single miss in bookings or foundry guidance can trigger 20–30% repricing across the long-duration names.

Given these mechanics, tactical positioning should be asymmetric: own the capacity winners while protecting against a policy/architecture shock. Monitor foundry utilization, HBM spot spreads, and OEM procurement cadence as primary leading indicators for revenue revisions over the next 1–4 quarters.

Key Decisions for Investors

• Long NVDA via calendar/LEAP call spread (buy Jan-2027 LEAP call, sell Jan-2028 higher-strike call) — target entry on a 5–12% pullback or after next earnings if guidance is stable. Rationale: capture multi-quarter structural demand while capping premium; size 3–5% of tech book, target 2–4x payoff over 12–24 months, max loss = premium.
• Pair trade: long NVDA / short AMD (equal deltas) for 6–12 months — express conviction in capacity-led share gains. Entry: initiate after foundry utilization print confirms >90% 5nm/4nm utilization; target relative return 15–30%, stop if NVDA outperforms by >20% or AMD reports upside market-share surprises.
• Directional TSLA trade: buy Oct-2026 call spread (bull-call) sized 2–3% of equity allocation to play continued software/AI upside while capping capital at risk. Trigger: enter on delivery beats or on any 10%+ pullback in stock; expected holding period 6–12 months, max loss = premium, upside 2–3x if feature monetization scales.