Sumco Company (OTCPK:SUMCF) This fall 2023 Earnings Convention Name February 14, 2024 1:00 AM ET
Firm Individuals
Takayuki Komori – Investor Relations
Mayuki Hashimoto – Chairman and Chief Govt Officer
Michiharu Takii – Vice President and Chief Monetary Officer
Hiroshi Ito – Govt Officer and Basic Supervisor, Accounting
Convention Name Individuals
Takashi Enomoto – BofA Securities
Takato Watabe – Morgan Stanley
Atsushi Ikeda – Goldman Sachs
Shigeki Okazaki – Nomura
Yuta Nishiyama – Citigroup Securities
Yu Yoshida – CLSA Securities
Mikiya Yamada – Mizuho Securities
Takayuki Komori
Thanks in your participation in the present day. That is the outcomes briefing for the Fourth Quarter of the fiscal yr ending December 2023.
Earlier than beginning the presentation, permit me to substantiate in the present day’s supplies, which consists of three gadgets: the temporary assertion on consolidated monetary outcomes for fiscal 2023, the announcement concerning revision to dividend forecast and the presentation deck entitled Outcomes for Fiscal Yr 2023. This will probably be a 60-minute briefing, which is able to finish at 5 p.m.
Subsequent, a disclaimer. The estimates, expectations, forecasts and different future data mentioned right here and proven in in the present day’s supplies have been ready primarily based on the data out there to the corporate as of in the present day and on sure assumptions and {qualifications}, together with our subjective judgment. Precise monetary efficiency or outcomes might differ considerably from the long run data contained on this materials as a result of threat components, together with home and world financial circumstances, traits within the semiconductor market and international change charges.
We could have displays in the present day from Consultant Director, Chairman and CEO, Mayuki Hashimoto; and Vice President, CFO, Michiharu Takii. Hiroshi Ito, Govt Officer and Basic Supervisor of Accounting can be readily available. Chairman and CEO Hashimoto will talk about our forecast and working setting to be adopted by an evidence of the monetary outcomes by CFO Takii. We now have put aside time for a Q&A session as effectively.
I’ll now hand over to Chairman Hashimoto.
Mayuki Hashimoto
I’m Chairman Hashimoto. I’ll begin with the overview on Slide 5 of the presentation.
This can be a abstract of the outcomes for the fourth quarter of the fiscal yr 2023. Earnings have been effectively forward of our forecast. The primary components have been a optimistic contribution of round JPY1 billion from the weaker yen, with the remaining roughly JPY3 billion coming from better-than-expected value reductions. Because of this, fourth quarter working earnings overshot our forecast by JPY4.7 billion, and unusual earnings exceeded our forecast by JPY3.7 billion.
Turning to the earnings forecast for the primary quarter of 2024. We venture a big sequential decline in gross sales. On the again of this, working earnings is projected to be JPY4.5 billion; unusual earnings, JPY4 billion; and web earnings, JPY2 billion. We assume a dollar-yen price of JPY145 to the greenback. The primary cause for the sequential declines is the anticipated drop in gross sales. I’ll go into extra element later, however our prospects are presently carrying stock ranges effectively above regular and are at a degree the place they’re not able to simply accept extra wafers as a result of an absence of accessible storage base, therefore, the adjustment by way of cargo volumes.
Please flip to Web page 6. This slide reveals shareholder returns or dividends per share. Reflecting the better-than-expected outcomes, we now have raised our steerage for our fiscal year-end dividend per share from JPY10 to JPY13 for a full yr complete of JPY55. This represents a dividend payout ratio of 31%. Word that we now have excluded subsidies obtained of JPY1.7 billion after tax from the calculation of the dividend as it isn’t acceptable to make use of prefectural subsidies to pay dividends.
Subsequent web page, please. fourth quarter market circumstances for silicon wafers, each logic and reminiscence prospects continued with manufacturing changes, resulting in a continuation of low cargo ranges for 300-millimeter wafers. Equally, for 200-millimeter and smaller diameter wafers, that are utilized in quite a lot of functions, reflecting the general weak point of market circumstances, prospects proceed to regulate manufacturing, miserable cargo ranges. That mentioned, prospects revered LTA value ranges for each 300- and 200-millimeter wafers. So SUMCO was in a position to obtain regular value will increase consistent with LTA costs.
On the outlook for 300-millimeter volumes in first quarter 2024, whereas there are some prospects which have seen slight recoveries, there are others which have seen an extra deterioration, so the image is blended. On an total foundation, circumstances have improved, however we predict that fourth quarter into first quarter is prone to be the underside with a gradual restoration to observe, though the momentum of the restoration is unlikely to be robust, in my opinion. It’s because thus far, prospects have complied with the phrases of the LTAs, persevering with to buy wafers, which has resulted in important buyer stock ranges. As such, we now have no selection however to cooperate with buyer requests to delay deliveries.
The entire volumes below the LTAs stay unchanged, however prospects are asking to barely push out the timing of deliveries. We now have agreed to such request to a sure extent. For 200 millimeter and smaller diameters, prospects proceed to regulate manufacturing ranges so shipments are anticipated to say no additional. Not like 300 millimeter, indicators of a restoration have but to emerge. Nonetheless, if smartphone handsets begin to recuperate, provided that 200-millimeter wafers are utilized in many smartphone functions, then we’d anticipate to see a restoration in 200-millimeter wafers as effectively. That mentioned, the restoration in smartphones continues to be weak. When it comes to costs, LTA costs proceed to be revered. However costs for spot, particularly for smaller diameters in sure areas for sure functions, are softer for commodity product. The exception is extra technologically troublesome functions the place costs stay comparatively agency.
Trying on the outlook, investments are prone to stay robust for information facilities for generative AI and agency for EV and energy-related functions. On high of this, the bottoming out of smartphone and PC demand is contributing to indicators of sunshine on the finish of the tunnel. Nonetheless, the restoration in wafers sometimes lags a restoration in finish merchandise. Specifically, this time round, prospects are carrying important ranges of stock, which suggests it might take time earlier than we see a restoration in wafers. I consider it’s prone to take till the second half of this yr.
That mentioned, as I’ve been saying at every of our current conferences, 300-millimeter epitaxial wafers are utilized in made-to-order merchandise whereas 300-millimeter polished wafers are used for reminiscence, which is often produced in anticipation of demand and, due to this fact, tends to overshoot. So it’s essential to consider not solely wafer stock ranges however finish product stock ranges as effectively, which look like excessive. As such, a restoration in polished wafers is prone to take extra time than epi in my opinion.
Web page 8, please. This slide reveals the wafer pattern for 200 millimeter by quarter. Sadly, volumes have fallen under 2016 ranges. We consider a restoration will take time.
Web page 9, please. That is the pattern for 300 millimeter. I talked earlier about circumstances for 300 millimeter. The primary half of 2024 will in all probability be the underside for wafers, however I believe we must always subsequently see a gradual enchancment. Clearly, there will probably be a slight distinction in pattern between epi and PW. I anticipate it’ll take longer earlier than we see a restoration in PW, however I think about we must always see a comparatively early restoration in epi.
Web page 10, please. I’d now like to speak concerning the future for silicon wafers. I’ll clarify the 4 themes proven right here so as.
Slide 11, please. I’ll begin with AI servers and the way they’re structured. You could already be aware of this, however an AI server consists of a GPU and HBMs, that are DRAMs. Fabricating DRAM HBMs is technologically difficult. There is just one participant that may fabricate HBMs in massive volumes. Though there are different gamers that may do small volumes, the highest participant produces considerably greater volumes. The chips are mixed on a silicon interposer as effectively.
As such, AI servers will devour an enormous quantity of silicon, driving a big improve within the space of silicon required. A single AI server is estimated to devour 1.8 wafers. In comparison with a standard server, an AI server consumes 3.4x as a lot silicon. Not too long ago, smartphones have begun to introduce generative AI. This, mixed with the emergence of 3D structure, counsel silicon consumption is prone to develop considerably in future.
Slide 12, please. This chart reveals the projected demand development for server volumes. Unit volumes are anticipated to extend as proven right here. As there was a big decline final yr, we anticipate to see a slight restoration this yr. Nonetheless, we consider a full-scale restoration will solely are available in 2025. We consider that the proportion of AI servers will improve on the again of the rise in total volumes. We venture a CAGR of 26% for AI servers in comparison with standard server quantity development of 5%.
Slide 13, please. This slide seems at how advances in AI will drive demand for modern course of wafers. AI servers, in fact, use modern logic at 3 or 2 nanometers, so development in AI servers will due to this fact drive demand for modern course of wafers. Additionally, AI servers devour 3.4x the silicon in floor space phrases in comparison with standard servers. Given this, we anticipate AI servers will rise to account for half of the demand for wafers utilized in servers within the close to future.
Slide 14, please. Turning now to smartphones. We’re seeing a sudden improve within the adoption of AI in smartphone handsets from this yr. We predict it’s affordable to anticipate that, over time, AI-enabled smartphones will come to account for half of complete quantity. Assuming that is the case, the variety of parameters or calculation quantity, which is presently 10 billion, ought to double by 2026 and proceed to extend subsequently. Given the necessity to course of enormous volumes of knowledge at excessive speeds, this could drive demand for modern wafers utilized in CPUs, GPUs and DRAM, successfully considerably boosting wafer consumption.
Slide 15, please. This reveals buyer wafer stock ranges for 300-millimeter wafers. The white bars signify buy volumes whereas the blue bars signify wafer enter volumes and the inexperienced bars are stock. Present buyer stock ranges are at document highs. Even now buy volumes proceed to exceed enter volumes, so prospects at the moment are at a degree the place they are saying they can not purchase any extra wafers. On the identical time, wafer purchases can not go to 0 both. So we’re engaged with our prospects in fascinated with the best way to return to extra regular ranges of stock. The impression of buyer stock changes is a significant component behind our forecast for a decline in first quarter income to JPY87 billion. It’s not a sign of a deterioration in market circumstances. As an alternative, the market is bottoming out in fourth quarter into first quarter. We don’t anticipate to see an extra deterioration of market circumstances.
Slide 16, please. This reveals the breakout of buyer inventories between logic and reminiscence. You’ll be able to see that wafer stock ranges are equally excessive.
Slide 17, please. This chart reveals traits for logic-use epi wafers. The orange line is stock. Stock is at document excessive ranges. Nonetheless, the grey line, which is value, has been rising steadily consistent with the LTAs. That mentioned, demand has fallen barely, though not too long ago, there are indicators of a attainable rebound after changes in first or second quarter.
Slide 18, please. This is similar chart for PW. Though there was discuss that reminiscence costs have been rising, this displays a decline in enter volumes, which has led to a slight tightening of the market slightly than costs rising on the again of elevated demand. Given this, it’s possible it’ll take a while for a restoration to emerge.
Slide 19, please. So by way of how we see the market over the longer-term, we anticipate the primary half of this yr to be very powerful, though we anticipate a rebound within the second half of the yr. On a full yr blended foundation, we’d hope to match 2023 ranges, though it could be barely weaker on a year-on-year foundation. We’re presently in a transition interval, which makes it powerful to learn. What we will say is that the market has bottomed. The energy of the restoration will hinge on the energy of generative AI-related demand in addition to what occurs with China, given the sheer scale of buying energy. The energy of substitute demand for smartphones and different finish merchandise in China may even be an essential issue.
This completes my part of the presentation. I’ll hand over to CFO Takii to speak about particulars of our earnings.
Michiharu Takii
I, Takii, will current the earnings in additional element. Please flip to Slide 21. This slide reveals the outcomes for fiscal 2023. On a year-on-year foundation, gross sales declined however consolidated gross sales, embrace gross sales of polysilicon from second quarter onward which suggests the declines in wafer gross sales, was really bigger. Working earnings additionally fell year-on-year to JPY73 billion. Atypical earnings was JPY72.6 billion. Nonetheless, we acknowledged a unprecedented achieve in first quarter associated to the acquisition of the polysilicon enterprise, so the magnitude of the revenue decline is smaller on the web revenue degree at JPY63.8 billion.
CapEx was up considerably year-on-year on continued full-scale Inexperienced Subject investments in each Japan and Taiwan on an acceptance foundation. Equally, depreciation additionally rose on the again of investments. The primary improve was in property. Because of this, EBITDA was down on a year-on-year foundation, however we now have been in a position to keep an EBITDA margin within the 30% vary. On a full yr foundation, ROE was 11.6%.
Please flip to Slide 22. That is the evaluation of modifications in working earnings. First, the sequential evaluation. As famous on the outset, fourth quarter earnings have been a lot better than anticipated. As proven within the evaluation, prices improved by JPY4.3 billion Q-on-Q. Initially, we had anticipated to see an enchancment of JPY1 billion in prices in fourth quarter associated to decrease electrical energy prices, however as well as, we have been in a position to obtain JPY3 billion in value reductions by means of inside efforts to scale back bills, decrease supplies prices and upkeep bills and different initiatives. Depreciation was in-line with plan at minus JPY2.5 billion. Gross sales-related variance was minus JPY7 billion. The explanation for the decline in sales-related variance, regardless of gross sales remaining largely unchanged, is a mirrored image of manufacturing slightly than gross sales.
I had beforehand talked about our plans to bear common upkeep at our mainstay crystal ingot vegetation in October and November in addition to at a processing plant. This depressed capability utilization considerably. As well as, associated to the common upkeep, we additionally selected to scale back work-in-progress stock considerably in anticipation of a considerable drop in shipments in first quarter. The weaker yen generated a optimistic of JPY1.3 billion.
Trying on the year-on-year change for the complete yr outcomes, as proven on the precise, working earnings fell JPY36.6 billion. Prices elevated JPY15.5 billion, which breaks down right into a JPY9 billion unfavorable for supplies unit costs, a JPY7 billion unfavorable from electrical energy prices and a rise in labor prices, notably in first half when capability utilization was excessive, of JPY2 billion. We have been in a position to offset the upper value with inside value discount efforts of JPY2 billion. Depreciation elevated JPY11.9 billion. Gross sales-related variance was a unfavorable of JPY22.8 billion, which breaks down right into a JPY22 billion optimistic from greater costs associated to the beginning of a brand new fiscal yr, offset by a JPY45 billion unfavorable because of decrease volumes. The weaker yen had a optimistic impression of JPY13.6 billion year-on-year.
Slide 23, please. Slide 23 reveals the steadiness sheet. I’ll contact upon this later, however free money circulation was considerably unfavorable in fiscal 2023. We tapped into money and time deposits, decreasing it by JPY100 billion. Uncooked supplies and provides rose, reflecting a rise in provides as part of initiatives to strengthen BCP. Tangible and intangible belongings elevated JPY250 billion. That is the impression of the Inexperienced Subject CapEx of greater than JPY300 billion on an acceptance foundation mentioned earlier much less depreciation.
Curiosity-bearing debt elevated roughly JPY83 billion year-on-year. This breaks right down to new borrowings at three way partnership FST of JPY50 billion and at SUMCO in Japan of JPY30 billion. If we take a look at the capital account, retained earnings elevated, reflecting the impression of revenue development. As proven on the backside of the desk, the fairness ratio declined barely, however fairness per share elevated. On the money circulation assertion to the precise, the sub complete of pretax earnings, extraordinary earnings and depreciation was JPY144 billion. Nonetheless, there was a rise in stock, and below Others, tax funds have been comparatively massive. The ensuing working money circulation was barely lower than JPY100 billion.
Capital expenditures have been JPY310 billion, break up between roughly JPY200 billion in Japan and JPY100 billion for FST in Taiwan. In response to the unfavorable free money circulation, we undertook new borrowings and tapped into money and time deposits, therefore the decline in money readily available.
Slide 25, please. On Web page 25, we present our forecast for first quarter. We anticipate a big sequential drop in gross sales. Working earnings is predicted to be roughly half Q-on-Q. The figures are as proven on the desk. Depreciation is projected to be JPY18 billion, down sequentially reflecting the beginning of a brand new fiscal yr. As you’ll be able to see from the assorted metrics, we predict a deterioration in profitability.
Slide 26, please. Slide 26 reveals the evaluation of change in working revenue. First, the primary quarter sequential change on the left. With regard to the Q-on-Q decline in working earnings of JPY6.7 billion, we predict a slight improve in prices, reflecting ongoing will increase in unit value of supplies. Depreciation, as famous earlier, is predicted to say no Q-on-Q, reflecting the beginning of a brand new fiscal yr. Gross sales-related variance displays the numerous anticipated drop in gross sales. This contains an anticipated drop in non-LTA spot costs of round JPY2 billion, primarily in smaller diameters. We additionally think about a JPY6 billion decline to replicate decrease volumes. On the again of decrease gross sales, we venture a deterioration within the profitability of gross sales. Our ForEx assumption is for a barely stronger greenback yen at JPY145 to the greenback. Close to-term, the yen is barely weaker, however that is topic to vary.
Trying on the evaluation of year-on-year change for first quarter, we anticipate important declines in each gross sales and earnings in first quarter. On prices, we anticipate to considerably scale back labor prices in first quarter for a year-on-year enchancment. Electrical energy unit costs are additionally anticipated to enhance, however this will probably be offset by a slight improve in supplies value. On a year-on-year foundation, depreciation is predicted to extend for a unfavorable of JPY3.6 billion. Of the sales-related variance, as famous earlier, spot costs are falling, so JPY2 billion of the decline is said to cost. The remaining roughly JPY20 billion unfavorable displays the general massive drop in gross sales volumes. We anticipate a optimistic impression from the weaker yen.
Web page 28, please. This reveals EBITDA and gross sales. As you’ll be able to see, we anticipate gross sales to fall sharply in first quarter. If we assume this would be the backside, we anticipate to see a reversal of the pattern in second quarter and third quarter of 2023. That is our assumption.
This completes my part of the presentation.
Takayuki Komori
Thanks. We’ll now open the ground to questions.
Query-and-Reply Session
A – Takayuki Komori
We’ll begin with Mr. Enomoto of BofA Securities.
Takashi Enomoto
Earlier than asking my query, I want to take this chance to thank Mr. Takii for his service. Thanks a lot. Transferring on to my query, are you able to speak about how we must always take into consideration costs in 2024? Beforehand, you had indicated that you simply anticipate unit costs would rise in 2024. However primarily based on the evaluation of change in earnings offered in the present day, you might have factored in a decline in spot costs and look like anticipating LTA costs to be flat. Is that this due to the pushouts in supply timing for LTAs? So for earlier deliveries which were pushed into 2024, are you anticipating costs is not going to rise? Please speak about how we must always take into consideration LTA pricing and why costs are usually not anticipated to rise.
Mayuki Hashimoto
At a excessive degree, though there’s variance in particular person contracts, LTAs might be categorized into two teams. These have been the costs fastened over a contract interval of, say, 5 years and people the place there’s a step-up characteristic the place the costs rise yearly, for instance. For fastened value contracts, a delay in deliveries has no impression for the reason that value is similar. For step-up contracts, nonetheless, if a supply initially scheduled for 2024 have been to be pushed into 2025, in an excessive case, all the supply will probably be made at a later date however on the 2024 value level. Clearly, that’s an excessive instance that’s extremely unlikely. Nonetheless, the cumulative impression of gradual pushouts thus far is having an impression on the blended value level. So costs, which ought to have risen this yr, received’t rise a lot. For FST, the proportion of LTAs isn’t as excessive, so the decline in spot costs is having an impression. Because of this, on a web foundation, costs are usually not anticipated to rise.
Takashi Enomoto
With regard to beforehand pushed out deliveries, can we assume that it is possible for you to to realize the unique value when the precise cargo is made? So for delayed shipments in 2024, when the cargo is definitely accomplished, can we anticipate you’ll obtain the upper 2024 value?
Mayuki Hashimoto
Sure, that’s appropriate.
Takashi Enomoto
Understood. I’ve excessive expectations for the completion of stock changes. Thanks.
Mayuki Hashimoto
Me too.
Takayuki Komori
Thanks. Subsequent is Mr. Watabe of Morgan Stanley.
Takato Watabe
Based mostly on in the present day’s dialogue, if the remainder of the trade tracks in line-with SUMCO’s first quarter gross sales forecast, it appears trade cargo quantity for 300 millimeter will possible drop under the extent of 6 million wafers per 30 days. Do you consider that this displays the magnitude of the stock changes? When you’ve got a picture of the magnitude of the stock changes, both on a quarterly or month-to-month foundation, might you share it with us?
Mayuki Hashimoto
SUMCO does enterprise with nearly all the trade gamers, and I consider that the purchasers deal with all of their suppliers equally and are usually not requesting cargo changes simply from SUMCO. In such circumstances, prospects are prone to be as equitable as attainable. So I consider that the trade as a complete is in an analogous place to us.
Takato Watabe
Is it troublesome to speak about particular numbers at the moment?
Mayuki Hashimoto
When it comes to a selected picture, there are some prospects that anticipate a restoration from April, so it’s exhausting to say. One factor that I can say, nonetheless, is that first quarter will possible be the underside by way of wafer volumes. Clearly, this might change if there are new developments. Not too long ago, there have been so many alternative geopolitical components which makes it very troublesome to foretell. However barring some unexpected circumstances, I consider first quarter needs to be the underside, and that is the view rising from our discussions with prospects.
Takato Watabe
Understood. Thanks. Additionally, many because of Mr. Takii.
Michiharu Takii
Thanks in your variety feedback. I didn’t take the chance to say it within the presentation, however thanks all.
Takayuki Komori
Thanks. Subsequent is Mr. Ikeda of Goldman Sachs.
Atsushi Ikeda
Are you able to present slightly extra colour on the LTA pricing, please? At the moment, prospects are carrying excessive ranges of wafer stock of round three months and are additionally going through challenges to profitability. In these circumstances, with the ability to keep flat pricing could also be the very best you’ll be able to hope for. Having mentioned that, there are contracts in place below which you’d hope to see costs rise. If the trade as a complete makes robust progress on stock changes, reducing stock to round 2x months, wouldn’t it be simpler to start negotiating for greater costs from second quarter? Moreover, can we anticipate the costs within the September quarter might rise by round 5%? Are you able to touch upon the steadiness between LTA costs and stock ranges?
Mayuki Hashimoto
LTAs might be usually grouped into two sorts, contracts with fastened flat pricing and contracts the place costs step up over time. Very simplistically, for the step-up contracts, utilizing a 3-year contract of three million wafers for instance, sometimes, there can be one value level for this yr, a better value level for the subsequent yr and so forth. Hypothetically, if the contract is for the acquisition of 1 million wafers this yr, though the push-outs usually are usually not as excessive, if we assume that supply of 500,000 wafers is pushed into subsequent yr, it’s troublesome for us to ask for the shopper to pay the 2025 value level for the pushed out 500,000 wafers, provided that we all know the shopper is struggling. In such circumstances, we ask the shopper to respect the worth level set out within the contract. Clearly, we’d not agree to chop costs, however agreeing to simply accept a push-out signifies that the anticipated improve in value additionally will get pushed out. That’s what’s taking place now. Nonetheless, there are numerous contracts the place costs are fastened for the lifetime of the LTA. There isn’t a lot impression from push-outs on these contracts. Yet another factor that complicates the image is the spot enterprise at FST in Taiwan. Though non-consolidated SUMCO is 100% LTAs, FST sells a few of its wafers at spot costs, notably when doing enterprise with associates. Negotiations with associates, the place FST is prioritizing the sale of wafers, might be difficult. In conditions the place associates are being requested to purchase, value factors can fluctuate. So, that is what’s taking place now. Nonetheless, if we take a look at the general image for the complete yr this yr, on a blended foundation, I don’t suppose that costs will fall on a year-on-year foundation. Costs might rise barely within the second half of the yr, however I wouldn’t anticipate to see full yr blended costs rise in the identical means as they did in 2023. That’s my view.
Atsushi Ikeda
Understood. So, costs might rise barely this yr, however might rise extra considerably by round mid-2025. If we take a look at costs on the finish of 2025 or within the first half of 2026, do you suppose they may revert again to the initially anticipated ranges?
Mayuki Hashimoto
Sure, precisely.
Atsushi Ikeda
Understood. I’ve excessive hopes for SUMCO. Thanks.
Mayuki Hashimoto
The way in which to consider it’s that all the wafers have value factors. So, there’s a value for the primary 1 million, one other value for the subsequent 1 million and so forth. The rise within the blended value is impacted by the timing of deliveries.
Atsushi Ikeda
Understood. Thanks.
Takayuki Komori
Thanks. Subsequent is Mr. Okazaki of Nomura.
Shigeki Okazaki
Thanks to CFO, Takii. I need to ask about gross sales quantity and manufacturing quantity. Three months in the past, you indicated that you’d be adjusting manufacturing to scale back your individual stock in fourth quarter following a cancellation, however urged gross sales and manufacturing quantity would return to regular within the March quarter. To what do you attribute the hole between what you might have mentioned final time and this time? Is it as a result of prospects are closely targeted on decreasing stock?
Mayuki Hashimoto
I don’t recall suggesting that stock volumes would rebound from January. What I mentioned was that we have been anticipating first quarter to be very powerful.
Shigeki Okazaki
What I understood was that after the manufacturing cuts you deliberate for the December quarter, you anticipated to revert to regular ranges within the March quarter.
Mayuki Hashimoto
When it comes to the manufacturing cuts, provided that complete quantity had fallen, we didn’t need to carry extra stock, so we lowered our stock to an acceptable degree by reducing manufacturing quantity. In the event you understood that to imply that we supposed to extend manufacturing quantity from January, then there might have been a misunderstanding. Fourth quarter into first quarter is the underside for the market itself. When it comes to what is occurring in first quarter, though there are prospects that decreased quantity additional, it’s extra that there have been finish customers which can be decreasing wafer consumption additional in first quarter. Total, first quarter might be flat to barely decrease on a sequential foundation. Epi, specifically, is down. So, first quarter into second quarter might be the underside in my opinion. We now have aligned our manufacturing to present circumstances.
Shigeki Okazaki
With regard to second quarter on a sequential foundation, as you talked about earlier, the image for the June quarter continues to be blended, is that appropriate, and also you anticipate a restoration in gross sales or manufacturing volumes to come back within the September quarter?
Mayuki Hashimoto
That’s true for epi. For SUMCO, the proportion of epi is greater, so I consider we must always see a strong restoration. PW, nonetheless, is prone to take longer to recuperate. However even for epi, there are a number of shoppers, a few of that are doing effectively whereas others are usually not. The winners are doing comparatively effectively, however there are some prospects which can be actually struggling. The general image is blended, so it’s troublesome to generalize. That mentioned, I consider that usually epi will recuperate within the second half of the yr.
Shigeki Okazaki
Understood. Thanks.
Takayuki Komori
Subsequent is Mr. Nishiyama of Citigroup Securities.
Yuta Nishiyama
I want to ask concerning the outlook for wafer demand. I can see on Web page 15 that wafer enter began to recuperate from the top of final yr. If I take a look at the supply-demand chart on Web page 19, SUMCO has revised up its forecast for 2024 by 7% relative to a few months in the past. As well as, the wafer trade, as a complete, is enterprise important cargo changes. This implies that the normalization of stock might occur sooner than was anticipated three months in the past, in my opinion. Are you able to remark in your view of those factors?
Mayuki Hashimoto
I utterly agree. When it comes to the normalization of stock, enter ranges ought to improve, though I’d view first quarter as the underside. Nonetheless, circumstances fluctuate considerably by buyer. The image is blended, which makes it troublesome to generalize. What I can say is that, total, first quarter is prone to be the underside and that we must always begin to see restoration from second quarter. Nonetheless, there will probably be prospects the place the restoration will probably be gradual whereas prospects with strengths in DRAM or AI are prone to see a strong rebound. Nonetheless, different functions, reminiscent of smartphones, are nonetheless weak. I’ll say it is rather troublesome to generalize.
Yuta Nishiyama
I see. If you speak about first quarter as the underside, are you referring to gross sales? Earnings are prone to be impacted by a rise in depreciation, I believe.
Mayuki Hashimoto
I’m speaking about gross sales.
Yuta Nishiyama
Understood. Thanks.
Takayuki Komori
Subsequent is Mr. Yoshida of CLSA.
Yu Yoshida
I need to ask about Slide 19 and the supply-demand steadiness. As talked about in a earlier query, you might have revised up your demand forecast for 2024 from a unfavorable development to flat this time. Please speak about the place you noticed the biggest change in your view by way of functions. If attainable, are you able to body your remark by way of epi. And for reminiscence, might you distinguish between DRAM and NAND? Are you able to additionally share your assumption for buyer wafer stock ranges as of the top of 2024, which underpins the flat forecast, i.e., what number of months of stock have you ever assumed?
Mayuki Hashimoto
We don’t have a digital learn of buyer stock ranges, so our evaluation doesn’t go that deep. The explanation why we barely revised our forecast to flat is as a result of we anticipate an earlier restoration in epi. We additionally factored in an early restoration in DRAM, given the numerous use of DRAM in generative AI.
Yu Yoshida
I see. Can I additionally ask what you’re assuming by way of the speed of change in manufacturing, both upward or downward, so as to obtain this forecast for flat shipments?
Mayuki Hashimoto
You imply the forecast for flat shipments, on an total foundation we anticipate it to be flat, though we predict a comparatively sharp enchancment within the second half of this yr. Clearly, I might be mistaken, however actually, we would want to see a comparatively strong restoration in smartphones to realize this. That mentioned, there will probably be a rise in smartphones enabled for AI. I additionally anticipate to see proactive development in AI information facilities. Nonetheless, as , the shopper’s manufacturing capability for these functions is proscribed. They’re working very exhausting to extend capability. Happily for us, we’re a significant provider to this buyer, so I believe it hinges on this buyer. I don’t anticipate to see a dramatic restoration in standard epi wafers, however do anticipate to see a speedy restoration in epi wafers for AI functions. I believe it’ll hinge on this.
Yu Yoshida
Understood. Thanks.
Takayuki Komori
Subsequent is Mr. Yamada of Mizuho Securities.
Mikiya Yamada
CFO, Takii, thanks for every thing through the years. I want to ask about how to consider LTA costs, apologies for being persistent. There was a dialogue of particular costs for 2024 wafers, however the cause why total costs are usually not at these ranges now could be due to the cumulative impression of earlier push-outs. So, in fascinated with value ranges for 2025 and 2026, ought to we anticipate an analogous pattern?
Mayuki Hashimoto
The unique contracts stipulate quantity and value for annually. So, as an example, the contract would say x variety of wafers for 2023 at a selected value. Nonetheless, the contract doesn’t state a value level for the remaining wafers within the occasion that not all the wafers for a given yr are delivered in that yr. When prospects ask to delay shipments, it’s as a result of they’re struggling so it’s troublesome for us to say that the 2023 value is not legitimate for delayed shipments and that the worth level is now greater. It’s troublesome to interact in any such dialogue until each events are doing effectively. Moreover, we now have been concerned in joint improvement up to now as effectively, so it makes it troublesome to take a purely self-interested strategy. I additionally acknowledge that the shopper in all probability desires to ask for a lower cost, however out of respect for his or her commitments, they don’t. They really feel dangerous about having to delay shipments. On this scenario, each side find yourself discovering center floor by sharing the ache. It’s not specified within the contract, however there’s an understanding that the portion of 2024 deliveries that’s pushed out will probably be priced on the 2024 value level. Push-outs within the subsequent yr will observe the identical sample. The reverse can be true. If the shopper ask to take extra quantity earlier, that portion can be on the greater value level. That is the unstated understanding that we now have with our prospects.
Mikiya Yamada
On this scenario, given that there’s continued progress being made on course of improvement, I’m assuming the diploma of issue for wafers in 2025 and 2026 will improve, does what you might have simply talked about imply that it’s essential to promote extra superior wafers at decrease costs?
Mayuki Hashimoto
That’s completely different, we set costs for extra technologically difficult wafers rent to start with, so that’s completely different.
Mikiya Yamada
In different phrases, the wafers are completely different, so the costs are completely different as effectively. A extra superior wafer is offered at a better value, is that appropriate?
Mayuki Hashimoto
Sure.
Mikiya Yamada
Understood. Thanks. I’ve excessive expectations for an early normalization of stock ranges.
Mayuki Hashimoto
Me too. Please help us.
Takayuki Komori
Thanks in your questions. Thanks to everybody for becoming a member of the fiscal 2023 outcomes briefing. We’re grateful in your participation in the present day. We’ll finish the assembly right here.
Mayuki Hashimoto
Thanks.
Michiharu Takii
Thanks.