Taiwan Semi builds chipsets for other companies like Nvidia and Qualcomm. It does so in its highly expensive, highly complex chip fabrication plants. These are called “fabs” for short.

b. Guidance & Valuation

Q3 2024 guidance was 1.6% ahead on revenue, 4.4% ahead on operating income & 180 basis points (bps; 1 basis point = 0.01%) ahead of 52.7% GPM estimates.

For the full year, TSMC now expects to do slightly better than its mid-20% Y/Y revenue growth guide offered last quarter.

TSMC trades for 26x this year’s earnings estimates. Earnings are expected to grow by 24% Y/Y this year and 29% Y/Y next year. Here’s how its next-12 month earnings multiple compares to historical norms:

Sub N7 technology made up 67% of its total wafer revenue vs. 53% Y/Y while high-performance compute (HPC) crossed 50% of total revenue for the first time. HPC is where the GenAI explosion shows most clearly in its product demand and overall financial success. Strong demand for the N3 and N5 technology was able to overcome some smartphone seasonality to facilitate the outperformance. Robust N3 shipment levels were also the source of its days inventory on hand decrease highlighted in the balance sheet section.

As a reminder, the ramps of its new nano-chip technologies always lead to gross margin dilution. It takes time for capacity to scale to a point of broader efficiency and higher profitability. Beyond this gross margin headwind, TSMC is also shifting some N5 production capacity to N3, which creates disruption and another 1 to 2 points of GPM headwinds for 2024. Both of these headwinds coincide with electricity inflation in Taiwan and its continued manufacturing expansion to higher-cost regions. Considering this, I found the 90 bps of Y/Y GPM contraction to be quite reasonable (and better than expected). Advanced tech scaling and the capacity shift will eventually turn to tailwinds towards the end of this year and into 2025.

The Q2 gross margin beat was due simply to better-than-expected capacity utilization vs. its forecast, which it expects to continue through Q3. “Better cost improvement efforts and productivity gains” were also GPM tailwinds during the quarter. The team was asked why it isn’t raising its 53% or higher long term GPM target, which it raised from 50% in 2021. All they were willing to offer is telling investors to focus on the “or higher” portion of its 53% or higher target. Wink, wink.

TSMC raised its CapEx guide for the year from $30 billion to $31 billion. About 75% of that will be for its advanced nano-chip process tech, 15% for specialty technologies and the remaining 10% for packaging, testing and other categories. This CapEx allocation leads us to an interesting idea: TSMC sees its broadening use cases greatly expanding its total addressable market (TAM) from $115 billion to $230 billion. It now sees overall foundry industry services including packaging and testing (not just creating an actual chip). Packaging includes storing and integrating chips with thermal protection, maintenance and connectivity tools too. Traditional foundry services make-up the actual creation of an integrated circuit (IC) or chip for a customer. The firm believes this new TAM definition “better reflects TSMC’s expanding market opportunities.” It had a 28% market share of this extended TAM definition in 2023, which it sees rising in 2024.

TSMC used to tell us that packaging margins would be inferior vs. its core foundry services. That is no longer the case as GPM for this segment approaches the rest of its business.

The team was also asked about potential geopolitical tensions, Trump’s comments on wanting the U.S. to control more chip manufacturing and Biden’s comments on more export restrictions. None of this has changed its overseas factory expansion plans in the slightest. Full speed ahead.

“The continued surge in AI-related demand supports strong demand for energy-efficient computing. As a key enabler of AI applications, the value of our technology position is increasing as customers rely on TSMC to provide the most advanced process and packaging technology at scale in the most efficient and cost-effective manner.”

CEO C.C. Wei

TSMC N2 technology is “progressing well” with “device performance and yield” on track. N2 delivers 10%-15% speed improvements at the same power or 25%-30% power improvements at the same speed vs. its best N3 chip – with 15% higher chip density.  As TSMC always does, it’s already iterating on this N2 technology with a version called N2P. N2P offers a 5% performance boost at the same power or a 5%-10% power benefit at the same speed compared to N2. This will support more HPC apps with volume production set for the 2nd half of 2026.

Its newest nano-chip technology is 1.6NM (N16) for HPC use cases. This comes with a separate power rail offering with backside power delivery (first of its kind). Traditionally, power is transferred to transistors from the front of the chip. By moving it to the back, energy efficiency and heat management are both improved. N16 offers an 8%-10% speed improvement at the same power or a 15%-20% power improvement at the same speed vs. the N2P product. It also offers 7%-10% chip density advantages over N2P. Production here will begin in the 2nd half of 2026.

This was another great quarter for a world-class company. There’s nothing to pick at fundamentally. Geopolitical risk will continue to be the focal point for this company and I don’t really see that positively changing any time soon. As long as the world lets TSMC operate as an independent company in a fair market, this should do quite well. But that’s far from certain.

Revenue outperformance this quarter was driven by outsized membership gains due to stronger acquisition trends and low churn. This implies net adds were ahead of Netflix internal estimates. Paid sharing continued to have a positive impact on growth as well.

For Q3, Netflix’s guidance missed by 0.8% on revenue, beat by 8.3% on EBIT and beat $4.72 GAAP EPS estimates by a comfortable $0.38. More Q3 items to note:

For the full year, Netflix raised its 14% Y/Y revenue growth guidance to 14.5% Y/Y. Membership and business trends were the source of the raise, which was DESPITE incremental FX headwinds now baked in. Strong in my view. Still, revenue guidance actually missed sell-side estimates looking for 15% Y/Y growth. Netflix also raised its 25% EBIT margin guide to 26%, which beat 25.1% margin estimates. Implied EBIT dollar estimates are also relatedly ahead of guidance. When asked why FCF guidance wasn’t raised from $6 billion alongside EBIT, the team cited uncertainty related to timing of content spend and taxes.

Netflix trades for 35× 2024 EPS estimates. EPS is expected to grow by 54% Y/Y this year and by 21% Y/Y next year. Here is how its earnings multiple compares to historical norms:

Netflix is considered the gold standard when it comes to recommendation algorithms matching viewers to relevant, delightful content. While that’s difficult to argue with, the firm is also not resting on its laurels. Netflix is launching a new TV homepage in beta this June in a bid to extend its lead here. Per the team, this is about laying the foundation for more rapid future split testing and iterating. To start, it will simply mean a prettier interface and more visible title descriptions with easier navigation.

The company thinks it needs to unlock an ability to go faster here, and this new home page’s architecture will fulfill that desire. Companies like Duolingo have mastered this concept to turn the most subtle of interface changes into material conversion and revenue gains. It’s a very cheap way to find more growth within your existing base of traffic. Finally, Netflix is also changing its signup page to diminish friction there as well. This should lead to material revenue gains in the coming quarters.

The advertising business continues to enjoy rapid sequential growth and “scale nicely.” Impressively, 45% of all new signups in eligible markets are opting into the ad-tier and Netflix thinks this lower price point is providing an easier top-of-funnel to drive incremental membership growth. 

While things are definitely going well so far here, there’s much more to do for Netflix to become an advertising juggernaut. It’s looking to address a lot of this needed work with its in-house ad tech platform that will be tested this year in Canada with a few main objectives. 

First, it wants to give prospective ad buyers integrations with the programmatic platforms they know and love. These include buy-side players like The Trade Desk and sell-side firms like Magnite too. Netflix simply wants to become a better ecosystem partner to allow big buyers to continue working with the companies they’ve grown to know and love. The Trade Desk and Google (another new partner) represent a massive amount of programmatic demand, which should make this change highly important. Netflix has built the audience scale that ad-buyers want to resolve that important bottleneck while partnerships resolve a second bottleneck. 

The final limiting factor for Netflix to create a giant ad business is creating better tools for buyers. The platform needs to make Netflix better at audience segmentation, targeting and return metric reporting. It also needs to make campaign creation more seamless and intuitive. It knows exactly what it needs to provide for ad buyers to embrace its inventory. As of now, advertising will be a material driver of revenue growth in 2025, but won’t become a potential centerpiece of growth until 2026 and beyond. The ad platform build-out costs are already part of 2024 guidance, which included yet another EBIT margin raise.

Netflix finished phasing out the cheapest ad-free plan in the UK and Canada. This went as well as expected, with the firm now poised to make the same change in the USA and France. Many, many streamers are pushing their customers to either more expensive ad-free plans, or their newer ad-supported packages.

In the past, Netflix has told us that ad-supported average revenue per member (ARM) was higher than its cheapest ad-free plan. This quarter, it told us that ad-supported ARM is lower than its premium tiers. This could be because it’s phasing out the cheapest premium tier, but that’s still available in almost all of its markets, so I don’t think that’s the case. Just something to note.

Despite paid sharing impacting eyeballs per account, Netflix still grew engagement with existing accounts Y/Y. This improved sequentially vs. flat Y/Y engagement growth with the same headwind. To be fair, Q2 2023 did include some impact from paid sharing, but not a full quarter. It enjoyed more view hours (per Nielsen) for films and series vs. all other streamers combined.

Leadership was actually quite kind to YouTube throughout the call. It all but called streaming a two-headed monster of an industry and doesn’t see YouTube as a primary source of market share going forward. It sees most of its market share coming from legacy entertainment providers within its current niche. Beyond that, it sees gaming, sports and live entertainment as allowing it to “win a larger share of the other 80%+ of TV time that neither Netflix or YouTube has today.” Interesting to hear them essentially concede that YouTube will be a big winner. High praise coming from a direct competitor. 

I personally think it will find significant incremental sharing in both live sports and gaming. 90%+ of the most watched titles in the USA are football games, and Netflix has built massive scale without ever having access to any of these rights. It just purchased 2 NFL Christmas Day game rights to get the ball rolling. It doesn’t think it will ever bid for full season rights with leagues, but does think these side deals provide compelling unit economics to make sports rational to pursue. In gaming, Netflix continues to find most of its success within interactive games that connect to actual Netflix characters and IP. In September 2023, the firm launched Netflix Stories as a library of interactive games. It will continue to build this library throughout 2024 with one new title per month starting in July. It’s also launching a new multiplayer game using the Squid Game series, which will launch around season 2’s premiere this year.

Per Nielsen, Netflix now has an 8.4% share of streaming vs. 8.1% as of March 2024. Nice sequential gains.

Despite recent rumors, Netflix isn’t intrigued by bundles with other, smaller streamers. Here’s an interesting quote from the letter:

“We haven’t bundled Netflix solely with other streamers like Disney+ or Max because Netflix already operates as a go-to destination for entertainment thanks to the breadth and variety of our slate and superior product experience. This has driven industry leading penetration, engagement and retention for us, which limits the benefit to Netflix of bundling directly with other streamers” – Letter

In Q2, India was its 2nd best nation for paid net adds and its 3rd best for revenue growth rate. Wonderful to hear from a market with sky-high potential that is quite difficult to profitably execute in.

I thought this was a great quarter. An annual revenue guidance raise despite rising FX headwinds, a large annual profit raise, more membership growth outperformance, more traction in new verticals… more winning. The stock has been on fire lately, which is likely why it’s not responding more positively to this news (alongside the small Q3 revenue guide miss). For the long term investor here, I think you should be pleased.