Is TSMC Stock a Buy After Q2 Earnings? What 22x Forward Earnings and 61% AI Revenue Tell Investors

TSMC stock at approximately 22 times forward earnings after a 52% year-to-date gain puts investors in a specific and uncomfortable position that the strong Q2 results do not automatically resolve. TSMC stock has already rewarded shareholders generously in 2026, and the question of whether buying after a 52% run still makes sense requires a different analytical framework from the one that justified buying at the start of the year. TSMC stock at current prices is not the same investment it was in January, and treating it as though the original entry thesis still applies without adjustment is the specific mistake that leads investors to buy quality companies at prices that do not generate satisfactory forward returns.

The buy decision today runs through three specific questions that the 22 times forward multiple and the 61% AI revenue figure each illuminate in different ways.

Why 22 Times Forward Earnings Is Both Cheap and Not Cheap Simultaneously

The forward multiple of approximately 22 times on next twelve months earnings is the number that most directly frames the buy decision and the one that most requires context to be useful.

Twenty-two times forward earnings is cheap relative to where the rest of the AI investment ecosystem trades. Nvidia trades at a significantly higher forward multiple on comparable revenue growth. ASML trades near 49 times forward on strong but not superior earnings trajectory. AMD trades at 73 times forward. On a relative basis within the AI chip supply chain, TSMC at 22 times is the most compressed valuation of any company with direct and demonstrated exposure to the AI hardware buildout.

Twenty-two times forward earnings is not cheap in an absolute sense for a company that is simultaneously managing the geopolitical risk of Taiwan concentration and the margin risk of overseas manufacturing expansion. A semiconductor manufacturer at 22 times forward earnings in a normal cycle would be considered fully valued. The argument that TSMC deserves more than a normal semiconductor multiple rests entirely on whether the AI chip demand cycle has the duration and magnitude that the 61% AI revenue contribution implies.

The resolution of that tension between relatively cheap and not absolutely cheap is what the buy decision at current prices requires investors to take a position on. There is no neutral ground between those two framings at 22 times forward earnings.

What 61% AI Revenue Actually Tells You About Duration

The 61% AI revenue figure is the most important number in TSMC's results for evaluating the forward earnings trajectory that the 22 times multiple is applied to.

Two years ago AI chips were approximately 6% to 8% of TSMC's revenue. The progression from single-digit percentage to majority revenue in under two years is the fastest structural shift in the composition of a major company's revenue in recent memory, and it happened not because TSMC pivoted its strategy but because its existing customers redirected their capital expenditure budgets toward AI hardware at a pace that exceeded every projection.

At 61% AI revenue and rising, the earnings trajectory that makes 22 times forward earnings a reasonable multiple depends on AI chip demand sustaining at levels that prevent the revenue concentration from becoming a single-point-of-failure risk. If AI chip demand plateaus or reverses, TSMC's revenue and earnings compress faster than a more diversified revenue mix would allow.

The evidence that AI chip demand is not plateauing is visible in the quarterly results themselves. Revenue growth of 33% year over year on a base that was already extraordinary is not the profile of a demand cycle approaching its peak. The CoWoS packaging bottleneck that is physically constraining AI accelerator delivery is not the constraint profile of a market where demand is softening. Demand that is constrained only by supply capacity is the demand structure that justifies owning a supplier rather than watching from the sidelines.

The $200 Billion US Investment and What It Does to the Valuation Framework

The additional $100 billion US investment commitment announced alongside Q2 results, bringing the total to $200 billion, introduces a specific dimension to the buy decision that a purely earnings-based analysis misses.

A $200 billion commitment to US manufacturing is not primarily a cost decision. At the higher labor and overhead costs of US semiconductor manufacturing relative to Taiwan, this commitment will compress margins in the facilities it builds rather than expand them. Management has been managing this through pricing power and high utilization rates, but the long-term effect of shifting a meaningful portion of production to higher-cost US fabs is a structurally lower gross margin profile than the 66% that Taiwan-concentrated manufacturing has produced.

The reason a sophisticated management team makes a margin-compressing commitment of this magnitude is that the alternative, remaining concentrated in Taiwan manufacturing without meaningful US presence, carries geopolitical risks that the $200 billion is buying insurance against. A TSMC with $200 billion in US manufacturing assets is a different geopolitical risk profile from a TSMC with $20 billion in US manufacturing assets, and that different risk profile is worth a margin premium in the form of lower investor required return on equity.

For the buy decision, the US investment commitment changes the valuation framework in a specific way. It reduces the Taiwan geopolitical risk discount that has historically kept TSMC's multiple below what its growth rate and competitive position would otherwise justify. A TSMC that is progressively building dual-location manufacturing capability is a TSMC that deserves a higher multiple than a purely Taiwan-concentrated manufacturer, even if the absolute margin profile is slightly lower. Whether the multiple re-rating from reduced geopolitical risk more than offsets the margin compression from higher cost US manufacturing is the specific analytical question the $200 billion commitment introduces.

The Competitive Position That No Multiple Fully Captures

One dimension of the TSMC buy decision that valuation multiples systematically understate is the competitive position that makes TSMC's earnings quality different from any other semiconductor company.

TSMC has no direct competitor at the leading edge of semiconductor manufacturing. Intel Foundry is progressing toward competitive parity on its 18A process but has not yet demonstrated it can attract and retain the customer relationships at volume that TSMC commands. Samsung Foundry competes for some customers but has consistently ceded the most advanced and most profitable AI chip work to TSMC because of TSMC's superior yield rates and process maturity at leading edge nodes.

A company with no direct competitor in its primary revenue-generating activity is a different investment from a company with a leading position in a competitive market. TSMC's customers cannot threaten to take their business elsewhere in any meaningful sense because there is no meaningful elsewhere to take it. Nvidia can express frustration with TSMC's delivery timelines. It cannot shift its Blackwell production to Samsung or Intel Foundry and expect comparable yields, delivery reliability, and advanced packaging integration.

That absence of substitution options for customers creates pricing power that is structural rather than cyclical. TSMC has been raising prices for its most advanced nodes because the demand at existing prices already exceeds available capacity. No competitor is waiting to undercut those price increases with equivalent technology at lower cost. The pricing power is real and durable in ways that a 22 times multiple does not fully reflect because conventional semiconductor company analysis assumes competitive pressure limits pricing power over time.

What the Buy Decision Looks Like for Different Investor Profiles

Rather than a single universal buy or hold answer, the TSMC stock buy decision at current prices produces different rational conclusions for different investors.

Long-term investors with a three to five year horizon who believe AI chip demand compounds at rates that sustain TSMC's current revenue trajectory and who have incorporated the Taiwan geopolitical risk into their portfolio construction find the 22 times forward multiple on a company with no direct competitor in the most important manufacturing process in technology to be one of the more defensible large positions available in the current market. The $200 billion US investment commitment is progressively reducing the geopolitical risk discount that has historically prevented TSMC from trading at the multiple its competitive position deserves.

Investors who bought TSMC earlier in 2026 and are evaluating whether to hold through the Q2 results have the specific question of whether the post-earnings price represents continued opportunity or the point where the easy money has been made. A 52% year-to-date gain has captured a meaningful portion of the re-rating from the start-of-year multiple to the current 22 times. Continued appreciation requires either earnings growth to sustain the current multiple or multiple expansion above 22 times as the geopolitical risk discount further diminishes.

Investors who have not yet owned TSMC and are evaluating entry after the Q2 beat face the classic dilemma of buying a quality company after it has already performed well. The honest answer is that buying after a 52% run requires accepting that a portion of the potential return has already been captured, and that the remaining return depends on the earnings trajectory and multiple re-rating continuing from here rather than having already delivered most of the available upside.

The One Risk That Every Buy Case Must Acknowledge

The Taiwan geopolitical risk is the specific concern that prevents TSMC stock from trading at the multiple that its competitive position and earnings quality would otherwise justify, and any honest buy case must engage with it directly rather than noting it and moving on.

TSMC's primary manufacturing base in Taiwan is approximately 110 miles from mainland China. The political status of Taiwan, and specifically the risk that any escalation in the relationship between China and Taiwan disrupts TSMC's manufacturing operations, is the single largest non-business risk in any TSMC stock position.

The $200 billion US investment commitment is the most concrete available mitigation of that risk, but it is a mitigation rather than an elimination. Even with $200 billion in US manufacturing, the majority of TSMC's most advanced manufacturing capacity will remain in Taiwan for the foreseeable future. The US fabs being built today will not reach production volumes that meaningfully reduce Taiwan concentration until the late 2020s at the earliest.

For investors who are specifically concerned about Taiwan geopolitical risk, TSMC stock should be sized as a position that can absorb a severe disruption scenario without creating portfolio-level damage. For investors who have evaluated the geopolitical risk and incorporated it into their position sizing, the business case at 22 times forward earnings with 33% revenue growth and an unmatched competitive position is as strong as it has been at any point in 2026.

For those looking to participate in global financial markets, having access to the right trading platform matters. WEEX offers crypto and stock trading products, covering major global markets including US stocks and digital assets.

Conclusion

TSMC stock at 22 times forward earnings after Q2 results confirming 33% revenue growth and 61% AI revenue concentration is a buy for investors who believe the AI chip demand cycle has the duration to sustain current revenue trajectory and who have sized the position to accommodate Taiwan geopolitical risk.

The case for buying is not that 22 times is cheap in absolute terms. It is that 22 times on a company with no direct competitor in the most critical manufacturing process in technology, growing revenue at 33% annually, and progressively reducing its geopolitical risk concentration through a $200 billion US investment is undervalued relative to the rest of the AI investment ecosystem where comparable and lesser businesses trade at significantly higher multiples.

The case against buying is that 52% of the year's potential return has already been captured, that the Taiwan geopolitical risk remains despite the US investment commitment, and that 22 times forward earnings leaves less margin for any demand softness than the compressed multiples of early 2026 provided.

What Q2 confirmed is that the demand softness scenario has not arrived. Whether it arrives before or after TSMC's US manufacturing base is substantial enough to provide genuine geopolitical protection is the specific uncertainty that determines whether buying today at 22 times generates the returns the competitive position and earnings quality deserve.

FAQ

1. Is TSMC stock a buy after Q2 2026 earnings?
At 22 times forward earnings with 33% revenue growth and no direct competitor in leading-edge semiconductor manufacturing, the buy case is strong for long-term investors who have incorporated Taiwan geopolitical risk into their position sizing. The primary consideration for investors who have not yet owned TSMC is that 52% of the year's return has already been captured, and continued appreciation requires earnings growth and multiple re-rating to continue from current levels.

2. What does 22 times forward earnings mean for TSMC stock?
It means TSMC trades at a lower multiple than Nvidia, ASML, and AMD despite comparable or stronger demand visibility and a competitive position none of those companies can replicate. The 22 times reflects a persistent geopolitical risk discount from Taiwan concentration that the $200 billion US investment commitment is progressively reducing but has not yet eliminated.

3. What does 61% AI revenue mean for TSMC's future earnings?
AI chips at 61% of revenue means TSMC's earnings trajectory is now primarily determined by AI chip demand rather than by consumer electronics cycles. AI chip demand driven by enterprise capital expenditure commitments has longer duration and higher floor than consumer electronics demand, which makes TSMC's earnings more durable than historical semiconductor cycle analysis would suggest.

4. How does the $200 billion US investment affect the buy decision?
The additional $100 billion commitment brings total US investment to $200 billion and progressively reduces the Taiwan geopolitical risk discount that has historically prevented TSMC from trading at the multiple its competitive position deserves. The cost is modest margin compression from higher-cost US manufacturing. The benefit is a different geopolitical risk profile that justifies a higher multiple over time.

5. What is the biggest risk to buying TSMC stock at current prices?
Taiwan geopolitical risk is the specific concern that prevents TSMC from trading at the multiple its competitive position would otherwise justify. Despite $200 billion in US manufacturing commitments, the majority of TSMC's most advanced manufacturing capacity remains in Taiwan for the foreseeable future. Position sizing that accommodates a severe disruption scenario is the specific risk management approach the geopolitical overhang requires.