How do you think about margins compressing as TSMC moves large amounts of production to the US where labour costs are far higher? If margins compress, earnings diminish and so too do valuations.
While this is margin dilutive, we see a path for TSMC to achieve 65% Gross Margins by 2030. This would continue a trend of higher margins over the past two decades. There are three core drivers of future gross margin expansion:
- Leading-edge pricing power and higher utilization
- Advanced Packaging margin improvement
- Removal of temporary gross margin headwinds
We estimate TSMC is currently running its Taiwan wafer fabs at +60% Gross Margin. TSMC should be able to leverage its quasi-monopoly status at the leading-edge to drive ASP increases 3% above cost increases. Pricing power and higher utilization should drive Taiwan wafer fabs to the high-60% range.
Advanced Packaging / Other Revenue currently accounts for 13% of total company revenue and is GM dilutive. CoWoS expansion, HBM base die relations, and CPO capabilities deepen customer reliance and give TSMC an edge in its goal of building a systems foundry. While likely a low-50% GM business today, we think the business can reach mid-60% margins with scale.
TSMC is currently facing 8-11 ppts of GM headwinds. Over half of the headwind is structural in our view, with new node ramp (N3), overseas fabs, and electricity costs cumulatively a 5-8 ppt headwind. However, proforma TSMC corporate margins are likely in the high-50%/low-60% range when adjusting for transient headwinds.
We think TSMC should be able to continue to leverage its effective monopoly at the leading edge to drive GM expansion through the end of the decade, as pricing power and scale more than offsets dilution from overseas fab buildouts.
One more question. With the advent of AI, there has been a huge spike in GPU demand, but these things aren't cheap and have a reasonable life span, so how much of future semiconductor demand do you think has been pulled forward and how much of a vacuum has that created in the years ahead? This is a cyclical business after all.
Do you view business operating model portability as a risk? I.e. can they replicate what they've done in Taiwan in the US, esp as it pertains to getting the proper talent (e.g. pick of Taiwan's best EE grads without competition from finance and software co's looking to hire)
How do you think about margins compressing as TSMC moves large amounts of production to the US where labour costs are far higher? If margins compress, earnings diminish and so too do valuations.
From SemiAnalysis:
TSMC has committed to invest $165b in the US. 💵 🚀
While this is margin dilutive, we see a path for TSMC to achieve 65% Gross Margins by 2030. This would continue a trend of higher margins over the past two decades. There are three core drivers of future gross margin expansion:
- Leading-edge pricing power and higher utilization
- Advanced Packaging margin improvement
- Removal of temporary gross margin headwinds
We estimate TSMC is currently running its Taiwan wafer fabs at +60% Gross Margin. TSMC should be able to leverage its quasi-monopoly status at the leading-edge to drive ASP increases 3% above cost increases. Pricing power and higher utilization should drive Taiwan wafer fabs to the high-60% range.
Advanced Packaging / Other Revenue currently accounts for 13% of total company revenue and is GM dilutive. CoWoS expansion, HBM base die relations, and CPO capabilities deepen customer reliance and give TSMC an edge in its goal of building a systems foundry. While likely a low-50% GM business today, we think the business can reach mid-60% margins with scale.
TSMC is currently facing 8-11 ppts of GM headwinds. Over half of the headwind is structural in our view, with new node ramp (N3), overseas fabs, and electricity costs cumulatively a 5-8 ppt headwind. However, proforma TSMC corporate margins are likely in the high-50%/low-60% range when adjusting for transient headwinds.
We think TSMC should be able to continue to leverage its effective monopoly at the leading edge to drive GM expansion through the end of the decade, as pricing power and scale more than offsets dilution from overseas fab buildouts.
This is very helpful. Thank you.
One more question. With the advent of AI, there has been a huge spike in GPU demand, but these things aren't cheap and have a reasonable life span, so how much of future semiconductor demand do you think has been pulled forward and how much of a vacuum has that created in the years ahead? This is a cyclical business after all.
Do you view business operating model portability as a risk? I.e. can they replicate what they've done in Taiwan in the US, esp as it pertains to getting the proper talent (e.g. pick of Taiwan's best EE grads without competition from finance and software co's looking to hire)