Value Spotlight

💰 News & Commentary: What's happening at Texas Instruments?

Recent performance in the semiconductor industry has been quite divergent. Anything related to data centers (NVIDIA especially) has seen an explosion in profits. Companies who produce memory chips (like Micron) are in a deep downturn with negative earnings.

The analog chip designers and integrated manufacturers ($TXN, Infineon, $STM, and others) are embarking down their own downcycle.

After 3+ years of explosive growth following supply chain shortages, the YOY trends are starting to decay, with each of the 5 top analog semiconductor companies, tracked by our friend and teammate Seobyoung Chil, forecasted to decline EPS further next quarter. I think his chart quite nicely illustrates how this is starting already (source):

Texas Instruments in particular has been stuck on the proverbial struggle bus. Their recent results have been near the lower end of the industry, with Revenues and Gross Margin contracting rather than expanding:

As Dave Ahern shared on a recent podcast episode (yet to be released), these problems were well communicated in advance by management. It's our job as long-term investors to remember those words, and apply them to our analysis today.

The biggest changes, in our opinion, which are likely affecting $TXN today include:

1. The investments in new fabs, in order to produce 300mm wafers which will be significantly more profitable than its 200mm counterparts
2. The strategic decision to consolidate its customers, allowing the company more direct control through higher direct sales
3. Market forces that created a short-term supply crunch (such as in automotive) mean reverting sharply

Notice that each of these developments, while painful in the short term, do not affect our long-term thesis (the Nick Sleep "destination in mind"). That is, a company that's driven to deliver Free Cash Flow growth, and is generous in returning that to shareholders.

Portfolio buy/sell recommendations and full research reports with your paid subscription

You may access it by signing into your account below. Additionally, click to access the portfolio's 💰 News & Commentary, 🎓 Peer Companies reports and 🎓 Letters to Investors. ​

1— Investing in 300mm capacity

Semiconductors are produced on a wafer, which is then sliced and packaged. If you're looking for a general overview of the basics of semiconductors, I'd recommend reviewing Issue #90. For great visualization of the process, I also like this video from CNBC.

Basically, 300mm is measuring the diameter of a wafer:

It makes sense that the bigger the wafer, the more semiconductors which can be sliced from it. And as Texas Instruments has discussed in their 10-K's (bold emphasis mine), "An unpackaged chip built on a 300mm wafer costs about 40% less than an unpackaged chip built on a 200mm wafer."

It's clear that the company continues to make heavy investments into 300mm wafers. From the latest 10-K released 02/02/24:

Texas Instruments (also known as "TI") first mentioned 300-millimeter wafers in their annual reports in 2014, so this is not a completely new development. They have been talking about this for a long time.

What's significant is the scale of these investments, with so many happening in Texas and Utah as you can see above.

Management called out the effect of these expansion of capacity investments on the latest earnings call (01/23/24):

With the major spends in capex ($5.1B in 2023) comes depreciation, which for $TXN impacts their Gross Margin. Depreciation expense is expected to be $1.5B - $1.8B for 2024, and $2B - $2.5B for 2025.

There's no exact timeline for when higher revenues from the capacity expansions will overtake depreciation expense and lead to operating leverage. Of course, analysts don't like that; there's also general worries about volumes and pricing of current sales.

As long term investors, we have to consider whether those worries are long-term or more shorter-term in nature.

To me, that answer is obvious and so I'm less worried about the recent margins trends. I see it as an exercise in patience.

2— Transition to direct sales

Selling to distributors is a natural strategy in a technologically matured industry like analog semiconductors, especially when you consider that a company like Texas Instruments has had over 100K customers for years.

With so many different products sold to so many different customers, it makes sense to offload some of that servicing to top electronics distributors like Arrow and Avnet.

There are some downsides to using distributors; a company gives up some profit margins and has less control over final delivery to the customer. A company using a distributor also doesn't generally maintain the customer relationship, making that relationship potentially disruptable by a competitor. Depending on the power dynamics in the value chain, this could also lead to reduced margins for the OEM and/or buying concessions.

Texas Instruments has seen an opportunity to go more direct with its sales, recently offloading one of its primary distributors over a several year period.

Direct sales as a percentage of total revenues have increased from around 67% in 2021 to 75% in 2023.

Without a doubt, this has led to lost sales and reduced revenues growth over the short-term, especially compared to peers who are more distributor-dependent.

However, TI is confident in its growing ecommerce presence through TI.com, as well as the relationships it has maintained, and its unique value proposition.

Today, TI offers customers the option of either buying direct, or using a single worldwide distributor (with a few region-specific distributors for order fulfillment). This convenience greatly simplifies business operations on both sides.

With the direct relationships, Texas Instruments is able to have more access to its customers than competitors, which allows for opportunities to sell more products into each design. They also get better insight and knowledge from the breadth and depth of these direct relationships, which a competitor who relies more heavily on distributors simply can't match.

Like many of TI's major decisions, the strategic focus is on the very long term.

This makes any short-term market share loss much more palatable for the long term investors who are willing to see strategies like this through.

3— Developments in TI's major end markets

The downcycle in analog has not been felt only at $TXN, although they have been among the obvious decliners from the Winter '23 earnings releases.

As managements at peer companies have stated, end markets are starting to turn over or have turned. Jean-Marc Chery at STMicroelectronics talked about how the company had a significant inventory correction in Industrial, which is expected to continue into the first half of 2024 as the company now has a much lower backlog than last year.

Microchip Technology, who has similar end markets to Texas Instruments with the exclusion of consumer, has described the current environment like this:

Dave Pahl of Texas Instruments talked about how all cycles are the same and yet different. In this latest one with semiconductors, there has been a bifurcation. Personal electronics weakened in Q2 last year, automotive is starting a decline now, and the other markets are somewhere in-between.

It appears that from a quarter-to-quarter basis, the impacts of a downcycle are impacting each company differently.

This makes sense, and it speaks to the wide variety of applications, customers, and products that each of these analog companies service.

As a result, comparisons of direct market share are tough to determine. In the attempt to be approximately right rather than precisely wrong, I've made some adjustments to what end markets data we have available publicly from some of these companies, and have observed some of these longer term developments:

It appears to me that Texas Instruments has been losing ground, for a while, in its more traditional end markets such as enterprise and consumer. But as Automotive and Industrial continue to grow at attractive rates, as many expect, the impact from $TXN's declining enterprise and consumer markets will drop lower than its already diminished proportions.

Many exciting technological advancements are leading industry experts to become very bullish on the future of analog semiconductors.

Things like the EV revolution and electrification of vehicle components, the emergence of AI and robotics, and the AR/VR for industry using 5G, are just a few of the examples which can drive heavy demand through increased IC content needs, even for those semiconductors with matured technologies, like in analog.

For a few leading examples, competitor NXP Semi sees +9% and +13% CAGR of the Industrial's and Automotive's TAMs through 2030 globally.

ASML, the leader in semiconductor equipment for the most cutting-edge manufacturing, also has very bright expectations for these end markets, as they've described on their latest annual report:

All-in-all, Texas Instruments is very well positioned to excel in some of the most exciting end markets for the next decade. Its specific and thoughtful business strategies should improve its competitive advantages in a very competitive industry.

The maturation of chip design enhancements through reductions in size has led much of these particular semiconductor products to become commoditized. In other words, shrinking these chips comes with diminishing returns for their functions, which makes these semiconductor companies different than analyzing cutting-edge designers like NVIDIA.

It is a big reason why decisions around manufacturing and distribution are so impactful to the company's long-term competitiveness, and I don't see anything yet which indicates that these strategies are not likely to work well.

Time to Buy the Dip?

Last month I received a great question from Matthew Ward about TXN and BAC:

Obviously, we now all know the fate of $BAC after selling it for $SBUX this month. As far as adding more to $TXN, I am still hesitant for now. While I believe in the long term investments the company is making, these have obviously suppressed Free Cash Flow, especially for 2023.

As a general default rule, I tend to buy a stock when it has more Free Cash Flow than less. Of course there is the very occasional exception to the rule, like when I recommended Costco in January 2023.

But because there is some uncertainty behind the effectiveness of TI's strategy over the long term, I hesitate to make exceptions for them at this time. For example, Costco raising their membership fees is very much an eventual certainty to me, which is why I was willing to make the FCF exception to the rule. For TI, it's not as obvious.

Hopefully, TI's strategy does pay off for the company. In that case, FCF should rise dramatically from here, making the value appear obvious once again.

If we really do have that decades-long mindset, and can't be certain how long this short-term uncertainty (and downcycle) will last, then I think patience probably pays here. Either way, we have a great company in the portfolio, and waiting and holding looks pretty win-win to me.

Portfolio buy/sell recommendations and full research reports with your paid subscription

You may access it by signing into your account below. Additionally, click to access the portfolio's 💰 News & Commentary, 🎓 Peer Companies reports and 🎓 Letters to Investors. ​

To get the best out of your subscription, I highly recommend downloading the Mighty Networks app to get access and be notified when new content is published throughout the month, which includes:

• Portfolio news and commentary
• Latest ranking of positions in the portfolio
• Deep dives of peer companies
• Deep dives of portfolio companies
• Letters to Investors

Download the app on the Apple App Store or Google Play. This will give you the best that your membership has to offer.

​

© Sather Research, LLC. All rights reserved.
​Privacy | Disclaimer | Refund Policy | Contact​