Good day. Thank you for standing by. Welcome to the NXP third quarter 2021 earnings conference call. All lines have been placed on mute to prevent any background noise. After the speaker's remarks, there will be a question and answer session. To ask a question, you will need to press star one on your telephone. If at any time during the conference you need to reach an operator, please press star zero. Now I would like to turn the conference over to Mr. Jeff Palmer, Senior Vice President of Investor Relations. Please go ahead, sir.
Thank you, Dexter, and good morning, everyone. Welcome to the NXP Semiconductors third quarter 2021 earnings call. With me on the call today is Kurt Sievers, NXP's President and CEO, and Bill Betz, our CFO. The call today is being recorded and will be available for replay from our corporate website. Today's call will include forward-looking statements that involve risks and uncertainties that could cause NXP's results to differ materially from management's current expectations. These risks and uncertainties include, but are not limited to, statements regarding the continued impact of the COVID-19 pandemic on our business, the macroeconomic impact on the specific end markets in which we operate, the sale of new and existing products, and our expectations for the financial results for the fourth quarter of 2021. Please be reminded that NXP undertakes no obligation to revise or update publicly any forward-looking statements.
For a full disclosure on forward-looking statements, please refer to our press release. Additionally, we will refer to certain non-GAAP financial measures, which are driven primarily by discrete events that management does not consider to be directly related to NXP's underlying core operating performance. Pursuant to Regulation G, NXP has provided reconciliations to the non-GAAP financial measures to the most directly comparable GAAP measures in our third quarter 2021 earnings press release, which will be furnished to the SEC on Form 8-K and is available on NXP's website in the investor relations section at nxp.com. Before we start the call today, I would like to remind everyone of our upcoming Analyst Day on Thursday, November 11, 2021. We are hosting a hybrid event. We will be in person in New York City, but also simulcasting the event from our website for virtual attendees.
After the event concludes, we will post the slides to the investor relations website. Now I'd like to turn the call over to Kurt.
Thanks very much, Jeff, and good morning, everyone. We appreciate you joining our call this morning. I will read you our quarter three results and then discuss our guidance for quarter four. Overall, our quarter three results were better than the midpoint of our guidance, with the mobile end market stronger than planned as a result of improved supply. At the same time, the trends in the auto, industrial and IoT, and communication infrastructure markets were all in line with our guidance. Taken together, NXP delivered quarter three revenue of $2.86 billion, an increase of 26% year-on-year, and $11 million above the midpoint of our guidance range. These are very good results given the constrained supply position we knew we would face entering the quarter.
We continue to view our channel and on-hand inventory metrics below our long-term targets. We exited quarter three with our distribution channel supply metric at 1.6 months, almost a full month lower than our long-term target, and we expect this to be the situation in quarter four again as well. Our non-GAAP operating margin in quarter three was a strong 33.5%, which is 770 basis points better than the year ago period and 50 basis points above the midpoint of our guidance. Operating profit dollars were $19 million better than guidance, driven by higher revenue and lower spend. Now let me turn to the specific trends in our focus end markets. Starting with automotive. Quarter three revenue was $1.46 billion, 51% up versus the year ago period and in line with our expectations.
In industrial and IoT, quarter three revenue was $607 million, up 18% versus the year-ago period, and again, in line with our expectations. In mobile, quarter three revenue was $345 million, up about 2% versus the year-ago period and above our expectations. Lastly, in communication infrastructure and other, quarter three revenue was $454 million, about flat versus the year-ago period and in line with our expectations. With this, let me move straight to our outlook for quarter four. We expect the midpoint of quarter four revenue to be $3 billion, up 20% versus the fourth quarter of 2020, within a range of up 17% to up 23% year-over-year. From a sequential perspective, this is up 5% at the midpoint versus the prior quarter.
We again anticipate demand outstripping available supply in our quarter four outlook. At the midpoint of this range, we anticipate the following trends in our business. Automotive is expected to be up in the high 20% range year-on-year and up in the mid-single-digit range versus quarter three 2021. Industrial and IoT is expected to be up in the high 20% range year-on-year and up in the mid-single-digit range versus quarter three 2021. Mobile is expected to be down in the mid-teens range year-on-year and up in the low single-digit range versus quarter three 2021. Finally, communication infrastructure and other is expected to be up in the high teens range versus the same period a year ago and up in the low single-digit range versus quarter three 2021.
Over the course of the last two quarters, investors continue to ask how to reconcile the revenue performance of NXP's automotive business with that of the global vehicle production numbers as they are reported by IHS. Specifically, NXP's automotive segment revenue is expected to be up over 40% in 2021. Against this, the auto OEMs continue to struggle to match supply to strong consumer demand, with the auto industry likely not able to meaningfully grow unit production versus 2020. Now let me make a few observations which may help you understand this divergence. First, the auto supply chain is very extended and complex, with multiple points of product transformation across the globe. This extended supply chain needs to coordinate the timing and delivery of up to 30,000 parts and up to 1,500 different semiconductors from hundreds of suppliers to build just one single car.
During normal periods, from the time at which NXP ships a finished component to when the final assembly is fitted into a finished car, it takes up to six months, and this is on top of the normal semiconductor manufacturing cycle times of three to six months. At each step of the transformation, thousands of parts move through a complex global network of suppliers. For the process to work efficiently, it is essential that all of the components needed to complete a car are available exactly where and when they are required. Now, we believe the extended auto supply chain significantly depleted on-hand inventory of all types of products, including semiconductors, already by the beginning in the second half of 2018 and continuing through 2019 and most of 2020.
That depletion was a result of global car production declining 6% in 2019 and another 16% in 2020. While NXP's auto business, despite content increases, declined by 7% in 2019 and by another 9% in 2020. Now let me illustrate the impact by using the NXP distribution channel as a proxy for overall auto supply-demand trends and how we have been directly affected. We consistently monitor and measure all component movements and inventory data at our distribution partners at the end market level. Throughout 2021, these metrics for automotive have been at record low levels, with on-hand inventory being about one month below our long-term target of 2.5 months. With demand in the intermediate term being consistently greater than our ability to rebuild inventory back to normalized levels.
In our and my personal daily discussions with our customers throughout the auto supply chain, we hear the consistent message that they want significantly more product. In some cases, tier ones are struggling to assemble full kits, and in other cases, the OEMs choose to build partially completed cars or halt their production lines altogether. For NXP, lead times for about 75% of our automotive products continue to be above 52 weeks. Against this backdrop, our customers are placing NCNR orders to assure long-term supply. We, in turn, are making long-term supply commitments to our supply partners. In summary, we think the automotive supply demand equation will continue to be out of balance through 2022.
In addition, as a learning out of the current material shortage situation, and in order to mitigate the impacts the auto OEMs are experiencing today, our tier one partners explicitly demand that more supply and inventory will be needed in the extended supply chain, which we believe cannot be broadly achieved before 2023. With this currently dysfunctional supply chain as a backdrop, there are very clear and very positive trends that have simultaneously increased the demand for auto semiconductors industry-wide as a consequence of content growth. We have seen multiple OEMs prioritize the production of premium vehicles, which require upwards of twice the semiconductor content from NXP and others.
Another clear and emerging secular content driver for the auto semiconductor market is the fast acceleration of full electric and hybrid electric vehicles, which combined have moved from 8% of global production in 2019 to about 20% of production in 2021. This is very impactful since the average semiconductor content in an xEV is about $900, which is roughly two times that of an equivalent ICE vehicle. These trends have resulted in industry-wide content per vehicle increasing at 10% per year over the last three years. On top of all of this, NXP is consistently gaining share in our focused growth areas and increasing content. These content gains include 77 GHz radar safety systems, multiple electrification system opportunities beyond just battery management, and new domain and zonal processing, as well as others.
Now for NXP, it's of course not just automotive driving our performance. Within the industrial and IoT market, we see our ability to provide complete turnkey connected edge processing solutions consisting of processors, connectivity, security, and analog, all leading to increased customer traction. These are all just a few examples that underpin our confidence in our company's specific growth. As Jeff mentioned earlier, we plan to go into much greater detail at our Investor Day on November 11, next week in New York. In summary, we continue to execute very well in a strong demand environment, notwithstanding the industry-wide supply challenges. From a company-specific perspective, NXP is experiencing very positive customer traction of our newest products and solutions.
Putting it all together, we are highly confident that the company's specific drivers within our strategic end markets will continue to build also beyond Q4 into Q1, as well as over the intermediate term through 2022. Now before we move to the financial details of the quarter, I'd like to make a few remarks in the context of our recent announcement of Bill Betz as our new CFO. I have personally worked with Bill as a business partner and one of Peter Kelly's key finance leaders for over eight years. Bill brings both a strong track record and career in the semiconductor industry, as well as truly intimate knowledge of NXP and consistency to his new role. I personally drove the evaluation and interview process, interviewing a wide number of external candidates as well as Bill.
I concluded Bill is the right person to lead our NXP finance organization, and I'm personally truly excited to work with Bill and drive NXP's profitable growth going forward. At the same time, I'd like to highlight the outstanding contribution Peter Kelly has played in the strategic evolution of NXP. Peter first came into the company in an operations role over a decade ago, and then quickly moved into the CFO role to drive the financial discipline our stakeholders have all come to expect. He has been a clear-thinking, strategic advisor to myself and a highly valued mentor to many on the NXP management team, obviously including both Bill and Jeff. We wish Peter the very best in the next phase of his life and hope he gets to spend more quality time with his family.
With that, I would now like to pass the call to you, Bill, for a review of our financial performance. Bill?
Thank you, Kurt, and good morning to everyone on today's call. As Kurt has already covered the drivers of the revenue during Q3 and provided our revenue outlook for Q4, I will move to the financial highlights. Overall, our Q3 financial performance was very good. Revenue was above the midpoint of our guidance range, and we drove an improvement of non-GAAP gross profit and non-GAAP operating profit, both of which were at the high end of our guidance range. Now moving to the details of Q3, total revenue was $2.86 billion, up 26% year-on-year and above the midpoint of our guidance range. We generated $1.62 billion in non-GAAP gross profit and reported a non-GAAP gross margin of 56.5%, up 640 basis points year-on-year and near the high end of our guidance.
Total non-GAAP operating expenses were $657 million, up $107 million year-on-year and up $31 million from Q2. This was $8 million below the midpoint of our guidance due to lower material and mass spend during the quarter. From a total operating profit perspective, non-GAAP operating profit was $959 million, and non-GAAP operating margin was 33.5%, up 770 basis points year-on-year. This was also at the high end of our guidance range as a result of better fall through on higher revenues. Non-GAAP interest expense was $94 million, with cash taxes on going operations of $86 million and non-controlling interest was $7 million. Taken together, the below-the-line items were $8 million better than our guidance. Stock-based compensation, which is not included in our non-GAAP earnings, was $81 million.
Now I would like to turn to the changes in our cash and debt. Our total debt at the end of Q3 was $9.59 billion, flat on a sequential basis. Our ending cash position was $2.3 billion, down $607 million sequentially due to higher CapEx and robust capital returns during the quarter. The resulting net debt was $7.29 billion, and we exited the quarter with a trailing twelve-month adjusted EBITDA of $3.92 billion. Our ratio of debt to trailing 12-month adjusted EBITDA at the end of Q3 was 1.9x , and our 12-month adjusted EBITDA interest coverage ratio was 11x . Our liquidity continues to be excellent, and our balance sheet is very strong.
During Q3, we repurchased $1.16 billion of our shares and paid $152 million in cash dividends for a total of $1.31 billion of capital return to our owners. Subsequent to the end of Q3, between October 3 and November 1, we repurchased an additional $300 million of our shares via a 10B5-1 program, resulting in a total of $4 billion return to our owners year to date. Turning to working capital metrics, days of inventory with 85 days, a decline of three days sequentially. Our DIO continues to be below our long-term target of 95 days. We continue to closely manage our distribution channel with inventory in the channel at 1.6 months, flat sequentially and below our long-term targets. Both metrics reflect the continuation of strong customer order rates and a tight supply environment.
We continue to believe it will take multiple quarters before we're able to rebuild on-hand and channel inventories to our long-term target levels. Days receivable were 31 days, down four days sequentially, and days payable were 83, a decline of nine days versus the prior quarter. Taken together, our cash conversion cycle was 33 days, an increase of two days versus last quarter. Cash flow from operations was $924 million, and net CapEx was $200 million, resulting in non-GAAP free cash flow of $724 million. Turning to our expectations for Q4, as Kurt mentioned, we anticipate Q4 revenue to be $3 billion ± $75 million. At the midpoint, this is up 20% year-on-year and up 5% sequentially. We expect non-GAAP gross margin to be about 56.5% ± 50 basis points.
Operating expenses are expected to be about $680 million, ±$10 million, consistent with our long-term model. Taken together, we see non-GAAP operating margin to be about 33.8% at the midpoint. We estimate non-GAAP financial expense to be about $94 million and anticipate cash tax related to ongoing operations to be about $100 million. Non-controlling interest will be about $9 million. For Q4, we suggest for modeling purposes you use an average share count of 270 million shares, which is down 15 million shares from a year ago period as a result of the consistent execution of our communicated capital return policy. Finally, I have a few closing comments I'd like to make.
First, demand trends continue to be strong across our target end markets and customer interest in our newest products continues to be very robust. At the same time, we are closely working with our customers and our suppliers to address order requests in a timely manner. Second, our Q4 guidance reflects the clear potential of our business model, both in terms of revenue growth as well as the significant profit fall through, which will enable us to consistently drive our non-GAAP gross margin above the midpoint of our gross margin targets. Thirdly, our business continues to generate significant free cash flow, and we are committed to our capital return policy and will return all excess free cash flow to our owners, so long as our leverage ratio remains at or below 2x net debt to trailing 12-month adjusted EBITDA.
I'd also like to take this opportunity to thank Peter Kelly for his significant contributions to NXP since he joined the company in 2011. Not only did Peter guide NXP from a financial perspective, but he truly helped lead the company and drive the strategy that made NXP what it is today. On a personal note, I consider Peter a great leader, a mentor, and a friend, and I will always be grateful for his support. Finally, I'd like to thank all my colleagues across NXP for their outstanding work and dedication. We shouldn't forget that we are all still working under strict pandemic protocols. With that, I'd like to turn it back to the operator for questions.
Thank you. As a reminder, to ask a question, you will need to press star one on your telephone. To withdraw your question, press the pound key. Again, press star then the number one to ask a question. Our first question comes from the line of Ross Seymore. Your line is open.
Thanks for letting me ask the question. Bill, congratulations on the new role as CFO. My first question is for Kurt, and it's one on customer behavior. You gave a great deep dive into the apparent disconnect between the auto revs you have in production and all those explanations. I really wanted to get into, are the customers changing their behavior? If so, is it just with longer visibility for you, greater certainty, or is the profitability that all the auto companies are gonna be able to garner from this sector starting to improve as well?
Yeah. Hey, good morning, Ross. Indeed, a couple of changes. First of all, very technically, and I just have to highlight that because it sits on top of my agenda every day in and out. There continues to be no change in pressure to get more product. Expedites, escalation calls, all sorts of weird actions to try to get product faster to their manufacturing locations has not slowed down by an inch. We see that continuing. From a more strategic and, say, learning perspective, I would say clearly the whole auto industry has realized now, and that's for me a step function and a reset at the same time, has realized now how big the significance and importance of semiconductors is for their future today, but also years out.
I mean, that's about innovation, but obviously it's also about being able to to build complete cars. That has led to much closer and better relationships between us and directly the OEMs in terms of understanding both their innovation needs, but also from a more tactical perspective, their product supply needs in the near and midterm. All of that leads to a number of significant, I would say quite significant changes. We are seeing longer-term orders, so we get these so-called NCNR orders, which easily reach out for the entire next calendar year. These are coming from the tier one customers of us, but of course, they do this since they get similar behavior from the OEMs.
We also get greater visibility into the longer term, which I find a very important perspective, which reaches then over to innovation because the intimacy which we can build with the OEMs now to think about future systems is a great help for us to construct the right roadmaps.
Great. Thank you for that color. I guess as my follow-up, just shifting gears quickly to a little bit of a nearer term, potentially question, and that's on the mobile business. It's good to see that it upside surprised in the third quarter versus your original guidance. I guess if I put the two quarters together, I know supply is tight and maybe you're just doing some allocation, but the mobile business is a bit weaker seasonally than we would have expected in the combination of the quarters. I had thought that was supposed to snap back a little bit in the third or in the fourth quarter, excuse me. Can you just talk about what's driving the relative weakness on a year-over-year basis in that business?
Well, it's clearly supply or the lack thereof, Ross. When we went into quarter three, I think we talked about a sequential decline, which we could avoid. We could pull in some supply, and that's the whole reason why we actually ended up in mobile better than guidance. There is no difference in sockets or anything. It was just that we had a somewhat better supply capability, and we continue to forecast now a similar thing into the fourth quarter. Overall, if you take the full second half of this year, we are unfortunately significantly constrained in supply into the mobile market. It is gradually getting better, which is why we have this performance I just quoted.
If you hold it just in total, it is pretty significantly suffering. Now, from a year-on-year perspective, Ross, it's a more difficult compare. Last year, Q3, we had the ban on Huawei coming into play, which artificially bumped up quarter three. Quarter four, one of our very large mobile customers needed a lot of product, which also had to do with the phasing of the quarters last year, which was later than normal. That's why the year-on-year comparisons, both for Q3 and Q4, need some bridging, given those two specific customer events.
All in all, I just wanna absolutely reconfirm we haven't lost a single socket there, so everything is going as we had planned. We are supply constrained, but the bigger drivers being the adoption of mobile wallet attach rates and also the early penetration of ultra-wideband secure ultra-wideband solutions are fully on track with earlier targets and earlier planning.
Thank you.
Okay. Next, we have Vivek Arya. Your line is open.
Thank you for taking my question, and good luck to both Peter and Bill. Kurt, on the automotive industry, if one had, you know, told you that auto production would be flat this year, but your auto sales would have been up 40%, would that have been predictable? If not, what do you think surprised NXP other than, you know, some of the bad weather and the fab shutdown effects? Was it mix that surprised you? Was it pricing? Was it content? What specifically, you know, was the surprise this year to create this big gap between production and sales?
Yeah. Hi, Vivek. I mean, very transparently and honestly, with the knowledge of today, of course, we could have predicted it. With the knowledge at the time, no. No, we didn't know. Now what you call surprises is what actually happened. A few things. One, clearly the mix change to premium vehicles, which is an optimization to higher profit on the OEM side, clearly is something we had not foreseen. I think nobody had anticipated that. That has a significant impact on semiconductor consumption, given that premium vehicles can easily have twice the semi content of a mass volume car.
On the same note, again, the xEV penetration, which I think is a result of the big focus in the world on CO2 targets, but also a lot of government subsidies in various countries to boost actually more electric vehicles. Those two elements, the mix to premium and the much stronger penetration or faster penetration of xEVs, those have significant positive impacts on the content increase for semiconductors. Even if the car production is flat, the content goes up massively. Again, think about that both of these elements, each one on its own right, are potentially bringing the semi content of a car to a factor of 2x. That's—I mean, this is just outstripping any car production by far.
The third element, which I would say, I use your language, surprised us, I would rather say we gained more insight, is the enormous depth and complexity of the extended automotive supply chain. Which means how many stages and companies sometimes are between us after we ship a product before that product ends up in a car. If that supply chain is enormously depleted, as it has been the case after 2019 and 2020, it becomes incredibly dysfunctional. We are still in the process of refilling that supply chain just to normalize levels. Really important, this is not about building any strategic inventory, it's just getting back to more normalized levels which allow for proper operation of that supply chain. That's that we didn't see either.
I would say we had no idea how far it has been depleted, how deep it has been depleted, nor how deep it is, which means how much elasticity is in there to be refilled in the first place to try to make it functional. Again, it is still dysfunctional at this stage.
Got it. For my follow-up, Kurt, just to kind of continue that line of discussion. You mentioned over the last few years that the content delta right between units and sales have been, for the industry, has been about 10 points with NXP doing better. If I look at the current IHS production forecast for next year, and I know that they keep on changing all the time, but right now it's about 10%-11% unit growth. Assuming that happens, what is the crystal ball saying in terms of that rough delta for the industry next year? Do you think it'll be that at 10 points, or it'll be closer right to the much bigger number we saw this year? Thank you.
Yeah, Vivek. First of all, you just mentioned another very important element, which is obviously our performance in all of this. I mean, what I described to your earlier question was how we think the industry surprised us all. The third element clearly is the NXP specific company specific content gains, where we grew share, which makes it even faster. Now, how that all goes into what it means for the next year and the next years from both a content growth perspective as well as a NXP specific revenue growth perspective, I just ask you to wait for a week. Next week when we have our Investor Day in New York, we will give you the detail you're looking for.
We will guide the next three years, both how we think the semi market and auto is gonna develop, as well as our performance relative to that. Please, hold your horses until next week.
Thank you.
We have a question from John Pitzer. Your line is open.
Yeah, good morning, guys. Thanks for letting me ask the questions, and congratulations on the solid results. Kurt, I think we all appreciate all the details you gave in your opening comments about the complexity around the auto supply chain. I'm kinda curious if you could spend a couple minutes just talking about your ability to grow supply from here, both internally, externally, both on the wafer side and on the back-end test. I'm assuming that the nice sequential jump in autos means that some of the weather events in the first half of the year have reversed themselves. But specifically, as you look over the next several quarters, would you expect your ability to grow supply to be relatively linear, or is it gonna be chunky, and is there a point in time where your ability to grow supply kind of accelerates in 2022?
Yeah, thanks, John. Let me give you a few pieces on this. The one is indeed the negative impact, which we had some, I think, early Q2 in Texas, is indeed behind us. I mean, to just put this to rest, both factories which were impacted are running very much at pre-storm levels, and we are good. Another element I have to mention here because it has impact on the industry. You've seen that over summer, there have been quite a few factory shutdowns of back-end testing assembly sites in Malaysia. I just wanna highlight that we have been in a very disciplined position to organize and manage our own factories in such a way that our shutdowns were very immaterial.
We had a few days, but by and large, we had very little negative impact from those shutdowns on our revenue. When you think about what we shipped in Q3, assume there was hardly any negative impact from COVID factory shutdowns from Southeast Asia and/or specifically Malaysia. Now going forward, we will clearly see improvements on the wafer side. We talked about earlier that we are entering into long-term supply commitments with our supply partners, and that's largely on the wafer side, which is gonna start benefiting us next year, but also the years after.
Again, this is in balance against the NCNR orders which we are collecting from our customers. The back end, test and assembly, which largely is in-house for NXP, we of course put all the CapEx and all the investments in place to make sure that any wafer we get our hands around will also find enough capacity in-house, to be tested and assembled into finished product. Think about the test and assembly increase in capacity as pretty gradual because it is something which we are putting in tune with the wafers coming in. On the wafer side, it is in a way more discrete, but it's not chunky, John, because we have obviously several wafer suppliers, and they don't improve all at the same time.
From that perspective, what comes out into the revenue, think about something which is gradual, which doesn't have huge step functions at any specific point in time. However, the result of all of this is that clearly our supply capability, as it has done through this year, will continue to grow also through next year. Now, finally, I should say this is clearly not a comment which is in any way or form specific to automotive. This is a broad comment because the supply shortages which we are facing is across all markets. We discussed earlier briefly mobile. Everybody knows about automotive, but we have the same negative impact on our industrial IoT business and partially in the comms infra business.
All of these improvements will help and will support revenue growth across all of our end market segments.
That's really helpful color, Kurt. Then maybe for my follow on, one for Bill. First, congratulations. Looking forward to working with you in your new role. I'm just kind of curious on the gross margin line. Another really solid gross margin quarter. Given some of the logistics inflationary costs out there, oftentimes it's hard to, within a quarter to get all of that right relative to pricing. Was there anything holding back gross margins from a cost side in the quarter? I guess more importantly, as you look out across these LTAs, can you help us understand, you know, how they're being constructed relative to kind of your gross margin goals? I would assume it's much easier to price to value in this sort of environment. Does that mean, you know, we could see some tick-ups here in gross margins over the next several quarters?
Yes. Thank you for your question. Let me give a bigger picture on the margin and more color about it. So as mentioned on our previous calls, our gross margin had improved drastically versus last year, driven by the higher volumes and improved internal utilizations from our factories. For example, if I try to remember, I think Q3 utilizations last year were in the mid-60s%, and today we're in the high 90s%. From a guidance standpoint, as you saw, we were slightly better and from a quarter-over-quarter basis by 20-30 basis points related to the improved product mix. Now as we move forward over the next several quarters, it's gonna be really difficult to have that additional fall through on our internal factories as basically they're running full out, I'd say.
To address your question on pricing, we are only passing on the increases of our input costs from our suppliers to our customers, and we're not padding any of our margins in this process as we value our long-term relationships with our customers, I'd say. We're really equally sharing the pain together. As you can see in Q4, we are guiding our gross margins to be flat quarter-over-quarter, and any additional margin expansion, I'd say, will come in the short term, will be driven by that continued product mix. And more longer term will be really driven by the expansion of our new NPIs. I'd say overall, we're very proud of achieving these levels, and we look forward to continue delivering, you know, toward that higher end financial state of model 57% as we go forward.
Thanks, guys.
Your next question is from Stacy Rasgon. Your line is open.
Hi, guys. Thanks for taking my question. I wanted to follow up on that pricing question. I know you are not trying to extract more, but I mean, to play devil's advocate a little bit, like given the shortages out there, why not? I mean, you're in a period of time right now where the supply chain is probably tighter than it ever has been and ever will be, and there's demand for your products that's off the chart. You're locking customers in, like at this point, like why is this not the time, even on a selective basis, to try to extract more, especially given it seems like some of your peers actually are going down that path? Like why not you guys?
Yeah. Hi, Stacy. There is a very clear answer to this, which is we are not in the commodity business. The very, very vast majority of our portfolio is application specific, and with that portfolio, we are in very deep longer-term relationships with our customers. What we do is, as Bill said, we pass on the input cost, which is already, in some cases, a pretty significant element to digest for our customers. We want to be transparent, and we want to have long-term relationships which are built on trust with our customers. In that context, we absolutely consider it the right thing to pass on the input costs, but not use it to pad on our margins.
I would personally say from being in different markets in the past, in a commodity market, that works different, but that's not the kind of market environment we are moving in. I'm very sure that this will pay off positively in the long term relative to the customer relations which we need to continue to build our business. In that context, it is also important to note that the whole environment clearly has a lot of input costs which is going up for our customers. I mean, it's not just on semis. A lot of other things are also moving. It's a tough environment for all the participants, where I think consistency, transparency, and trust into the long-term relationships is a big value, and that's something where we do differentiate.
Got it. Thank you. I guess for my follow-up even to follow up on that a little bit, you know, the historical pricing practices, at least at a high level, especially in automotive, was to have long-term price downs that were sort of built in, built into the contracts, and you'd be sort of fighting that every year. I guess understanding that we're, you know, maybe coming off a higher base now, at least on raising price because of the input cost increases, do you foresee like a similar type of long-term kind of pricing behavior that are built into the long-term contracts? Are you still gonna be fighting, you know, a couple of hundred basis points of margin compression every year just off the higher base, like as part of these long-term contracts?
Is the general pricing structure like long-term of these contracts different than it was in the past?
Yes, Stacy, that's indeed the other side of this discussion. I mean, I can't discuss any customer or area segment-specific trends here, but the overall situation indeed is a reset, and that is important and positive. With reset, what I mean is that the step function which we have to take is something which is here to last. That comes with, I think I made the comment earlier, with the realization of all the industries we are serving, that's not just for automotive, of how important semiconductors are. It just represents a different value to them. From that perspective, I think we make a step here, which is a reset, which is here to stay. How that goes then in individual cases on individual prices year on year is hard to anticipate, but I don't think it will go backwards.
Got it. Thank you.
Your next question is from William Stein. Your line is open.
Great. Thanks for taking my questions. First, I think you just spoke about this a moment ago, but can you talk about the magnitude and duration of your purchase commitments to Foundry and your customer purchase commitments to NXP?
Well, Bill will fill in maybe with a bit more detail, but I'd say in general, it really varies. I mean, we have some of them on the NCNR side, which is the customer orders. They would typically easily cover next year. Now getting an order which stretches all the way until the end of next year. If you think then on the other side, which is the agreements with our foundry and supply partners, I think we actually published a new number, which if I remember right, is between $4.3 billion-$4.4 billion of agreed and signed supply commitments, which we have entered into with our supplier base. The vast majority of that is obviously in wafers. Mind you, this is not just for one year. This stretches out over multiple years to come. $4.4 billion to give you a feel.
That helps. Thank you. Bill, I wanna offer some congratulations in your new role. Of course, you have some pretty big shoes to fill, given Peter's very strong performance, and the company's strong performance in the last 10 years or so. I'm hoping you might give us a preview of what you might say next week, at least in terms of your priorities as the new CFO. Thank you.
Yeah. Thank you for your question. First off, I'm very honored and thankful for the role. I've been in the semi industry for about 20 years. Prior to NXP, I've worked with Fairchild, LSI, and Agere Systems. Really no change as you have to remember, I've been with NXP for about eight years supporting the company's financial strategy alongside with Peter, Kurt, and Jeff and the entire management team. I'd say my style is a bit different than Peter's, but our principles are the same focus on delivering, like you mentioned, those commitments and driving long-term value to our customers and employees and shareholders. Related to Analyst Day, yes, I'll share that next week, related to our long-term financial model. We really like to save that for next week and not address that in this call.
We have a question from Chris Caso. Your line is open.
Yeah. Thank you good morning. First question one if i could just ask about those long terms agreements you are signing with you customers now. Could tell us what role of commitments have your customers made you under those agreements. It sounds like you are using those agreements near customers to back stab agreements that you are making with your suppliers is that worth specific volumes of commitment made over certain amount of time and is pricing commited to in those agreements as well such that you have agreement on both time and pricing
Since I really can't go into great detail on this, as you will understand, Chris. But the headline is indeed in line with what you are saying. Those customer agreements would typically cover volume and price, say, for example, a period until end of next calendar year. They even fix it by quarter. It's not just one number for the whole year, but in many cases, it even goes to binding it to quarters.
Very helpful. If I could follow up pricing as well. You know, clearly you know, over the past couple of quarters, you've been passing along, as you said, those price increases you've gotten from suppliers. Can you give us a sense of the magnitude of the pricing benefit you've seen out of the sort of 26% year-on-year increase you've seen in revenue? How substantial has price been in that calculation? Then as you go forward, you know, are you expecting as you go into next year, there still to be a pricing tailwind as the input costs rise?
Well, it generally, as we discussed before, we will continue to match the rising input cost with adjusting the prices accordingly. Such that, yes, since input costs will continue to go up also next year, there will also continue to be a pricing adjustment into next year, in general terms. For the rest of your question, I really cannot and don't want to go into greater detail in how much is what piece. It's really just important that the leading concept and philosophy of all of this is that the price increase is just matching the input cost increase, and that's it. That is something which indeed we have this year and which is gonna follow us next year too.
Thank you.
Next question is from Blayne Curtis. Your line is open.
Hey, thanks for taking my question, and I'll offer my congrats to Peter and Bill as well. Just going back on the auto side, I just wanted to be clear. You've heard from some other companies that they are seeing their customers start being more selective, trying to get kits together. Thanks for all the detail, Kurt, on that supply chain. Seems very complex. I'm just kinda curious, are you seeing that same behavior and maybe seeing offsetting strength elsewhere? Or are you a bit different given the company specific, you know, the weather in Texas, etc , and some of the secular growth that you're not seeing that selective behavior on certain parts? Thanks.
Blayne, I think what you tried to ask me is, am I seeing a slowdown? The clear answer is no, we absolutely don't see a slowdown. You can imagine that especially the car companies with a pretty significantly negative impact to their car production numbers of this year and the ambition to grow them next year by, I don't know, 10%+ over this year, they clearly need more semiconductors. That is not limited to one product or one technology. I think I also made the very open comment earlier that 75% of the product portfolio in our automotive business still has a lead time of above 52 weeks. I think that says it all.
That's an amazing stat. Thanks. No one asked on the industrial IoT, maybe just kind of discussion about what the areas of strength are for you and maybe you could describe the supply chain as well for the back half of the year here. Is it just as tight? That'd be helpful. Thanks.
Yeah, Blayne. It is very tight too. I would say no different to automotive. Also, by the way, the impact of not shipping enough product is at least as significant as in automotive. It just catches less headlines in newspapers. If you think about the impact on industrial automation, which is huge machinery, which cannot be built if there is not enough semiconductor supply, or in the smart home, it's equally drastic as it is in automotive, just catching less headlines. Now, our big differentiator, more and more, is our capability to offer complete solutions around the processor.
We've had that leadership in a very wide range and portfolio of processing solutions into industrial, but it's now very nicely complemented by connectivity, security, and analog attach, which really differentiates us from our competitors. I would dare to say it's a very unrivaled position and offering here, which, by the way, is also gonna be very much in the spotlight of our Investor Day next week to give you much more detail, what we have and why we think it makes us grow so much in the time to come. Now from a supply chain or channel perspective, a lot of that business indeed is going through distribution, which is because it goes to thousands, literally thousands of small customers.
This is exactly why the solution capability of NXP makes such a big difference. Because many of these companies which we are serving there through distribution don't have the capability themselves to deal with the complexity of these connected edge applications. When and if we can offer them a complete solution, then that has a huge advantage for both time to market, but also from an R&D perspective for them. That's actually what I think is really a very significant differentiator for us. Now, another element in the equation is we do have a relatively strong exposure to China. If you think about the industrial business, a lot is going through distribution and a lot is going into China. Just to anticipate the question, we do not see a slowdown there in China.
Again, mind you, it's also that we continue to be supply capped. As I said earlier, also in industrial, we could certainly run higher revenues if only we had more supply. Good news is also there, our supply capability, as we discussed with John earlier, is improving every month, so that over time also there, we hopefully get into a better place.
Okay.
Okay, our last question from-
I guess we'll take one last call.
Okay. Our last question is from C.J. Muse. Your line is open.
Yeah, good morning. Yeah, thanks for squeezing me in. I guess first question for you, Kurt. In your prepared remarks, you talked about auto supply demand balance probably not achieved until 2023. Curious, as you contemplate that, what kind of assumptions are you making in terms of what kinda the future inventory management will look like, you know, both at the tier ones and OEMs? I imagine there's more of a just in case as opposed to just in time. Would love to hear your thoughts on kind of that evolution that is likely coming ahead for the auto industry.
Yes, C.J., absolutely. I did indeed say that I don't think there is, broadly speaking, enough supply next year to already start building those additional inventories, because that's exactly what you're asking for. Yes, we clearly see a demand by the OEMs to the tier ones and other participants in the extended supply chain, explicitly not the semi companies, but in between us and the OEMs, a demand and a requirement to build more inventory. A typical, if you wanna size this, a typical ask is somewhere between three and six months, sometimes longer. It has a very simple reasoning. That's the manufacturing cycle time for a semiconductor product.
In a very simple way, they say, "Okay, if this is the additional inventory, then that puts us maybe on the safe side for more flexibility." Yeah, this is something which I don't see the industry has the capability to build that next year, but that's them for 2023.
Very helpful. I guess my follow-up question, you know, obviously we've seen positive mix shift on the EV side, which you spoke to earlier. I'm curious, you know, as we're seeing more and more OEMs target kind of full platforms moving over to EV or hybrid as well, I'm curious how that has kind of played with your full solution, you know, which kinda I would think would fit nicely into that. Would love to hear your thoughts there.
We love it, C.J. I mean, we talked a lot about our capability actually also from a solution perspective in battery management solutions, which is greatly benefiting from this faster and accelerated penetration. We will open the kimono next week a little bit more in the Investor Day and give you insight in what else we do in electrification, because we are not only with BMS exposed to this accelerated trend. It's a very good thing for NXP, and it goes beyond battery management. While battery management itself obviously is benefiting very greatly. Kurt, I guess with this we come to the end of the call, which maybe gets me in a position just to summarize very briefly.
I mean, it's a turbulent time out there, but I'm really proud that we could deliver what I think is a very good quarter three, have a strong guidance into quarter four. Even more so, we continue to have excellent momentum also into next year, which means continued strong demand across our end markets, no inventory build out there, and we see that our supply capability is getting better and better and better to start matching that demand in the future. With that, I thank you very much for your attendance today, and I hope to hear and see quite a few of you also next week when we have our much more extended Investor Day in New York. Thank you.
Ladies and gentlemen, this concludes today's conference call. Thank you for participating. You may now disconnect.