TSMC reports February revenue growth year-over-year, sequential decline from January

TSMC reported February 2026 consolidated net revenue of NT$317.66 billion, +22.2% YoY but -20.8% MoM, and NT$718.91 billion for Jan–Feb 2026, +29.9% YoY. The SEC filing shows loans to subsidiaries of NT$10.96 billion (TSMC Nanjing) and NT$1.88 billion (TSMC Washington) and guarantees of NT$2.60 billion (TSMC North America), NT$203.19 billion (TSMC Global) and NT$341.62 billion (TSMC Arizona). Derivatives exposure includes NT$290.99 billion notional in forwards not applying hedge accounting (MTM value NT$1.80 billion, cumulative unrealized profit NT$4.82 billion, realized losses NT$3.62 billion) and NT$456.40 million notional applying hedge accounting with an MTM loss of NT$2.31 million. All figures are per the company's press release filed with the SEC.

Analysis

TSMC’s public disclosures around off-balance-sheet support and active forward hedging reveal the firm is managing a multi-front transition: capitalizing on structural demand for advanced nodes while simultaneously financing and de-risking a geographic capex pivot. Large guarantees to US projects create a meaningful contingent liability pathway — if US build schedules slip or financing costs rise, earnings volatility could arrive through higher interest or project delays rather than product demand misses.

The derivatives footprint (big notional, small mark-to-market, but realized losses on expiries) is consistent with frequent short-dated roll activity that mops up FX and commodity mismatches but leaks P&L volatility quarter-to-quarter. This churn suggests near-term EPS growth can be lumpy even as underlying demand holds; importantly, persistent realized hedging losses are an early warning that the firm’s mix of currency and interest-rate exposures may be biased toward outcomes that penalize earnings in a low-volatility or appreciating domestic-currency regime.

Second-order beneficiaries include ASML/Lam/Applied-equivalent capex suppliers that win on multi-year node transitions and local US construction contractors who capture outsized near-term cashflows from Arizona projects. Conversely, regional OSATs and small fabless players that prioritized China-based capacity could see pricing pressure as TSMC rebalances allocation to higher-margin, advanced-node customers in nearer-term capacity tightness.

Contrarian synthesis: the market is likely discounting short-term seasonality while underweighting a persistent margin tailwind from advanced-node scarcity and US-fab pricing power. That creates a two-phase trade — tolerate near-term hedging/capex noise for asymmetric multi-quarter upside as node mix and geographic diversification begin to lift ASPs and reduce customer-side bargaining power.

Key Decisions for Investors

• Buy TSM (TSM) Jan-2028 LEAP call (approx. 12–18 month horizon): target 30–50% upside on node-driven ASP recovery; max loss = premium (size position at 1–2% notional). Rationale: capture multi-quarter margin rerating while limiting downside to option premium.
• Buy-and-collar on TSM stock for 6–9 months: long shares, buy 6-month 10% OTM puts and sell 6-month 25% OTM calls to finance ~60–80% of put cost. Expect limited upside cap but protect against >10% downside from hedging/capex shocks; use collar if you own stock and want convexity reduction.
• Relative-value pair: long TSM / short SMCI, 12-month horizon, notional-neutral. Target 20–30% relative outperformance if foundry ASPs firm and server/AI hardware cyclicality lags. Stop-loss: 10% absolute move against pair or re-assess on material revision to demand data.
• Volatility tactical: avoid selling near-dated straddles into earnings given hedging churn; instead buy 30–60 day straddles only if realized vol begins to structurally exceed implied vol by 25% (event-driven). This captures episodic volatility from derivative roll losses or project announcements while controlling premium spend.