Texas Instruments’ Q3 was similar in some ways to its Q2: the company reported strong quarterly numbers (that significantly beat expectations, if ignoring some one-offs) only to follow these with somewhat “weak” guidance. As expected, the market did not like this much and the stock dropped considerably in after hours trading to $165:
We still have to see what the stock does today, but it’s undeniable that Texas Instruments is suffering several headwinds simultaneously. I would summarize these as follows:
Slower recovery than previously expected: customers seem to have stopped depleting inventory (good), but they are running on low inventory rather than building inventory to more normalized levels
The Chinese anti-dumping investigation (discussed in more detail in this article)
Restructuring and lower factory loadings (fresh news from this quarter which have a significant impact on profitability)
Expectations of a weaker automotive sector should China’s rare-earth limitations go into effect (still TBD)
All of the above have introduced significant uncertainty and therefore have caused Texas Instruments’ stock to be quite volatile since April. The stock dropped to almost a 5-year low following the first tariff tantrum in April ($147), then skyrocketed up to all time highs barely three months later ($221 or +50% from the lows), only to subsequently fall back to the $160 range (for almost a 30% drawdown). While it’s tough to envision a scenario where the value of the company is changing as fast as its stock price, there’s no denying that management has not done a great job in managing the market’s expectations (mainly due to their lack of visibility).
Texas Instruments comfortably beat revenue expectations in Q3 (+2%), with revenue coming in at $4.74 billion (compared to analyst estimates of $4.64 billion). The good news did not carry over to EPS. Texas Instruments missed EPS estimates by 1 cent ($1.48 vs $1.49 expected) for its first miss in a while and also considerably missed EPS guidance. However, the Q3 miss was due to a restructuring charge that made EPS come 10 cents lower than previously expected. Management included two things in the restructuring…
Closing of two 150mm fabs as the company fully transitions to 300mm fabs
Consolidation of several R&D sites that were not providing the expected returns
Both things are a one-time hit to Q3 but should deliver better profitability in 2026 and beyond (through both lower COGS and Opex). What seems evident is that the recovery is still underway, even if at moderate pace:
Even though TI continues to grow year over year, the pace of growth moderated somewhat and revenue growth decelerated sequentially. This needs some context. First, it demonstrates that Q1 and Q2 might indeed have enjoyed some sort of a pull forward that is now normalizing as the cycle struggles to accelerate (comps are also getting tougher):
There was also an interesting discussion around seasonality. Analyst Stacy Ragson said the following during the call:
Your Q4 guide is down about 7% sequentially off the slightly higher than expected Q3 base. My message is that down 7% or so is pretty much seasonal, like on a pre-COVID basis. I know post-COVID seasonality has been all over the place, all over the place. Pre-COVID, it typically was down, call it high single digits. You seem to be on a seasonal trend now, and maybe that’s consistent with customers no longer draining inventory. How do I think about normal seasonality, like pre-COVID levels for Q1? My feeling is that it’s typically down sequentially. What is, I’m not asking you to guide it, but just what is normal for Q1, at least on a pre-COVID basis, if we’re running more of a seasonal pattern from here?
To what management basically confirmed that, “yes, seasonality seems to be back”:
Stacy, before we talk about Q1, let me just add a little bit more color on Q4. As you said, I look at it as a roughly seasonal guide, as you said. The reason is there is a recovery, but it’s a very moderate pace, right? That’s what guides our, call it, seasonal view into Q4. I also mentioned, that’s what we’re seeing. This is part of the way we do business these days. More customers are direct, more customers are on consignment. Customer inventories are low, and I think they’ve gone through this depletion process. That’s behind us. We are going to be just seeing it real time as it comes, and hence our guidance.
How I interpret this is that, in a scenario of a more violent recovery, TI should not be experiencing seasonality just as they didn’t experience seasonality on the way down. The reason would be that customers would be ordering to build inventory, not just to satisfy their end market demand. However, in the scenario of a mild recovery where customers are not building up inventories, the company is indeed experiencing seasonality. Note that, considering that TI’s lead times have come down a lot, and the fact that the company has $4.8 billion in inventory probably doesn’t help the cause of an inventory build up (i.e., why would I carry that inventory if my end markets are not exploding and I know TI is going to hold it for me?)
All the above conversation is interesting because I believe it impacts the numbers vs. expectations “game” and helps us frame better what has happened to Texas Instruments this year. The company beat Q3 revenue estimates by $100 million, but the midpoint of the guide was $60 million below the market’s expectations ($4.44 billion vs. $4.5 billion expected). This means that adding Q3 and Q4, TI is around $40 million ahead of the street in terms of revenue. Of course, this is not how the market and expectations work, but nonetheless something to think about.
What’s undeniable, though, is that EPS numbers have come considerably below estimates if we add both Q3 and Q4. The market expected $2.9 in EPS for Q3+Q4 and, if TI meets the midpoint of its guidance, it will deliver $2.74 (around 5% lower). This honestly makes me think that the drop is not unwarranted, but at the same time that the market is struggling to look ahead.
Based on the above you might have guessed that profitability continued to be a lowlight. TI is not enjoying the “promised” operating leverage as the cycle recovers for one main reason: factory loadings. The management team pressed too hard on the inventory pedal expecting a more violent recovery, but now that they’ve seen the recovery is “mild,” they have stopped growing inventory (and even hinted at lowering inventory days going forward). This means that factories run at lower loadings/utilization (and therefore at lower margins because it’s a fixed-cost business). This is evidently not great news for profitability.
What surprised me the most about the market reaction was that there was not much information in the quarter that was unknown. Management alluded to the quarter playing out as expected (without nasty surprises like in Q2) and reaffirmed two things that we already knew:
The cycle is not recovering at the pace they expected
They might be in the low end or even under the low end of their 2026 revenue scenarios
Neither of these things should’ve taken anybody by surprise considering that (1) they had already told us, and (2) market estimates for 2026 revenue are currently below the $20 billion mark ($19.6 billion). All this said, it’s evidently not great to see all the headwinds the company is suffering, but that’s already to an extent reflected in the current drawdown the stock is experiencing. I’ll consider all of this later in my valuation exercise.
One of the reasons TI seems to be restructuring some of its R&D organization is that Haviv doesn’t seem to be entirely happy with it. He believes the company has to do more work on core markets:
When I think about industrial, automotive, data centers, the amount of opportunity to expand our portfolio is high. We have a lot of good investments to make there, and we plan to continue to grow our portfolio in these three areas. We care about all markets, all five markets, but these three will have a long-term growth opportunity ahead of them, and Texas Instruments can do more to sell to these markets. I expect to see that in 2026 and beyond.
One of these markets is data center. Management (consistent with the rest of the semiconductor industry) mentioned that this is the only market where they are really experiencing significant growth. It’s still a small market for TI at a $1.2 billion run rate, but it’s growing 50% year over year. Expecting that it will become a more significant market for the company, Haviv mentioned that they will disclose it separately starting Q1 2026. The data center revenue is currently disaggregated between enterprise, communications, and even industrial. There’s a lot of power/electricity going into data centers and TI can be a significant player here. They don’t seem to be standing still considering that, barely a week ago, the company published the following press release on the topic.