Pricing "The Compute Theory of Everything", Texas Instruments Hitting Inflection Point

Believe it or not, the iShares Semiconductor ETF (SOXX) was essentially flat on the year as of March 30…less than four weeks ago. Since then, semiconductor stocks appear to have chugged something with serious kick: the sector is up 47% YTD(!!), and given Intel's print last night, it's probably adding to that today.

I am a tad bit surprised that after three and half years since ChatGPT was released in November 2022, there was still so much money to be made in the “Picks and shovels” of this AI revolution. I’m not sure people are looking at Semi’s valuation that much anymore given that consensus seems to be clustering around the view that we may simply have not enough compute as far as the eyes can see. Given such “reassurances”, SOXX ETF is currently trading at valuation multiples not seen at least since 2001 (I don’t have data prior to 2001; I’m also taking LTM instead of NTM because KoyFin doesn’t have NTM data that far back).

Imagine for a moment you are in early 2000s and you just read Hans Moravec’s seminal paper “When will computer hardware match the human brain?” Let’s say the paper is compelling enough for you to start believing in “the compute theory of everything” and you then decide to buy SOXX to express such view through a diversified instrument. Even though your broader view on the unrelenting value of more compute would be correct, you would experience a disorienting volatility over the next two decades. At this point, we are indeed perhaps starting to price “the compute theory of everything” given today’s stratospheric valuation demands fundamentals to remain rosy for years to come for investors to make an acceptable return from here.

Having said that, I will acknowledge that MBI Deep Dives has utterly missed the boat on the picks and shovels in the last 3-4 years. As a generalist, I knew very little about semis when ChatGPT was released and even though I tried to make amends by studying them since early 2024, the market moved at too frenetic a pace for a novice semi observer like me to feel comfortable to act. To add insult to the injury, I did buy one semi stock in 2024: Texas Instruments (TXN) and after 18 months of holding the stock, I decided to sell it late last year to re-allocate the proceeds in Meta and Alphabet.

It now appears the TXN stock was “waiting” patiently for me to sell and it has only gone up since then.

Self-deprecating jokes aside, Texas Instruments’ Q1 does seem to have reached an inflection point as analog revenue has exceeded 20% mark for the first time since 4Q’21.

This growth was primarily led by industrials which was up 30% YoY and ~20% sequentially. Data center revenue was predictably also up ~90% YoY.

TXN management reminded despite this inflection in revenue, they are still noticeably below the previous peak. From the call (emphasis mine):

“we had a very good quarter in industrial…but still 15% lower than the peak that was back in 2022. And as I say many times, there is a secular growth continuing in industrial. So we deserve a higher peak, right, 4 years later. So I think there is a lot of room to grow. The encouragement I would have on industrial this time is that I see it as a broader application.

So all of them, not only the data center-related energy infrastructure or power delivery, not only aerospace and defense and we know the geopolitical tensions and the market is establishing new peaks every quarter. I saw it across all sectors in industrial and also across all customers in terms of regions but also the size of customers. It’s the first quarter where we saw the broad market, basically the tail starting to wake up again after a long hibernation period, I would call it. So I am encouraged about the fact that we are seeing growth over there. But I think there is, I mean, I want -- I would like to see a secular growth in industrial continuing and then higher peaks establishing in 2026 or later versus the 2022 peak. So in that sense, trend line are suggesting we still have room to go.

However, management also mentioned last year also started with a similar buoyant sentiment which turned out to be bit of a head fake. Indeed, the latter half slowdown in 2025 fooled me thinking 2026 would be a rather slow recovery. Again, from the call:

We had a similar, let's say, strong beginning of the year last year. Maybe the year-over-year growth last year was a little lower but it was still in the teens and it looks like it was getting stronger. But it was whatever you want to call it, a head fake, a false start or whatever. We had a good year in analog but it did not accelerate in the second half. It actually slowed down a little bit, right? So I think we need to be cautious…There is geopolitics. There is a macro that we are watching.

Nonetheless, the demand now appears to be strong enough that even pricing lever is on the table which makes me think this inflection may not be “head fake”. TXN management sounded pretty upbeat while discussing the pricing environment during the call (emphasis mine):

right now, the demand signals are strong, if demand continues to be strong, and we are monitoring the market price and there is definitely at least an average price increase in the last several months across the analog market. I think it’s likely that prices may go up in the second half of the year. Again, this is going to be a case-by-case discussion in our case but that’s the pricing environment as I see it right now.

And again, it’s always a function of supply and demand and the unknown for me right now is the sustainability of demand. So I want to see it playing out one more quarter and then we’ll figure out for the second half. So high level, not immediate support on growth, both sequentially and year-over-year on pricing. Now what we have seen is just breadth of demand, right, what I said before, multiple sectors or all sectors, all regions, all type of customers, small or large and supported by a data center market where we do pretty well.

Given the revenue inflection, predictably margins have inflected as well. Gross margin increased to ~58% in 1Q’26 (+212 bps YoY), but still significantly below last cycle’s peak of ~70%. Similarly, operating margin expanded to ~39% (+669 bps YoY), but well below last cycle’s peak of ~52%. So, TXN should have a lots of headroom for further margin expansion if the cycle continues.

Consensus EBIT margin estimates are in the low ~40s in 2026. At ~30x NTM EV/EBIT, buy-side may be penciling a higher margin inflection than consensus numbers imply at the moment.

In addition to “Daily Dose” (yes, DAILY) like this, MBI Deep Dives publishes one Deep Dive on a publicly listed company every month. You can find all the 67 Deep Dives here.

Current Portfolio:

Please note that these are NOT my recommendation to buy/sell these securities, but just disclosure from my end so that you can assess potential biases that I may have because of my own personal portfolio holdings. Always consider my write-up my personal investing journal and never forget my objectives, risk tolerance, and constraints may have no resemblance to yours.

My current portfolio is disclosed below: