Would you please explain the context for leaving the FY2019 financial estimates announced in April unchanged? Several customers appear to have pushed out their CY2018 investment plans. Is it because such investments would still fit within the time span for TEL’s FY2019 (ending March 2019) and consequently have no impact on the financial estimates? Or is it because the financial estimates are carefully examined half yearly rather than quarterly and consequently not changed at this time?
Whereas we announced an outlook for the CY2018 WFE*1 market of 15% YoY growth in April, we have now changed this to a 10 to 15% YoY increase. This is attributed to customers pushing out their investment plans due to a slight delay in progress for miniaturization and higher yields. However, it is not as though demand has declined, and our April forecast for an active WFE market over the medium- to long-term has not changed.
Furthermore, we believe that financial estimates should be changed when necessary but not changed without reason. There are past examples of us making changes when releasing the Q1 financial announcement. Careful examination is not done only semi-annually.
Should we assume that the 5-point reduction in the outlook for the CY2018 WFE market means there is the same level of fluctuation risk in TEL’s FY2019 results as well? In that case, which applications are more at risk?
Investment has been pushed out or pulled in for all applications. This was taken into consideration when leaving the financial estimates unchanged.
The chart on Page 21 of the Presentation Materials indicates the forecast composition ratio for non-volatile memories has increased from when it was announced in April. In which generation of 3D NAND is investment increasing?
The chart is not of the WFE market but indicates the forecast composition ratio by application of TEL’s sales. Results are starting to show in areas where we have been focused on such as etch systems and cleaning systems for 3D NAND, and we expect this will contribute to sales.
Would you provide the quarterly forecast of the second half sales for the semiconductor production equipment business? Conjecturing based on past trends, will Q4 be higher than Q3?
We do not release quarterly sales forecasts, yet we do not expect the trend in FY2019 to differ greatly from the past.
What are the quarterly forecasts for the second half sales in the flat panel display (FPD) production equipment business?
I would like to refrain from commenting on quarterly forecasts.
From a medium-term perspective, we expect the FPD production equipment market for TFT array process*2 to remain firm throughout CY2018 to CY2020. We will focus on sectors in this market that generate high added value, taking initiatives to enable operating margins to reach 20%.
You have changed the CY2018 forecast for the WFE market from the original $58B to between $56 and $58B. Does this mean we can expect the $2B drop in CY2018 to result in a boost to the CY2019 forecast from the previous $61B to $63B? Or, have customer investment plans themselves disappeared?
We have forecast a lower limit of $56B in anticipation that customers will substantially push out their investment plans. Since our close examination for this Financial Announcement was only for CY2018, we have not revised the originally announced forecast for CY2019.
Note that customers have pushed out their investment plans for periods of about three months to half a year. This is mainly due to technological progress such as the yields in leading-edge miniaturization and increased number of layers, and the plans have not disappeared.
The forecasted WFE market for non-volatile memories in CY2018 has been increased from April. Is this increase in 3D NAND or other non-volatile memories?
The composition ratios for 3D NAND and other non-volatile memories have not changed from April.
The lower forecast for the CY2018 FPD production equipment market for TFT array process is attributed to customers pushing out their investment in G10.5. Is this due to technological issues or delays in building factories?
Some customers have revised their capital investment plans.
Is there high probability of the problem with customer yields being resolved in three months to half a year? Would you please explain any specific technological hurdles?
Although I am not in a position to be able to respond on our customer’s technological issues, I have heard that this is a matter of process tuning and the like, and not of having to considerably revise materials and structures. At the present, we think this will take about three months to half a year.
Are you concerned of the possibility of the balance of demand and supply problems arising around CY2019 with the gradual increase in supply of semiconductor devices as the customers’ current technological problems are resolved?
Demand from hyperscale data centers and such is very high in this age of big data centered on IoT. Consequently, there is also demand for technological innovation in areas such as large capacity, high speed, increased reliability and lower power consumption. Few customers can deal with such cutting-edge technological requirements and those that can have a large market share, and they understand the market trends well because they occupy a high share of the market. Since customers formulate their production plans while closely monitoring the market trends, we do not believe there is much need for concern about the possibility of future large changes in the balance of demand and supply.
What is your outlook for the price and profitability of NAND devices for FY2019 and beyond? Will customers continue to invest even if the price of NAND falls?
The focus of future NAND investment will be on 9X generation 3D NAND. While some customers will invest in new buildings, others will convert existing production lines to the 9X generation.
Setting aside emerging market customers in China, major semiconductor manufacturers basically establish their investment plans according to profits, EPS and operating margins. Even if customers adjust the timing of investments by about a quarter so that profit ratios and EPS do not fall when there is a sharp drop in the device price, considering the large growth in the data center market in the age of big data, the need for concern about bit demand appears to have disappeared for the near term.
What sort of circumstances would cause customers to pull in their investment plans?
I would like to refrain from specific cases, but investment could be accelerated if there were faster than anticipated improvement in yields for example. In addition, there could also be a quick response to diversifying semiconductor demand due to IoT.
Why are TEL’s sales so highly seasonal being low in Q1 and Q3 and high in Q2 and Q4? Is this intentional? The trend differs from your U.S. competitors.
We do not claim to have a full understanding of the state of other companies, but in general, we think U.S. companies record sales close to the time of shipment. Since shipping volumes probably do not change much from month to month, sales are also likely to be growing in a straight line.
However, in our case, sales are recorded on completion of set-up and testing of equipment, so in many cases, sales are recognized in a lump sum for each production line at a customer’s factory. Since sales recognitions are concentrated in a particular week or month, quarterly sales fluctuate up and down. This is not intentional.
*1 WFE (Wafer fab equipment): The semiconductor production process is divided into front-end production, in which circuits are formed on wafers and inspected, and back-end production, in which wafers are cut into chips, assembled and inspected again. Wafer fab equipment refers to the production equipment used in front-end production and in wafer-level packaging production
*2 TFT array process: The processes of manufacturing the substrates with the electric circuit functions that drive displays
* The above content is a summary of question and answers session.