ABC summary: TSMC's Q4 2024 earnings call

The to-the-point summary of the Q&A at TSMC's most recent earnings call. I used AI to simplify questions, abridge answers, and bundle by key theme for accessibility and readibility.

TSMC Q4 2024 earnings call analysis

Financial performance and outlook

Q: What were the key financial metrics for Q4 2024?

• Revenue increased 14.3% sequentially in NT dollars

• Gross margin increased 1.2 percentage points to 59%

• Operating margin increased 1.5 percentage points to 49%

• Full year 2024 revenue increased 30% to USD 90 billion

• Full year EPS increased 39.9% to TWD 45.25

Q: What is the guidance for Q1 2025?

• Revenue expected between USD 25-25.8 billion

• Represents 5.5% sequential decline

• Projects 34.7% year-over-year increase at midpoint

• Gross margin expected between 57-59%

• Operating margin projected between 46.5-48.5%

Context: TSMC's financial performance shows strong growth momentum despite industry cycles. The growth projections are particularly significant because they're coming off an already high base in 2024. The sustained high growth rate, especially in AI accelerators, suggests TSMC sees no near-term ceiling for AI chip demand. The company is notably confident about maintaining strong growth even after 2024's exceptional performance. The company maintains high margins even while expanding globally, though there are some headwinds from overseas expansion costs.

---

Capital expenditure & investment

Q: What are the 2025 investment plans?

• 2025 capex budget: $38-42 billion

• Allocation breakdown:

  • 70% for advanced process technologies

  • 10-20% for specialty technologies

  • 10-20% for advanced packaging, testing, mask making

• 2025 depreciation: High single-digit percentage increase YoY

Context: This represents a significant increase in capital spending from 2024's $29.8 billion, indicating TSMC's strong conviction in future demand. The heavy weighting toward advanced processes suggests the company expects continued strong demand for leading-edge nodes, particularly from AI applications.