Good day, welcome to the Texas Instruments Capital Management Strategy Call. Today's conference is being recorded. At this time, I would like to turn the conference over to Dave Pahl. Please go ahead, sir.
Thank you. Good morning, thank you for joining our 2018 Capital Management Call. This call is being broadcast live over the web and can be accessed through our website at ti.com/ir. A replay will be available through the web. 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. During today's presentation, we'll begin with a quick recap of our capital management strategy and our scorecard for 2017. We'll provide a historical summary of our capital allocation and take a deeper look into specific areas of investment and our free cash flow per share performance.
Finally, we'll wrap up with a review of our cash returns. We believe the key points that investors can take away from our discussion today are, first, we remain focused on consistent execution of our capital management strategy. Second, our business model is designed around four sustainable competitive advantages, and we continue to strengthen and leverage those advantages with a view towards the long term. Third, our disciplined allocation of R&D and investments in our initiatives is delivering growth from the best products, analog and embedded, the best markets, industrial and automotive. Our 300-millimeter analog manufacturing strategy is a unique advantage and will provide benefits for a long time in the future. Lastly, we remain committed to returning all free cash flow to our owners. These key points are consistent with what we've been doing over the last several years.
As a reminder, we believe TI is in a unique class of companies that can grow, generate, and return significant cash. We're focused on the best products in the best markets in the semiconductor industry. We believe analog and embedded are the best products. They are large markets with fragmented competitor bases and are used across a diverse set of applications and customers. Moreover, these products have demonstrated decades of generating profitable returns and significant amounts of cash. We believe industrial and automotive are the best markets and will drive growth in our industry and at TI. These markets are the fastest growing due to their increasing semiconductor content, a trend fueled by products becoming more intelligent, more connected, safer, and more efficient, especially as mechanical and electromechanical features are replaced with solid-state electronics. Electric vehicles are a great example of this potential.
The chip content in the powertrain of an internal combustion engine is about $25 per car today, and focused on mechanical components like cylinders, cranks, and carburetors. In an electric vehicle, that chip content could increase tenfold to more than $250 per car as battery management, motor control, and power electronics come into play. This is an example of one system in an electric automobile. Imagine similar increases across thousands of industrial and automotive application. That's why we believe these markets will outperform the overall semiconductor market, and why we have made them a priority. Our strategy is designed around four sustainable competitive advantages that, in combination, provide tangible benefits that are difficult to replicate.
The first is manufacturing and differentiated technology, the broadest portfolio of analog and embedded products, the reach of our market channels, diverse and long-lived positions, which result in a high terminal value. With that, I'll turn it over to Rafael, and he'll review our capital management strategy. Rafael?
Thanks, Dave. As we have said, our objective is to maximize long-term growth of free cash flow per share, which we believe is the best metric to judge our financial performance and to drive higher intrinsic value for the owners of the company. We do this in three ways. First, by having a great business model that is built on our four competitive advantages in which we're continuing to invest and make even stronger. Second, by being disciplined on how we allocate our resources, focusing on the best product opportunities, as well as what strengthens and leverages our competitive advantages. Third, by striving to constantly increase our efficiency, which is generally about achieving more output for every dollar of input.
As a reminder, our capital management strategy begins with a great business model that focuses on analog and embedded, the best products in semiconductors with a demonstrated history of cash generation. We have designed a company that is capable of growing, generating, and returning significant amounts of cash to our owners. It takes into consideration a strong balance sheet and investments to strengthen our competitive advantages. The tax landscape has changed significantly due to the U.S. Tax Reform Act passed in December. We applaud the reform to U.S. corporate tax law because it enables U.S. headquarter companies like TI to compete more effectively on a global basis. The new law recognizes and rewards companies for exporting and for having manufacturing, R&D, and intellectual property in the United States. There are many important financial implications with this change.
We gave a lot of detail on this in our recent earnings call. I will just highlight three items here. First, investors should assume an ongoing 18% annual operating tax rate starting in 2019. This rate comprehends the 21% statutory corporate rate and the benefit of exports and having manufacturing, R&D, and intellectual property in the U.S. This rate does not include an estimate for stock-based compensation benefit, which we provide at the start of each year. Second, for 2018, investors should assume a transitional 23% annual operating tax rate before the benefit of stock-based compensation. The five percentage points of transitional tax effects are mostly non-cash charges. Lastly, in the fourth quarter of 2017, our tax expense included approximately $800 million of tax expense that was primarily related to the recently passed Tax Reform Act.
While there was no impact to free cash flow in the fourth quarter of 2017, we will pay this out over the next eight years. I'll refer you to our fourth quarter 2017 earnings call for more details, which may help with your longer-term models. Again, we believe the reform to U.S. corporate tax law will enable U.S. export companies like TI to compete more effectively on a global basis. Importantly, with more cash available on an ongoing basis, we will continue to have a strong balance sheet to invest to grow our business, strengthen our competitive advantages, and return all free cash flow to the owners of the company. Our capital management scorecard is one that we have shared with you every year since 2013. Consistent with prior years, in 2017, we again met our multiple objectives.
We are pleased with the consistency that our business model and strategic decisions have allowed us to deliver in these results. Note, with the recent changes in tax law, the metric on cash owned by U.S. entities is no longer relevant moving forward. This year, to provide more clarity and insight into how we think about our business, we have updated our scorecard in two ways. First, we have included descriptions of our long-term objectives for each metric. Second, we have adjusted or simplified how we will judge several of the metrics while eliminating those that are no longer relevant. The long-term objectives should provide more insight into how we actually run the business, as opposed to a number that reflects only a single data point. For example, a long-term capital expenditure objective is to invest to support new technology development and revenue growth.
We also want to extend our low-cost manufacturing advantage, including 300-millimeter, while maximizing long-term free cash flow per share. We recognize CapEx may run higher in any period if there is an opportunity to extend our long-term manufacturing advantage. We continue to believe that about 4% of revenue is still the right long-term capital expenditure target. Additionally, we're adjusting how we will judge a few of these objectives. First, we have updated our expectation for free cash flow generation to 25%-35% of trailing 12-month revenue, primarily driven by the impact of U.S. tax reform. Second, we have increased the range for inventories to 115-145 days. This reflects our continuing plans to own, and therefore control, more of our inventory.
Increased control of inventory results in improved visibility to customer demand, as well as improved customer service, since we can rapidly direct the inventory to the demand. This is consistent with the past few years as we have increased our consignment levels. Third, our cash balance allows us to provide necessary liquidity through all market conditions with about 10% of trailing 12-month revenue and our 12 months of dividend payments. Finally, we're simplifying how we communicate to our targets. The first is to say that we will return all free cash flow over time, and the second is for dividends to be 40%-60% of current year free cash flow. Neither simplification changes the overall objectives. They are just a simpler way in how we describe them and an easier way in how to think about them.
Overall, our objectives remain consistent, and we hope these additional insights give you a better understanding of how we run our business. In summary, our capital management strategy continues to serve the owners well. Free cash flow per share continues to grow steadily while we continue to strengthen and leverage our competitive advantages. Now, I would like to provide additional insight into how we allocate our capital. I'll give you updates on several key investment areas. Over the last 10 years, we have allocated about $72 billion of capital. Given that magnitude, you can quickly appreciate why capital allocation is a job we take quite seriously and one that has significant impact on owners' returns. You can see our largest category of capital allocation is investment in critical areas that drive organic growth, such as R&D, sales and marketing, capital expenditures, and inventory.
With our approach of funding strategies, not projects, we spend significant time ensuring these investments are delivering long-term competitiveness and generating returns greater than our cost of capital. The second-largest category is share repurchases. Here, our objective is the accretive capture of future free cash flow for the long-term owners. We focus on consistent repurchases when the present stock price is below the intrinsic value, using reasonable growth assumptions. Next is dividends, where our objective is to appeal to a broader set of investors, and we focus on their sustainability and growth for obvious reasons. Finally, potential acquisitions are evaluated through two primary factors that have remained unchanged. First, it must be a strategic match, meaning catalog analog focus with high exposure to industrial and automotive.
Second, it must meet certain financial metrics, such as generating a return greater than a weighted average cost of capital within about four years. For simplicity, we have not included changes in net debt, which over this period was $2.1 billion. With that framework set, let me now examine our investments in several specific areas. I will first focus on our R&D investments that we allocate to higher value growth opportunities. As you may recall, our broad portfolio of analog and embedded products is an important competitive advantage. This breadth of portfolio brings more visitors to ti.com each year than to our competitor's website. It is also critical that we continually grow and strengthen this portfolio with highly differentiated products that are developed with an eye on the best market opportunities over the next 10 years.
At the highest level, we see good opportunities in all of our markets, but we believe that industrial and automotive will be the best opportunities over the next decade. This is primarily because semiconductor content in industrial and automotive applications will significantly increase as companies make their equipment smarter, more connected, safer, and more efficient. This chart summarizes the direction of our R&D investments across end markets and also provides the revenue breakout for 2013 through 2017. In industrial and automotive, we continue to increase investments broadly across sectors and product categories. We're excited to see the continued progress on revenue growth as these markets comprise about 54% of TI revenue, up from about 42% in 2013. Personal electronics is an important market, and while investment level in total is down, we do invest selectively.
In communications equipment, we announced several years ago that we were reducing our embedded investments. We continue our investment in the analog growth opportunity. Our investments in enterprise systems and other have been flat and at low levels. We will now talk through our manufacturing advantage. We have two 300-millimeter wafer fabs supporting our analog business. We made good progress on this plan again in 2017 as we continued to increase factory loadings in both of these fabs. In 2017, we utilized about 50% of the combined capacity, which supported about $4 billion of revenue. This was up from about $2.5 billion in 2016. We remain committed to strengthening and leveraging this competitive advantage over the long term.
As a reminder for those not as familiar with the semiconductor industry, a chip, meaning an unpackaged product made on a 300-millimeter wafer, costs about 40% less than a chip built on a 200-millimeter wafer, the size used by most of our competitors. This translates into a great competitive advantage. The reason it is an advantage is because a 300-millimeter wafer has 2.25 times more area, which in turn means we can get about 2.3 times more chips, but it does not cost 2.3 times to process that larger wafer. This translates into a structural cost advantage. To understand how a 40% less expensive chip impacts gross margin, it is easiest to use an example of a part built on 200-millimeter compared with 300-millimeter. This example shows a theoretical part that sells for $1 with a gross margin of 60%.
The chip itself would cost about $0.20 built on 200-millimeter, and this would reduce to $0.12 on 300-millimeter. In this example, the remaining cost of assembly and test are the same regardless of the size of the wafer. The net result is that gross margin improves by eight percentage points. As this simple example illustrates, our 300-millimeter manufacturing capability and the resulting cost structure provide a unique competitive advantage for TI. The best part of our 300-millimeter advantage is that it is just beginning. In 2017, we continued to make steady progress with about 40% of analog revenue built on 300-millimeter and a lot of room to expand. Moving forward, we believe that the majority of our incremental analog revenue will be built on 300-millimeter. We will now go through our free cash flow growth and outlook.
As we described at the beginning, our overall objective is to maximize long-term free cash flow per share. We believe this is not only the best metric to judge our performance, it is also the one that owners ultimately care about. Since 2004, our free cash flow per share compounded annual growth rate has been 13%. In 2017, free cash flow margin increased 70 basis points to 31.2% of revenue. We reduced share count by 1.3%, and free cash flow per share increased 15.7%. Last year, we had three drivers of free cash flow per share growth. Top-line revenue growth driven by analog and embedded, incremental free cash flow margin increases, and additional share count reductions. We believe these three drivers of free cash flow per share growth will continue. Our analog and embedded segments have a proven track record of growth.
Combined, they have recorded both five and 10-year revenue CAGRs of 8%. Part of that growth is market share gains, where we have seen on average 30 to 40 basis point improvement annually over time. With analog and embedded comprising 90% of our revenue, we expect that we'll be able to continue to drive the top-line growth. Let me change gears and talk about cash returns. Our free cash flow margin also stacks up well against the S&P 500. Over the last 12 months, our free cash flow margin of 31.2% places us in the top 15% of the S&P 500. Generating cash is obviously important, but it is only valued if it is invested wisely, which we have already discussed, or it is returned to investors.
We have an equally strong benchmark in how well we return cash through dividends and stock repurchases, where we are in the top 10% of the S&P 500 in terms of cash returns as a percentage of revenue. As our cash return to owners has grown, so too has our dividend as a percentage of the cash return. We continue to believe a sustainable growing dividend is an important element of our capital management strategy. Our objective in repurchasing shares is the accretive capture of free cash flow for long-term investors. We focus on consistently repurchasing shares where the intrinsic value of the company exceeds its market value. By using realistic discount factors and reasonable growth assumptions to calculate the intrinsic stock value, we're aiming to have confidence that investments made in stock repurchases are, in fact, earning rates of return greater than our cost of capital.
While the ultimate assessment of ROI depends on the future cash flow stream, the track record of this approach is encouraging. If you make the very conservative assumption that TI's future free cash flow remains flat, we will be earning about 10% annualized return on all stock repurchases made to date since 2004. We have reduced shares outstanding 43% since 2004, including the 1.3% reduction in 2017. We ended 2017 with $9.2 billion in open authorizations, having bought back $2.6 billion of stock in 2017, including $706 million in the fourth quarter. As we commented earlier, our objective with dividends is to appeal to a broader set of investors, and our focus is on both growth and sustainability. We have now raised the dividend for 14 consecutive years, including a 24% increase in fourth quarter 2017.
We have increased the dividend at a compounded annual growth rate of 24% over the last five years. Our consistent growth of free cash flow resulted in our dividend in 2017 consuming only 45% of free cash flow, supporting our objectives of sustainability and growth of dividends. Let me now wrap up my prepared remarks with a few summary comments. TI is in a unique class of companies that can grow, generate, and return cash. Our business model is designed around four competitive advantages that deliver tangible benefits unique to TI and are difficult to replicate. First, manufacturing and differentiated technology. Second, breadth of products. Third, broad reach of our channels. Finally, diversity and longevity of products, markets, and customer positions. We will continue to focus on growing free cash flow per share.
It is the ultimate objective, we believe focusing on it will deliver the highest growth in the value of the company. In the coming years, we believe that we will have three drivers contributing to free cash flow per share growth. Top-line revenue growth, free cash flow margin expansion, and share count reduction. Top-line revenue growth will be driven by our position in the best products, analog and embedded, and in the best markets, industrial and automotive. Our 300-millimeter analog manufacturing strategy is a unique advantage and will provide benefits for a long time. Finally, owner returns will continue to be driven by dividends and share repurchases. We have a disciplined culture and processes to ensure that we are strengthening our competitive advantages and generating the maximum return for the investments we make. Thank you. With that, I'll turn it back to Dave.
Thank you very much, Rafael. Operator, you can now open the lines up for questions. In order to provide as many of you as possible the opportunity to ask a question, please limit yourself to a single question. After our response, we'll provide you an opportunity for an additional follow-up. Tracy?
Thank you. Ladies and gentlemen, if you wish to ask a question, please press star, then one on your touch-tone phone. If you are using a speakerphone, please pick up the handset before pressing the corresponding digits. Once again, please press star one at this time. We'll go first to William Stein with SunTrust.
Great. Thank you for taking my question. Gentlemen, I'd like to ask something that's maybe a little bit off focus, and that's how your capital allocation, let's say, not strategy, but tactics might shift somewhat, given different positions in the cycle. In particular, M&A and inventory builds, how your approach to those might shift as we see different parts of the cycle. Thank you.
Yeah. Let me get a start on that, and then I'll hand it over to Dave to give his comments. Frankly, the bottom line is that I don't see any change in our strategy because of that. Over many years now, we've been very consistent on how we allocate capital. Let me just comment on a few of the specific things, and that goes through any cycle, by the way. If you look at our free cash flow trends, they actually have been pretty steady during the cycles, which actually goes to show the stability and the strength of the business model to begin with. Let me comment on a few of the things you mentioned. On M&A, our strategy hasn't changed. We have two criteria.
First, it has to be a strategic fit, and as I mentioned before during the prepared remarks, it needs to be analog-focused, differentiated products, and then we have a focus on industrial and automotive, and then the price has to be right. That hasn't changed. Then on inventories, we spelled out in our objectives in our scorecard how we think about it. Our objective, as I said earlier, is to maintain high levels of customer service, minimize inventory obsolescence, improve manufacturing utilization. Some of these things will vary based on things like consigned inventory, where we control the inventory, and we think we can deploy it very efficiently to ultimately drive higher long-term free cash flow growth. We will continue to do that. Thank you.
Yeah. If I could just add a couple of comments, Will. I'd say, as Rafael laid out, the strength of our business model allows us not to have to focus on the cycle. I think it's something that you do have to respect and certainly be prepared for. We've got a playbook that we've been through many cycles, both Rafael and I, through our careers, and there's four things that I'd highlight of that playbook. For us, the first thing is getting the operating plan correct. Depending on the magnitude, we bring down wafer starts. That's probably the first thing operationally that's most important. We have a list of long-lived, low-volume products that we know we would build inventory on. I think that you could expect to see that. We don't want to put infinite dollars into working capital.
You want to set that operating plan so that if it's a shallow correction, you've got inventory positioned on the other side of it. That's probably the first and most important step. The second is when you look at R&D, our investments are three- and five-year type horizons. In the short term, demand changes and fluctuations in the short term won't curb our enthusiasm for the opportunities that we see both in industrial and automotive. The third thing on SG&A, there's places I think when things weaken that you can be prudent. You've seen us do those types of things in the past. There's also places that we're investing to strengthen our competitive advantages, like in ti.com and our demand creation, so we wouldn't change that. The last one is the more interesting one, and that's where other opportunities might present themselves.
Whether that's assets that we can buy to strengthen one of our four competitive advantages in manufacturing and technology, those types of things can come available. That's what you saw us do in the 2009 downturn with purchasing assets. Somewhat of a long answer to that question, but it was a good one. Thank you, Will. Do you have a follow-on?
Very robust answer. I appreciate it, Dave. Thanks, guys.
We'll go next to Blayne Curtis with Barclays.
Hey, guys. Thanks for taking my question. Just wanted to follow up. You obviously have been allocating more to dividends here. Just kind of curious, you talked about reasonable growth assumptions. Just your perspective, obviously, it was a knockout year in 2017, and just kind of curious what you're thinking about growth in the market as you go forward.
Let me give you my take on that, and Dave, if you want to chime in. From a growth standpoint, I think your question is on revenue growth. You started asking about dividend, but I think you're going on revenue growth. My answer is going to be long term. We view this long term, not any particular quarter or even any particular year. Longer term, we see our semiconductor business tied to global GDP. That's kind of the starting point. Think of that over the long haul, maybe around 3% or so. Semiconductors should grow faster than that. On top of that, the markets where we're focusing on industrial and automotive, those are outgrowing the overall semiconductor.
If you add 3 points GDP plus maybe 1 point on semiconductor, 1 to 2 on that, then 1 to 2 on industrial and automotive, that gives you the range of where we think the revenue growth over the long haul should be in the markets where we play and for our analog and embedded businesses.
Yeah. If I just added on that to Blayne, I think automotive, it's easier to drop into a spreadsheet and see that the growth first. You just go down to a showroom and you can viscerally see the electronics content going into a car and contrast that to five or 10 years ago. If you look last year, there's probably $375, $400 of content in a typical car. I think most industry analysts will have that number doubling. Some have it tripling over the next 10 years. Usually, the difference between those two forecasts are the rate of expectations of electric vehicles take up on that. Electric vehicles average $1,100 to $1,200 of content in a car. If it doubles, that means we're looking at somewhere in the upper single digits of growth over that period of time.
We're not really worried about whether it doubles or not, or whether it triples or not. We just know that there's good growth there. That'll drive those investments. The most exciting thing, I think with industrial, is you can imagine that type of growth. You can't drop it into a spreadsheet the same way. There is content growth going into a lot of things. We've got 14 sectors there and hundreds of end equipment. That's what makes us so excited about those two markets.
Let me just add to that, because particularly in the industrial market, the really exciting thing is, when you look at those end equipments, it's not a coincidence that they perfectly match the capabilities that we have in our portfolio. If you look at, we have power management, we have sensing, we have signal chain, we have embedded processing, we have connectivity. That talks to the breadth of our portfolio, one of our competitive advantages. The other one is the diversity of the marketplace in terms of number of sectors and equipments and customers. Just highlights the importance of the channel reach advantage, which is another competitive advantage that we have. The entire industrial market really plays well to how we have positioned Texas Instruments and the competitive advantages that we have to play in this environment.
Okay. Thank you. Blayne, you have a follow-on?
Yeah, sorry. I might have intermingled my two questions on, because you had talked about the decision between dividends and buybacks and looking at the intrinsic value of the stock. Maybe just talk on the dividend side. You gave this range of 40 to 60. You've been more at the lower end of that range. I know you want to grow dividends. Can you just talk about the decision to allocate more to dividends than buybacks? Do you expect to get more into the midpoint of that range?
Yeah. Let me step back a second. Our intent is to return all free cash flow to the owners of the company. We have been doing that for many years, and we will continue to do that. The question is how we do that. In fact, in 2017, 45% of the returns were dividends. Dividend as a percent of free cash flow. Buybacks were actually more, were 55 or so percent, because we returned $4.7 billion, and free cash flow was $4.7 billion. We return everything. That's our model. On the dividend side, we talk about, as we spelled out on the prepared remarks, we want to be between 40% and 60%. That's the range for a reason, right? We don't want to be tied to any particular end of that range. We do what makes sense.
Within that, as we have said, our intent is the dividend needs to be sustainable, and it needs to be growing. We understand that being able to grow the dividend is important. Being in any particular end of that range, that just depends on various factors, one of which is the stock price. We think the model is robust enough to allow us to both have a sustainable and growing dividend and take advantage of buybacks. As we do in both cases, the ultimate intent is to return all free cash flow to the owners of the company.
That's great. Thank you, Blayne. We can go to the next caller, please.
Thank you.
We'll go next to Ross Seymore with Deutsche Bank.
Guys, thanks for letting me ask a call. Just had a question on your free cash flow margin target update. You took the lower and the upper end up by 5 points. If the tax benefit is the primary reason why, and at least by next year, the tax benefit should be at least 10 points. Is there a reason why you only increased the free cash flow margin by 5?
Let me first maybe clarify something. We increased that margin primarily due to tax reform. I'm going to let you do the math on our ongoing tax rate had been 31%. In fact, it was 31% for 2017, and now we said that starting 2019, it'll go to 18%. Actually, in 2018, from a cash standpoint, it'll be close to that 18% right away in 2018. If you do the math with that, you probably get about 5 percentage point increase on the margin, not 10. I just wanted to clarify that. Beyond that Remember, our ultimate objective here is to drive free cash flow growth. It's not a percent, right? We don't pay dividends with a percent. We don't buy back shares with a percent. We do it with dollars.
We just want to drive those dollars of free cash flow per share growth higher and higher over the long haul.
You have a follow-on, Ross?
Yeah, just a quick one. I know you guys didn't talk about OpEx as a percentage of sales, I assume you're going to keep the 20%-30% range that you've long held. Within that OpEx number, last year was a good example of R&D going up, I think it was about 10%, and SG&A coming down by about 4%. Can you just talk about how you approach OpEx and more specifically those two lines, any balance between them?
Yeah. I think you asked a similar question during the earnings call, I'll go ahead and try to answer it. We're pleased with our level of OpEx. Our focus is to wisely allocate our money to the best opportunities and ultimately to drive higher free cash flow per share. Remember, that is the ultimate objective. How we look at this is it helps to look at it in absolute terms rather than percentages, right? In 2017, for example, we spent $3.2 billion of OpEx. I know you asked about the mix between R&D and SG&A. We like to look at it combined, because there are good things in both, in R&D and SG&A, in terms of things that we're investing to strengthen our competitor advantages. By the way, that $3.2 billion increased 3% from 2016, pretty reasonable.
We will just continue to look at opportunities. If we have opportunities to invest more and drive a higher free cash flow per share growth, we will do that. If they're not, we will not invest more, right? That's how we think about it. Now, as a guide, if you want to look at OpEx as a percentage of revenue, that is a helpful guide for some things. We ended the year at 21%. That's inside our range that we have talked about. In fact, it's towards the lower end of the range. Did that help or?
Yep.
Okay.
Yeah, it did. Thank you very much.
No problem.
We'll go to the next caller, please.
Okay. We'll go to Stacy Rasgon with Bernstein Research.
Hi, guys. Thanks for taking my questions. I was wondering if you could comment a little bit on how you see the potential M&A landscape as it stands today, viewed through the filter of your two criteria, both strategic fit and financial return. Also maybe if you could comment how maybe the valuation of your stock currently might impact that second criteria.
Yeah. M&A, our criteria is the same, like I talked about. The criteria itself didn't change or hasn't changed. Valuations up until last week, they had gone up significantly, and they're still high compared to historical standards or compared to the last two years. We have to take that into account. Frankly, as you saw earlier in the presentation, where we spend the bulk of our capital, where we allocate our capital is on organic growth. We think that's where a lot of the great opportunities are and have been. We'll continue to focus on that. If there is an M&A opportunity that makes sense, we'll consider that and we constantly see opportunities, but very rarely do we pull the trigger. Dave, do you have a comment on that?
Yeah, I'll just add, I think as we look at a potential target strategically, National is just a great example of what we would look for. An analog-based company that has high-quality catalog products that are finding very diverse places in the marketplace. One way to measure that quality is by higher gross margins. Not the only way. There could be some structural problems or whatever, but we're not interested in finding commodity-based products as a contrast. We want to have a company that has product teams that are continuing to put out high-quality products. We don't want to buy that asset for what they did 10 and 20 years ago. We want to be able to take that asset and have it continue to grow longer term.
I think the other thing that we would want from a potential target is, does it strengthen one of our four competitive advantages or does it take advantage of those four competitive advantages? As you walk through that list and think of National, I think they fit that very nicely, and that's what we'd look for. Then you've got the financial piece that Rafael talked about that's layered onto that.
Let me comment on the second part of his question. Stacy, you asked about repurchases. As I said earlier in the prepared remarks, the goal there is, the objective is the accretive capture of future free cash flow for long-term owners. We look at our free cash flow. We make realistic projections of the growth of free cash flow for the next X number of years. Let me stress that they're realistic. They're not conservative, they're not aggressive, they're realistic. Then based on that, we come up with a valuation, and as long as the stock price in the market is less than the valuation, we'll be buying back. As we said earlier, we bought back over $700 million worth of stock in the fourth quarter.
The other thing that I'll point out is clearly tax reform was a good thing, as I talked in my prepared remarks. It's putting U.S. headquarter companies in a better position to compete effectively. We're obviously taking that into account as we make those realistic assumptions of where free cash flow can go.
You have a follow-on, Stacy?
I do. Thank you. I wanted to ask a bit about the increase in the inventory targets. Is this increased consignment, or is this owned inventory? I guess why are you doing this? Are you actually seeing issues with your service levels, or is this just an opportunity to improve above and beyond where you were before?
Sure. It's both on the first part of your question. Let me step back and talk about the objective, just to remind everyone in the call. We want to maintain high levels of customer service, minimize inventory obsolescence, improve manufacturing utilization, and this will vary depending on various factors, and one of which is consigned inventory. Ultimately, we want to have control of the inventory. That just makes it more efficient for us to deploy. One good example of that is consigned inventory, where even though it could be at a customer's warehouse or at a distributor's warehouse, it's our inventory, so we can actually redeploy it, or we can keep the inventory levels as we need them to support that particular customer, versus just having it in the distribution channel non-consignment, where it may not be as efficiently used.
Another way, as I mentioned during the earnings call, is to have buffers of low-volume inventory. That's another way where we can own and control the inventory, and we can redeploy it or deploy it very effectively to ultimately drive higher free cash flow growth.
Yeah. I'll just add, as we think about allocating capital, working capital as a piece of that puzzle, there's strategic value in owning and controlling that inventory going forward. That's why you see differences in what we do and what we're trying to do there, perhaps versus others. Thank you, Stacy. We'll go to the next caller, please.
We'll go next to C.J. Muse with Evercore.
Good morning. Thank you for taking my question. I guess first question, in terms of CapEx, was elevated in calendar 2017 above that 4% target. It sounded like on the last call that will stay elevated in calendar 2018. Curious when you would expect to normalize and I guess more color around that 4% as being sustainable as you fill out your 300-millimeter factories.
It all depends. It depends on many factors, one of which are expectations of revenue growth, of course. We will see how that develops. I do want to emphasize, as Dave talked about earlier or during the prepared remarks, that if we see opportunities of adding assets opportunistically, manufacturing assets, 300-millimeter analog assets, we will do that. That could drive that percent higher. Keeping in mind that over the long term, 4% is the right way to look at it in any given year. Any one of those years, especially as we drive higher revenue growth or the capacity to be able to handle higher revenue growth, we will not be afraid of pulling the trigger and going higher than that percent.
You have a follow-on, C.J.?
Dave, quick follow-up. In terms of taxes, you talked about GAAP, I guess, at 18%. Curious what we should be thinking about long term for cash tax rate.
Let me clarify. The 18% is cash and GAAP, it's both, for 2019 and beyond. For 2018, the GAAP rate is 23%, but on a cash basis, we'll be very close to that 18%. On a cash basis, you can think of 18%, and that's the operating tax rate. That's the ongoing one. That does not include discrete items. As you have seen over the last few years, we've been getting a benefit. In fact, for 2018, what we're projecting and we have on the website, it's about a one-point benefit.
Whatever GAAP or tax rate you want to use following our guidance for 2018, subtract about a point to get to the all-in effective tax rate.
Those that didn't hear, we put a lot of detail on that, just as a reminder, in our fourth quarter call. We also have that information on our website by quarter because we're expecting about $55 million in discrete items, and it will vary by quarter fairly significantly.
Let me just clarify two points. One is a lot of companies are reporting this, and you may be hearing all kinds of numbers. Just remember that we report GAAP. Many companies out there, they also report GAAP, of course, but they focus on non-GAAP, so that could be a little confusing. Also that the percentages that I told you about, the headline numbers that I told you about are operating, so the ongoing, not the all-in. As I said, the all-in, you would have to subtract about a point. Many companies talk about the all-in and only that one. The other one is stepping back. What is the big deal about Tax Cuts and Jobs Act? Is that we are now much better able to compete effectively on a global basis.
That is a really good thing, and that's going to enable us to be that much stronger going forward.
Great. We'll go to the next caller, please.
We'll go next to Mark Lipacis with Jefferies.
Hi. Thanks for doing this call and for taking my question. It seems an important driver of the free cash flow margin growth has been the expansion in gross margins, which have increased by about 1,200 basis points over the last four years. One of the ways that we try to determine if that is going to continue is to look at the incremental gross margins. You've characterized your gross margin fall through around 70%-75%, but this year, the incremental gross margin was 87%. I was wondering if you could help us understand what were the drivers of that incremental gross margin above what you would characterize your normal fall through? Is it mix? Is it 300-millimeter? Is it less pricing pressure? Because that gross margin fall through of 70%-75% looks a bit conservative, at least after this year. Thank you.
Sure, Mark. I'll go ahead and start off on that. I think that we've talked about 70 to 75, and I think with that and with any of the other targets and objectives, what we're focused on is growing free cash flow per share over a long period of time. 70 to 75 is just a guideline. Could it be above that? Could it be below that in any given period? Sure. It's a lot easier to look at it when it's above it. Both sides of that coin are certainly possible. Our focus is really just growing free cash flow per share over a long period of time.
Yeah. I'll just add that there were several factors that have contributed to that increase over the years, but probably the most important one, consequential one over the long term, especially on a go-forward basis, is the 300-millimeter factories that we have. That is a unique competitive advantage, where we're the only one in the space that have around 300-millimeter factories. As I talked during the prepared remarks, why is that a big deal? Because that gives us a structural cost advantage. We have a 40% advantage versus 200-millimeter, and we can deploy that in any number of ways.
You have a follow-up, Mark?
Yes. Thank you. I guess most people believe, if you look at your vertical market distribution, that the profitability of the industrial and automotive markets would be higher than the other ones. Would you be so kind as to confirm that, and can you characterize that or quantify that in any sense with a range? Are these markets 500 basis points better, or how should investors think about those particular vertical markets when we think about mix? Thank you.
Sure, Mark. I'll start off on that one. I think we like those markets for multiple reasons. We don't want to quantify the margins by market or by customers or any of those types of cuts. I'll tell you why we like them is the diversity of those markets. Inside of industrial, we've got 14 different sectors and hundreds of customers that sit inside of that. Even inside of automotive, you think of the diversity that we've got there. We've got five sectors, we could cut it by that. We're seeing good growth out of all five of those sectors. You could cut it by technology, by product level, by region of the world. Lots of different cuts to be able to show that diversity.
We have hundreds of customers that even inside of automotive that we're shipping to, that represent tier 1s and tier 2s. The fact that these products tend to live a long time, that provides lots of benefits, right? When you have a product that can live five years or 10, sometimes 20 or 30 years, the financial returns on those products are very high. All those things combined are the reasons why we like those markets. Thank you, Mark, and we'll go to the next caller, please.
We'll go next to John Pitzer with Credit Suisse.
Yeah, guys. Thanks for letting me ask the questions. A lot of mine have been answered. Rafael, in the past, as you've looked at profitability between your analog and embedded processing businesses on the operating line, you've kind of mentioned that there's really no reason over the longer term why embedded processing margins can't approximate analog. If you look at the last quarter, there's still probably 12-13 percentage points gap between the two. I guess my question is: Is that still a valid statement? As you think about closing that gap, is it a function of embedded finally starting to look like analog, or would you expect maybe with higher levels of investment that that analog might come down as you close that gap?
Yeah. I'll get it started if Dave wants to chime in, but let me just remind everybody that the objective is to grow free cash flow per share, and nothing else. It's not any particular margin. While embedded has increased their margins a little faster than analog, they're looking like they're catching up, that's not the objective, right? The objective is free cash flow per share growth. Embedded has been and continues to be, and we expect it going forward to continue to be a great contributor to that.
Yeah. I'll add, John, I think by way of demonstration, go back to 2010. If you look at the operating margins of embedded, we were running in the Mid-20s from an operating margin standpoint. At that point, we decided to significantly increase our investments in that business. We wanted to repurpose the application processors and connectivity products so that we could use them in both automotive and industrial. There's a quarter transition there that you can see how those investments impacted operating profits. We're not running around with an objective to optimize that. We made that decision so that we could build, over time, a high-quality business that has diverse products, diverse markets, and will contribute to that free cash flow growth over a long period of time. You've seen us take decisions like that that'll play out over a longer period of time.
Again, it's really that focus back on growing free cash flow per share. You have a follow-on, John?
This is my follow-on. When you guys look at your core analog and embedded processing markets, you've consistently grown share at a measured pace every year. I'm kind of curious, just relative to the investments you're making, the end markets you're investing in, your 300-millimeter strategy, and sort of what you're doing around inventory and your inventory strategy, should we expect or do you expect that the historic share gains you've seen should accelerate from here? Again, I guess the real question is, relative to outsize growth to the market, how important is that relative to growing your free cash flow per share?
Well, yeah, I'll start off if Rafael wants to add anything. It's important revenue growth and share gains, we believe, are important long term to growing free cash flow. They're an important component of it. The great news is that we'll have three contributors to it. Long term, investing in our four competitive advantages, whether that's manufacturing and technology. If I just take that one for an example, we spent some time going through the 300-millimeter example of that competitive advantage. The other is that we own those assets. Over time, the control of those assets, the ability to improve the technology differentiation over time through that ownership as the years go by, that will make a very significant difference in the types of products and the differentiation of those products over time.
Other investments in channel advantages, the investments in ti.com, making it easier for engineers to find our products, and when they find those products, to be able to have the opportunity to recommend and sell other products to them is very important. I think those types of activities, John, it's not something that we can drop into a spreadsheet and say we'll see an inflection point. Would we like it to be higher? Absolutely. I think if we just continue to make those investments over time, we're going to like those results. Thank you, and we've got time for one more caller, please.
We'll go next to Harlan Sur with J.P. Morgan.
Good morning, thanks for hosting this call. As a follow-on to the prior embedded question, as you think about this segment and the overall goal of driving free cash flow per share, how is the team allocating R&D to embedded versus analog? Is the op margin expansion in embedded more a function of revenue growth and leverage going forward, or are there other initiatives you're focused on to continue to drive margins higher?
Harlan, I'll start off with that. I think when we allocate R&D, you can think of it as three major buckets. The biggest bucket that we've got goes to product development. Inside of analog, just like inside of embedded, our structure across the company, we have around 70 or so product lines. Each product line manager will have design resources, product engineering teams, marketing teams, application engineering teams, and operations. Basically everything that you need to run a business, we measure that at a full P&L at those levels. Those product line managers will have, at any given time, a number of products in development and then a backlog. As we judge that backlog and starting new projects and the number of projects that that product line manager starts, we look at the quality of that opportunity.
How many customers do we expect them to use? How many markets, how many sectors? How long will that product be in the marketplace? What's the differentiation of that product, and how long will we have it? Those are more qualitative judgments that our product line managers and the business unit managers, and the senior managers that make all of those judgment calls, and we're constantly trying to find the best places where to allocate that R&D. If you sit through those reviews and you walk out and say, "Hey, we're leaving on the table more opportunities than we can get to," that's when you actually take up the amount of spend that we would allocate in that bucket. The other two buckets, real quickly, the second one is a centralized R&D function.
That's kind of the teams that put together the tools that the engineers use, whether that's SPICE models and the development backplanes and those types of things. The good news with that is if we had seven product lines or 70, our investments there really would be the same. We get some leverage off of that just because of the scale we have. Lastly, is over the horizon investments, and we do that in what we call Kilby Labs, named after Jack Kilby, the TI employee who invented the integrated circuit. Some examples of that, in analog is our gallium nitride products. In embedded, it is our millimeter wave products that have come. Multiple years of investments, a high degree of differentiation in products, and certainly things that will contribute to free cash flow.
That's how we think about allocating R&D across the board and specifically in embedded. Do you have a follow-on, Harlan?
Yes. The team always thinks longer term. Assuming a normalized demand environment and above-market growth for the team, you'll probably start to get to kind of max capacity on your 300-millimeter assets in about 5-7 years. How are you guys thinking about 300-millimeter capacity expansion to support the continued margin expansion? Do you guys have land to expand the area footprint for DMOS6 and RFAB?
Good question, Harlan, and we'll finish with this question. Let me tell you how we're thinking about 300-millimeter capacity, and we just gave you all the numbers that you need to do the math you mentioned, right? How much of our 300 we're currently using, how much is left, our analog business. As you project that growth forward, you can figure out how we continue to fill that capacity. We are looking at adding additional, because obviously we want to continue to support the long-term growth of revenue and free cash flow. The preferred option would be to buy a used facility, and we've done that before in multiple places. The reason that's the preferred option is, one, it would be immediately available. Two, generally, we can get that at a good price when you're buying a used facility that somebody else doesn't want.
It comes fully equipped. When you buy a used facility, that's generally a running factory, so it already has a lot of the equipment. Not everything that we need, but a lot of what we need. That is not the only option. The other option is for us to build a new facility. Keep in mind, the cost of the building itself of a new facility, relatively low compared to the overall cost once you include equipment. We could build a new facility, an actual shell, a building, and then we fill it with equipment, and that equipment would be mostly used. Not all, but mostly used.
We're looking at both of those options and have been, depending on various factors, obviously demand growth being a big one, we then decide when it makes sense to pull the trigger on either one of those options. Dave, you want to comment on-
I think that's good.
All right.
If you would like to wrap this up.
Yeah. Just to finish up, I want to thank all of you for taking time today to go through our capital management strategy. Let me just emphasize a few points. We remain focused on consistent execution of our capital management strategy as we have done for many, many years. Our disciplined allocation of R&D is delivering growth from the best products, analog and embedded, and the best market, industrial and automotive. We have great diversity across all the sectors within this market. 300-millimeter manufacturing advantage, as we talked about during the call, is a unique advantage and will continue to benefit TI for a long time to come. We remain committed to returning all free cash flow to the owners of the company. Dave?
Okay. Thank you all for joining us. A replay of this call will be available on our website, as well as the slides we used today. Good day.
This does conclude today's conference. We thank you for your participation. You may now disconnect.