SK Hynix's (HXSCL) Management on Q2 2016 Results - Earnings Call Transcript

| About: SK Hynix, (HXSCL)

SK Hynix, Inc. (OTC:HXSCL) Q2 2016 Earnings Conference Call July 26, 2016 8:00 PM ET

Executives

Cha Seong-Geun - Head of IR

Kim Jun-Ho - President, Head of Corporate Support

Analysts

Peter Seicheol - NH Investment Securities

Yoo Jong Woo - Korean Investment and Securities

Ricky Seo - HSBC

Do Hyun-Woo - Mirae Asset Securities

Nicholas Gaudois - UBS

JJ Park - JPMorgan

Simon Woo - Bank of America Merrill Lynch

Operator

Cha Seong-Geun

[Interpreted] Good morning and good afternoon and evening to those calling in from abroad. This is Cha Seong-Geun, the Head of IR at SK Hynix. Welcome to the SK Hynix 2016 Second Quarter Earnings Release Conference Call.

Before starting the conference call, allow me to introduce the executives present here with me today. First, President Kim Jun-Ho, Head of Corporate Support; Senior Vice President and CFO, Lee Myeong-Yeong; Vice President Pang Me-Hak [ph], in charge of the DRAM Marketing Group; and Vice President Kim Young-Ne [ph], in charge of the NAND Marketing Group.

Let me issue a disclaimer that all outlooks presented by the Company are subject to change, depending on the macroeconomic and market circumstances.

With that, we will now begin SK Hynix 2016 second quarter earnings release conference call. Let me now turn over to President Kim Jun-Ho to present the first quarter earnings and the Company's plan and outlook.

Kim Jun-Ho

[Interpreted] Good morning. This is Kim Jun-Ho, in charge of SK Hynix Corporate Support. Let me begin with a review of our financial performance in the second quarter of 2016.

Even as the global economy and financial market remained volatile in the second quarter, demand for memory products recovered since the first quarter, leading to a higher-than-planned bit shipment growth in both DRAM and NAND. Despite a slightly stronger Korean won and continued slump in the ASP, sales grew from the previous quarter. Second quarter consolidated sales totaled KRW3.941 trillion, an 8% increase from the previous quarter.

DRAM bit shipment rose by 18% quarter on quarter on the back of strong mobile demand and recovering demand for computing DRAM that started from the end of the second quarter. ASP fell by 11% quarter on quarter, a slower drop than in the previous quarter, although it continued to fell across most products.

For NAND Flash, bit shipment rose 52% quarter on quarter, demand for mobile embedded products recovered, while demand for SSD continued to increase. There was also a base effect from the first quarter, but ASP fell by 11%, the same pace as in the previous quarter as prices for solution products fell and the sales portion of the lower-priced TLC-based products rose.

Sales portion rose to 21%, the highest ever in a quarter. Chinese smartphone makers continued with inventory build-up and adoption of higher-density products.

In terms of profitability, our operating profit in the second quarter was KRW453 billion, a decrease of 19% quarter on quarter, with operating profit margin of 11%. Cost of goods sold increased 15% quarter on quarter due to increase in DRAM and NAND bit shipment. SG&A rose 6% from the previous quarter, owing to the rise in R&D expenditure.

For DRAM the drop in ASP could not be completely offset by cost reduction because of the increased portion of the higher-cost non-PC DRAM products and high cost requirement in the initial stage of migration to new products. For NAND, however, profitability improved, thanks to cost reduction offsetting the ASP drop. High bit shipment growth more evenly distributed the fixed cost and there was increase in the portion of relatively low-cost products such as TLC.

Depreciation and amortization expense for the quarter was KRW1.125 trillion, a slight increase from the previous quarter. EBITDA was KRW1.578 trillion, representing an EBITDA margin of 40%.

There was a net non-operating expense of KRW70 billion from transaction loss resulting from the recovery of account receivables as the Korean won appreciated in the second quarter. There was recognition of KRW97 billion in corporate tax for the second quarter. Our net profit after corporate tax for the quarter was KRW286 billion, with a net profit margin of 7%.

Consolidated cash balance as of the quarter-end was KRW3.433 trillion, a decrease by approximately KRW824 billion from the previous quarter. There was payment of corporate tax for the fiscal year of 2015, as well as dividend payout. Our interest-bearing debt stood at KRW4.355 trillion, a decrease by KRW26 billion from the previous quarter. Debt-to-equity ratio at the end of the second quarter was 20%, the same as in the previous quarter, and net debt to equity ratio was 4%, a slight increase from the previous quarter.

Next I would like to discuss the Company's market outlook and its plan for the next quarter.

For DRAM there is now recovery in the PC DRAM spot price which had been leading the ASP drop. Unlike in the first quarter when demand remained sluggish, oversupply in PC DRAM is easing, thanks to reduced inventory levels among PC and server providers, especially towards the end of the second quarter.

Particularly for mobile DRAM, we experienced robust demand, thanks to inventory build-up and adoption of higher-density products by Chinese smartphone makers and resumed purchasing by major smartphone makers to launch new models.

We look forward to the positive trend continuing into the third quarter. PC clients appear to be on an active inventory build-up to be ready for the seasonal demand in the second half. Such demand/supply dynamics in PC DRAM is likely to have a positive impact on the server DRAM market as well.

For mobile DRAM, the robust demand coming from Chinese smartphone makers from early this year is expected to continue. It is likely to be accelerated by the increase in their content per box. Seasonality is expected to have a positive impact on mobile DRAM given that the third quarter traditionally sees a jump in customers demand for new smartphones.

From the supply side, bit shipment growth is expected from seasonal factors and with increase in the portion of 20 nano. However, along with suppliers' product mix shift from PC DRAM to mobile and server DRAM, we foresee demand/supply to balance out in the second half.

In the NAND market, a positive environment that took shape in the first quarter improved further in the second quarter as the spot price for major products continued to rise. Demand is gaining traction for SSD, not only in client SSD but also in enterprise SSD for storage providers. This is thanks to the improved price competitiveness of SSD as NAND price steadily declined.

In addition, there are strong moves by Chinese smartphone providers to build up inventory just like in the DRAM market. And the expectation that NAND content per box will increase in the new mobile products to be launched in the second half appears to have had a positive impact in the market in general.

Demand is expected to be even more favorable in the third quarter than in the second quarter, thanks to seasonality. 2D NAND supply will remain limited, while increase in 3D NAND supply remains uncertain. In the second half, there will be new mobile product launches, increase in the content per box, and increase in demand for high-capacity SSD following the launch of new platforms for PC and robust demand for enterprise SSDs.

DRAM bit shipment growth is expected to be in the high single digit in the third quarter, which will make the shipment growth for the year to be low to mid 20%, in line with market growth.

Production for 2z nano increased from the second quarter as planned, led by computing DRAM products. 2Z nano production also started for mobile DRAM from the late second quarter, to be ready for the demand for new products in the second half. The pace of increase in 2z nano will be dependent on the growth in demand for DDR4 and LPDDR4 expected to reach around 40% of total DRAM production by the end of this year.

For NAND Flash, bit shipment increased in the third quarter and is expected to be around low-teen level, led by increase in the content per box in new mobile products and increasing demand for SSD. We will keep increasing the portion of 14 nano products which started in the second quarter. Thanks to the overall strong demand, bit shipment growth in the market is expected to be around low 40% level for the year. The Company expects to outperform the market with a high 40% level bit shipment growth.

Meanwhile, we have been expanding the high-density eMMC [ph] generation 3D NAND products for [ph] mobile in the second quarter. Third-generation products will be developed and made available for sale within the second half of this year. We plan to be ready for the new 3D NAND market by securing a 3D capacity of 20,000 to 30,000 wafers by the end of this year.

For the remainder of this year, SK Hynix will keep striving to improve our technological competitiveness. We kindly ask for our shareholders' and investors' continued interest and support.

This concludes my report. Thank you.

Cha Seong-Geun

[Interpreted] And with that, we are now ready to take your questions.

Question-and-Answer Session

Operator

[Interpreted] Now Q&A session will begin. [Operator Instructions]

The first questions will be presented by Mr. Peter Seicheol from NH Investment Securities. Please go ahead, sir.

Peter Seicheol - NH Investment Securities

[Interpreted] Now I have two questions, one is about the DRAM, the other is for the NAND market. First, about the DRAM market, it seems as if the market seems to be relatively stable these days, especially with the spot prices rising. Now I wonder what SK Hynix's strategy is amidst this environment. So, are you going to, especially for the DRAM, are you going to focus more on profitability or increasing market share?

Unidentified Company Representative

[Interpreted] That's a tough question, I must say, but basically SK Hynix's strategy remains that we are going to keep trying to increase market share while putting our priority on profitability.

Peter Seicheol - NH Investment Securities

[Interpreted] Now the second question is about NAND. So as you have mentioned, the demand for SSD is growing and there is also a faster spread and development for 3D NAND as well. And since as if there is growing investment not only domestically but also outside Korea as well, so I wonder what the development phase for a 3D NAND is in SK Hynix and what is your migration to the next-generation 3D NAND going forward.

Unidentified Company Representative

[Interpreted] Now I would like to field the question about the 3D NAND. Now the 3D NAND development is underway, and so as we have explained earlier, the second-generation 3D NAND products, mostly in the form of MLC products, have gone into mass production in the second quarter, and we have already started shipping these products for mobile.

And then the third-generation product, the TLC, we believe that we will be able to complete the development of TLC products within the second half of this year to start shipping for mobile in the second half.

And then about the CapEx for 3D NAND, as President Kim Jun-Ho mentioned in his report, our plan is to secure the capacity for 3D NAND up to 20,000 to 30,000 wafers by the end of this year.

And investment and production will focus on 36-stack product until the third quarter. And then from the fourth quarter we will shift our focus to the 48-stack product.

Operator

[Interpreted] The next questions will be presented by Mr. Yoo Jong Woo from Korean Investment and Securities. Please go ahead, sir.

Yoo Jong Woo - Korean Investment and Securities

[Interpreted] Now I have two questions about DRAM. First is that, it seems that there are some lingering concerns in the market about the customer qualification and supply for the mobile to second-generation products. So I would like to just ask about the status about your customer qualification and supply for 2Z nano products -- sorry, 2Z nano products?

Unidentified Company Representative

[Interpreted] Now for the mobile 2Z nano products, for example, the 4D eMCP, now they are ready, so, qualification is currently underway. And for most of the major clients, our mass production has already begun.

Yoo Jong Woo - Korean Investment and Securities

[Interpreted] Now next question is about the inventory. So it seems as if there has been a reduction in inventory in the second quarter compared to the first quarter for the first time in a long time. So perhaps that is because your bit shipment exceeded the production bit in the second quarter. So now then, in the third quarter, what is your projection for DRAM? So, do you believe that again the bit shipment is going to outnumber production bit?

Unidentified Company Representative

[Interpreted] Yes, it is true that the inventory level, which stood at around three weeks at the end of the first quarter fell to about -- around two weeks in the second quarter, and that is because, as you have mentioned, the bit shipments outnumbered the production bit. And we believe that this trend is likely to continue into the third quarter, meaning that there is going to be more bits sold than produced, and this will further lower the inventory level.

Operator

[Interpreted] The next questions will be presented by Mr. Ricky Seo from HSBC. Please go ahead, sir.

Ricky Seo - HSBC

[Interpreted] Now my question is regarding CapEx. You mentioned last time that there was a possibility that the CapEx might be about 10% lower than your projection in the early part of this year. So I wonder now then, does this projection still stand, because it seems as if for the 3D NAND the capacity build that is not going to be by much, so, is your projection for the CapEx still valid for the remainder of this year as well?

Unidentified Company Representative

[Interpreted] Yes. Our CapEx plan remains largely unchanged from the early part of this year, so we believe that it is going to be about 10% lower than last year's level, to be around KRW6 trillion. In the first half, about 50% has already been spent and KRW3 trillion will be additionally spent in the second quarter.

Ricky Seo - HSBC

[Interpreted] Now next question is perhaps more for the long term, and that is for the automotive memory. So if you are making investment already into this, then what is the projected scale for both DRAM and NAND? And also if there are, let's say, for automotive memory, if there are any contracts, then, are there any specificities just for the automotive memory? So for example, do the clients require some specific criteria for -- in terms of the processes or in terms of the contract provisions? And what is your overall preparation for automotive products?

Unidentified Company Representative

[Interpreted] Now in terms of automotive products for SK Hynix, we are currently conducting business, mostly on the car infotainment market, where the market already has been shaped. So for the DRAM, out of the total consumer products, it's a little over 10%. And for NAND, the portion is slightly lower than that but it is catching up very quickly to the DRAM level.

And now for the newly-emerging markets of ADAS and self-driving cars, now for both these segments, we are already mass producing in -- starting from 2015. In terms of the -- in terms of the quality, now we are not there yet for the -- so, not for the self-driving yet, but then for the AECQ 100 [ph], we are already fulfilling this standard. And then also for the ISO 26262 and ASILBQ [ph], so once these standards are standardized, then of course we will try to fulfill them as well. And in such a case, it will be mostly about safety, so, safety and also reliability and performance under high temperature.

And our strategy is to focus on self-driving cars and ADAS because they seem to present the most promising growth potential.

Operator

[Interpreted] The next questions will be presented by Mr. Do Hyun-Woo from Mirae Asset Securities. Please go ahead, sir.

Do Hyun-Woo - Mirae Asset Securities

[Interpreted] My question is about the current computing trend and how this is going to impact the demand for server DRAM, because what we are seeing now is that, for the computing work, for the, let's say, for computing tasks, that will require dozens of servers. Now the trend today is to replace this with just one server using CPU or FPGA. So if that is the case, and I believe that it would also have a negative impact on the server DRAM, so, what is Hynix's take on the current new trend and how do you plan to respond to this?

Unidentified Company Representative

[Interpreted] Now for the server system, in addition to the current CPU, we are also adding the GPU in order to assist quicker, higher performance computing. But then this is not replacing memory, this is just being added, the GPU is just being added to the existing memory. So we do not believe that this is going to reduce the demand for DRAM memory.

Operator

[Interpreted] The next questions will be presented by Mr. Nicolas Gaudois from UBS. Please go ahead, sir.

Nicholas Gaudois - UBS

[Interpreted] Yes, good morning. Thanks for taking my question.

First one is on your progresses for 48 layers 3-bit cell in NAND. You said you will be moving towards mass production by later this year. How do you assess your progresses and the key milestones ahead for, number one, the actual NAND technology itself [inaudible]? And number two, your microcontroller solutions, since I understand that for 48 layers TLC, unlike 36 layers, it will start to use your own microcontrollers going forward. Thank you. And I got a follow-up on DRAM. Thanks.

Unidentified Company Representative

[Interpreted] Now about our NAND overall, now for mobile, the 36-stack product has been launched already, and then in addition to that, the 1Z nano product we have already security yield for mass production and now it has already been - started shipping in the second quarter.

And about the 48-stack, we believe that the first placeholder for this would also be mobile. And now as we have explained in the last quarter, it would -- the timing of the mass production of the 48-stack product would depend on how soon we can secure the yield for mass production. And we believe that that is going to be in the third quarter, or, at the latest, by the fourth quarter. So we are confident that we will be able to secure the yield for mass production for the 48-stack products by the fourth quarter at the latest.

Now for the controller for the 48-stack and the 36-stack products, now for the 36-stack controller, for mobile, we have already adopted the controller for the 2D and it seems as if there are no problems in using this for mass production. And therefore, the 48-stack product, likewise, we are going to modify the controller used for mobile and the SSD and we believe that this will be enough to support the 48-stack products as well.

Cha Seong-Geun

Your second question please?

Nicholas Gaudois - UBS

Thanks so much. On, second question is on DRAM. Just to clarify, when you said you started mass production already for most of your large customers for 20nm, LPDDR and DDR4 in particular, should I assume you have started production for your largest customer in particular? And if so, how much of a further [ph] shift could you see in volumes from PC DRAM to mobile DRAM into the third quarter versus Q2?

Unidentified Company Representative

[Interpreted] Now for the PC and mobile, yes, we have started mass production in the second quarter for major OEMs, so I don't know what you mean by the largest customer. But then we have started the mass production for major OEMs. And then for server products, the qualification is currently underway for the OEM clients and they are about to be finished, so we believe that we will go into mass production in the third quarter.

And then I think your -- the other question was pertaining to the projected portion between the PC and the mobile DRAM. Now for the year, for the year of 2016, PC DRAM will account for about 20% or less than that. So, no more than 20% out of the production, while -- whereas mobile, it's going to be over 40%. And I believe that the gap between these two is going to become even bigger in the second half.

Cha Seong-Geun

[Interpreted] We will take two more questions because of the time constraint.

Operator

[Interpreted] The next questions will be presented by Mr. JJ Park from JPMorgan. Please go ahead, sir.

JJ Park - JPMorgan

[Interpreted] I have one question each for DRAM and NAND. Now for DRAM, in the second quarter, the margin has dropped, and it was explained that the cost for the non-PC DRAM is higher. Now then in terms of the profitability in the second quarter, now even the PC DRAM prices have fallen. So, does this mean that the profitability is still better than the non-PC DRAM products?

Unidentified Company Representative

[Interpreted] Now I will try to answer the question to the extent that I have understood it. Now in terms of the profitability by application up to the second quarter, I would say that the non-PC side is higher than the PC side, but then this operating margin gap is reducing as we move along from the second quarter to the third quarter.

And looking at the blended margin then, we see that the -- as of the end of the second quarter, now, we do have the margin that we have seen, but then as we move into the third quarter and the remainder of this year, we believe that the fall in the price will be lesser than the cost reduction that we can achieve, meaning that we will be able to improve on the operating margin going forward.

JJ Park - JPMorgan

[Interpreted] Now my next question is about the NAND. So, what is SK Hynix's overall strategy for 3D NAND? And you mentioned that your target capacity by the end of this year is 20,000 to 30,000, which seems quite small-scale compared to the other players, and so what is your target capacity for 3D NAND by the end of 2017? And then another question in relation to that is, now, there is prediction that the NAND market size is going to be bigger than DRAM market by next year, but it seems as if SK Hynix's CapEx so far has been mostly on the DRAM side. So I wonder what is your -- what is the size of your CapEx in NAND. And also for -- as we move more towards SSD, then it seems as if the controller solution will become more important. So I wonder, for the SSD going forward, are you going to use your self-developed proprietary controller or are you going to keep using the controller solution from Company S [ph]?

Unidentified Company Representative

[Interpreted] Now three questions in all it seems. Now first, about the 3D NAND. Our plan for next year was the question. Now for this year, we -- it seems as if we are going to outperform the industry in terms of the bit growth. And if it is thanks mostly to the increase in the portion of TLC and 1Z nano, then that means that -- so we were able to improve the -- increase the bit growth over the other players, thanks to that. That means that we will continue with our bit growth into next year, mostly led by the 3D products.

And although I cannot tell you the exact portion of 3D NAND, but as for the end of 2017, our projection is that over 50% of the output will be in 3D NAND.

And now for the controller, on one hand, yes, it is true that the applications for NAND continue to change and increase, then also for the same NAND products, we see NAND products with different architectures. So for example, 3D NAND, which also requires different types of controllers. But now 3D NAND is still NAND. So, on one hand we plan to make active use of the existing NAND controllers. But then we do not plan to keep using the 2D NAND controllers at the sacrifice of performance or reliability, but we believe that up to fourth-generation products, we will be, for many of the SSD controllers, we will be using the Company's self-developed controllers.

And next was about the -- our planned CapEx for NAND. Now it was mentioned already that by -- our plan is to increase up to 30,000 capacity for 3D NAND by the end of this year. And for us to go beyond the capacity would require additional space. So in the second half we will work on trying to secure more space for additional NAND capacity. And once we do that, then, in the first half of next year, there would be additional CapEx into 3D NAND.

So for this year, out of the CapEx, there is also the construction cost included for -- in order to build out the space for NAND. So, overall the CapEx is about seven to three, meaning seven for DRAM and three for NAND. And for next year, the construction for NAND is not included, but even if we exclude that, the NAND would still account for about 30% to 35% of the CapEx.

Cha Seong-Geun

[Interpreted] This should be the last question.

Operator

[Interpreted] The last questions will be presented by Mr. Simon Woo from Bank of America Merrill Lynch. Please go ahead, sir.

Simon Woo - Bank of America Merrill Lynch

[Interpreted] Now first of all, congratulations on the good performance despite the adverse condition. Now my question is about the DRAM and it seems as if your mass production to -- for 2Z and 21 nano, in terms of the, let's say, the migration to finer [ph] technology, it seems your mass production of 2Z and 21 nano, let's say, compared to one year ago or six months ago, by SK Hynix, or even compared to other players like those in the U.S. or domestic, it seems as if the mass production of the 2Z and 21 nano is not -- still not very high. So what do you believe are the barriers against it? And when do you believe that those areas would be eased and resolved so that you will be able to reach 40% of the DRAM by the end of this year? And when do you believe that you could reach up to a 60% or 70%? And also, when do you believe that you can also move on to 1x node?

And I believe that you would have to inevitably use the EUV, so, when do you think you would be able to reach 100 per hour in order to use this on the critical layer as well, up to like, for example, 10 nano? So, do you have a roadmap for this as well? Can you also share that with us? Thank you.

Unidentified Company Representative

[Interpreted] Now in terms of the delay in the ramp-up of the production for 2Z nano, of course, on one hand, yes, there was, let's say, some technological gap at the company itself, but I would say that the bigger cause is the change in the market demand or the market environment, meaning that in the beginning we were focusing on developing the 2Z for PC DRAM, but then, now the demand is shifting more towards the server and the mobile side, which actually requires technology for more challenging products. So in order for us to do the mass production for these server and the mobile side, which are technologically tougher, this is actually requiring more time.

So one evidence of that is, now, of the 2Z nano, which is planned to take up about 40% by the end of this year, initially our plan in terms of the order or the volume of production was PC, server and mobile. But then now the order in the volume has changed from -- to mobile and then server. And also, looking at the total percentage, then out of the total, 20 nano would take about 40%, but then out of mobile it will be over 50%.

About the EUV, now up to 1x nano, the Company plans to do the development without the EUV by the -- until the end of this year for 1x nano. And then for the 1y nano, so we are also planning to start the development for the 1y nano. And for this, now the EUV that has been used for R&D, we might use this for some of the backup processes for some of the layers, but then whether the EUV will be adopted in full for 1y nano, that is something that we cannot say at this point. But then one thing is for sure, and that is we will use EUV for mass production of 1Z nano, and that is going to be in 2019.

Simon Woo - Bank of America Merrill Lynch

[Interpreted] Thank you very much for the very clear answers. And my last question is about the M14 wafer capacity. Is it up to 20,000 now? And by the end of next year, is it going to be up to 50,000, 100,000? What is the plan?

Unidentified Company Representative

[Interpreted] For M14, some of the facilities are new, some have been relocated from M10 and are being rearranged. So the capacity at M14 is up to 30,000 now. And as we migrate to 2Z nano, we believe that the M14's portion is going to increase.

Cha Seong-Geun

[Interpreted] Thank you. This concludes the SK Hynix 2016 second quarter earnings release conference call.

Copyright policy: All transcripts on this site are the copyright of Seeking Alpha. However, we view them as an important resource for bloggers and journalists, and are excited to contribute to the democratization of financial information on the Internet. (Until now investors have had to pay thousands of dollars in subscription fees for transcripts.) So our reproduction policy is as follows: You may quote up to 400 words of any transcript on the condition that you attribute the transcript to Seeking Alpha and either link to the original transcript or to www.SeekingAlpha.com. All other use is prohibited.

THE INFORMATION CONTAINED HERE IS A TEXTUAL REPRESENTATION OF THE APPLICABLE COMPANY'S CONFERENCE CALL, CONFERENCE PRESENTATION OR OTHER AUDIO PRESENTATION, AND WHILE EFFORTS ARE MADE TO PROVIDE AN ACCURATE TRANSCRIPTION, THERE MAY BE MATERIAL ERRORS, OMISSIONS, OR INACCURACIES IN THE REPORTING OF THE SUBSTANCE OF THE AUDIO PRESENTATIONS. IN NO WAY DOES SEEKING ALPHA ASSUME ANY RESPONSIBILITY FOR ANY INVESTMENT OR OTHER DECISIONS MADE BASED UPON THE INFORMATION PROVIDED ON THIS WEB SITE OR IN ANY TRANSCRIPT. USERS ARE ADVISED TO REVIEW THE APPLICABLE COMPANY'S AUDIO PRESENTATION ITSELF AND THE APPLICABLE COMPANY'S SEC FILINGS BEFORE MAKING ANY INVESTMENT OR OTHER DECISIONS.

If you have any additional questions about our online transcripts, please contact us at: transcripts@seekingalpha.com. Thank you!