Texas Instruments Incorporated (NASDAQ:TXN) Capital Management Off-Cycle Call August 20, 2024 11:00 AM ET
Company Participants
Dave Pahl – Head of IR
Haviv Ilan – CEO
Rafael Lizardi – CFO
Conference Call Participants
Timothy Arcuri – UBS
Vivek Arya – Bank of America Securities
Ross Seymore – Deutsche Bank
Stacy Rasgon – Bernstein Research
Joshua Buchalter – TD Cowen
William Stein – Truist
Chris Caso – Wolfe Research
Tore Svanberg – Stifel
Tom O’Malley – Barclays
Chris Danely – Citibank
Joe Moore – Morgan Stanley
CJ Muse – Cantor Fitzgerald
Dave Pahl
Good morning, and welcome to the Texas Instruments August 2024 Capital Management Call. I’m Dave Pahl, Head of Investor Relations, and I’m joined by Haviv Ilan, our Chief Executive Officer; and Rafael Lizardi, our Chief Financial Officer. This call is being broadcast live over the web and can be accessed through our website at ti.com/ir. In addition, today’s call is being recorded and will be available via replay on our website along with the complete presentation and prepared remarks for your convenience.
This call will include forward-looking statements that involve risks and uncertainties that could cause TI’s results to differ materially from management’s current expectations. We encourage you to review the notice regarding forward-looking statements contained in our most recent earnings release, as well as our most recent SEC filings for a more complete description. In addition, for more information regarding our overall capital allocation strategy, scorecard and business overview, you can access our February capital management call materials and our investor overview on our website.
With that, let me turn it over to Haviv.
Haviv Ilan
Thanks, Dave. Let me start by welcoming you to our off-cycle capital management call. As we are now more than 60% through a capital intense period, this is a good time to share more details on our progress and where we are headed, which we believe is helpful for the investment community to understand. The investments we are making will uniquely position TI over the next 10 to 15 years to deliver dependable low-cost 300-millimeter capacity, scalability of CapEx and free cash flow per share growth across a range of market conditions.
During today’s presentation, I will start by providing additional insight into our growth expectations. Then I will provide more granularity on our 300-millimeter wafer fab investments as we continue to make progress on our plans. And finally, I will share a framework of revenue scenarios and associated ranges for CapEx and free cash flow per share.
Before we get started, you may have seen that on Friday, we announced that we have signed a preliminary agreement with the U.S. Department of Commerce for up to $1.6 billion in proposed direct funding under the CHIPS and Science Act. This proposed funding, coupled with an estimated $6 billion to $8 billion in investment tax credit will help TI provide geopolitically dependable capacity for our Analog and Embedded Processing semiconductors. I want to point out that in today’s materials, we’ve incorporated the benefit we expect to receive for the investment tax credit, but we have not included any assumptions on the proposed direct funding.
Now let me start with a reminder that our strategy is designed around four sustainable competitive advantages that, in combination, provide tangible benefits and are difficult to replicate. During today’s presentation, we will spend more time on the advantages from our broad portfolio of Analog and Embedded Processing products, as well as our foundation in manufacturing and technology. If you haven’t already, I encourage you to review our investor overview, which provides additional insight into our business model and competitive advantages and it is also available on our Investor Relations website at ti.com/ir.
Moving on, let me provide some insight into our expectations for growth. As we consider our investments, it is important to be mindful of where we are in the semiconductor cycle. This chart shows semiconductor units shipped, excluding memory, on a trailing 12-month basis over the past 30 years as reported by WSTS. You can see here that the current market environment remains weak with global semiconductor shipments now below 2019 levels.
We believe it is helpful to consider the last three peaks as we prepare for the opportunity ahead. But the more important element is that the gray line, which shows the long-term trend grows consistently over time. Our approach is to have a disciplined plan that incorporates a range of potential market recovery scenarios with a gray line in mind.
Let me show how TI’s portfolio and market positions have evolved and strengthened through the last several market cycles. During the last decade, we have worked hard to focus our product portfolio on Analog and Embedded Processing. As a result, our other business now primarily consists of DLP products and calculators and is small and stable. In addition, during the past several years, we have driven changes in our embedded business to better position the business for growth. Our belief and expectation are that both analog and embedded will contribute to future growth.
While we have opportunities in all markets, we biased our investments to the fastest-growing markets, industrial and automotive and our work during the past decade has further strengthened our position. This includes investments in process technology, package technology and the expansion of our product portfolio. In addition, our work to build closer direct relationships with our customers allows us to better service tens of thousands of customers in the industrial and automotive markets. These two large markets, which combined have grown double digits through the last two cycles will likely continue to be the fastest growing markets over the long term, given the secular content growth.
In summary, the efforts to strengthen our portfolio and market positions make us well positioned for continued growth. Our broad portfolio is able to address an increasing number of semiconductor opportunities. With more than 80,000 products, we have the most comprehensive portfolio of general-purpose analog products as well as a full breadth of application-specific analog and embedded processing products to support our customers.
I’d like to share a few examples within the industrial and automotive markets that highlight how our technologies are addressing the opportunity. If you look at the left side of the slide, starting with industrial, you can see an example within factory automation motor drives and robotics. This is an area where semiconductor content continues to increase as our customers design smarter and more efficient automation applications. For TI, this sector represented about $2 billion of revenue in 2023, growing about 10% per year since 2013. As one example, a collaborative robot or a cobot can often have more than $400 of TI content in one system for across more than 200 chips. This includes a mix of application-specific products such as vision processors, radar, C2000 real-time control and GaN as well as general purpose products such as Ethernet 5, data converters, isolation, amplifiers, Hall effect sensors and voltage regulators to name a few.
We also see an increasing opportunity in the automotive market. As shown on the right side of the slide, our products play an important role in advanced driver assistance systems, helping our customers design systems for a safer, more automated driving experience. For TI, ADAS represented about $1.5 billion of revenue in 2023, growing about 20% per year since 2013. A single vehicle can have more than $350 of ADAS-related TI content across more than 300 chips. This includes application-specific products like RADAR front end, [indiscernible], functional safety power and vision processors as well as numerous general purpose products with a power and signal chain portfolio.
Moreover, the opportunities we see for increased semiconductor in body electronics and lighting as well as ADAS systems applied to all vehicle types, including internal combustion engine, hybrid, plug-in hybrid and full battery electric vehicles. And while I only shared a few examples, these general trends are not unique to these sectors. Our product portfolio is addressing an increasing number of sockets across all sectors, giving us further confidence in our plans.
To summarize, our market exposure has grown from about 40% to about 75% revenue in industrial and automotive markets, which experienced double-digit growth through the last two peaks. Second, we have a stronger product portfolio. The breadth of our Analog and Embedded Processing products, combined with our investments in process and package technologies have strengthened our portfolio offering. As a result, we are positioned to grow our exposure to large, fast-growing markets and our strong product portfolio position us to capture the opportunity ahead.
Now I’d like to provide more details on our capacity investments in 300-millimeter wafer fabs. As I mentioned earlier, our investments will uniquely position TI for the next 10 to 15 years in several ways. First, we can provide dependable low-cost 300-millimeter capacity to meet our customers’ demand, given the increasing semiconductor content in industrial and automotive especially in a world of rising geopolitical tensions. Second, our CapEx is scalable, including capacity modularity, which means we can handle strong markets or times of persistent market weakness. And third, we are positioned to deliver free cash flow per share growth across a range of market conditions.
There are three phases to our 300-millimeter wafer fab investments which are important to understand. I’ll walk through each one in more detail. The focus of Phase I is to support 300-millimeter transfer and incremental growth by equipping RFAB2 and LFAB1 to support growth in the coming years. We are continuing to equip RFAB2 with tools to provide the capacity needed as we execute our legacy 150-millimeter shutdowns. The devices that were running on 150-millimeter are being redesigned into modern, more efficient 300-millimeter process technologies that will run in RFAB2.
At LFAB1, we are completing technology and customer qualification, which allows us to transfer external foundry wafers internally, providing us with substantial cost advantages as well as dependability of supply. We expect this work to be completed by the end of 2026. We Phase 2 is all about new fab preparation and primarily includes long lead time work that spent several years before a fab can produce a single wafer.
First, we are constructing LFAB to have the cleanroom ready for first production. LFAB2 will be connected to LFAB1 and these clean rooms will operate as a single factory. Because of this, LFAB2 will not require a pilot line or customer qualification and enables us to add tools according to demand. Phase 2 also includes getting our Sherman site ready where we will be running our most advanced analog process technologies. We are constructing the SM1 building and clean room as well as completing the pilot line. A pilot line is necessary to begin technology and customer qualification which will take approximately a year before we can begin shipping qualified product to customers.
We are also completing the SM2 shell to be weather tight but without a complete clean room or required equipment. This approach provides efficiency benefit from joint construction with SM1 while also eliminating future construction lead time. Just like with RFAB2, SM2 will be connected to SM1 and these two cleanrooms will operate as a single factory. Therefore, SM2 will not have the same technology or customer qualification lead time as SM1. This work at LFAB2 SM1 and SM2 is necessary in order to prepare these fabs for future tool installs. The preparation in Phase II is allowed us to be in a position of modular capacity expansion. Phase 3 will allow us to equip and ramp fabs according to customer demand without any qualifications, and we will be able to quickly respond to what the market needs. We will scale CapEx according to demand and ultimately deliver free cash flow per share growth across a range of market conditions.
In summary, our 300-millimeter wafer fab manufacturing investment spend three site in Richardson, Lehi and Sherman. Here, you can see more information across each fab. Let me highlight a few important points. First, these wafer fabs represent an important element of our overall manufacturing footprint and position TI to deliver dependable capacity in the 28-nanometer to 180-nanometer process technologies that are optimal for our product portfolio. For each fab, the maximum revenue support that shown assumes optimal mix and full utilization.
The CapEx shown is growth and doesn’t include benefits related to the U.S. CHIPS Act. You can see that RFAB2 and LFAB1 will benefit from the investment tax credit through 2026 as these sites were already under construction when the CHIPS Act was passed in 2022. Sherman and LFAB2 will continue to benefit from the investment tax credit through the first half of the next decade. As I mentioned earlier, we don’t have any further details related to the timing or breakdown of the proposed direct funding for SM1, SM2 and LFAB2.
More importantly, the execution across all of these projects is going very well, thanks to the hard work from our manufacturing and facilities teams. We and our customers remain very pleased with the progress. RFAB2, which was built entirely through the COVID-19 pandemic continues to ramp with new tools. We are now approximately 50% ramped with new tool installations continuing every day.
In LFAB1, we have qualified products on 65-nanometer and 45 nanometer. In addition, we are developing our 28-nanometer process technology and just outside of LFAB1 construction is underway for LFAB2. In Sherman, we just installed the first tools on the floor at SM1, and we are beginning technology qualification. The construction of SM2 is progressing well. This great execution is enabling us to accelerate several of our key metrics, which I will show in a minute.
With that, let me transition to share more about how we are positioned to deliver free cash flow per share growth across a range of market conditions. Here, we are providing four revenue scenarios for 2026 that represent a range of market convenience. You can see how these scenarios compare to TI’s performance during the last several market cycles in 2014, 2018 and 2022 and the respective growth rates.
The higher revenue scenario of 7% CAGR versus 2022 indicates a rapid market recovery from today and represent a similar level of growth from the prior cycle. The zero growth scenario would indicate a persistently weak market, which likely means that the next industry peak is delayed. We remain confident in our ability to gain market share through the next cycle. Our confidence is a reflection of the combination of the growth in semiconductor content particularly in Industrial and Automotive, our portfolio offering and position in these markets and the continued strong customer demand for our geopolitically dependable capacity.
Our capacity build-out plans and investments remain important to position the company for growth. We plan to spend about $5 billion in CapEx per year through 2025, which supports Phases 1 and 2 of our 300-millimeter wafer fab investments. For 2026, CapEx is expected to be between $2 billion to $5 billion, depending on revenue as I’ll show in the next slide. These levels of CapEx allow us to meet the strategic objectives to support transfers, incremental growth and prepare our new fabs to then be able to modulate capacity based on demand.
It also balances other considerations, such as benefits related to the U.S. CHIPS Act investment tax credit. In 2027 and beyond, you’ll see a range of CapEx levels. as this will be determined based on revenue and expected revenue growth. Finally, at the bottom of this slide, we have highlighted several key metrics that this role that we deliver, which we have accelerated. By 2026, we now expect to have more than 90% of our wafers sourced internally with more than 70% on 300-millimeter as well as more than 85% of our assembly internal.
By 2030, we now expect more than 95% of our wafers to be sourced internally. Overall, this will provide us with lower cost and greater control of our supply chain across technology and manufacturing. We have often said that the best measure to judge a company’s performance over time is the growth of free cash flow per share as that is what drives long-term value for our owners. Here, we are showing our 2004 to 2022 free cash flow per share trend line continuing at the same rate through the end of the decade.
As we consider the four revenue scenarios provided, free cash flow per share will begin to approach the trend line in 2026 as growth returns and CapEx begins to moderate. You can see here the range of CapEx in 2026 as well as the associated free cash flow per share results of $8 to $12. As such, we are prepared to support a rapid market recovery or persistently weak market. As a reminder, these estimates incorporate the benefit we expect to receive from the investment tax credit, but exclude any proposed direct funding from last week’s announcement. The impact of that proposed funding would represent additional free cash flow per share benefit.
Beginning in 2027, free cash flow per share growth will be driven by revenue growth and our CapEx strategy. This underscores the strength of our business model, including the scalability of CapEx with modular capacity. This will allow us to deliver free cash flow per share growth aligned with the long-term trend line. And finally, long-term free cash flow per share growth will continue to guide our decisions.
In summary, we are more than 60% through a 6-year elevated CapEx cycle. We are executing well to the plan and we remain confident in the opportunity ahead. When completed, this will uniquely position TI for the next 10 to 15 years to deliver dependable, low-cost 300-millimeter capacity, scalability of CapEx and free cash flow per share growth across a range of market conditions.
With that, I’ll turn it back to Dave.
Dave Pahl
Thanks, Haviv. Operator, you can now open up the lines for questions. In order to provide as many of you as possible an opportunity to ask your questions, please limit yourself to a single question. After our response, we will provide you an opportunity for an additional follow-up. Operator?
Question-and-Answer Session
Operator
Thank you. We will now be conducting a question-and-answer session. [Operator Instructions] Our first question is from Timothy Arcuri with UBS. Please proceed with your question.
Timothy Arcuri
Thanks a lot. Haviv, I’m wondering why you are excluding the grants and the loan support. I think you also — you got the $1.6 billion in grants, but you also have $3 billion worth of loan support. That’s part of CHIPS Act. So why is that being excluded? Is that because you haven’t — you don’t have final terms yet on that. Can you just give more details on that?
Dave Pahl
So Tim, just let me jump in real quick. This is Dave. I just want to point out that the slides are available. And if anyone wanted to use those or refer back to them, they are on our website at ti.com/ir. So with that…
Haviv Ilan
Thank you, Dave. Ans good morning, team. So yes, just to answer your question, I’ll let Rafael also add a little bit regarding the loans. This is — this PMT or preliminary terms agreement was only decided end of last week. And remember, this is only a PMT, meaning, we do have to get to final terms and exact details of how to earn the proposed direct funding. Usually, there are several months between the PMT signature, that milestone. And as such, we want to go through the process, make sure that due diligence is fully completed. And then we will be able to have more specificity about how much and when.
I will mention that the proposed direct funding is specifically tailored towards the construction of SM1 and the first kind of line of production. The LFAB2 cleanroom and the SM2 Shell, these are all belong more or less to Phase 2, as I described earlier. So it gives you more or less a time frame, assuming everything lands as planned, the timeframe that we expect to see the grants, again, the specificity of how much every year will come later. And I’ll let Rafael mention a few points about the loans.
Rafael Lizardi
Yes. So on the loans that we just stress, first, it’s an option or an obligation. And if you think of the loans that just corporate debt that we would issue. It’s just that it should be, it would be at lower interest rates. So the benefit would be with lower interest payments. That’s what we showed through operating cash and free cash flow.
Dave Pahl
Do you have a follow-on, Tim?
Timothy Arcuri
I do. Yes. Can we — can you just give some view on what your share is? I mean part of — if we just look at your share, your share of the analog market was basically 19% pre-COVID, it’s come down on a trailing 12-month basis to about 16% now. So when you think about coming off the bottom, how much of that share are you assuming that you regain? I know you’re not making a firm assumption on 2026 revenue. But doesn’t it stand to reason that you’re going to regain much of that share. So can you just talk about how you think about where your share used to be in Analog versus where it is now? And if you’re seeing signs that, that share is actually coming back? Thanks.
Haviv Ilan
Thanks, Tim. Let me take that. So again, if you think — and we said it many times, we like to look at the share over the long term. And as you know, the cycle is still playing out. We are seeing it very clearly in the business that is coming in every quarter. The synchronous behavior of the cycle is not done. I think some of the markets have already established the bottom. Some are still working towards that. And also geographies are also behaving in a different way. We just mentioned in the last earnings call, how China is starting to show strong recovery, grew 20% sequentially in Q2, while other geographies like Japan or Europe are still struggling.
So to me, I want to let that process complete. And we believe the best way to look at it is the, call it, the next peak. If you think about our market share sometimes in the beginning of the decade, I call it kind of close to the previous peak. It was around 18%, 19% on the analog side, and I do expect it to fully recover. That’s the expectation I have from the team. That’s how we are attacking the market. I think our product portfolio is ready to compete for each and every socket. Our capacity that — to be very transparent, we were lacking during the last peak is now ready and our inventory is in a good position. So I give ourselves a good chance to continue to get back to the areas where we used to be on market share. And of course, as you said, we are not making a specific assumption on market share if we think about different scenarios.
Dave Pahl
All right. Thank you, Tim. We will go to the next caller please.
Operator
Thank you. Our next question is from Vivek Arya with Bank of America. Please proceed with your question.
Vivek Arya
Thanks for taking my questions. Haviv, I’m curious what is driving the lower CapEx for 2026, right? You change it from $5 billion to — $2 billion to $5 billion, and TI has had several calls in the last year consistently talking about the $5 billion CapEx target for 2026. So what changed in the last few weeks to drive this CapEx lower? Should we interpret that as more conservatism about growth? I’m just curious as to what has created this big change?
Haviv Ilan
Yes. And again, I don’t think it’s a big change. And I think Rafael and I have addressed it previously on some analyst conferences that we’ve been at. We are now at 60% inside the process of investment. When we put the spend together several years ago, we still had a very complex, nongranular plan as we can talk to you today as we are now through, I would say, almost two-thirds of our investment plan. I think the level of risk is dramatically done. We have executed according to our plan and sometimes pulled in our plan. And therefore, we are ready now to provide a more specific number for 2026.
We also — I would also say that if you look at — when you start such an investment cycle, you have to prepare for each and every scenario, and you always have to assume the worst case, because that will require the maximum level of capacity you have to build. As we get closer to 2026, we want to think about the different scenarios. We have introduced them today, and you see a wide spend from $20 billion on our presentation all the way to $26 billion. And therefore, we have this flexibility to tune down the investment after Phase 2, specifically in 2026. So that would be my high-level answer on this one, Vivek.
Dave Pahl
You have a follow-on, Vivek?
Vivek Arya
Yes. Thank you. The second question I have is you’re using 2022 as part of the historical trend line, but that was sort of in the peak of a lot of the COVID induced shortages. So is that data point really representative of the trend line? Can it really be used to drive any forward conclusions? I mean, if I ask the question in a different way, Haviv, are you saying that the consensus expectation, which is about $20 billion in TI sales for 2026, is that very low relative to how TI expects to grow, because that would still imply double-digit, well above trend growth for the next several years. So that’s why I’m just curious, why did you use 2022 as an important part of the trend line given how abnormal that year was?
Haviv Ilan
Yes. And again, I’ll talk specifically about 2022 at the end of my response, but let me just say a little bit about the framework we have built here. I’m not trying to predict when the next peak will be. I mean I’ve just shown three examples of 2014, 2018 and 2022. Therefore, 2026 is a good time to think about what the business could be. And I don’t want to assign probabilities, because it all depends to the way the cycle will behave. We’ve seen a tremendous amount of inventory buildup at different markets at different phases. It’s starting to play out, but when it’s all done, I expect the recovery to be fast.
Now when exactly does that happen? I cannot predict today. I just want to be ready to the point when it comes. Now does that imply that the next cycle is later than 2026? It could be. It all depends. I do believe, and this is the most important part that we would be able to address each and every scenario. Coincidentally, in 2026, we are more or less done with Phase 2. And we’ll have our clean rooms ready. Our sites qualified. We will take an assessment of what revenue wants to be in 2026. And then we have a couple of options, either we continue to run fast to support the demand that comes our way. Or we slow down CapEx as we showed before, and we let the free cash flow fall through.
And the last point I would say, if I envision the next peak for the company, I would be extremely disappointed if it ends up at $20 billion. That’s not my expectation. That’s not the signature I see as we compete for market share today. I think I’ve mentioned in the past, during the 2022 pandemic cycle, yes, the demand was very, very high, but we could have done more. We had to take some very tough decisions, specifically on the nonindustrial and automotive markets, where we have to give away positions we had. Today, we already have the capacity. We have the inventory, and we are fighting back for our market share, and our customers are responding very well. So to your direct question, absolutely, if TI picks the next cycle, the $20 billion, I will be disappointed, yes.
Dave Pahl
All right. Thank you, Vivek. Now we will go to the next caller please.
Operator
Thank you. Our next question is from Ross Seymore with Deutsche Bank. Please proceed with your question.
Ross Seymore
Hi, guys. Thanks for letting me ask question and thanks for the presentation. It’s great to see the scenario analysis here to clarify things. Thanks. Haviv, for my first question, I wondered it’s great that you have significantly greater flexibility in 2026 and beyond. But exactly how early do you need to make the bet if at the beginning of the year, you think it’s going to be $20 billion? Do you just slash it to the $2 billion in CapEx? Do you have that visibility on revenue? And how far ahead of time do you really need to adjust the CapEx?
Haviv Ilan
Yes, I think it’s a great question. And this is part of the playbook that we are developing. Of course, I mentioned in my previous response, it’s very hard to predict what the cycle will do. But as you go through the quarters, you start to see the signature by market, by geographies and you have a better understanding. My expectation is that somewhere in the next 12, maybe, to 15 months, probably be the second half of 2025, we’ll have pretty good visibility. We know about our ramps that are coming. We also know what — at what phase of the market, our customers are at.
And when I think about it, I think about geographies, I think about different markets and that’s when we will be able to take the call. And I think it allows us enough time with lead times and slots to modulate our 2026 capacity investment. As you can see in the scenarios, there is kind of a floor of investment at around $2 billion, I think we said $2 billion to $3 billion just because we do want to complete the Phase 2 investment, meaning, Sherman 1, including the qualification phase, Sherman 2, the cleanroom or the shell. And of course, to be ready for growth — sorry, LFAB2 to be ready to grow as LFAB1 completes its transfers and we’ll need some more capacity in order to enable growth for both analog and embedded.
Dave Pahl
Great. Ross, you have a follow-on?
Ross Seymore
One for Rafael about the CHIPS Act of things, and I appreciate that it’s not in the numbers here. But the $1.6 billion, how theoretically does that flow through? What’s the depreciable life? How many years do we divide it by? Is that the max number? Or when you do Sherman 3 and 4, does that come into the equation? So the $1.6 billion could be larger? Just trying to figure out when it flows in, how much and for how long?
Rafael Lizardi
Yes. So unfortunately, Ross, I’m not going to have a great answer for you because, as Haviv mentioned, it’s still a preliminary agreement. So we have to still work out the details of the full agreement. And when we have that, we’ll share it. What I could — what I can tell you is that if, in fact, we reached an agreement, that $1.6 billion will reduce the asset value of SM1, SM2 and LFAB2 and of course, we’ll reduce the depreciation because the asset value will be lower. The timing of that is the part that I don’t have a good answer for. But as soon as we have a better understanding, we reach agreement, then we’ll share those details.
Dave Pahl
All right. Thank you, Ross. We will go to the next caller please.
Operator
Thank you. Our next question is from Stacy Rasgon with Bernstein Research. Please proceed with your question.
Stacy Rasgon
Hi, guys. Thanks for taking my questions. First, I wanted to ask about the revenue rate. So $20 billion to $26 million. So I get the idea that the low range is no growth off of 2022, but off of the lease where things look likely to be in 2024 even the low end would be like a 13% CAGR for the next couple of years to get there. And the high end 2026 would be almost 30% CAGR. Both of those, I could argue, if we don’t get some big cyclical down stake it could still be aggressive. I mean, what are the plans if 2026 revenues are lower than $20 billion? I guess, I would ask, you see that as a possibility and where would CapEx and other things go in 2026 if our revenues are lower than $20 billion? Is that $2 billion of floor on CapEx?
Haviv Ilan
Yes, I’ll take that, Stacy. But again, we’re trying to help me predict the shape of the recovery of the market. I do know one thing. I know that in 2024, today, the units shipped into the market on a trailing 12 month basis is below 2019. I also know — as I meet customers and I look at generation-to-generation end equipment, the amount of addition of parts per system is tremendously higher, okay? So something has to fix itself. And it’s very hard for me to tell you how quickly it will happen. But it will come back to the trend line. Of course, it’s very easy to sit here at the trough of the market and think that we’ll never get there. I think our responsibility as a supplier to our customers is to be prepared when it does and it will.
So this is where the discipline of our company comes in. Does it come in 2026? Does it comes later? We’ll be ready for each and every scenario, including lower than $20 billion and higher than $26 billion. We are prepared to solve each and every one of them. I will say that the one that we have presented today, kind of give us a spend between different cases of higher quick recovery and return to trend line and a little bit of a slower one. And I think it represents the realm of possibilities of the future.
Dave Pahl
Do you have a follow-on, Stacy?
Stacy Rasgon
Yes, I do. You didn’t mention our depreciation targets. I know you’ve given us those in the past. Do those depreciation targets change at all based on anything that you’ve presented today? Or — it sounds like the big change is in 2026. So maybe it would take time for any of that change if it’s lower to roll through. But how do we think about depreciation going forward versus what you’ve said before?
Rafael Lizardi
Yes, sure. Let me address that. And of course, what I’m about to say does not include the potential benefit from the direct funding. It’s only the CapEx and the ITC. So for 2024 and 2025, it’s the same as what I said at the last earnings call. For 2024, we expect the depreciation to be $1.5 billion to $1.6 billion. For 2025, we expect depreciation between $2 billion to $2.3 billion. Now I’m going to go ahead and give you a 2026 expectation on depreciation, which is $2.3 billion to $2.7 billion.
Stacy Rasgon
Got it. That’s helpful. Thank you.
Dave Pahl
Thank you, Stacy. We will go to the next caller please.
Operator
Thank you. Our next is from Joshua Buchalter with TD Cowen. Please proceed with your question.
Joshua Buchalter
Hi, guys. Thank you for taking my questions. I wanted to ask about some of the assumptions in your scenario analysis for 2026. It seems like — if I do back the envelope math, they’re all landing at around high 30s to low 40% free cash flow margin and sort of low to mid-double digits CapEx to sales ratio. Anything in between those metrics you can give us to get us confident in the free cash flow per share floor, whether it’s across OpEx or working capital or any incremental details on CHIPS, which I guess you’re not going to be able to share. Thank you.
Rafael Lizardi
Yes. So I’ll take a shot at that. And that back of the envelope is not a bad place to end up. Let me address maybe the free cash flow margin first. So of course, our focus is to grow free cash flow dollars per share over the long term. And as Haviv talked in the prepared remarks, our goal is to get back to the trend line, we just show you a path to get to that. The free cash flow per share — the free cash flow margin our objective of 25% to 35% that we’ve had, but we ran above that before the elevated CapEx period that we’ve been in. And obviously, we’ve been below that during that elevated CapEx period, but there’s — it’s feasible that we get back above that 25% to 35%. I think you talked about capital intensity as well.
Let me maybe address that real quick. Our gross capital intensity, that’s before ITC. We expect that to be roughly 1.2 times the revenue growth. And this is after when we get to Phase 3, right? So you take your expected long-term trend on revenue growth, you multiply that by 1.2, so 20% higher, you get to the gross CapEx intensive. And of course, there’s ITC that subtract that at the free cash flow level.
Joshua Buchalter
Thanks for the [indiscernible] Rafael. I might actually ask one I asked in the last capital management call. I know that here and give him a shot answer. Have your priorities or thoughts on M&A changed as the capacity investments have been ongoing, but revenue hasn’t really followed and that gap has widened as the ROI decisions on any potential deals where you could bring something on to your manufacturing network increased meaningfully. And I fully recognize that a lot of users have been done, but I would be curious to hear your updated thoughts on potential M&A.
Haviv Ilan
In a high level, our thoughts have not changed. Same methodology, as we’ve mentioned before, analog-centric, centers around catalog, industrial and automotive markets and the numbers have to make sense. Regarding the numbers have to make sense, I will make one point. We do believe — and this is over the long term, it’s very hard to get convinced about it right now at the trough of the market where supply is well ahead of demand. But we do believe our geopolitically dependable capacity will grow in importance, but also in value.
And I do believe when I go see customers, especially in the markets of industrial and automotive, but I’ll add also think about cloud and AI and what we run on these type of sectors and end equipment, there is a growing importance that the capacity comes from the right places, from the right companies and that people can count on it for the long term. As such, I do believe that the value of our capacity and our front end and back end side will grow, and we’ll have to wait and see that maturing, but that’s our conviction moving forward.
Rafael Lizardi
Let me just address the first question, one piece real quick. That capital intensity that I just talked about, it is lower than what we have said before. And as Haviv mentioned, it’s a complex plan. Now that we are 60% through, we have more details, more granularity on what the plan looks like. So that’s why we are updating that and the CapEx numbers on the slide that we presented show that and the 1.2 times figure that I gave you also reflect that.
Haviv Ilan
Just to be a little bit to provide more color about what Rafael just said, look, this equipment that we are now installing in our past, we are discovering the efficiency and really the throughput that we can get out of the fab today versus the older used equipment we used to install in our factories, also the advancements of information data are allowing us to take the throughput numbers to a higher level, therefore, that helps on the capacity intensity moving forward.
Dave Pahl
Great. Thank you. We will go to the next caller please.
Operator
Thank you. Our next question is from William Stein with Truist. Please proceed with your question.
William Stein
Great. Thanks for taking my questions. First, I just want to make sure I understand. Maybe you can dumb this down for me a little bit. Relative to the capital management call we heard earlier this year, I think there’s sort of two changes. One is the over change in the 2026 spend plan going from five previously to two to five now. I think there was some expectation among investors that perhaps this call would introduce a more flexible spend in the next couple of years based on demand. And what I’m perceiving as the change in today’s call is now what you’re showing us is sort of the dependencies — the operational dependencies and the reasons why you can’t spend less through at least the early part of 2026. Is that the right way to characterize this update?
Haviv Ilan
I’ll take a shot at it. Rafael, please chime in. So again, I think the 2026, you’re right. I think we are not changing anything. We’re just adding more scenarios, right? I think we talked about the maximum preparation in the last call. Now we want to introduce several revenue scenarios for 2026 and associated CapEx. I think it’s very clear why we want to keep a steady hand on our CapEx investment through 2024 and 2025. Basically, I think we said before. I mean you see the good reasons in that phase approach that we have mentioned hopefully, that allows investors to further understand the decision-making of Texas Instruments. I will say one more thing that is implied in the slides, we show in the CapEx slide, kind of a high level — the highest level, we see a $5 billion a year CapEx plan through the end of the decade. That represents the higher end the growth. Think about what we talked about before, about 10% CAGR. The message — and I think Rafael addressed it in the previous response, if we need to support a very high level of revenue growth, we can now do it at a lower level. I think we indicated levels above $5 billion last time. Now with our efficiency of the capacity and the output we get for the fab, we can support a 10% CAGR through the entire decade. It’s just a lower CapEx of $5 billion a year. That’s the high case.
Dave Pahl
Do you have a follow-on, Will?
William Stein
Yes, that helps. And I do have a follow-on, Dave. Today, you reiterated an underpinning sort of expectation for TI and that is that auto and industrial should be the fastest-growing end markets over the long term. I wonder when would the last time you reevaluated that in a really deep, meaningful way? Some might argue that based on what’s going on in new AI-oriented data centers could make that faster-growing market. We’re sort of in the midst of that today, maybe not over the next 20 years, but who knows. And I wonder to what degree this is evaluated in a sort of thorough way. Thank you.
Haviv Ilan
Of course, it’s evaluated in a very thorough way. We are not ignoring the opportunity in enterprise and cloud, computer, they call it or you can call it AI, but it’s beyond AI. There is also CPUs or whether that needs support and many, many other communication equipment that need power, that need connectivity, that need sensing, and we are playing in the market. So first, let me just reiterate what I said in the industrial and automotive market. These are very large, tens of billions of dollars for our Analog and Embedded Processing portfolio, okay? And they are fast growing. I do agree that cloud or AI will grow faster. It’s simply very small today. The beauty of AI, I think the third for data and bits and storage will just go — will grow exponentially. And I don’t see when it will saturate as you said.
So of course, we are well prepared to attack that market. I think I’ve mentioned in one of my previous conferences, our current business peaked in 2022 close to $1 billion, it was around $800 million. During that year, it experienced a strong inventory correction for us in 2023, but we are seeing already this market recovering and similar to consumer or personal electronics kind of fast following the recovery there. I am excited about the future there. Today, we serve the market with many, many parts, similar to examples I’ve shown before, our catalog portfolio across analog and embedded is vast.
But of course, the biggest opportunity over there is on the power side. We are working diligently to build, I think, a great solution for the application-specific powering that V-core of high-performance compute, AI included. Our process — our BCB process developed in SM1 that I mentioned before, we believe, is the best in the world. Our customers do care about capacity coming from the U.S. So we are well positioned as we ramp towards the end of 2025 and 2026 and our product portfolio is getting developed there.
I am very bullish about our opportunity in this market. And you’re right, it’s going to be maybe another growth opportunity for the company in the future it’s simply a very small today just because the market is still new.
Dave Pahl
Thank you. We will go to the next caller please.
Operator
Thank you. Our next question is from Chris Caso with Wolfe Research. Please proceed with your question.
Chris Caso
Yes. Thanks you. First question is regarding some of the timing of these decisions. And you provided the revenue scenarios, but obviously, you would need to make decisions in advance of that. I guess as you’re contemplating what ultimate scenario you’ll face for 2026 CapEx, what will the signals be along the way? And I guess how far in advance do you have to make those calendar 2026 decisions before finally making a decision?
Haviv Ilan
Yes, I’ll repeat what I said, Rafael, if you want to chime in, please do. As I said before, we are following like you guys, the development of the cycle. As we — as I said before, and I think we said on earnings, I think the personal electronics market is [indiscernible], and we are shipping and it’s growing nicely. I believe that other markets such as enterprise comms are following. And industrial and automotive, that order came to their decline later and partly will correct later. You always have to remember the complexity of the automotive market as we have such a strong secular growth over there and also a higher — a better position with our product portfolio. So that also has to all play in. We’ll have to let it run its course.
Now as I said before, we believe that a checkpoint or a good time will be more or less 12 to 15 months from now. So I think about it as kind of end of Q3, end of Q4 that allows us plenty of time to modulate our capacity towards 2026 for the equipment that we can push out a little bit to exactly when we need it. Anything else?
Rafael Lizardi
Yes, just as Haviv alluded to, the difference between $5 billion and $2 billion is entirely equipment and clean room manufacturing equipment and clean room equipment, not construction, okay? So the construction, you can think of it as in the $2 billion base, right, and what we’ve done before. So as Haviv alluded to, the lead times of equipment and the way we work with suppliers will give us the flexibility to it sometime middle to late 2025 to make the call and be able to flex that CapEx to the high or the low end of that range.
Dave Pahl
Follow-on, Chris?
Chris Caso
I do, and that was helpful. Thank you. Just with regard to CHIPS Act grants again and realize tough position because it was only announced last week. But Rafael, just in terms of the timing of when you might see some benefit to that asset value and therefore, lower depreciation, you mentioned that it was primarily directed towards SM1. Does that mean it could have a fairly immediate impact, meaning perhaps next year? Or would it be more likely we would wait for 2026 before we had some of that benefit?
Rafael Lizardi
Yes. So again, similar to what I answered to Ross earlier. Unfortunately, I don’t have a lot of details to give because we’re still — we only sign a preliminary term, not a full. But just to clarify, the proposed direct funding is specifically for — to construct and build SM1 clean room and complete the pilot line, to construct and build the LFAB2 clean room for first production and to construct the SM2 shell.
And what I could tell you beyond that on the numbers, obviously, if we do get a contract and agreement with the government, a full agreement, then we’ll get $1.6 billion of cash at some point over some years. The asset value will be lower by 1.6 and depreciation will be lower by 1.6. The question is timing. And that’s the part that we’ll get — we’ll need to get to a full sign agreement before I can share any details.
Dave Pahl
Right. And that is part of Phase 2, and that is planned to be done by 2026.
Rafael Lizardi
So those three items.
Haviv Ilan
The execution, yes, but the whole — the way the CHIPS office will distribute funds and all that, that to us is still unknown. And we just have to let it play of course, again, I think, as I said, it usually takes three to six months. We’ve seen examples between the PMT and the final terms. So we’ll just have to be patient with that. Thank you.
Operator
Thank you. Our next question is from Tore Svanberg with Stifel. Please proceed with your question.
Tore Svanberg
Yes. Thank you for all the detail. I have a bigger sort of picture question here. When I look at the free cash flow per share chart that you have on Page 19, it just seems that you would not be able to get back to trend line without the ITC help First of all, is that implementation correct? And if so, obviously, it puts much more pressure on the top line. And I was just wondering, other than targeting, obviously, industrial and auto, which is the fastest-growing segments of analog, how do you intend to accelerate the growth rate and take that share back up into the 20s?
Rafael Lizardi
Well, let me address the ITC comment and then, Haviv, you want to add anything. But ITC, as we’ve said before, is — we said during the call, we expect $6 billion to $8 billion of ITC benefit for the next few years, about $1 billion a year through about 2027 or so. And we’re already — we started receiving cash on that last quarter. And this year, we expect $1 billion. But hey, that was comprehended in our thinking when we put together this plan and that is an offset to the CapEx that we’ve been spending. So — and then beyond the 2027 time frame, the reason that we have additional benefit is, as we said during the call, in the case of Sherman and LFAB2 because those factories were started after the bill was signed. The benefits go through the first half of the next decade.
Haviv Ilan
Yes. Regarding — I’ll just again add regarding slide — I would like to talk about the market share question specifically. And why is our confidence high? Just go back to Slide 6 in our presentation. Look, I’ve been in this company for more than 25 years. If I think about our portfolio, it’s never been ready to attack and take share, okay? We had to go through a lot of changes, especially on our other business in the last decade. And fortunately, this work is behind us. I’d also say that during the 2020 time peak or cycle, we could have done more. We had more positions. And unfortunately, we could not supply to all of them, and we had to take some tough choices on the nonindustrial and automotive market.
So when I think about the next peak or at least the next decade, let’s talk about long term in 2030, our portfolio is much stronger. Our Embedded business, which I mentioned in my prepared remarks, is now ready to go and be an equal contributor to the Analog business. That was not the case from 2018 to 2022. If you look at that Embedded business during that time, it actually did not grow, okay? So I will say that based on our investment in our product portfolio, based on the changes we have driven on the strategy of embedded, based on the fact that we are five years into that process and based on what I see when I go and meet customers in the wins that are coming in for the Embedded business, I see it now we have two engines that are represented representing 90% or more of the company ready to take share. Fortunately, for the next peak, we will have the capacity. We are making the responsible investment to be ready for that. And I just have a very high confidence that you will see that translating into market share gains.
Dave Pahl
Follow-up, Tore?
Tore Svanberg
Yes. Thank you, Dave. My follow-up goes back to sort of the market priority — end market priority. Obviously, industrial and auto is the priority. The industrial part, I absolutely get. On the auto side, I’m just wondering if anything has changed there because obviously, it’s a market that today is increasingly dominated by China, and we all know China obviously is trying to insource more and more. So I’m just wondering, as, again, we look to the next 10 years, is automotive still sort of the best opportunity for another companies to go after.
Haviv Ilan
Yes, I think the short answer is yes. I think about China, they are 20% of all GDP. They wait for them to play the game, and I came from China just a couple of weeks ago, is to play the global game. These guys are very motivated to be able to sell their cars [indiscernible] beyond China and they wanted the best automotive solutions. And they turn to companies like us to go solve their problems. So our design momentum in China is strong. Our revenue in Q2 grew very fast also in the automotive market. Our designing momentum very strong, and we will continue to play the global game in automotive, the EV game that you mentioned in China, but also the traditional players in Europe, in the U.S. are as important. We are seeing momentum across all type of powertrains, as I mentioned in my prepared remarks and all geographies as well.
Tore Svanberg
Very helpful. Thank you.
Dave Pahl
Thanks, Tore. We will go to the next caller please.
Operator
Thank you. Our next question is from Tom O’Malley with Barclays. Please proceed with your question.
Tom O’Malley
Hey, thanks for adding me on and appreciate all the detail here. I just had one question. It’s kind of a multi-partner for Haviv. And I just wanted to reference to the slides you kind of showed here today. One is just the unit down [indiscernible] over the past couple of years and just kind of the trend line over the last couple of decades and then also kind of the TI revenue projection. So I think when you talk to investors, I think really what people want to understand is just I think the recovery of that market is understood over the course of the next several years, but there are two factors: One, kind of the pressure from China; and two, pricing that some feel may vary the recovery time frame. Something that could be helpful, maybe, and it’s kind of a question [indiscernible] statement here. But could you just help give some confidence that you guys have considered both the impact of China in that recovery, just thinking about the correlation between those units upticking and TI revenue upticking with it?
And then also just on the pricing side and how is that kind of factored into your forecast on the CapEx side over the next couple of years? Just any commentary around that, I think, would be really helpful.
Haviv Ilan
Okay. Let me say a few comments about China and the competition, and then I’ll let Rafael talk a little bit about the model and what we assumed in our — moving forward. So first, I would say that the competition, especially now when — it’s always been the case. When the market goes through a down cycle, competition intensified. It’s always been the case, and this is not a different one. The China market is important for us. And as I’ve said many, many times, we selected and we can compete in China as well as any other part of the world. I do believe that a company that competes in China and win market share in Shenzhen has a good chance to do it elsewhere as well. We are seeing a growing number of local competitors in China. They have access to, again, subsidize the capacity investment.
And sometimes, the price, call it, to win a new socket could be lower than it used to be two, three years ago. It’s very natural. The good thing is that our cost structure and our competitive advantage, especially our investment in our own manufacturing and technology allows us to compete in each and every socket. In some of the sockets, the local competition in China do not have yet an answer. Maybe they’ll have it in the future, not today. In the other areas where they do have a competitive part, we can go head-to-head and we usually win. And we — and I have a high level of confidence that we can continue to do that in the future. You are right that during the previous cycle, prices in the market went up, and we do expect them to go and return to previous trends of kind of low single digit decline in the long term, but they are still at a higher level than 2019, okay?
So the last point I will add, TI has been our prices as well with the market went up during the up cycle, but we have not done it in an exaggerated way. Our customers fully realize it, which allows us to continue and compete. Rafael, a few words about the model and what it assumes for the future?
Rafael Lizardi
No, I think you alluded to all the points already. So I’ll just mention on the pricing. As Haviv Aviv said, we’re assuming a low single-digit decline, and that’s what we’re seeing right now. In aggregate, of course, there are some places where it’s more, there’s some places where it’s less, but in aggregate, is at low single digit. That’s on the price side. And then on the cost side, as Haviv said, we have the best cost structure. So we welcome the competition, and we welcome some of those spaces that are more and more competitive because we can win there, just like we also win on the less price-sensitive areas.
Dave Pahl
Do you have a follow-on, Tom?
Tom O’Malley
Yes. I think it was asked in a roundabout way earlier, but I just wanted to understand into 2026, the downside. Let’s just say a scenario plays out kind of below the low end of your expectation. Can you just help us get a framework on the CapEx side, but the downside you’ve obviously expressed multiple times in this call that you guys can be flexible. But is there kind of a baseline where your CapEx can go that you need to kind of support ongoing business operations. Where would the CapEx go kind of the worst-case scenario in your eyes?
Rafael Lizardi
Yes. So we gave you four scenarios to work with and CapEx for those, for 2026. Obviously, as Haviv mentioned, if it turns out that the recovery is delayed, then we’ll — we can adjust from there and the goal is to grow free cash flow per share over the long term in an environment. We’re well positioned to do that.
Haviv Ilan
Let me just add just a little bit — and I’m not sure it’s exactly what you’re asking. But in 2026, there is kind of a floor on the CapEx, right? The reason is we do have to complete the long lead time investment in capacity. So we are going to complete SM1, including qualification and pilot line. We are going to complete the clean room of LFAB2. We are going to complete the share of SM2. So that gives you kind of a floor, if you will, in 2026. And I think it’s on this slide, it’s somewhere between $2 billion and $3 billion, okay?
Now if the environment continues to be persistently slow for some — for whatever reason, it doesn’t have to stay there, okay? Then you have some buildup capacity that you’ve completed in 2026. And I think even the slide indicates that. That 2027 number can go as low as almost you want other than maintenance CapEx, okay. So there is no floor beyond 2026, if you will, depending on the market environment.
Rafael Lizardi
Yes. No. So I agree 100%. That’s the — yes, to be very transparent that in a low revenue scenario, the CapEx related to building would be essentially zero and then it’s all modular equipment…
Haviv Ilan
Beyond 2026.
Rafael Lizardi
Beyond 2026, right? So it’s all modular scalable flexible equipment. So it just purely depends on the revenue. And obviously, we’ve built a base that can handle the revenue at that point, and we’re not going to spend any more just to spend it, right? So CapEx in an extreme scenario a very low revenue scenario can go pretty low.
Dave Pahl
Great. Thank you. We will go to the next caller please.
Operator
Thank you. Our next question is from Chris Danely with Citigroup. Please proceed with your question.
Chris Danely
Thanks, guys. I’ll be quick. So it seems like CapEx is peaking. Does that mean the depreciation would peak five years out and, let’s say, 2029 or 2030? And would it be substantially above the 2026 number you guys gave?
Rafael Lizardi
Yes. So let me try to give you an answer on that where you’re going. First, on depreciation, I gave the numbers earlier. So for 2026, that’s the new number that I gave today, we expect 2.3 to 2.7. We’re not giving you a number beyond that. But of course, it’s going to depend on CapEx. So depending on the scenario, that depreciation could continue to increase, but that pretty much will only happen if revenue is really strong and then we have to spend more CapEx to support that. Now where I think your — the heart of your question is really on gross margins. So let me give you an answer on that. On gross margin, we expect the revenue fall-through, that’s excluding changes in depreciation, to be in the range of 75% to 85%. The model works best when you do a full year to full year, okay?
And in addition to that, I would tell you for the next few years because we have some tailwinds, we expect to be at the high end of that range, so closer to 85% than the 75%. For 2026 and beyond under most revenue scenarios, that would put gross margins in the low to mid-60s, okay? So as you model that, think about that, and that’s about the place where you want to be.
Haviv Ilan
I will say also, Chris, that if we see depreciation continue to grow, as you mentioned, that means we are in a very, very good market. You can connect that to kind of a 10% CAGR CapEx scenario that I’ve shown over there in that slide. In all other scenarios, yes, it picks and then you don’t need to run in the same intensity. Remember that the elevated level of CapEx was done really — it’s really to get more of our manufacturing internally and to get ahead, right? So once that completes, you choose your revenue CAGR you want. But in a high level, I can tell you the CapEx slide that we showed kind of the top of that funnel shows about 10% CAGR. And that bottom 1 corresponds to kind of a mid-single-digit revenue CAGR for the company. It gives you kind of the framework of how CapEx can change between these two scenarios. And I just don’t believe that this decade will run at below mid-single digits, and that’s why we have not predicted it right now on the slide for the second half of the decade, if you will.
Rafael Lizardi
Chris, the other thing I would add, remember, gross capital intensity, if you want to model beyond 2026 in the longer term, we have given an answer earlier, 1.2 times revenue growth, and that’s the growth capital intensive before we get benefits from ITC. And the other thing I would mention is all those numbers that I gave you and the model that does not include the potential upside of $1.6 billion from the proposed direct funding. We don’t know the timing of that. I don’t know how the depreciation — the contra depreciation would flow. But clearly, it’s going to be — if we get that signed, it’s going to be a benefit.
Dave Pahl
Do you have a follow-on, Chris?
Chris Danely
Yes. Just a quick one. So Putting all this stuff together, you guys have given your free cash flow margin goals in the past and like you said, you got above there. As we’re going through the next five, 10 years, is there anything that would prevent you structurally or otherwise from getting back to that previous free cash flow margin peak?
Rafael Lizardi
The short answer is no. It is quite feasible that we get back there.
Dave Pahl
Thank you, Chris. We will go to the next caller please.
Operator
Thank you. Our next question is from Joe Moore with Morgan Stanley. Please proceed with your question.
Joe Moore
Great. Thank you. As you think about 2026, can you give us a sense for the new generation potential that you have? Or maybe just the factory utilization that you’re implying with the scenarios, your PP&E will have roughly tripled from the low. I know that’s not capacity, but there’s buildings and stuff in there, but the PP&Es up a lot, if the revenue is kind of at that 2019 level and like-for-like pricing is up. Just curious like what kind of fab utilization is implied by these numbers.
Haviv Ilan
Yes. So I’ll — again, you’re talking about the peak case, right? We represented $20 billion to $26 billion in 2026. So if we need to support $26 billion in 2026, that would not be at a full utilization, but I would call it a healthy utilization meaning there is still some more spare to go. What we’ll probably modulate and even if it wants to be as high as $30 billion, I think we can do it, we’ll have to run full. And probably our internal versus external ratio will have to change. So right now, I think the probability of $30 billion in 2026 is low. So we haven’t put it on the slide for today. But if the market wants to boom, and we will be able to do it, but relying a little bit heavier on our foundry partners.
Joe Moore
Yes, that’s helpful. Thank you. And then I wonder with the rent money. How do you think about the loans? How are you thinking in terms of the longer-term capital structure. You obviously still lot of cash generation potential in the business. Do you need to think about government-sponsored loans as a value creation opportunity.
Rafael Lizardi
Yes. So that’s the capital structure question and when it makes sense to take on debt, given the opportunities that we have to do with that cash and — versus the cost of that debt, then it can make sense to take on that. As you’ve seen us over the last several years, going on seven, eight years now, taking on debt pretty much every year, and now we have roughly $14 billion of that. But given our cash flow generation or EBITDA, we have plenty of room for additional debt. If that makes sense and the loan program from the government makes it incrementally better because it’s a slightly lower interest rate potentially. So we’ll take that into consideration as one of the variables.
Joe Moore
Great. Thank you.
Dave Pahl
Thank you, Joe. And operator, we have got time for one more caller please.
Operator
Thank you. Our final question is from CJ Muse with Cantor Fitzgerald. Please proceed with your question.
CJ Muse
Yes. Good morning and thank you for taking the question and thank you for hosting today’s event. I guess first question, as it pertains to your flexibility in 2026 for CapEx, I’d be curious how to think about ITC, which I believe is expiring at the end of the year, how that plays a role in kind of your thinking about how aggressive to spend or not?
Rafael Lizardi
Yes. No. So as the law works, it expires in 2026 for projects that were started before the law was signed in August of 2022. So that is, in our case, RFAB2 and Lehi 1. So we will maximize most likely, we will maximize equipment purchases through 2026 for those factories. But for Sherman 1 through 4 and Lehi 2, the way that works is a 10-year period, if the fabs were started after the lowest sign, there’s a 10-year period from when the construction began. This is why for those factories, the ITC benefit goes through the first half of the decade — of the next decade. 2032, in the case of Sherman. 2034, in the case of Lehi 2.
Haviv Ilan
Yes. Maybe just a quick addition, CJ, on the capital allocation decision here. It’s kind of an easy decision. It’s 25% through 2026 as Slide 14 says and LFAB1, for example, it’s a transfer fab mainly. So the revenue right now is being built by our foundry and we will have still great opportunity between now and the end of 2026 to transfer some or most of that revenue inside Lehi 1. So that makes it a very easy decision for us to go through completely equipping LFAB1. LFAB2 has similar economic benefits. We are doing the same on the 150-millimeter wafer fabs and it also host our latest and greatest product technology is 130-nanometer on the BCD node for analog. And that runs very nicely right now on both industrial and automotive. So it’s an easy decision to make to go and fully build RFAB2 and LFAB1 by the end of 2026 and get the benefit of the 25%.
Dave Pahl
Follow-on [indiscernible]
CJ Muse
I do, Dave. I guess going back to the capital intensity, 1.5 time moving to 1.2 times seems like a big deal. So would love to hear maybe more granularity on those efficiencies. It sounds like you’ve got better throughput, so higher productivity. But are there other drivers that are enabling that kind of seismic shift in intensity?
Haviv Ilan
I’ll start and Rafael, you can — I’m not just sure exactly we have given a very exact number before. I think the $1.5 billion was earlier in the process. These fabs are not running yet in production like they do now. So I think we just know more. But — the biggest benefit is the new equipment, right? We think it’s best to see enough of [indiscernible], they operate as one fab. But you see how the new equipment runs just faster, okay?
So all our models have been recalculated as we — as we operate that joint fab. And I think there is more to do. So we are working for us to be an internal manufacturer doesn’t mean just we do it internally. We have to be very competitive on the way we run the cost of the fabs. There are a lot of innovation in data, in information, in quality TI and we think there is more opportunity to go. The tools enable us to do so. So that’s the biggest difference. Also us now moving two, three years or 60% through the investment, we just want to be — we can be a little bit less, I would say, guarded with our assumption. Anything else to add Rafael on that perspective?
Rafael Lizardi
No.
Haviv Ilan
Okay. Thanks.
Dave Pahl
Okay. Well, thank you all for joining us today. A replay of this call will be available on our website. as well as the slides that we used in the call. Have a good day.
Operator
This concludes today’s conference. You may disconnect your lines at this time. Thank you for your participation.