Hello. I am Ken Hsiang, the Head of Investor Relations for ASE Technology Holdings. Welcome to our fourth quarter earnings release. Thank you for attending our conference call today. We apologize for not having our earnings release in our typical physical format. As with most people in Taiwan, we are acting with an overabundant caution in regards to minimizing the chance of exposure to the coronavirus. Please refer to our safe harbor notice on Page 2. Our lawyers have worked very hard on this disclosure. So please give it your utmost attention. All participants consent to having their voicing and questions broadcast via participation in this event. Please refer to our safe harbor notice. I would like to remind everyone on this call that the presentation that follows, may contain forward-looking statements. These forward-looking statements are subject to a high degree of risk, and our actual results may differ materially. For the purposes of this presentation, our dollar figures are stated generally in New Taiwan Dollars, unless otherwise indicated. As a Taiwan-based company, our financials are presented in accordance with Taiwan IFRS. Results presented using Taiwan IFRS may differ materially from other accounting standards. For today, I will be going over our business update and financial results. Afterwards, we will have a Q&A session with Joseph Tung, our CFO. As a reminder, because ASE Holdings was jointly formed on April 30, 2018, as a legal entity, our SPIL subsidiaries results are consolidated only as of that date going forward. For the sake of comparability, our full year 2019 results will be compared against a pro forma set of results for full year 2018 as it still was a subsidiary and consolidated during the entire year of 2018. This set of results is labeled or will be labeled pro forma. Please refer to Page 3. Let's go through an update of our 2019 goals. For our SPIL combination, as you know, we have been limited in our ability to tap the full potential of our combination with SPIL during the restrictive period imposed by the Chinese antimonopoly bureau. It is our understanding that the assigned third-party regulatory auditor has completed and submitted its report pertaining to our compliance during the restricted period ended November 24, 2019. We continue to expect to receive an official acknowledgment, accepting that a satisfactory audit has been conducted and reconfirmation that the restructuring period has been completed without incident. We also continue to expect to receive such an indication during this quarter. As of today, we have not received any additional information regarding the status of this process. However, given the current situation in China, the timing of this notification may understandably be subject to some delay. During 2019, our EMS unit achieved record revenues on the strength of SiP products. We continue to believe that our EMS entity provides a crucial component of SiP, a system-level understanding. During December of 2019, we announced our intention to acquire Asteelflash, the second largest EMS company in Europe. We are in the process of clearing all relevant regulatory hurdles, and we expect to complete the combination during the third quarter of 2020. Unfortunately, we will not be able to give an outlook in terms of how much such an acquisition will contribute to our results during 2020. But we can state that Asteelflash had revenues of around $1 billion. In addition, our organic EMS business should also continue to grow during 2020, driven by our portfolio of SiP products and across our computing, industrial and automotive segments. Along with the acquisition, we also expect to have an expanded geographical footprint, which could make a difference for current and new customers. There's much to look forward to from our EMS business. Exiting the year, our test business outgrew the rest of our ATM business. Given the context that the semiconductor industry shrank during 2019, we were pleased that we were able to grow our test business by 7%. And hold our assembly business relatively flat. Looking out into 2020, we continue to see strong momentum for our test business. We are looking for our test business to pick up speed and grow greater than twice the logic semiconductor growth rate. Group SiP grew at 13% during 2019, with $230 million of new SiP products scattered across multiple customers. For 2020, total SiP growth should continue to expand with SiP adoption accelerating. At this time, we see accelerating momentum for our SiP business in 2020, driven in part by 5G-related products. Finally, in 2019, fan-out revenue was ahead of our target of $50 million for the year. As we see increasing input/output density for shrinking die, we see fan-out becoming a necessity and a path to become the mainstream packaging method on the next foundry node. Such packaging adoption should accelerate gradually during 2020 and pick up more aggressively during 2021. We currently estimate fan-out revenue to grow more than $50 million during 2020. During this year, we will continue to invest in this capacity, along with panel level fan-out. During 2019, we were selected to be included in the Dow Jones Sustainability World Index for the fifth consecutive year. And more impressively, within this index, we have been the worldwide leader within all semiconductor and semiconductor equipment companies for 4 consecutive years, which is considered a record. The Dow Jones Sustainability Index is widely considered the global standard for measuring advancing corporate ESG practices. During 2019, we also were selected for inclusion and the CDP A List for climate change for the third time. CDP recognized ASE as a pioneer for action on climate change. CDP scored over 6,800 companies with a rating of A to D minus, only the top 2% made the A List. We were also the first Taiwanese company to be included in both the Dow Jones Sustainability World Index and the CDP A List. We're extremely proud of the company's ESG accomplishments. Next, a little bit on the coronavirus. As with all entities in Asia, we're in the midst of trying to run our business under the impact of the 2019 novel coronavirus. As of today, the impact to us from this virus generally relates to China's lack of available labor. Much of China's workforce are migrant workers. These workers often travel thousands of miles between their hometown to their job. A large portion of these workers go home during the Lunar New Year holidays. During the 2020 Lunar New Year holidays, China has taken aggressive actions trying to curb the spread of coronavirus, including extending the duration of the Lunar New Year holiday and restricting the movements in and out of a number of effective regions. Today, transportation in China is significantly more complicated than before the coronavirus measures were put in place. Many workers have difficulties rejoining the workforce due to transportation routing limitations or an inability to leave their hometowns. Even when these workers complete the logistical challenge of getting back to a factory, they must go through a health screening and a quarantine process. The competition for available basic workers is quite strong right now. Even with our factories running over time to make up for these labor deficits, our China-based factories are all running at varying levels of suboptimal reduced staffing. The current situation is very dynamic and much depends on the ability of the world to control the spread of this virus. Overreaction is probably the most prudent course of action. With the information given to us now in terms of availability of labor, we can provide you the following information. Within ATM, roughly 15% of our revenues are from China-based factories. Within EMS, more than half our revenues are from China-based factories. From a seasonality perspective, the first quarter is typically our trough quarter, which provides some cushion to the overall impact to these factories. Nevertheless, during the Lunar New Year, our China factories still run at 60% to 70% staffing. We expect to return to 80% to 85% staffing by the end of February. And we still currently expect to be able to get near full staffing by the end of the first quarter. Page 4. On this page, you will find our fourth quarter consolidated results at the holding company level. We will generally defer business explanations of these results to our ATM and EMS P&L discussions. Intercompany transactions between our ATM and EMS businesses have been eliminated during consolidation. For the fourth quarter, we recorded fully diluted EPS of $1.47 and basic EPS of $1.50. Consolidated net revenue was $116 billion. This represents a 1% decline quarter-over-quarter and a 2% improvement year-over-year. ATM revenues of $66.8 billion were flat quarter-over-quarter and up 6% year-over-year. EMS revenues of $48.7 billion declined 4%, both quarter-over-quarter and year-over-year. We had gross profits of $19.8 billion, with a gross margin of 17.1%. Our gross margin improved by 0.8 percentage points quarter-over-quarter while improving 0.7 percentage points year-over-year. A sequential increase in gross margin is primarily the result of stronger ATM loading and higher ATM product mix. Our operating expenses increased by $0.4 billion during the fourth quarter to $11.1 billion, primarily the result of increased operating expenses from our EMS business unit. Our operating expense percentage of 9.6% represents an increase of 0.4 percentage points. This increase was primarily the result of higher operating expenses within our EMS entity. Operating profit was $8.7 billion versus $8.4 billion in the third quarter. Sequentially, operating margin improved 0.4 percentage points to 7.5%, or being flat year-over-year. During the quarter, we had a net nonoperating loss of $0.1 billion. This amount includes net interest expense of $0.9 billion, offset in part by our foreign exchange and financial instrument activities. Tax expense for the quarter was $1.8 billion. The effective tax rate for the fourth quarter was 20.7%. Net income for the quarter was $6.4 billion, representing an improvement of $0.6 billion from the previous quarter and $0.9 billion from the same period in 2018. On the bottom of the page, we have again provided here key P&L line items without the inclusion of PPA-related expenses. Consolidated gross profit, excluding PPA expenses, would be $21 million, with an 18.1% gross margin. Operating profit would be $10.2 billion, with an operating margin of 8.7%. Net profit would be $7.8 billion, with a net margin of 6.8%. Basic EPS, excluding PPA expenses, would be $1.84. Page 5, here is the 2019 consolidated full year results with pro forma legal entity comparison. Again, we will defer to the business unit P&L to make business commentaries. For 2019, consolidated net revenues grew by 4% as compared with 2018. EMS revenues grew 9% annually and ATM revenues stayed flat. Gross profit for the year increased by $0.3 billion year-over-year, primarily as a result of higher EMS revenues. Gross profit margin for the consolidated entity declined 0.5 percentage points, primarily attributable to higher EMS product mix. Operating profit for the year declined by $3.2 billion, primarily because of higher operating expenses. Operating margin declined by 1 percentage point. Our nonoperating expense was $0.2 billion for the year, including $0.9 billion of interest expense. Nonoperating expenses declined as a result of higher income from financial instruments and currency hedging activities. Net income increased by $1 billion from $15.9 billion to $16.9 billion. Fully diluted EPS for the year was $3.86, while basic EPS was $3.96. Removing the effect of PPA depreciation and other transaction-related costs, our gross margin would be 16.7%, operating margin would be 7.1%, and our EPS would be $5.35. On Page 6 is our ATM P&L. It's worth noting here that the ATM revenue reported here contains revenue eliminated at the holding company level related to intercompany transactions between our ATM and EMS businesses. For the fourth quarter, revenues for our ATM business were $69.3 billion, up $1.4 billion from the previous quarter and up $5.2 billion from the same period last year. This represented a 2% increase sequentially and an 8% increase year-over-year. Our ATM revenues came in slightly ahead of our expectations due to stronger-than-expected seasonal demand. For our industry. We believe the sequential increase during the fourth quarter to be somewhat unusual as we typically see a slight seasonal decline as the holiday season build, winds down. We believe this to be a positive sign and the result of a general business recovery. Gross profits within ATM were up $1 billion quarter-over-quarter and $1.7 billion year-over-year to $15.7 billion. Both of these gross profit improvements are the result of incrementally higher loading. Gross margin for our ATM business was 22.7%, up 1 percentage point sequentially and up 0.9 percentage points year-over-year. Margin improvement sequentially and year-over-year are primarily the result of higher overall loading and a higher mix of test revenue, offset by negative foreign exchange impact. We believe that our margin recovery is just starting and believe a small year-over-year increase of 0.2 percentage points through a downturn to be a significant achievement. During the fourth quarter, operating expenses were $8.3 billion, flat with the third quarter, but up $0.6 billion from the same period last year. The year-over-year increase is primarily related to increased R&D expenses related to higher NPI costs. Our operating expense was 12%, down -- sorry, down 0.2 percentage points sequentially and up 0.1 percentage points year-over-year. We would like to make some comments in regards to what is becoming a more significant portion of our operating expenses, new product introduction or NPI costs. Our R&D NPI expenses sometimes significantly precede their associated revenues. As we've previously mentioned, leading-edge packaging methods such as fine-pitch bumping, Copper Pillar and fan-out require longer and more significant NPI efforts. These NPI expenditures often precede their associated revenues by more than 3 to 6 months. As such, our operating expenses will reflect higher and somewhat imbalanced R&D-related expenses when there are significant amounts of advanced packaging products in the pipeline. In short, more NPI costs now lead to more products to run down the road. We believe our focus related to these expenses should be towards better monetizing them once the associated business comes into mass production. During the fourth quarter, operating profit was $7.4 billion, representing improvements of $1 billion quarter-over-quarter and $1.1 billion year-over-year. Operating margin was 10.6%, increasing 1.2 percentage points from the third quarter and increasing 0.8 percentage points year-over-year. This improvement was primarily attributable to better gross profit performance from higher loading during the quarter. Without the impact of PPA-related depreciation and amortization, ATM gross profit margin would be 24.4% and operating profit margin would be 12.7%. On Page 7, we have our ATM full year P&L with a pro forma and legal entity comparison. For our ATM business, we, along with most of the industry, had a soft first half of 2019. The back half was marked by a significant recovery. And even though our performance was one of the most resilient within the semiconductor industry, the sluggish first half was a significant drag on our full year results. On a full year pro forma basis, our ATM business in total increased by 1%, composed of our packaging business being flattish, materials business being down 1% and our test business being up 7%. During this downturn year, we were able to increase our gross profit by 1%, while holding gross margins to a 0.1 percentage point decline. This was primarily driven by softer loading in our assembly business, partially offset by higher test product mix. Operating income declined 10%, primarily due to higher operating expenses. The higher operating expenses primarily consisted of higher R&D costs and employee compensation. The R&D costs are primarily associated with higher NPI and advanced project costs, the employee compensation costs related to employee stock options issued after completing our combination with a spill. Operating margin declined by 0.9 percentage points. Please turn to Page 8. Here, you'll find a graphical presentation of our pro forma ATM P&L. We're not going to talk much about that. On Page 9 is our pro forma ATM revenue by market segment. We do see the Communications segment continuing its recovery here. On Page 10, you can see our pro forma ATM revenue by service type. During the fourth quarter, our revenue mix continues to transition towards bumping flip-chip and SiP. As stated earlier, we continue to expect our test business to outgrow our assembly business as we continue to expand our turnkey business model. On Page 11, you can see the results of our EMS business and its associated revenue by application. During the fourth quarter, we had revenues of $48.8 billion, representing a decline of $1.8 billion or down 4% sequentially and a decline of $1.9 million or 4% year-over-year. EMS revenues came in slightly ahead of our expectations as a result of higher demand as the seasonal build was slowing down. As mentioned during our previous earnings release, the seasonality of our EMS business accelerated forward a month or so. This seasonality adjustment had some business pulled in from the fourth quarter into the third quarter. As such, we had a relatively stronger third quarter, which led to a relatively softer fourth quarter. The year-over-year decline was also driven by this seasonality adjustment. Our EMS gross profit was $4.3 billion, declining $0.2 billion sequentially and $0.3 million year-over-year. The gross profit decline was driven primarily by softer loading. Gross profit margin for the EMS business unit came in at 8.9%, flat with the previous quarter and declining 0.2 percentage points year-over-year. The year-over-year decline in gross profit margin was driven by softer loading and product mix. Our EMS business unit's operating expenses closed the quarter at $2.8 billion, increasing $0.4 billion sequentially year-over-year. Operating expenses increased primarily as a result of planned expansions of business, resulting in higher employee-related expenses, transaction-related expenses and NPI costs for future products. Operating profit for the quarter was $1.6 billion, which is a $0.5 billion decline sequentially and a $0.6 billion decline year-over-year. Our operating margin came in at 3.2%, which is a 0.9 percentage point decline sequentially and a 1.1 percentage point decline year-over-year. For the full year, our EMS business continued to grow, driven by strength within our SiP business and Communications customers. For 2019, we had EMS revenue of $165.9 billion, representing a 9% annual increase from 2018. This was primarily driven by growth within our SiP-related products. Gross profit increased $0.3 billion or 2%, primarily as a result of higher revenues. Gross profit margin declined 0.6 percentage points, primarily as a result of product mix changes and initial ramp-up of costs of sites outside China. Operating profit increased to $4.9 billion, with operating margin declining 0.8 percentage points to 2.9%. The operating margin decline was driven by higher operating expenses related to higher expansion costs for sites outside of China, higher NPI costs related to future projects and higher transaction-related expenses. On Page 12, you will find a graphical representation of our EMS revenue by application. You will see that our Communications segment picked up while our Consumer applications declined. These moves are seasonally driven. On Page 13, you will find key line items for our balance sheet. At the end of the quarter, we had cash and cash equivalents and current financial assets of $65 billion. Our interest-bearing debt decreased $6.9 billion to $220.7 billion. Total unused credit lines amounted to $225.4 billion. Our EBITDA for the quarter was $22.5 billion. On Page 14, you will find our pro forma equipment capital expenditures, machine and equipment capital expenditures for the fourth quarter totaled $457 million, of which $227 million were used in packaging operations, $205 million in testing operations, $18 million in EMS operations and $7 million in interconnect material operations and others. For the full year of 2019, we spent $1.575 billion for capital expenditures, $798 million in packaging operations, $689 million in testing operations, $69 million in EMS operations and $19 million in interconnect material operations and others. For our ATM business, we had a strong fourth quarter, one that outpaced the third quarter in every aspect. Everything picked up. All major product lines are running well. A quarter ago, we said that the recovery was well on its way. And that the business looked healthy. We saw that demand even seem to be persisting into the first quarter. This, we believe, would lead to a stronger than seasonal first quarter. For the whole year, 2020, we saw 5G being the biggest story of 2020, driving all our business units towards better years. This view has not changed. It remains backed up by our customer forecasts. A shallower trough than a typical first quarter followed by strong quarter-over-quarter growth. For our EMS business, during the first quarter, we usually see a significant cliff as the holiday seasonal build concludes. This also continues to be the case. Customer forecasts have remained relatively consistent. So if you look out into the rest of the first quarter, I do know what we'll be selling extremely well, surgical face masks. All kidding aside, this is the business environment in which we love trying to understand. Relatively few people have any control of the macro factors that will impact our business. The situation is pretty much in the hands of sovereign government officials, when will the coronavirus be controlled and what measures will be necessary for that control to happen. It's the effectiveness of those measures that will ultimately determine the availability of labor for ASE's China-based factories, the health of the overall supply chain and eventually the health of consumer demand. And even for me, I fundamentally believe the science that indicates the current coronavirus is similar to the flu. And I believe the transmission method of this virus is via droplet format and not airborne. Rationally, we should just avoid crowded areas, wash our hands and avoid touching our faces. Rationally, I shouldn't be using an N95 mask and hoarding hand sanitizers. But when the stakes are ultimately high and include your very own life, probability becomes meaningless. Fear trumps rational thinking. This virus is a negative lottery, and everyone is doing whatever they can not to win. So the fear that is gripping the world the overabundances of caution at a personal, company and sovereign government levels are completely understandable, but the impacts of -- to our business are totally unpredictable. Supply chain risk is also troublesome and extremely difficult for us to evaluate. It doesn't matter if we can manufacture our components if the company ahead of us cannot provide the raw materials or the company below us cannot assemble the end product. This environment will test supply chains. But remember, in times of uncertainty and fear, product manufacturing risks are amplified. Customers in turn choose to stick with the leaders because the leaders have the best capability and flexibility to get the job done. Unnecessary product risk could be the difference between a product that gets made and one that does not. So for our first quarter, we have adjusted our expectations given our operational constraints, and we've even adjusted for higher labor costs. But please remember, that our expectations are made in an environment in which the macro factors, which are completely out of our visibility, easily outweigh the micro factors, which are immediately visible to us. We actually had a significant internal debate over whether we should provide an outlook at all. But we felt that not guiding would contribute to the atmosphere of fear. We decided that highlighting the strength of our business should not be outweighed by uncertainties exogenous to our entire industry. But with that said, please regard our outlook with a higher potential for variance than normal, and we would appreciate it if you would all assign a broader range of potential performance. So for ATM in NT dollar terms, ATM first quarter 2020 business should be between second and third quarter 2019 levels. ATM first quarter 2020 gross margin should be slightly above second quarter 2019 levels. For EMS, in NT dollar terms, EMS first quarter 2020 business should be similar to first quarter 2019 levels. EMS first quarter 2020 operating margin should be slightly lower than first quarter 2019 levels. These are at the end of the prepared remarks, we can go to the Q&A session at this point.

[Operator Instructions]. The first to ask questions is Randy Abrams, Crédit Suisse.

Yes. I appreciate you're still giving out the guidance I wanted to ask the first question just on, as you're factoring in the first quarter guidance, it sounds like you're implying more of the production side impact. I'm curious from the demand side, if you've seen any customer revisions at this stage from the demand side. And within that guidance you gave, is that more profile of how orders look now? Or have you taken a discount to factor in any demand-side weakness?

I think what we are -- whatever we are providing now is really based on the situation at our factory level. So it's more on supply side, the of constraint that we have. And we've made those kind of the best guess adjustments into our first quarter outlook. But in terms of demand, I don't think we are in a position to comment on the overall demand situation because there's still a lot of uncertainties in front of us. But indirectly, I think this demand situation should be reflected into the forecast that we're getting from our customers. So whatever we're providing at this point is really looking at our own situation and looking at the forecast that we're still getting from our customers, maybe the overall estimation of what kind of impact it will have on the first quarter and remaining of the year.

Okay. And if I could ask on the investment side, you're running at a high level through second half, I think, in -- ahead of some of the good drivers you were expecting for this year. How are you thinking about maybe the CapEx relative to what you were doing the last few quarters, holding -- or if you have a full year CapEx, knowing that it could change based on what's happening?

At this point, we're still very optimistic about the 2020, and we will continue our CapEx as required. But at the same time, there is some of the CapEx being made earlier in 2020, what we call the upfront investments that we were already making. So for 2020, we're expecting for CapEx now to exceed whatever we had spent in 2019.

Okay. It won't exceed 2019, okay. And if -- okay. If I could then ask on the -- and maybe if I missed it, the SiP business, you had the 13% growth last year. Do you have a potential, what you're kind of seeing at this stage in terms of SiP, how much -- or what the outlook could be for this year? And then maybe on the broader business, factoring in like the virus gets contained. Do you have kind of a view on maybe full year for the overall company versus industry? Maybe a rough view what type of growth could come through?

I'm sorry, the latter part of the question?

Okay. Yes, the first was more micro, just more looking at the SiP, what you expect for this year. And then second was, if you have kind of a broader baseline view, maybe going into the year, if the virus is contained within a quarter or so, how your business and also how the industry looks on a full year growth?

Well, I think in terms of SiP, we had a very good run-up in the -- in 2019. And I think that momentum will continue. I think Wu was mentioning it that the new projects revenue exceeded our target of $100 million incremental revenue in 2019, and we're expected to get momentum to further strengthen in 2020 as well. So there will be a fairly healthy growth in our SiP business in 2020 as well. As for the whole year, it's taking the virus situation aside, I think we still remain very optimistic about the whole year situation. And we're expecting a healthy -- a very healthy growth in terms of overall revenue, particularly on the ATM side of this. And first quarter, we are still expecting an above-seasonal quarter. And for the remaining of the year, we're still expecting sequential growth on a quarterly basis overall.

Okay. And last question, if I can ask the margins in IC ATM. Maybe the factors you mentioned, just the start of the margin improvement. And it looks like test may be one factor. But if you could go through the areas, giving confidence that margins and maybe how much you might be seeing in that front? And if you could give an update then on the EMS, where first quarter looks like a lot of impact on the production side, but maybe the -- how second half could look for margin versus the second half last year, like except for -- through the virus setback?

From a consolidated point of view, we do expect a healthy margin improvement in the year. Because the -- first of all, higher loading and also a lot of the investment that we're making in 2019 will start to pay off in 2020. We are expecting a better control over our OpEx ratio. We are expecting, although, I cannot quantify at this point, but we do see room for more efficiency improvement after -- with the restriction period is lifted. So overall, I think the margin situation in 2020, we'll see some healthy -- quite a bit of healthy improvement. That's both on the ATM and also on the EMS part of the business as well.

Next question we're having Bill Lu from UBS.

A lot of information given. So I'm just hoping for some clarifications first. I think Ken said that test would grow faster than 2x logic. I think -- did I hear that right? And if so, what is the -- what's your expectation for the logic growth this year?

People are expecting somewhere around 7% to 8%.

Okay. So can you just talk a little bit about the key drivers for test this year? And also, I don't know if I'm heard, besides that, was there also a guidance given for total ATM growth this year over last year? C. S. Chang: No, no. we didn't give out total ATM for 2020.

Okay. Can you try to help me a little bit around that?

Just as I mentioned, we are expecting a healthy growth, particularly in the ATM part of it because a lot of the new initiatives that we're taking will start to pay off in 2020. We're seeing a strong momentum in our SiP business. We're seeing strong movement in in our test business as well. So overall, I think we're going to have a pretty -- if taking all the uncertainties aside, particularly in the this virus situation, whatever we're getting at this point, whatever information we're getting at this point leads to quite a healthy year for us in 2020.

Okay. Maybe I'll ask that slightly differently because you have said that if you look at the growth this year, 5G is one of the bigger drivers. If I look at 5G, a lot of it is China. I guess my question is, is there way to quantify how exposed you are to China and to 5G within that, the optimistic outlook for ATM, IC ATM?

What percentage of our revenue will be 5G related, is that the question?

Yes. That would be very helpful if you could answer that.

Well, that's one question I cannot answer. I think we -- of course, 5G will be a major driver for growth for us, including 2020 and beyond. But in terms of what percentage of 5G -- what percentage of revenue will be 5G related, I think that, that's a very difficult number to come by because everything is interconnected, some can be 5Gs, direct 5G or some 5G related, some can be riding out of 5G implementation. But I think the key for us is really to maintain our first-mover advantage in terms of the 5G development or deployment, and this is driven by the -- our overall -- our business -- by this coverage over all the 5G players. So whatever players, whoever that comes out with 5G more applications or products, I think we want to maintain our first-mover advantage, and we think we'll be -- stay as a dominant player in the 5G field.

Great. Sorry, can I sneak in one more? So you talked about production out of your China-based factories, and how that's going to be back to normal by the end of the quarter. But assuming issues do not get resolved, what's your ability to shift manufacturing away from the Chinese factories?

I think the -- or right now, the overall capacity in our China factory. In terms of ATM, it's about 15%. And I think most of the products can be taken over by our Taiwanease side or other areas as well. So I think this is the point Ken decided to make earlier when the -- when there is a crisis, I think, the leader with the higher scale and flexibility tend to be the -- a safer bet for all our customers. So I think that's -- a silver lining is here, if there is a crisis in front of us, there is some challenges for -- in terms of overall strategic positioning for ASE. You could look at it as a plus for us.

The next to ask question is Gokul Hariharan, JP Morgan.

First of all, could you talk a little bit about CapEx, looking at 2020. It seems like 2019, eventually CapEx ended up being higher than the expected range. What do we expect for 2019. And secondly, I think you talked about some of the NPI investments, which drove OpEx higher in Q4. What are the time frame roughly in terms of when we see these investments come through in terms of revenues? And how should we think about OpEx in 2020? Are we expecting OpEx to start -- OpEx as a percentage of sales to start coming down, given that we are expecting meaningful revenue growth? And I had a follow-up question on the guidance as well.

Well, as mentioned, we are expecting. This is not a budgeted number, but as -- but the forecast we're getting from our customers, we're expecting a relatively steady CapEx compared to 2018 -- sorry, 2019. And -- but the combination of that CapEx would be somewhat different because in 2019, up to 44% of the CapEx was spent on test, but going into 2020, that ratio will come down to below 30%. So more resources will be put in assembly and packaging.Also, a larger CapEx will be spent on EMS. So some of the new SiP project that we put the NPI in already in the second half of -- or fourth quarter of 2019.

Okay. Could you talk a little bit about the OpEx as well and the NPI? Yes.

In terms of OpEx, I think this year, the overall OpEx ratio grew from 9.4% in 2018 to 9.9%. And this is because of the -- we have additional ASP-related expense. We have more -- we spend more on R&D. And also, we are expanding some of our sites -- some of our non-China sites given the trade war situation. So we are looking at 2019 as a year of investments. And I think a lot of this front line investment will start to pay off in 2020. So that gives us a better margin outlook for the -- for 2020 as well. In terms of the overall control of the OpEx, our target for this year is really to come back to 2018 level of life 9.4% or 9.5% level.

Okay. Just a quick question on the guidance. Could you talk a little bit about how you handicap the guidance for any demand shortfall? I know that it's very early for customers' feedback to you in terms of any potential demand risk on -- by the products, et cetera, in China. So could you talk a little bit more in detail about, are you already handicapping some of those potential demand -- near-term demand disappointments in your Q1 guidance? Or the caution in Q1 guidance is primarily related to supply disruption and not really counting on any potential demand related to China 5G?

Well, as I mentioned, whatever outlook numbers that we -- our guidance we've provided is based on the forecast that we're getting from our customers. So without directly commenting on the demand -- over demand situation, we would like to say that, that is somewhat reflected in the forecast that we're getting from our customers. And with that forecast, we look at our own capacity situation and made the necessary adjustments in -- particularly in the first quarter, outlook. So that's how we come up with the overall outlook that we provided.

Right now, we're having Charlie Chan, Morgan Stanley.

So first of all, can I ask how you're going to split the [indiscernible] with AiP revenue between ASE and subsidiary, USI, both revenue and profits?

So what?

The [indiscernible] with AiP, how are you going to split the revenue and the profits between you and the USI?

No. We're not commenting specific parts of that performance.

Okay. And then, Joseph, you mentioned about this year, you're expecting higher margin, better control of OpEx ratio, does that discount the merger synergy already? Or there's a kind of separate factor?

Right now, I don't think we are -- we are not putting a lot of weight on the synergies that can be created, particularly on the operation after the merger with [indiscernible]. At this point, we're not in a position to comment on that because the restriction hasn't been listed yet. So no, we're not budgeting this into the overall.

Okay, okay. And then on that front, any time line or any reason that, that restriction is still there? When that is going to be removed?

Well, I think the -- we have already filed the -- I think the appointed reviewer has already filed their last report to the antitrust bureau in China. And we are waiting for them to review such reports and give us a clear state of health. But because of the virus situation, I think it's kind of put -- I think if there is some delay, that is well understandable. But at this point, we're still hoping that by the end of the quarter, we should be able to get the full clearance.

Okay. And lastly, maybe also related to your ATM business growth in 2020. So if you can compare your ATM business versus the industry growth in 2019 and 2020, do you think you're outgrowing or under growing the industry? And I think that associated to another question, whether you think local, Chinese back-end foundries are gaining share? And do you think Chinese semi customers, they have any preference between using ASE or the local vendors?

Well, I think we will be able to outgrow the industry, overall, if you're talking about the logic of it. I think you were pretty confident that we will be outgrowing not just the industry, but also all of our competitors. And I think -- what's the latter part of the question?

Yes. Because China is pushing this semi localization, right? So I'm not sure you've said us the change customers' preference. I mean for those local China semi customers, do they -- do you see the shift from your China factory to those local vendors? And I'm not sure that if there's a such kind of a preference, does those local vendors need to -- still need to discount your price versus yours?

Well, I think, preference is one thing about the capability is the other. So they can prefer their local vendor. But if their local vendor cannot provide service or the quality or the products, as a whole, I would think our customer have -- really have as a choice. So yes, we will review the overall situation or competition very seriously, and we will try to -- try our best to accommodate whatever our customer demand is, or requirement is. And that includes ramping up our China factory as well, if local manufacturing is preferred by our customer.

Right now, we're having Rick Hsu, Daiwa Securities.

The first question, as usual, is housekeeping for modeling purpose. So can I know your utilization rate in Q4 across the packaging, testing and bumping? And also, what's your view on first quarter loadings?

Fourth quarter, we are at a range of 80% to 85% utilization and going to across board. And first quarter, we're looking at 75% to 80%.

75% to 80%? Okay.

Yes.

Okay. Then second question, now this a little bit of hypothesis, assuming there's some impacts from the virus in your first half business, then would you be able to make up the shortfall in second half?

What's the impact of the virus?

Okay. Yes, because I'm just thinking -- I'm just trying to resemble the situation to the SARS back in the [indiscernible].

Well, I think -- first of all, I think within reason, of course, we can now -- we can make it up. And we do have the scale or the flexibility to handle that unless there is a total collapse of the whole situation. Then I think right now, what we're seeing is, things are still within a manageable level.

Okay. That's fair enough. The final question is about the 5G. I think a couple of days ago, [indiscernible] when they put out their business outlook, they actually say something about a slowdown in the 5G bill this year. So would you worry about that the demand given by your customers for the 5G bill this year a bit too bullish? Do you see any risk on that?

Oh, I think there is always risk whatever forecast or whatever market assessment there is, but we -- the forecast that we're getting still suggests that we are having a year. And we'll try to analyze the situation better and make the proper adjustments on whatever we're seeing, i.e., if there is any overbooking or any over optimistic kind of forecast that we're getting, and we'll try to make some more read into -- because we do have a very diversified customer base. So there's some level of check and balances we can get among our different customers. So I think we have been managing our business for over three years. So we should have the experience or knowledge, how to view the overall situation.

The next one to ask questions, Bruce, Goldman Sachs.

So my question is regarding for the -- your USI, you just mentioned that you have 85% production in Taiwan for ATM. How about your production in China? Are you able to move those AIC -- SiP productions back to Taiwan?

That will be a larger scale of work to do. Yes. But I don't know how to answer this question, actually.

Okay. So basically your assumption is based on that -- your labors and everything is going back to your factory around like 80% or 90% by the end of the first quarter? Is that your assumption for the USI business as well?

Yes.

I see. Okay.

Well, I think to answer your question -- your first question. If we can't do it, then I don't think anybody else can do it either. So that will be a real problem for [indiscernible].

Well, I don't -- well because you -- that's why we need to ask you, right? Okay. That's fine. I'll ask a different question. So regarding to your profitability for the ATM, so if you look at your profitability for ATM for 2020, you are looking at much better years, with high inflation rate and better product mix. And the profitability for ATM in third quarter is still lower than the historical level or before the ASE-SPIL opposition? So what kind of profitability or what kind of the new norm of profitability we can expect in 2020 or 2021 onwards? This one should be a lot easier.

Yes. It's a lot easier, but then I won't to give you the answer though.

You didn't give me the first one, you said -- you have to give me this one.

No. I want to get into trouble with the SEC here. If I give you any numbers, then I need to put out a full blown, what do they call it?

Right, [indiscernible]. Yes.

And we're not allowed to do that.

Well, the way the [indiscernible] they use is what they call it a structural profitability.

A what?

They call it structural profitability.

Structural probability, what do you...

Yes. No, that's what they -- what -- so in what terms do you think? Instead of getting into trouble.

I don't understand.

No. I mean, so for me it's that we are expecting some profitability improvement. So historically, ASE still are able to generate somewhere along by 25% or 28%, even 30% gross margin at the peak quarter. So are we expecting those kind of numbers in the foreseeable future?

Well, let me do this. Let me do this. I am -- as a CFO, I am asking for at least 2% improvement.

I see. In 2020?

Yes.

I see. I believe everyone listens to you, right?

Huh?

I believe all the staff listens to you and will follow the instruction.

Okay. You believe what?

I believe all your staff will follow your instruction in delivering results.

Well, remember, they're not all my staff. It could be my wishful thinking, right?

True, true.

Right now, we're having Sebastian Hou from CLSA.

Yes. My first question, just to -- want to clarify some of the numbers that I didn't hear very clearly. So I think Ken mentioned that the utilization rate in your EMS fab in factory in China was like 8% in Chinese New Year and expect that to go up to 4 percentage point by end of February and then fully utilized by the end of this quarter. Forgot the numbers. Could you repeat that again?

Yes. All right. We are looking at around a -- so 60% to 70% staffing during the Lunar New Year. And then we expect to return to 80% to 85% by the end of February. And then, hopefully, to get close to full by the end of the quarter. That's the more optimistic outlook at this point.

So these are the okay scenario?

Yes. That is the okay scenario.

Okay. But what's the usually passing rate increase a month after -- or in 2Q which is after CNY Holiday? Let's say, an attractive yield?

Normal year?

Yes.

Normal year, it's back to full.

Okay. So basically, you are saying that it's 80% to 85% by end of February, but the normal years by end of February, should have already gone back to 100%, right?

That's correct.

That's like 20 -- almost 20% below?

That's correct. But during this time frame, we're also going through basically the trough seasonal quarter, right? So that gives us a little bit more cushion to work with. We can also work with a little bit more over time planning workers' schedule and such. But in a positive scenario, this is what we're seeing.

Got it. I have a question, a follow-on, on this because typically, your EMS business is a -- as you -- as your definition, the trough quarter, with the low season in 1Q. So even if you can't get back to the quite fairly decent staffing rate, would that still be a bottleneck to your manufacturing facility with the full year customer demand? After all, it's low season in Q1.

It still has some impact.

Still has an impact. Or this year is somewhat different? Like, you see a better than typical Q1 seasonality, so that's why this will matter?

Even during -- so we're basically trying to get to an optimal percentage, right? So we're trying to get to 100%. But that 100% is 100% of the seasonal quarter.

Okay, got it. And my second question is, about your EMS business. I look at your 2018 to '19 change, you have the 9% revenue growth, but the operating profit dropped 14%. OP margin contracted by 0.8%. Can you explain, again, what drives that OP margin -- operating income decline? And also, what's your outlook for 2020?

So our EMS entity is in the process of some different type of expansion. There's geographic expansion, there is also trying to expand the current site. And then they're also buying Asteelflash, right, and doing a number of other transactions that have been talked about. These things, these expansion plans are actually generating additional costs within OpEx this year. Hopefully, these are not -- we don't see these necessarily a structural in nature for OpEx at this time.

Okay. Sure. But has most of the expands -- you will likely -- or do you expect we -- these will mobilize this year?

To a much lesser degree, yes. There will continue to be some ramp-up, and in particularly in Poland, in Mexico, in Taiwan, from -- in terms of EMS. In terms of NPI, it really depends on what new projects we'll be getting so -- but most of -- I think that most of the -- the costs for expanding non-China sites from EMS perspective is done in 2019. And things are starting to get much more streamlined. So there will be additional expenses put in. But in terms of overall percentage of revenue, I think, it's back to a more normal level.

Okay. So to the conclusion for the EMS business outlook this year is that the top line will grow and margin -- OP margin will stabilize here or will potentially improve.

Yes.

The next question is Szeho Ng from China Renaissance.

I just want to touch base on the EMS acquisition, Asteelflash. When will that start contributing to the company's top line?

We don't have -- we're not giving out the -- this year's number for Asteelflash because we don't -- the transaction is not closed yet. But as a reference, last year, they had a revenue of $1.1 billion. And if you look at their first and second half split, it's about 45%, 55%. And they had a net income of around $53 million for last year. So you could probably take that as a reference. And we're expect -- the schedule is to close the transaction by early third quarter.

Okay. All right. So is it fair to assume that, that business, the net margin, at least, right, is better than our existing EMS business?

Yes.

Any the reason for that? Is it 1/4? Or it's because of the mix issue?

I think it's the -- it's the different nature of business. I think what they're focusing on is smaller volumes, but larger variety. And it's -- a lot of it is not the traditional 3C type of business. So I think the biggest benefit to -- out of this transaction is 3-fold. One is, it does expand our overall geographical coverage throughout Europe. And the second is that the business model is different. They are serving a lot of the new application, smaller volume. And once those business turn into mass mass production, then it's very natural for the -- for Asteelflash to move these to USI. Clearly that's -- yes. I think there is the -- really, the customer portfolio is very, very complementary. I think there's very little overlap between their customers.

Okay, sounds great. And my last question regarding the second part. Is it fair to assume, at least for this year's dividend, will be flat from last year's level?

The what?

The cash back -- cash dividend will be at least flat compared with the last year?

Well, I think -- well, we'll be paying off 60% to 65%, seeing we're not changing it at this point.

I mean, from the dollar perspective -- dollar amount perspective.

Almost remain the same. Remains similar -- the EPS we've made this year, they're going to last. So dollar-wise, I'd say -- the amount was...

Okay. Good.

Yes.

Anything else you have? Do we have any more callers? Okay. Thank you for attending ASE's fourth quarter earnings release.

One of the things, I did not give any guidance on margin.

All right. Thank you very much. See you next quarter.