TSMC and ASML Results Underscore Strong AI Fundamentals, but Investors Continue to Fear the Party Will Soon End

Gene Munster, Brian BakerApril 19, 2026

The fundamentals of the AI trade continue to mostly exceed expectations, yet investors remain unimpressed. In the two trading days following better-than-expected earnings, ASML shares fell approximately 7%. Similarly, following its own beat, TSMC saw shares drop about 3% in a single day, all while the Nasdaq rose 2% over the same period. The bottom line: AI chip fundamentals are largely outperforming projections, but investors aren't rewarding these companies for their results. The reason? They fear the party is about to slow down or end entirely. The question I’m struggling with is: What will it take for these companies to win back investor favor?

Key Takeaways

TSMC beat March expectations by 2% and guided the June quarter 4% above consensus. Conversely, while ASML also beat March by 2%, it effectively lowered guidance for the remainder of the year by 1%.

The weak post-earnings stock reactions underscore that good results and in-line to favorable guidance alone continue to not be enough to move AI silicon stocks higher.

I’m struggling with what it will take to change the narrative on these AI chip stocks and get them to perform better. Ultimately, I believe that stocks follow fundamentals.

TSMC and AMSL Results and Guide

Regarding fundamentals, the takeaway from the TSMC (April 16) and ASML (April 15) earnings is that the AI silicon trade remains intact.

TSMC: Defying the Law of Large Numbers

March revenue increased by 35% y/y, which was 1.6% above Street consensus (FactSet). Furthermore, management guided June revenue approximately 4% higher than Street expectations, implying year-over-year growth of 34%. To put that in perspective, this follows 35% growth last year and 32% in 2024. The key takeaway here is that growth rates are not slowing despite the “law of large numbers,” which is a testament to the sustained demand for AI silicon.

ASML: Navigating the “Lumpy” Business

March revenue increased 13% y/y, which was 2.0% above the Street. For the June quarter, management essentially guided revenue to 16% growth (factoring in their typical reporting patterns versus their guidance range), which is in line with consensus. For FY26, they guided to 19% growth, slightly ahead of the Street’s 18%.

The negative read causing investor hesitation is that they did not raise guidance for the remainder of the year. When you factor in the 2% upside from the March quarter, they effectively lowered the revenue outlook for the final three quarters by approximately 1%. Keep in mind this is a lumpy business.

This lumpiness is a byproduct of ASML’s business model, where individual machines cost between $20M and $400M. What’s important is yearly growth rates have been going up.

2024: Flat y/y
2025: Up 12%
2026 (Projected): Up 19%

Despite the 1% guide down for FY26, the trend shows a clear acceleration from over the past couple of years.

Investors Want More

As mentioned, the fundamentals of the AI trade continue to largely exceed expectations (the exception was the 1% ASML FY26 guide down), yet investors remain unimpressed. In the two trading days following ASML’s earnings, shares fell approximately 7%. Similarly, in the single day following TSMC’s report, shares dropped about 3%, notably trailing the Nasdaq, which rose 2% over that two-day period. Over the past month, the trend persists: TSMC is up 7% and ASML is up 3%, both lagging the Nasdaq’s 8% gain.

The simplest conclusion is that investors are no longer willing to pay a premium for confirmation that AI demand is strong for the current year. Instead, they’re looking for indicators that next year’s already elevated numbers can be exceeded.

Changing the Narrative

I’m struggling with what will change the narrative on these AI infrastructure and chip stocks. At Deepwater, we continue to own many of them because of our belief that growth in these companies will be higher than investors currently expect for the next several years. That said, even as these companies have been marking time with better-than-expected results and guidance (setting aside ASML FY26 guide was 1% below the mark), shares continue to be under pressure.

Nvidia: On March 18th at GTC, Jensen suggested CY27 will grow at greater than 40% over CY26, while the Street was looking for 30% growth. Since then, shares of NVDA have traded up by 8%, essentially in line with the Nasdaq’s up 7%. I would summarize the stock reaction as a disappointment, given its lack of outperformance relative to that improved outlook.

In the end, while we have made changes, we remain long the AI infrastructure trade based on our belief that, eventually, stocks follow fundamentals.

Disclaimer