TSMC posts 35% jump in revenue to new record high as AI chip demand stays strong

TSMC reported first-quarter revenue of NT$1.13 trillion ($35.6 billion), topping the NT$1.12 trillion analyst estimate and rising 35% year over year; March revenue alone jumped 45.2% to NT$415.2 billion. The results were driven by sustained AI-chip demand from customers like Apple and Nvidia, though supply-chain disruption risks from Middle East conflict remain a watchpoint. Management does not give full quarterly profitability guidance until April 16, but an analyst said TSMC should easily exceed its 30% annual growth target.

Analysis

TSMC’s print is less about a single-quarter beat and more about pricing power at the bottleneck of the AI supply chain. If advanced-node capacity remains structurally tight, the incremental profit pool migrates upstream to the foundry rather than staying with the designers, which means gross margin expansion can persist even if end-demand normalizes. That is a subtle but important reversal of the usual AI trade, where investors overpay for the “visible” semis and underwrite the manufacturing choke point too cheaply.

The second-order winner is the ecosystem around leading-edge capex: ASML should benefit if TSMC’s utilization stays high and node transitions remain on schedule, but the more interesting angle is that every incremental custom chip program from hyperscalers still has to clear the same manufacturing gate. More chip design by cloud and AI firms does not necessarily dilute TSMC; it can actually widen its pricing umbrella because fragmented demand reduces customer bargaining power and raises the cost of switching. By contrast, legacy players with weaker process access face a widening strategic gap, especially if AI-related allocation crowds out non-AI silicon.

Near term, the key risk is not demand collapse but margin compression from mix or policy: if smartphone/PC weakness deepens faster than AI volume scales, the headline revenue growth can stay strong while profitability disappoints. Over 1-3 months, supply-chain shocks or export-policy changes could also disrupt order timing, but over 6-12 months the real threat is capacity normalization as peers and customers push harder into vertical integration and alternative foundry diversification. The market may be underestimating how much of this upside is already monetized through pricing actions rather than pure unit growth, which makes the next catalyst more about margin confirmation than revenue acceleration.

The contrarian view is that consensus may be too bullish on the downstream AI names and still too cautious on the toll-collector model. If AI inference matures into a more fragmented custom-chip market, TSMC can remain the primary beneficiary while a lot of the design-IP stories get competed away. The stock may also be trading against an earnings-quality narrative: as long as pricing holds, fundamentals can keep surprising even if revenue growth decelerates modestly.

Key Decisions for Investors

• Long TSM into the Apr. 16 earnings print; use a 3-6 week horizon and favor call spreads over stock if implied volatility is elevated. Thesis: margin confirmation and pricing commentary can re-rate the name even if revenue growth slows modestly.
• Pair trade: long TSM / short a basket of higher-beta AI design names that depend on future volume expansion (e.g., ARM vs. TSM on relative basis). Hold 1-3 months; risk is that custom silicon adoption accelerates faster than expected.
• Add to ASML on any post-earnings weakness over the next 1-2 weeks. The setup is a capex-duration trade: if TSMC keeps utilization high, lithography demand should remain resilient despite macro noise.
• Underweight INTC on a 3-6 month view as the foundry gap widens and the market prices in a longer timeline for credible leading-edge competitiveness. Risk/reward is favorable because upside requires multiple execution milestones, not just one good quarter.
• For hedged AI exposure, prefer TSMC over NVDA for the next quarter: TSMC offers a better risk-adjusted way to own AI infrastructure with lower product-cycle risk and more direct pricing leverage.