Good day, everyone. Welcome to the Microchip Technology third quarter and fiscal year 2018 financial results conference call. As a reminder, today's call is being recorded. At this time, I would like to turn the call over to Microchip's Chief Financial Officer, Eric Bjornholt. Please go ahead.
Good afternoon, everyone. During the course of this conference call, we'll be making projections and other forward-looking statements regarding future events or the future financial performance of the company. We wish to caution you that such statements are predictions and that actual events or results may differ materially. We refer you to our press releases of today, as well as our recent filings with the SEC that identify important risk factors that may impact Microchip's business and results of operations. In attendance with me today are Steve Sanghi, Microchip's Chairman and CEO, and Ganesh Moorthy, Microchip's President and COO. I will comment on our third quarter fiscal year 2018 financial performance. Steve and Ganesh will give their comments on the results and discuss the current business environment, as well as our guidance. We will be available to respond to specific investor and analyst questions.
I want to remind you that we are including information in our press release in this conference call on various GAAP and non-GAAP measures. We have posted a full GAAP to non-GAAP reconciliation on the investor relations page of our website at www.microchip.com, which we believe you will find useful when comparing GAAP and non-GAAP results. I will now go through some of the operating results, including net sales, gross margin, and operating expenses. I will be referring to these results on a non-GAAP basis prior to the effects of our acquisition activities and share-based compensation. Net sales in the December quarter were $994.2 million, modestly higher than the midpoint of our guidance and down 1.8% sequentially from net sales of $1.012 billion in the immediately preceding quarter. We have posted a summary of our revenue by product line and geography on our website for your reference.
On a non-GAAP basis, gross margins were at the high end of our guidance range of 61.4% in the December quarter. Non-GAAP operating expenses were at 22% of sales, which was a record low and below the low end of our guidance range, which was 22.2%. Non-GAAP operating income was a record $391.7 million at 39.4%, above the high end of our guidance of 39.2%. Non-GAAP net income was $341.2 million and was up 38.4% as compared to the same quarter last year. Non-GAAP earnings per diluted share was $1.36, which was above the midpoint of our guidance. On a GAAP basis, gross margins, including share-based compensation and acquisition-related expenses, were a record 61.1% in the December quarter. GAAP gross margins include the impact of $3.5 million of share-based compensation.
Total operating expenses were $361.8 million and include acquisition and tangible amortization of $121 million, share-based compensation of $20.5 million, $1.2 million of acquisition-related and other costs, and special charges of $0.2 million. GAAP net income was impacted by a $439.8 million discrete tax event in the quarter, primarily associated with the Tax Cuts and Jobs Act, and we also incurred a loss of $2.1 million on the retirement of our 2037 2.8% convertible bonds. After these adjustments, the GAAP net loss was $251.1 million, or $1.7 per diluted share. The non-GAAP tax rate was 8.5% in the December quarter, and we expect the rate to be about the same in the March quarter.
Due to the Tax Cuts and Jobs Act, we expect about $300 million of taxes to be paid over the next 8 years related to the tax incurred on foreign earnings that were permanently invested offshore, referred to as the transition tax. We estimate the tax payments to be approximately $25 million in years 1 through 5, $45 million in year six, $60 million in year seven, and $75 million in year eight. We will continue to evaluate the impact from the Tax Cuts and Jobs Act, and these estimates may change as we complete our analysis. We are working through the impact of recent U.S. tax reform on Microchip's longer-term effective tax rate. There is no impact on the fiscal year 2018 rate, as you can see from our March quarter guidance provided in today's earnings release.
Excluding the transition tax from the Tax Cuts and Jobs Act, we expect our ongoing long-term cash tax rate to be below 9%, which is what we are providing to investors and analysts for cash flow and operating model forecasting purposes. We will continue to refine this over the coming months prior to providing guidance for Q1 of fiscal year 2019. Moving on to the balance sheet. Our inventory balance at December 31st, 2017, was $487.1 million. Microchip had 115 days of inventory at the end of the December quarter. Our targeted inventory levels are between 115 and 120 days. Inventory at our distributors in the December quarter were at 34 days and up three days from the prior quarter, and in the normal historical range for distribution inventory. The cash flow from operating activities was a record $365 million in the December quarter.
As of December 31st, the consolidated cash and total investment position was $1.985 billion, of which about $550 million is domestic cash. Due to the Tax Cuts and Jobs Act, the majority of our offshore cash could be brought back to the U.S. without incurring any material additional tax cost. We bought back the remaining amount of our 2037 2.8% convertible bonds in the December quarter for $58.3 million in cash. Capital expenditures were $66.4 million in the December 2017 quarter. We expect about $50 million-$60 million in capital spending in the March quarter, and overall capital expenditures for fiscal year 2018 to be about $200 million-$210 million, which is at the lower end of the guidance that we provided last quarter.
We are aggressively adding capital to support the growth of our production capabilities for our fast-growing new products and technologies, to bring in-house more of the assembly and test operations that are currently outsourced. These capital investments will bring significant gross margin improvements to our business, particularly for the Atmel manufacturing activities that we are bringing into our own factories. As mentioned last quarter, our capital spending in fiscal 2018 also reflects three new building projects we are constructing in Arizona, India, and Germany, which will give us meaningful lease cost reductions and avoidance in the future, as well as allow us to cost effectively scale our future growth. Depreciation expense in the December quarter was $32 million. We also want to provide an update on the tax treatment of our dividends.
In fiscal year 2017, our dividends paid to shareholders were treated as return of capital, we were expecting similar treatment in fiscal year 2018. The Tax Cuts and Jobs Act, which passed in December 2017, has changed this outlook. Due to the U.S. taxation in fiscal year 2018 of our foreign earnings, which were historically permanently invested offshore, Microchip is taxable in the U.S. in fiscal year 2018, as a result, dividends in the June, September, and December 2017 quarters are fully taxable to our shareholders, dividends to be paid in March 2018 will be fully taxable to our shareholders. We believe the impact of U.S. tax reform on Microchip's U.S. taxable income in future periods will result in dividends being fully taxable to shareholders in fiscal year 2019 and beyond.
We will continue to provide updates to shareholders on this topic in the future if facts and circumstances change. I will now ask Ganesh to give his comments on the performance of the business in the December quarter. Ganesh?
Thank you, Eric, good afternoon, everyone. We performed better than we expected in the seasonally slow December quarter, with a sequential revenue decline of 1.8% and year-over-year growth of 12.8%, all organic growth, as there was no contribution from acquisitions in the last four quarters. The Microchip 2.0 transformation continues to make good progress, especially in terms of new design opportunities as we enable our clients' innovation with the very best smart, connected, and secure solutions. Taking a closer look at microcontrollers, our microcontroller business performed well in the December quarter, with revenue down sequentially only 0.6% as compared to the September quarter. On a year-over-year basis, the December quarter microcontroller revenue was up a very strong 18.9%. Eight-bit microcontroller revenue actually set a new record again, edging out the September quarter performance, while 16-bit and 32-bit revenue continued to have strong year-over-year performances.
Our microcontroller portfolio and roadmap has never been stronger, and we are seeing continued growth in our design-in funnel, which we expect will drive future growth as these designs progress into production over time. Microcontrollers represented 66.5% of Microchip's overall revenue in the December quarter. Now moving to our analog business. Our analog revenue was sequentially down 3.2% in the December quarter as compared to the September quarter, and up 1.5% on a year-over-year basis. As we highlighted at our last conference call, our publicly reported analog results continue to see some revenue classification headwinds resulting from a deliberate shift in strategy we made several quarters ago. The change in product line strategy, as we reported last quarter, is that we have, for some time, been adding microcontroller cores to several of our more complex analog products, especially those that provide our clients with smart connectivity solutions.
The addition of microcontroller cores to these analog products enables us to sell a higher value and more defensible total system solution. These smart connectivity products are growing nicely, and as they replace older products in new designs, our revenue classification for these new products has shifted from the analog product line to our microcontroller product line. Transitioning to more sticky and higher margin smart connectivity revenue is the right Microchip answer, but does impact some of the product line reporting that analysts are interested in as some of the revenue growth shifts into our microcontroller product line. In the longer term, as the revenue from analog attached design wins continues to ramp, we fully expect that the analog product line revenue will grow nicely. Our analog products represented 23.3% of Microchip's overall revenue in the December quarter. Moving next to our licensing business.
This business was up 6.8% sequentially in the December quarter, and up 15.5% on a year-over-year basis, setting a new record in the process. We are seeing the fruits of having licensed several foundries and independent device makers for several years on multiple process technology nodes as they manifest in our results as the licensed processes ramp volume and generate royalty revenue for many years to come. Finally, our memory business was sequentially down 7.6% in the December quarter as compared to the September quarter, and up 3.4% on a year-over-year basis. Let me now pass it to Steve for some general comments about our business and our guidance going forward. Steve?
Thank you, Ganesh, and good afternoon, everyone. Today, I would like to first reflect on the results of the fiscal third quarter of 2018. I will provide guidance for the fiscal fourth quarter of 2018. Our December quarter financial results were good. Our net sales were slightly above the midpoint of our guidance, which itself was slightly better than five years of seasonality. Our non-GAAP gross margin of 61.4% of sales was at the high end of our guidance. Operating profit of 39.4% of sales made a new all-time record, and earnings per share of $1.36 exceeded the midpoint of our guidance by $0.01 per share. I want to thank all the employees of Microchip worldwide for delivering an outstanding quarter in calendar year 2017. On non-GAAP basis, this was also our 109th consecutive profitable quarter. Now I will provide you with an update on Microchip 2.0.
We are continuing to experience an enormous customer preference to design with our microcontroller solutions in all 8-bit, 16-bit, and 32-bit customer applications. On top of that, our various acquisitions have now built a powerful, diversified product line through which we are able to provide total system solutions to our customers. We are winning incremental design wins with multiple products in the same boards of our customers. We have a robust design win funnel, and we feel very optimistic that Microchip 2.0 is working and increasing the organic growth of Microchip. Based on SIA numbers, our microcontroller market share in calendar 2017 was at a record 15.7%, up 220 basis points from calendar 2016. To be fair, calendar year 2016 only had three quarters of Atmel, while 2017 had all four quarters of Atmel.
I calculated market share for three quarters of 2016 to three quarters of 2017, which will be April 1 to December 31 of both years. Based on that nine-month comparison, our market share gain was 91 basis points higher than 2016. Also, our microcontroller market share grew in each of 8-bit, 16-bit, and 32-bit product lines. Let us go into the guidance for March quarter. Our inventories at Microchip at the end of December 2017 were 115 days, which is right at our target. We were able to get to this target a quarter or so ahead of what we were thinking. This is due to enormous progress we made on the manufacturing side and bringing capacity online and decreasing lead times.
Our book-to-bill ratio has also continued to moderate from 1.11 in both March 2017 and June 2017 quarters. The book-to-bill ratio was 1.05 in September 2017 quarter, 1.0 in December 2017 quarter. This moderation of book-to-bill reflects the improvements in lead times that we have seen, or we have been able to achieve over the last several quarters. We have indicated before, our seasonality for any given quarter will change as we integrate acquisitions and end up with a new blended seasonality. The seasonality of the Atmel business is driven by a larger percentage of consumer-oriented business, which has a sharper decline in the March quarter. In fact, the historic Atmel seasonality in the March quarter over the five years prior to the acquisition was a 7.8% sequential decline when they were an independent company.
Microchip's historic seasonality in the March quarter in the five years prior to acquiring Atmel was a 2.75% sequential growth. The blended average March seasonality results in a minus 1% sequential result, so 1% sequential decline. You can also see the contribution from Atmel's consumer business in the stronger-than-seasonal blended results Microchip experienced in the June, September, and December quarters of 2017. We will compare our March quarter guidance to this combined seasonality of new Microchip. Let's go into the non-GAAP guidance for the March quarter. We expect total net sales in the March quarter to be down 3% to up 1% sequentially, giving us a midpoint of the guidance at minus 1%, which is right at seasonality that I explained above. In September quarter last year, our net sales were up 15.8% from a year-ago quarter.
In December quarter, our net sales were 12.8% from a year-ago quarter. In March 2018 quarter, we are guiding the net sales to be up 9% from a year-ago quarter, clearly demonstrating the soft landing scenario that we have been talking to the investors. Our total fiscal year 2018 sales ending March 31, 2018, with midpoint of the guidance, is expected to be up 13.2% from fiscal year 2017 sales, beating most of our competitors for the same time window. Regarding gross margin, we see a steady improvement in overall gross margin of the company based on Microchip 2.0 margin drivers. We expect gross margin for the March quarter to be between 61.3% and 61.7% of sales. We expect overall operating expenses to be between 22% and 22.4% of sales.
We expect operating profit percentage to be between 38.9% and 39.7% of sales. We expect earnings per share to be between $1.30 and $1.39 per share. I want to remind investors that our long-term financial model is a non-GAAP operating profit of 40%. You have seen that we are relentlessly marching towards this model. Given all the complications of accounting for the acquisitions, including amortization of intangibles, restructuring charges, and inventory write-up on acquisitions, Microchip will continue to provide guidance and track its results on non-GAAP basis. We believe that non-GAAP results provide more meaningful comparison to prior quarters. We request that the analysts continue to report their non-GAAP estimates to First Call. With this, operator, will you please poll for questions?
Certainly. If you would like to ask a question, please press star one on your telephone keypad. If you're using a speakerphone, it may be necessary to mute the handset before making your selection. Once again, please press star one and please allow yourself one question per turn. We'll take our first question from Craig Hettenbach with Morgan Stanley.
Yes, thank you. Steve, thanks for the color in terms of book-to-bill and lead times. When you think through the March quarter seasonality with Atmel and Microchip combined, can you help us as you go toward June in terms of what you typically see for the June quarter?
Well, the June and September were stronger quarters for Atmel as well as for Microchip. You have seen pretty strong June and September quarters for the last couple of years since we have had Atmel in both quarters. You've seen the December quarter slightly stronger than our historic seasonality, as we just reported. What that results is a significant difference in the March quarter seasonality, where Microchip was historically up for 2.5%-2.75%, Atmel was down a whopping 7.8%. That's really, when you combine it together, that's what results into a minus 1% seasonality.
Got it. Then maybe just a follow-up for Eric. As you approach the target model and given the margin expansion you've seen, can you talk about just some of the initiatives this year and maybe on the intermediate term for both gross margins and op margins?
On the gross margin front, it's more of the same of what we've been talking about over the past year in terms of investing in the assembly and test and bringing more of that in-house. We're making progress on that front. We've deployed a lot of capital over the last couple of quarters. We have more coming in this quarter, as we talked about in my prepared remarks on the $50 million-$60 million of capital coming in. That will continue to add to the gross margin improvement over time. We've got good utilization in our wafer fabs, all those things are making our cost structure very favorable.
At a midpoint of guidance here of about 61.5% gross margin, we've got about a 1% improvement to go to get to the target model. We're confident that we're going to get there with all these initiatives that we have.
Got it. Thanks.
Same thing on the operating margin. We just reported 39.4%. Our long-term model is 40%. We're fairly confident that most of that gross margin will fall through and we're going to be in that range.
Got it.
We'll go next to William Stein with SunTrust.
Great, thanks for taking my question. I'm wondering if the inventory build is something that you had expected at the start of the quarter, or if that occurred more as a result of perhaps orders not coming in with turn rates as fast as you might have expected when the quarter started.
The inventory was pretty much right on target, what we had guided to and what we achieved. The turns rate in December quarter was not soft. We beat the December quarter from the midpoint of our guidance. I don't think there was any issue in turns or bookings in December quarter, which will have changed the inventory.
Yeah, you can see from the guidance that we had on our release, we think we're going to continue to be within our inventory target range of 115-120 days in the quarter that we're in right now. I think inventory is well managed.
That helps. One follow-up, if I can. You might have mentioned this, I apologize if I missed it, you've been talking about supply constraints on the Atmel portfolio, in particular in backend tests, I think you've noted that you expected to be supply constrained there for, one time you said for the next year, maybe now we're at either six or nine months more. Any update on that relative to the capacity you've been adding? I see the book-to-bill coming down and lead times coming down. Is that specifically related to the Atmel backend test? Any update there would be helpful. Thank you.
Atmel had many business units. It was just not one business. There was microcontrollers and memory and defense and aerospace and RF and automotive and wireless and other businesses. We had broad-based challenges on the Atmel business. Every quarter, certain products, certain business units get healthy on that. We have been guiding from the very beginning that we expect this to take till about June of this year for all business units to get healthy, and that really hasn't changed. As we speak right now, there's largely, I would say, one and a half business units that are still having significant challenges. Most of the others are in the very tail end, where a large amount of problems have been resolved. This last one and a half business unit largely chose the testing platforms to be the one that you cannot acquire any more of.
These were very old testers of 1970 and 1980 vintage that should have been obsoleted prior to the year 2000. 20 years later, they are still running, but they're not running. They're not running well. The capacity on them is going down, not up, because every quarter you got to take one of it down to use for spare parts because you cannot buy anymore. Therefore, the capacity is decreasing, not increasing. We are bringing that testing up on those products on more modern testers or Microchip platforms, but that process is painful and then qualification of customers, many of those are automotive customers.
That is the process we've been going through for the last year and a half, and I think during which we have grown the business, results have been great, lead times have been slowly coming down, but some of the remaining challenges remain, and we think we need another five months to get it all behind.
Thanks for the details.
Welcome.
We'll go next to Harsh Kumar with Piper Jaffray.
Yeah. Hey, guys. First of all, congratulations on a solid quarter. 9% growth is pretty impressive, Steve. I wanted to ask about your outlook for March. In those markets, your broad end markets, auto, industrial, consumer, which one do you think is going to outperform, in your opinion, the most?
Well, if the question is just for the March quarter, the consumer obviously is the weakest. Automotive should do very well. Industrial should do very well. In some markets, you have kind of brand-new budgets coming in in the new year. The only market that is soft, which is driving a minus 1% guidance, is because of the Atmel seasonality, which is usually down almost 8%, and that is driven by some of the consumer markets. Everything else should be good.
Yeah, and maybe I'll just add to that. On a geographic basis, this will be a strong quarter for Europe. It always is. Just many more shipping days in the quarter without the holidays. Asia is the most challenged just because of the Chinese New Year, and Americas should do pretty well.
Steve, as my follow-up, you guys always run the company for the long term, and historically in lean times, even you've grown inventories. Your target level of inventory is 115-120 days. You just hit 115 days now. By all metrics, end markets are still pretty good. Does it actually make sense for you, can you make the case that you should actually be growing your inventory levels even further and beyond the 115 days? Why stop here?
If our inventories get below 100, then we don't have usually all the right product in the right place and all the permutations and combinations and SKUs and flavors available, and then you go unsupported and you're expediting, and you're kind of having a lot of manufacturing challenges. When you are in the 150, 120 days of inventory, then I think we can manage the business reasonably well. I wouldn't distinguish between 115 and 120 or a few days less, a few days even more than that on the outer side. We don't want to keep going, like you said, to have 135, 140, 150 days of inventory because then you're not really making the right use of that inventory.
We don't need that much inventory. You have the opposite risk of obsolescence where certain products, customer changes the flavor they want to use or want to use a different package than we packaged it in. We want to keep most of the inventory in the die inventory. When you do that, then its dollar value doesn't grow as much versus if you put it in the finished goods. I think that's well tried over the years. That's really the right kind of number. When we are lower than that, we're trying to grow inventory. When we are higher than that, we're trying to decrease inventory, and that's kind of still the right number.
Understood. Thanks, guys, and congrats.
Thank you.
We'll go next to Chris Caso with Raymond James.
Yes. Thank you. Good evening. Just to start, if you could talk a little bit about where lead times are in general now. You talked about them coming down. Would you consider them at normal levels now? Just following on from that, Steve, you had talked about the soft landing that you were trying to achieve. You've got lead times down now, you got the book-to-bill at one. Would you consider that, with what you're seeing now, that soft landing is sort of what we're seeing now, or are there more steps that need to be taken as you go into next year in order to achieve that?
Let me have Ganesh answer the lead time, then I'll pick up on your soft landing question.
Lead times are running, if I look at the December Quarter, in the 4-week to about 14-week window. If you remember, we had it at the peak of this, we were between 4 and 20 weeks. We have started, as the capacity has come on board, as we are able to get our manufacturing to respond, we have been bringing that down. We fully expect that it gets a little bit farther down in the March Quarter, maybe closer to about 10, 11 weeks at the top end, and then gets down to between 4 and 8 weeks, which is what we would consider to be normal on about 90% of our standard line items by the June Quarter. Right about on the same timeline we established about a couple of quarters ago.
Picking up on the soft landing scenario, I think, I would say we pretty much demonstrated it. We are on the tail end of it. March Quarter is right at seasonality, combining Atmel and Microchip together. If June Quarter grows from here, June, September are stronger quarters, then we kind of just went through a soft landing and getting back to kind of seasonality and March Quarter is up 9% year-over-year. Right in the upper end of that range that we had given you, and over time, it gets into the middle maybe or something around that. We are not -10%. We are not correcting like some of the people thought we will.
Okay. Thank you. Just as a follow-up on the analog business and some of your comments on that, and I think I understand what you are saying, that there is growth in analog within the microcontroller market, which is affecting what gets classified the analog segment. What does that mean? I guess two questions. One is, why are we seeing that now, in that are there particular product cycle transitions happening which would cause that to show up in the numbers now? And then what does that mean for the analog business going forward? Should we expect some flattening of the growth on the analog business over the next several quarters as that trend plays out?
The transition we are going through is some of our analog business was a connectivity business, when it either got originally created or if it came to us from an acquisition, was largely all analog in its functionality, and therefore classified as analog. As time went on, if you look at a customer system, we then begin to see what are the other functions that are on that customer system. Often there are microcontrollers that are there to be able to take advantage of that analog. What we did, some time back, was create a new set of products which effectively combine the microcontroller and the analog into a single product. As we go win new designs, and by the way, sometimes that microcontroller was ours, in other cases, the microcontroller was somebody else's.
When you take a complete solution, which is a single chip, which has the microcontroller and the analog, particularly in some of these smart connectivity solutions, and there's a fair amount of software that we provide. The combined solution is a far more defensible, far higher margin, and far more sticky business that we have. So it creates the right answer, and we don't think ahead of time, should we create products that fall into analog or fall into microcontroller. We're doing what we think we need to do to be able to grow, protect, and have high profitability on the products we have. So that's one part of the headwind you're seeing, so to speak, by classifying it as microcontroller versus analog.
The other side of the tailwind that is coming, and I think we will see it progress along as the standard analog products that we've had, as they get attached around our microcontroller on new designs, as we have talked about the total system solution and the whole Microchip 2.0 approach of going to market. The remaining analog products will, over time, get back some of that tailwind and have that growth. And there's a bit of cross currents between some of the analog that is moving to microcontroller for the right reasons, and then new analog designs that will come in and give us growth. In the long run, we fully expect that analog will be a part of the growth story for Microchip.
By the way, if a pure analog company adds a microcontroller core to one of their complex product to improve performance for programmability or for any other reason, they would still call it an analog product. If the exact same product existed in Maxim or Linear or Intersil, they'll call it an analog product. The same product, if it has a microcontroller core, we put it in microcontroller. It's a classification difference. It doesn't change the value proposition. At the end, we have a better total solution for the customer, more sticky socket, higher ASP, higher margins, and overall growing the revenue.
Sure. Just to be clear then, with that trend and the microcontroller growth on a year-on-year basis already up, as we model your analog business going forward, that trend should persist until, I guess, you said you see that attach rate from the analog components start to rise. Is that a correct interpretation?
It certainly, as Ganesh described, has some headwind because of that. Of that classification, it puts a little tailwind on the microcontroller, a little headwind on the analog. How it will manifest itself longer term, next quarter, the quarter after, and all that, we don't really yet know. It's a new phenomenon. We manage the company through an overall growth, we don't really care where you put it.
All right. That's fair. Thank you.
We'll go next to Mark Delaney with Goldman Sachs.
Yes, good afternoon, thanks for taking the questions. First question relates to Atmel. The company did a fantastic job of improving the margin structure of Atmel, I had thought that maybe some of that margin improvement was adjusting which end markets are being targeted with the historical Atmel products, maybe more towards automotive and industrial. Can you just talk about to what extent that was a factor, or is consumer still a pretty big part of the mix? I ask it just to better understand the comment around Atmel seasonality in March versus the historical rates.
Well, we've had Atmel for seven quarters. Six quarters plus is the lead time to get a new product in production. If you thought a lot of this improvement has come from because we largely changed the flavor of which markets the products were sold, I think you miscalculated there. What we had said was Atmel had a focus on, what terminology we used, home runs, right?
Yeah, swinging for the fences.
Yeah, swinging for the fences. They had a focus on large account and consumer segment, swinging for the fences, and some you win, some you lose, and gross margins are lower and all that. We very largely changed the focus on where our future attention is. True, it's automotive, it's industrial, and other areas, and less on the consumer accounts. It didn't change the existing flavor of those parts in the last several quarters till those designs go obsolete and the new designs we are focused on come to fruition. A large majority of the margin structure change has come from improvement of pricing, disciplined pricing, dramatic cost reduction in the fab, assembly, test, bringing a lot of those parts inside, reducing large portion of the expenses, just doing the whole job overall better and rationalizing R&D, rationalizing manufacturing.
You cannot change the flavor of the business regarding what segments of business is in seven quarters. You barely touch it in seven quarters.
Got it. That's helpful. For a follow-up on capital allocation, I know the company had talked in previous quarters about potentially engaging in M&A now that the balance sheet has been improved. Curious for an update on how Microchip's thinking about that at this point, especially with the increased flexibility in your cash post the U.S. tax reform.
We are really not thinking any differently. Post-U.S. tax reform, the cash becomes easily movable from offshore that really was not easily movable before because it was permanently invested offshore. Our priorities have not changed. We continue to believe our priorities are investing in M&A first, dividend second, and stock buyback only opportunistically. Those have not changed.
Thank you.
We'll go next to Christopher Danely with Citigroup.
Thanks, Steve. Just to talk about the, I guess, the soft landing that we're all trying to engineer here. Can you give us a little more color on specifically what would be the combined typical June seasonality for a Microchip plus Atmel? Could we be above or below that? I guess, how can we be sure that we've, I guess, landed yet in terms of the soft landing?
I don't have the number for June quarter. It's a good question. We should have thought about it. I don't have it off the cuff. You have seen strong June quarters for the last two years. You've seen strong September quarters for the last two years. You have seen stronger than historical December quarters. March quarter is right at seasonality. That's what I'm saying, where it has landed. If the June quarter is below seasonal, if that's your concern, we think the June quarter should be pretty good. June quarter should be strong.
Got it. How would you compare this slowdown or return to normal or soft landing or whatever the hell we want to call it, versus prior slowdowns that you've seen. Maybe talk a little bit about the linearity of bookings, how your customers feel, how the distis feel, inventory.
Well, I think when I look at cycles, no two cycles are really different. One, many of the prior cycles have crash landed. This one certainly is not crash landing. We engineered the soft landing here. The other factor was in some of the prior cycles, some sort of second catalyst happened. In 2014, it was a dramatic slowdown in China that we saw at first. Four months later everybody saw it and stocks saw a significant sell-off. In some prior cycles, they were driven by other factors, a tech bust in 2000 or whatever, now we're going way back. I don't see any other second catalyst this time. It's just driven by getting our capacity in line, bringing the lead times down, going from a high book-to-bill ratio to a very moderating one book-to-bill ratio.
I'm going to add one book-to-bill ratio, you can grow the business forever. Just take incremental chunks every quarter and the business can grow. For, I think 90% of our business life, the book-to-bill is around one, right?
Yep, that's right.
That's where it should be. That means the lead times are short. People can get their product. We still have some challenges on the Atmel side in about one and a half business units like I talked about, and those are built in our analysis, and that part of it we still need to soft land. It's largely there. I think the plane is right over the runway.
Okay. Thanks a lot.
Yeah. Welcome.
As a reminder, please allow yourself one question per turn. If you would like to ask a follow-up, please re-queue. We'll go next to Craig Ellis with B. Riley FBR.
Yeah, thanks for taking the question, and Steve, for giving us the numbers that really show the 10 percentage point difference between Atmel's first quarter calendar seasonality and what was Microchip to make sense of things. What I wanted to do is shift over to margins, and ask you about the target margin model. Because when I look at the business, I see that with gross margins we're, I think within 100 basis points of target. With operating, I think we're within 60 basis points. The team has done a spectacular job improving margins over time, and you've raised the targets multiple times, I believe. The question is, as you get this close to targets, is there anything that you would see that says the business would have a ceiling at those targets, or is there an opportunity to operate above those targets?
I think what we do not want to do is on any parameter when we are above the target, or in the case of operating expense, where we are below the target, then we suck that in and make that as a new target and keep thinking that the other one has to go higher. Then the operating margin goes over 40 and then we say, "Well, higher is the target." The peak can't be the target. The business will oscillate over good times and bad times and very strong environments and weaker environments. 40% is a long-term target. Can it temporarily in good times go over 40? The answer is yes, but there'll be equal times when business is in recession, it's softer, it'll be less than 40.
If you're really doing your math like you did, you're saying, well, gross margin is about 100 basis points away from target and operating expense is already at target. Wouldn't our operating margin go higher? The answer to that is yes. That's not long term. That's temporary. It might go there. During huge growth we have had last year, fiscal year 2017 over 2018, we grew 13.7%, or some number like that. During that sometime expenses do not keep up, and you get on the lower end of the expenses. When the business moderates sometime, the expenses catch up a little bit, because otherwise you're not able to hire people enough in a very strong time.
Again, you got to give it a flex, and as managers, we don't want to be in such a tight box that every little slack, you suck it in, and then we're in such a small box, then we're just missing point one here and point two there. Think of that way. On a pure math basis, can the operating margins go higher? Yes, temporarily. We're not willing to yet admit that longer term margins could be higher than 40%. The second piece of that is still the M&A strategy. I have never bought a business that makes 40% operating margin or 39.5 where we are. Whenever we buy a business, it's usually much lower. You have to understand the flavor of its business. Can we improve its gross? Can we improve its operating expense? Where does it go?
A blended margin comes up and usually most acquisitions result into a blended margin to be slightly lower. For example, Microchip's operating margins are still several hundred bips above the Atmel flow of the business. It's 39.5 total, but Microchip is well above 40, and Atmel is much lower than 40, and we're still improving it. Some of the Atmel may never get where Microchip is, and we were quite honest from the beginning and say, "We'll improve it substantially." We have gone from Atmel breakeven to in the mid-30s. It can't match mid-40s where Microchip might be. I'm just putting fat brackets here. I'm not giving you where the breakdown is. I don't know whether that helps, but don't keep making our box smaller and smaller, please. We need some room to operate.
It's also the balance of-
Yeah
profitability. We got to make sure the investments for growth are there so that it's not just profitability, but without any growth.
Speaking of growth, Ganesh, I'm going to ask a question on a business I don't think I've ever called out on the quarterly call before, but you did note the very strong sequential performance of the licensing business, and underneath that, you highlighted some of the recent agreements that have been made with Foundry and the growth that's been driven off that. As we look at that strong sequential growth, what does that mean, and what do the recent business signings mean for the performance of that business intermediate term, and what kind of beneficial effect does that have on the company's margin structure?
I want to be clear that the improvement in the results are not from recent signings. Signings took place multiple years ago, after which you have to go install the technology, qualify the technology, allow the technology to ramp, and then have that ramp generate royalty revenue. Royalty is the substantially larger portion of the revenue that we get from the licensing business. That continues to layer in as by either Foundry or IDM that we license it to by node that it is licensed. It can also change based on how a given process node succeeds or doesn't succeed. Much like in our IC designs, where you have to get those designs early and then work with the customer to realize the revenue, it can take 24, 36 months.
Licensing has a similar kind of thing, where what you're seeing in today's benefit are things we did one, two, three years ago. We continue to have new licenses that we are getting signed off that will create future growth. I think we expect licensing will contribute to the overall growth that Microchip has and of course, at a much higher level of profitability.
I would say, though, that on the effect of licensing on the margin, licensing is a $100 million business. A $100 million business at a higher gross margin than Microchip's overall, but Microchip is at a $4 billion run rate, roughly. It's just really a small needle mover. There are equal number of other various business units where certain products are much higher gross margin, certain other products are less gross margin, and the mix moves forward, and sometimes one has a stronger growth, the other has stronger growth. True, licensing business is accretive to the gross margin, but at $100 million out of $400, it doesn't dominate.
Got it. Thanks, guys.
We'll go next to Vivek Arya from Bank of America Merrill Lynch.
Hi. Yeah, this is Adam Gonzalez on for Vivek. Thanks for taking my question. Just wondering how we should think about your CapEx moving into fiscal 2019 in the context of some of the views you laid out on the demand environment and the relative supply-demand balance. Thanks.
We haven't gone through our annual operating plan process for fiscal 2019 yet. We're not ready to give a number on that as it is. We have made significant investments this last year. We have goals to bring the assembly and tests that we do internally higher, and that will require investment. I don't think we're ready to provide the street with a number yet for next year's CapEx. Still going to be a small % of overall net sales.
Got it. I guess just one quick follow-up. I know this is a bit nitpicky on the seasonality issue. In the March quarter, you guys, it's down 1% quarter-on-quarter. What's different this year versus last year? Because you had Atmel for a full quarter then, and sales still grew 2% sequentially. Just what was better than seasonal in March versus this year?
That's kind of what threw everybody off. Last quarter was very noisy. Not the last quarter, March quarter last year was very noisy. We had just gone go live on January 1. There were a bunch of orders that were pushed out into the quarter. There were a lot of moving parts, and there was a large amount of delinquency we had carried in the prior quarter of December. Some of it we were able to meet. We were new to that business. It was the first March quarter we had had. Atmel had also left a fair amount of pricing and all that on the table. We talked about it. We were seeing a significant impact of price increases we had done, which masked some of the seasonality. There were just way too many moving parts.
The environment was stronger in March last year.
Yeah. I mean, not that the environment is weak now, though, but it's certainly not as strong as a year ago. I think that's what threw everybody else off.
Okay, thanks very much.
I think this time there's no such noise, and this is a true seasonality with Atmel.
Thanks. Got it.
We'll go next to Kevin Cassidy with Stifel.
Thanks, congratulations on the quarter. Going to the Microchip 2.0 strategy of platforms, how is that affecting your pressure on ASPs? You had mentioned before, Steve, that industry consolidation, you're seeing less pressure, is this helping maybe even lift ASPs?
Well, the answer to that is yes. I don't know if the ASPs continuously go up, we are severely resisting the pressure to give year-over-year declines, and in large majority of the cases, we're succeeding and are able to hold our ASPs. Plus, as we add microcontroller intelligence to analog, as we attach analog around microcontroller, it'd be more important to the customer to have a total solution. In general, you can bundle and it's more sticky, and it helps ASP. All that is happening. Semiconductor ASPs do not constantly go up. I think if we can have them not go down, that will be good enough.
It avoids commoditization.
Yeah
of the individual products when you can take a complete solution and show the customer the value in doing that.
Right. Maybe even on the OpEx side is, are your platforms, the designs, robust enough that you can win designs without Microchip employees or FAEs being involved?
Well, across 115,000-plus customers, vast majority of the customers, we don't have Microchip FAEs involved. We just can't have that many employees covered. There is a large amount of effort in us building reference designs, documenting, training on the web, training in person. We train 30,000 engineers around the world in various masters conferences, seminars, and webinars, and on-site customer training. Microchip is huge. Microchip is a very vast training organization teaching our customers how to do it. Then there's a lot of self-start help and all that where customer design product. Vast majority of Microchip's customers, you do not have a Microchip FAE involved in helping them in their design. There's a Microchip FAE involved in overall general training with which they do the design. Otherwise, we wouldn't be able to have enough people.
Okay, thank you.
Welcome.
We'll go next to Harlan Sur with J.P. Morgan.
Good afternoon, thank you for taking my question. From a geographical perspective, on a year-over-year basis, you're driving growth in all of your geographies, demand looks pretty broad-based. One of the geographies that kind of stands out is the Americas region. Growth was up 5% versus the rest of the geographies that were up at least double that rate. Any reason for the slower growth in the Americas region?
Well, the problem is, we count the revenue where the product is shipped, not where the product is designed. Manufacturing largely keeps moving out of Americas, I don't know whether it will change in the coming years. When you look at 5% Americas growth, it artificially understates it when internally we look at the revenue in Americas, counting back the revenue which was generated in America and shipped overseas, then Americas growth was great. It was just as good as any other geography. It wasn't weak.
Yep. Thanks for the insights, Harlan Sur.
What we report actually is where we shipped it.
Got it. Seems like the team is gaining quite a bit of traction with these analog products, with the integrated MCU cores for these, what you guys call smart connectivity solutions. Can you guys just give us some examples of where these type of connectivity solutions are used, applications and markets? When you say connectivity, are you talking about both wired and wireless connectivity integrated with your MCUs?
Yeah, go ahead.
It's both. Let's take one example, which probably you have, but you're not familiar with. Your car has a USB hub in it. Depending on the vintage of your car, that is a hub that has progressed over time from, at one time being largely an analog-only solution to today being a smart connectivity solution where that hub has a way to recognize what is being attached to it, has a way to either be a slave or a master. It has a way to control your display. Of course, power the devices connected. It's a far smarter hub today than it was several years ago. Several years ago, we would have called that an analog product. As the new designs have gone into production, those we would call as a microcontroller product.
Great. Thank you.
You could take a different market. Take a thermostat, a large number of thermostats nationwide, you can access them now through your phone or pad. I know I can access the thermostats in my house on my phone remotely while traveling, turn the temperature off or down or up, or if I'm going to be on business for longer and wasn't anticipated, I can keep the house turned off. Those are all connectivity solutions where you're doing it wirelessly.
It would have a micro Wi-Fi or a micro and a Bluetooth, for example, on the same chip rather than two chips.
Large number of speakers in the home, you don't connect them anymore. You simply Bluetooth it from your phone into the speaker, the sound is coming. You don't really access the phone, access the music through a box or something anymore these days. You can still do that.
Right.
Most times you access the music through the phone and just Bluetooth into any of the speakers in your home. That's connectivity wirelessly.
You can take those same general examples and apply it to Ethernet, to CAN, to various forms of wireless and so on and so forth.
Yeah.
Yeah. Thanks for the insights. Thank you.
Okay.
We'll go next to Gail Alexander with Careful Investments.
Good evening. When you look at your business, do you still see revenues growing at 7%-9% over the next few years?
Yes, as we are driving our business, we're modeling our business and looking at the impact of Microchip 2.0, those are the targets we're driving towards.
As you look at the business now, how strong do you find the Asian business and the European business?
The last year, all three businesses were very strong. If you look at just on a quarterly basis, March quarter has the Chinese New Year, so Asian business in the current quarter is always weak. European business in the March quarter, they're the strongest geography in the March quarter. America is normal. If you look at the December quarter, usually with all the holidays and all that, Europe and America businesses were weak, and Asia business was okay. Each quarter, it's kind of slightly different. When you go to June quarter, Asia business rebounds strongly from Chinese New Year quarter of March to June. I don't know if that helps.
I thank you.
Yeah. Thank you.
We'll go next to (Raj Vanda, Gail) with Needham & Company. Due to no response, we'll move on. We'll go next to John Pitzer with Credit Suisse.
Yeah, guys. Thanks for squeezing me in. Steve, congratulations on the solid results, especially on the operating margin. Steve, one of the questions I had, over the last several months, as you talked about sort of the soft landing scenario, you kind of alluded to the potential that you could keep every quarter this calendar year in that sort of high single digit long-term revenue growth target. Is that still kind of the way you see this calendar year playing out? I guess secondly, if it is, do you get there on just seasonal growth sequentially, or will it take some above-seasonal quarters?
John, I don't know. I think this March quarter Atmel seasonality last year kind of threw everybody off, I would say, including us. As we really dove into it and looked at our backlog and looked at the business and then analyzed prior five years of Atmel business, and the behavior of these large swimming for the fences accounts and design wins, the last year's performance kind of threw it off, and those accounts still had a problem last year. It was kind of overcome by the general halo of price increases and a lot of under-marketing of AVRs and all that Atmel had done and where we improved the position, stronger distribution, and things like that. The number of things last year kind of threw it off, and I guess all of us together somehow didn't quite capture the blended March seasonality.
That's what we're adjusting to. Give us a little time to get our feet back on the earth and really see what effect it has on the rest of the quarter. I mean, our economy is not bad. Bookings are not bad. The business is not bad. We're getting strong turns. Our inventory is right. We're not in a massive inventory correction mode of any kind. There's really no problem in business. We're doing well. When you do the calculation like we have shared, we're right on target.
Steve.
Go ahead.
Well, just anecdotally, for me, one of the things that's always important as we kind of orchestrate this soft landing, as book-to-bill and lead times come in, are you seeing any sort of unusually high cancellation patterns or different order patterns from your customers, or are they responding to your ability to meet demand in an orderly fashion?
Yeah. They're just responding to us shipping it in a more orderly fashion. If you recall, last March to last June, the bookings were not any stronger. The book-to-bill was exactly the same because we informed the customers of lead time. They responded with higher orders out in time to respond to their lead time. As the lead times have come in, customers are adjusting to that lower lead time and don't have to place outer bookings, which is lowering the book-to-bill ratio, and now it's at parity and it's moderating. Customers are essentially giving us as many bookings as they need and don't have to give huge bookings out in time. This is almost perfect landing, I think, so far. The only difference, I think, Street had higher expectations for March, looking at the March seasonality last year, which was kind of typical Microchip.
Basically, there were lots of other factors in the business, pricing and others, which masked that seasonality. This year, December to March, we can't.
Perfect. That's very helpful. Thank you, Steve.
Obvious result is that some of the other quarters have to be stronger, and I think they will be, and hopefully they will be. Certainly last year they were. March, June, September, December, they were all stronger. This year should be the same. Next couple of quarters should be stronger because of that March seasonality.
Perfect. Thank you very much.
Yeah, welcome.
We'll go next to Hans Mosesmann with Rosenblatt Securities.
Thanks for squeezing me in. Hey, Steve. Can you give us a rundown on the competitive dynamic in eight, 16, and 32 bit? Have things changed over the past several years and so on? Thanks.
The number of competitors, it continues to be a reasonable number of good, strong competitors. We have continued to move up that ranking. We have been gaining share in all of these product lines, including the overall microcontrollers as well. Some of these competitors have had various challenges relative to either their capacity, in some cases, whether they're being acquired or not. There have been those kind of things that are taking place there. Microcontrollers remains a competitive product line, competitive market. I think there are more rational actors today than they were perhaps five years ago, and they are responding, in some cases more so, or some cases less so, to the consolidation that's taking place, and that certainly helps as well. Nothing dramatic to report that is different about competitive dynamics in microcontrollers.
I'm sorry.
I was just going to say.
Microcontroller business up by, I think, almost 19% last year, was higher than Microchip's growth and higher than the industry's growth in that segment. Based on FAE numbers, we gained share in all three, eight, 16, and 32, and I shared with you the overall share gain by 90 basis points almost. I think things are good.
Okay. Would it be safe to say that the average microcontroller actor out there is not using price as a way to gain share, it's more a value proposition?
I don't really know if just gaining share by price has ever worked in microcontroller. These are complex parts, and over time we have shown surveys when you ask the customer, what do you consider for choosing a microcontroller? The most frequent answer is the ecosystem, the tools, the support, the quality, the reliability, the customer, the samples, the development tools and everything else. Price is one of the factors. These are not commodity products. It takes two years for a customer to go to production after they have made the device selection. Price has never been really the way to gain share. There were players in the market that were unresponsive. They will drop the price too quickly and not stand up and fight to really maintain their price. Some of those have disappeared, or some of those have become better and more responsible.
Overall, I don't think price has been a mechanism to gain share in the market ever.
Okay, good. Thank you very much.
You're welcome.
With no further questions in the queue, I would like to turn the call back over to Steve Sanghi for any additional or closing remarks.
Well, thank you everybody for joining our conference call, and we'll see some of you on the road at various conferences we might go to. Thank you very much. Bye-bye.
This does conclude today's conference. We thank you for your participation. You may now disconnect.