晶科能源 (JKS) 2021 Q3 法說會逐字稿

Hello, ladies and gentlemen, and thank you for standing by for JinkoSolar Holding Co., Ltd. Third Quarter 2021 Earnings Conference Call. (Operator Instructions) As a reminder, today's conference call is being recorded.

I would now like to turn the meeting over to your host for today's call, to Ms. Ripple Zhang, JinkoSolar's Investor Relations Manager. Please proceed, Ripple.

Thank you, operator. Thank you, everyone, for joining us today for JinkoSolar's Third Quarter 2021 Earnings Conference Call.

The company's results were released earlier today and available on the company's IR website at www.jinkosolar.com as well as our Newswire services. We have also provided a supplemental presentation for today's earnings call, which can also be found on the IR website.

On the call today from JinkoSolar are Mr. Li Xiande, Chairman of Board of Directors and Chief Executive Officer of JinkoSolar Holding Co., Ltd.; Mr. Gener Miao, Chief Marketing Officer of Jinko Solar Co., Ltd.; Mr. Pan Li, Chief Financial Officer of JinkoSolar Holding Co., Ltd.; and Mr. Charlie Cao, Chief Financial Officer of Jinko Solar Co., Ltd.

Mr. Li will discuss JinkoSolar's business operations and company highlights; followed by Mr. Miao, who will talk about the sales and marketing; and then Mr. Pan Li, who will go through the financials. They will all be available to answer your questions during the Q&A session that follows.

Please note that today's discussion will contain forward-looking statements made under the safe harbor provisions of the U.S. Private Securities Litigation Reform Act of 1995. Forward-looking statements involve inherent risks and uncertainties. As such, our future results may be materially different from the views expressed today. Further information regarding this and other risks is included in JinkoSolar's public filings with Securities and Exchange Commission. JinkoSolar does not assume any obligation to update any forward-looking statements, except as required under the applicable law.

It's now my pleasure to introduce Mr. Li Xiande, Chairman and CEO of JinkoSolar Holding. Mr. Li will speak in Mandarin, and I will translate his comments into English. Please go ahead, Mr. Li.

(foreign language)

[Interpreted] We are very pleased to have delivered total shipments of approximately 5 gigawatts, total revenues of USD 1.33 billion and gross margin at 15.1%. Total shipments were impacted by the delay in sales revenue recognition caused by logistical issues and blockage. The release of new cell capacity significantly reduced cell production costs, partially offsetting the pressures on production costs inflected by the high prices of polysilicon and other materials.

Logistics costs have further increased compared with the second quarter, and module prices hit a new high in almost a year. However, due to the transition to renewable energy in most regions of the world, the increase in electricity prices, financing support and other favorable policies, clients are more willing to accept higher module prices. Despite all the challenges, global installation this year has not differed from expectations set at the beginning of the year as the resilience of solar demand is gradually increasing. Currently, in its most severe shortage, we expect polysilicon supply will gradually return to sufficient levels starting next year. And as a result, installation demand is expected to increase significantly.

(foreign language)

[Interpreted] Due to high material prices, we accelerated the pace of cost reduction with upgraded technology. At present, we have reduced the thickness of our mono wafers by nearly 15% compared to the beginning of the year, which saves on polysilicon. In terms of cells, the 7-gigawatt high-efficiency PERC capacity put into production in the second quarter finally reached full capacity in the third quarter, causing cell production costs in the third quarter to drop by more than 10% sequentially.

The company's large-size module products accounted for nearly 50% in the third quarter, a significant increase from less than 20% in the first half of the year. The global demand for solar remains strong, but installation costs are rising over a period of time. We are committed to providing customers with the best solutions based on technological innovation and product competitiveness.

(foreign language)

[Interpreted] Modules at power generation products with a life cycle of 20 to 25 years and product performance is crucial. Recently, our high-efficiency N-type monocrystalline silicon solar cell reached a maximum conversion efficiency of 25.4%, setting a world record yet again. Based on our continuous leading R&D capabilities and 2 years mass production experience, we are quickly expanding N-type cell production capacity.

We are preparing for approximately 16 gigawatt of 10 -- N-type cell production capacity to be operational in the first quarter of 2022, contributing about 10-gigawatt output for the full year. This will help make up for the lagged cell production capacity and, hopefully, will lead the industry into the era of more premium and high-efficient N-type products.

Recently, we launched a new series of N-type modules with maximum power output of 620 watts. Next year, we are planning to increase our global market share by enhancing our sales and promotions of N-type products and achieve at least 50% growth in annual shipments.

(foreign language)

[Interpreted] We continued to improve our global supply chain infrastructure. So far, we have signed polysilicon supply agreements with a number of overseas polysilicon manufacturers and have strategically invested in Tongwei Sichuan Yongxiang and the Inner Mongolia Sichuan high-purity polysilicon production project.

Our 7-gigawatt monocrystalline silicon wafer plant in Vietnam will commence production in the first quarter of next year. After that, we will have approximately 7 gigawatts of integrated mono wafer-cell-module manufacturing capacity overseas. A sound and diversified global industrial chain infrastructure will enable us to be more flexible in terms of order production and customer delivery as we continue to provide integrated services for our global customers.

(foreign language)

[Interpreted] Before turning over to Gener, I would like to go over our guidance for the fourth quarter of 2021. We expect total shipments to be in the range of 7.3 to 8.8 gigawatts, including module shipments to be in the range of 7 to 8.5 gigawatts for the fourth quarter of 2021. Total revenue for the fourth quarter is expected to be in the range of USD 1.8 billion to USD 2.2 billion. Gross margin for the fourth quarter is expected to be in the range of 13% to 16%.

The annual mono wafer, solar cell and solar module production capacity is expected to reach 32.5 gigawatts, 24 gigawatts and 45 gigawatts, respectively, by the end of 2021. With impact from less shipments to the U.S. market and the global logistics situation, we are lowering the guidance for our full year 2021 shipments, including wafers, cells and modules, to be in the range of 22.8 gigawatts to 24.3 gigawatts.

Thank you, Ripple.

Total shipments in the third quarter were 5 gigawatts, with module shipment increasing 18% sequentially to 4.7 gigawatts. In terms of module shipment by region, Europe and Asia Pacific were the main contributors. Driven by China's combined goal of [30 60] and the efficient energy transition, shipment to Chinese market doubled sequentially. Shipments to emerging markets increased significantly both year-over-year and sequentially, a sign that markets are benefiting from policy support.

The People's Bank of China recently launched a carbon emission reduction support tool. As large-scale and distributed projects start construction, next year's total pipeline in Chinese market is expected to exceed 100 gigawatts. Looking forward to 2022, market demand in China, U.S., India, Europe and Australia will continue its upward trajectory.

Module prices remain high, which was accelerated the penetration of products with lower LCOE and the penetration of the distributed generation business, which is less price sensitive. Our distributed business accounted for approximately 35% to 40% in the third quarter and over 50% in some markets such as Australia, Japan and Brazil.

Clients have been favorable towards our premium-quality products such as Tiger Pro and a 2-millimeter product, which was specifically designed for residential C&I distributed generation facilities. Meanwhile, our global brand awareness and our global marketing team strengthened the market competitiveness for our BIPV products.

We have recently won the bid for the new Dubai Electricity & Water Authority's headquarter building projects, which will become the world's largest and tallest single building equipped with BIPV system. The BIPV system uses advanced N-type cell technology and modules with a transparency range between 3% and 16% and can provide a power output of 500 watts to 708 watts.

As one of the directions of our core developments, the proportion of shipments in the distributed generation market will further increase next year. Shipments of the Tiger Pro products that can bring lower LCOE accounted for nearly 50% in the third quarter and are expected to exceed 70% in the third -- fourth quarter. In 2022, we will optimize and match the capacity advantage of Tiger Pro and Tiger Neo product, solidifying our leading position in high-efficiency products.

In terms of contract performance and pricing in the face of uncontrollable factors such as the severe volatility in upper stream raw material prices, energy quota and the carbon emission control, we have established a market forecast and short-term dynamic review mechanism. For newly signed orders, we are securing materials in advance and introducing a floating price clause in contracts in order to avoid negative impact of market volatility to some extent. In short, we are very optimistic about the market demand. Short-term problems such as raw material supply, logistics, et cetera, are gradually being sorted through and remitting effort. We'll focus on winning the market with the best products and services.

With that, I will turn it over to Pan.

Thank you, Gener. Let me go into more details about this quarter now.

Total revenue was $1.33 billion, an increase of 8.1% sequentially, mainly benefiting from substantial growth in module shipments. Gross margin was 15.1% compared with 17.1% in the second quarter this year.

Total operating expenses in third quarter was $184 million, an increase of 18.2% sequentially. The sequential increase was mainly attribute to an increase in selling and marketing expenses due to rising shipping costs under a tight global logistic market. We remain flexible and adjusted both domestic and overseas shipments, mitigating the negative impact on profitability from volatile shipping costs by establishing long-term shipping agreements with logistic companies.

Total operating expenses accounted for 13.8% of total revenues in the third quarter this year. Excluding impacts from shipping costs, the ratio of operating expenses was stable compared with the second quarter this year.

EBITDA was $89 million compared with $143 million in the second quarter of 2021. Taking into consideration the significant increase in benefits from change of fair value due to an decrease in the company's stock price in the third quarter this year, net income was $30.1 million, a significant increase both sequentially and year-over-year. Non-GAAP net income was $2.5 million, and diluted earnings per ADS was $0.05.

Net foreign exchange loss, including change in fair value of foreign exchange derivatives, was approximately $1 million, a significant decrease from a net exchange loss of $9.4 million in the third quarter of last year and flat sequentially. We will continue to optimize our hedging against foreign exchange risk to reduce impact of foreign exchange volatility on our operating results.

Moving to the balance sheet. At the end of the third quarter, our balance of cash and cash equivalents was $1.1 billion, which improved from the second quarter. Accounts receivable due to -- due from third parties was flat sequentially, and AR turnover days were 65 days compared with 62 days in the second quarter this year. As shipments to the U.S. and global transportation routes were hampered by disruptions, inventory turnover days was 171 days compared with 138 days in the second quarter of 2021.

Total debt was $3.69 billion at the end of the third quarter compared to $3.12 billion at end of the second quarter this year. Out of the total debt, $68 million was related to international solar projects. Net debt was $2.5 billion compared with $2.1 billion at end of the second quarter this year.

This concludes our prepared remarks. We are now happy to take your questions. Operator, please proceed.

(Operator Instructions) First, we have Philip from ROTH Capital Partners.

The first one is on your Vietnamese wafer capacity that you're bringing online. So 7 gigawatts by Q1, and with that, you'll use German poly, I believe. When would you expect shipment volumes or shipments to hit the U.S. that are using either German poly or perhaps even U.S. poly? And then the Vietnamese wafers and then Southeast Asia cell and module, could we see the volumes to the U.S. start to restart in Q2? And then how much -- how many gigawatts do you expect to ship into the U.S. in 2022 as a result of this new supply chain?

Thank you, Philip. Great question. I think we have made a very strategic move to establish our [Venice] wafer factory together to secure long-term contracts with multiple, let's say, non-China polysilicon suppliers. According to -- based on the current schedules, we'll start to ramp up our Vietnam wafer factories by early Q1, let's say, January, February. And we are expecting a full value chain established outside China will start to fully utilize by end of Q1.

So ideally, the products start to arrive in U.S. market by late first half, let's say, May or June. And the massive volume will be definitely second half. So total volume-wise, we are looking to around -- production numbers, we're looking at around 5 gigawatt of production. And considering the uncertainty of logistics, we are not 100% sure how many of them will arrive at U.S.

Great. And then when you think about the anti-circumvention risk, we saw the current case go away, but I think the petitioners through their law firm and Tim Brightbill were talking about potential other cases. Do you have a view on what those cases could be and how they could impact this capacity? Or what's your view in general about that?

Thank you. I think these, let's call it, trade wars will stop. That's our long-term view. That's why we have made the decision to establish a fully, what we call, integrated value chain or capacities in both Vietnam, Malaysia, including U.S. So that should be our ultimate solution to solve all those potential trade issues. And at least that's the best solution we have right now. In long term, really, we don't have a crystal ball to see it, but we are expecting the solution Jinko can bring to the market should be the most -- one of the most reliable ones.

Right. Okay. And then another one here, on the Section 201. We recently saw a surprise exemption reinstated for bifacial modules in the U.S., which means, I think, the modules can come in tariff-free. And then also there might be substantial refunds for modules that have been already shipped in. So I was wondering if you might be able to quantify what the Section 201 refunds could be, the timing of it. Have you received any thus far, for example? And then what -- if this decision gets reversed yet again, what would the process be to perhaps return that money to the U.S. government? Have you thought through the longer-term picture for the Section 201 bifacial situation?

Thank you, Phil. I think from the cash flow-wise, we are not expecting to receive any refund in any short term. We have read your report about the uncertainty of the potential appeal from the government side. We believe that's highly possible, and that's why we still put a question mark there. We do not expect such kind of lawsuit will reach an end pretty soon. So -- but if we can get some potential refunds, then definitely, that will be very positive for the company. But we have to wait for quite a while to see the final results, so we will wait and see. We haven't planned it in our cash flow yet.

Okay. One last one and I'll pass it on. As it relates to margin, you gave us, I think, some capacity numbers for 2022. Maybe can you talk about how you see margins trending in '22? Especially if the assumptions that you make now kind of persist, do you think margin expansion is in -- is potentially available, especially with flexible pricing based on input costs as we get through '22?

This is Charlie. And looking into next year, 2022, I think we are more confident the margin will be improved year-over-year. And this year is very special and challenged, particularly the supply chain shortage and high input cost, including the, let's say, the logistic issues as well. And looking into next year, we think the supply chain will get more friendly and -- compared to this year. And on top of that, we will increase our production integration level, and we are expanding the N-type cell capacities.

And next year, we -- or roughly, we estimate the capacity will reach to wafer, 40 gigawatts; cell, 40 gigawatts; and module, 50 gigawatts; and roughly 80% integration level. And we are expecting the -- we can get the more, let's say, net income contributions from the N-type. And because we are more confident and we have the advantage of this next-generation technology in terms of the, let's say, production level and the cost advantage, which is the product is very popular in the market and we can get the price premium.

Next, we have Gary from Credit Suisse.

So my first question is, can management elaborate a little bit more on our upcoming N-type cell capacity? So I think we are relatively kind of early in terms of the relatively large-scale kind of N-type capacity. So can management maybe share with us more on your expected cost competitiveness of this new capacity? And how do we think of the kind of the 2 technology trends, the TOPCon and the [JT]?

Gary, this is Charlie. And we are investing 16 gigawatts N-Type TOPCon capacities. And we have the 2-years experience in running roughly 900 megawatts capacities in our factories. And now the production, let's say, metrics and the cost advantage versus the PERC. And we are more worried in, I think, leading the industries. And for next year, when we ramp up, we expect to ramp up the capacity in the first quarter and, I think, in full operations roughly in the second quarter.

And in terms of the integrated production cost for the N-type, which is the PERC, for the medium term, we expect the costs will -- the difference versus the PERC is very small and even the same. And more important, if we look at the conversion efficiencies, the N-type versus -- as well as the electricity -- additional electricity generations to our customers, and we believe with a very good leading market product, we can get the price premium and get more profit contributions for this product.

Yes, okay. My second question is on the module shipment. So I think some of our peers are relatively cautious on the 4Q kind of demand, but I think our guidance still suggests meaningful kind of a Q-on-Q shipment increase. So just one, if management touch more on the -- how can we achieve this kind of still very strong Q-on-Q increase?

Thank you for your question, Gary. This is Gener. I think quarter-over-quarter, we have pretty strong Q4 shipment expectations, mainly it's driven by combined demand from China and the non-China market. And I think after the relatively weak quarterly shipment in the first 3 quarters of '21, I think most of our customers, let's say, non-Chinese market customers are gradually accepting the market balance of the new prices. And they are trying to take the time, try to close the project or close their pipelines.

And for China demand, everybody knows there are quite a strong demand in China in the last 2 months' time. So we successfully secured quite significant volumes and to supply pretty strategic projects across China market. So combine those 2 factors together, I think we can achieve a pretty strong shipment in Q4.

Okay, yes. So my last question, and I'll pass on, is on the whether -- can management share with us the latest update of our subsidiary Asia listing?

For the IPO process for our subsidiaries, we are on track to get the public data nearly done. And by the end of September, we got approval from the Shanghai Stock Exchange. Now we are in process in the registrations with Chinese securities regulators. And we expect to -- roughly in the next 1 and 2 months, we can -- to get everything ready and to get the IPO done.

(Operator Instructions) Next, we have Rajiv from Sunsara Capital.

My first question is about the potential shipments that were lost because of logistics issues in the third quarter and might be lost in the fourth quarter as well. Could you help us give us a sense of how much in terms of megawatts you might have lost in terms of shipments not happening because of logistics in Q3 and what that impact might be in Q4? And also help us understand if these lost shipments, is that a temporary timing thing where the shipments get pushed out into 2022? Or this -- that market opportunity was -- is lost forever?

Thank you. I think -- let me take your first question. And in terms of [quantities] of the delays, I think it's -- we are looking at, let's say, a 100-megawatt level of delay. I think the delay between Q3 and Q4 is not because we haven't shipped or we haven't produced the module or its customer doesn't want the module. It's because of the logistic delay. There's a lot of traffic jams across the ports like, say, in Europe or in U.S.

So accounting-wise, we cannot recognize those modules, actual delivered module into our customers' sites or in their warehouse. That's why we have to delay the revenue recognition, I think, from -- partially from Q3 to Q4 by approximately 100 to 200 megawatts, let's say, several hundred megawatts. But it's -- the module itself has been produced. The contract is signed, and the price has been secured. So we just need to wait until this logistic bottleneck had been gradually solved, and we can recognize revenues in Q4.

And just to follow up on that. In terms of Q4, are you experiencing similar logistics issues or maybe even worse because a fourth quarter Christmas, all the competition you have from the rest of retail? So is the problem for the fourth quarter segment even worse?

I don't think so. At least Q4 will be slightly better than Q3. The reason is the China market, domestic market, demand is pretty strong. So we can divert the risk, not only relying on the shipping line companies. We can recognize more revenue in domestic market. So that's one of the positive side.

And also, after these port issues across Europe or U.S., the government side has taken many actions to push for the port itself to speed up the loading proceed or unloading speed to make sure everything can go faster. So we have seen this improvement from the logistics side as well. So combine those 2, we believe the situation will be much, much improved compared with Q3.

Okay. My next question is about the average selling prices. And you mentioned that there was a small increase in the selling price, but obviously, it did not offset the increases in material costs. Could you talk about to what extent your shipments in the third quarter were still subject to long-term contracts where you had to honor pricing that was negotiated sometime back or close to pricing that was negotiated sometime back versus pricing that is -- versus modules that are shipping into prices that are currently available in the spot market?

Thank you. I think from contract execution-wise, we have to find a mutual solution together with our customers. So we -- since we signed a contract, we cannot expect our customers to swallow all this potential increase because of raw materials. We have to find mutual solutions together. So there are quite several different kind of solutions we can reach agreement with customers. And so I won't spend too much time to illustrate everything.

But at the end of the day, I think over 75% of our contract delivered or executed in Q3 are all-time heritage. So -- and also we have to find lots of solutions, combined with short-term and long-term solutions together with our customers to find a win-win solution for long run because we are not doing business in spot market. We are doing business in power sector, which needs a very long-term sustainability and win-win solution with the customers for long term. There will always be market up and down. We cannot just ruin the customer relationship because of one-time issues.

Right. So can I -- just so I understand. You mentioned that 75% of shipments were into long-term contracts, and these long-term contracts have been somewhat modified to reflect the realities of the cost pressures. Can you give us a sense of what percent of -- going into 2022 and learning from the experience of these higher prices that we are now having to live with or higher cost of materials that we are having to live with, what percentage of shipments might be -- if you do 30 to 35 or 32 to 36 gigawatts next year, what percentage would be long-term contracts that have already been signed? Or are there -- or versus contracts where pricing has not yet been determined?

So we have approximately 25% to 30% of order book being signed or committed to the customer end for 2022. Some of the contract is based on the secured price and terms. Some of the contract is just a framework contract, which means we keep the commercial part flexible up to the market. So -- and we also have some of the framework contract which has index-based pricing mechanism, which allows both our and our customers to have flexibility when the market goes up and down. So basically, we have learned a lot of lessons from 2021 and we carried on, and we established some new business model with our customer end for '22 and longer term.

So what you're saying is that the framework agreements will give you more flexibility on pricing if the cost of material moves in -- or shipment cost moves in, in an unfavorable way. That -- is that also the reason why you are confident that the gross margin next year should improve because you are embedding a higher gross margin than was achieved in 2021?

Well, the gross margin will be a separate topic. I think the gross margin improvement mainly comes from a much improved vertical integrated capacity expansion by end of this year, so which will contribute a lot of margin improvement. Another important factor will be the -- our new product, Tiger Neo product, will contribute a lot of premium because of the nice parameters and extra power output such product is -- can generate.

And the third one will be our portion of, we say, large-dimension products. For example, our 2-millimeter product portion increased from, first half, around 20% to 25% to, end of this year, around 60% to 70%. And next year, it might go even higher, so which give us a very good competitive edge. So I hope that answer your question.

Yes, yes. And one last question for Charlie. Charlie, the general and administrative expenses have been much higher this year than last year in the first 3 quarters. And first 2 quarters, they went up quite substantially in the third quarter as well. Can you give us some sense of why that increase is happening? And in particular, how much of this increase in cost is related directly to the IPO in China?

Thanks for your questions. And I think the operating expense is relatively high. It's contributed by, I think, firstly, the -- it's the shipping cost. The shipping cost is pretty high. And we have very high exposures to shipments on China. Second one is the IPO cost. IPO cost, I don't believe it's very significant, roughly maybe USD 2 million for the third quarter.

And on top of that, I think, third quarter, we -- our shipment is not so big relatively. And by fourth quarter, we have very big jump on shipments, which will dilute the operating costs. And excluding the -- I think excluding the, let's say, the variable cost, particularly the shipping cost, the warranty cost and the fixed cost for operating expenses, it's fixed and relatively stable.

Actually, Charlie, I'm sorry, I missed something that you said about what amount the IPO costs were. But the question that I'm trying to grapple with is the G&A, which normally should not increase that much. Last year, it was running at around $50 million a quarter. This year, it started at around $55 million to $60 million, and it was almost $70 million in the third quarter. So that's a pretty substantial increase. Can you tell us what is driving this increase and what we should expect for Q4?

So the G&A cost is administrative cost. I think it should be stable. Looking for the future, it should be stable. And it could -- we have some IPO costs. And we have some, I think, the quarter adjustment for the, let's say, the provision for the accounts receivable. And excluding that, I think it's relatively stable.

(Operator Instructions) Next, we have Alan from Jefferies.

My first question is about -- so just to check, because the inventory has increased by quite significantly, so just to check, is this tied to the shipment in the sea to overseas market?

So I think 2 aspects. One is the inventories on transit to overseas markets and the challenge by the logistic and the delay. The second one is we are expecting, back to third quarter, the material cost is still on the upside, so we intentionally put more resources and get more materials to make sure we have relatively advantage to -- for the production in the fourth quarter.

Understood. And so when we come to the fourth quarter, I've noticed that the company was guiding for 12% to 15% on gross margin over the last 3 quarters, but the guidance has increased from 12% to 15% to 13% or 16%, while the upstream polysilicon cost has increased a lot. So first of all, I would like to learn from the management, is it because of the increase in percentage of large-size module and the cost-cutting efforts so that the company is confident to make a probably higher margin even when upstream poly cost is higher?

Yes, it's a combination. One is you are talking about the large-size percentage. And we -- our capacity is really 7 gigawatts we talked in the press release. Second one is percentage. And the third quarter, the large size, 182, roughly 50%. And the fourth quarter may reach to 70%. And on top of that, we are expecting that our ASP will continue to increase, I think, particularly because we have shipments -- more shipments in China. And China ASPs, the spot market, the price is pretty high. And the Q4 versus Q3, the ASP has, I think, increased. It's roughly $0.01 to $0.02 per watt.

Understood. I guess -- okay. So it's a combination of these factors. And on the N-type new-generation cells, so I suspect the company is using the TOPCon technology and wonder if the yield rate is -- is it like able to disclose the yield rate of the production line? Because it was a concern on the yield rate because TOPCon has more production steps.

We are reaching -- for N-type TOPCon, we are reaching almost 99%. So I think that's a leading number in this industry. That's why we have our confidence to expand our capacity to such sizable volume to make -- to extend our competitive advantage as well.

Understood. That's very impressive. I think it's one of the leading indicator in the market.

Next, we will be taking one last question from Brian Lee from Goldman Sachs.

I might have missed this or maybe I'm misinterpreting something, but could you help reconcile? I think you mentioned 100 to 200 megawatts kind of revenue recognition issue Q3 to Q4, but the actual shipment guidance is changing by 2-plus gigawatts versus what you had previously. So is that being pushed even further into the early part of 2022? Or what's the, I guess, the disconnect between the 100 to 200 megawatts you're calling out specifically 3Q to 4Q and the bigger kind of shipment guidance change that's happening here for the full year?

Yes, sure. I think my previous answers is between these delays between -- revenue recognition delays from Q3 to Q4, it should be like, let's say, somewhere around 200-megawatt level. So -- and the delay will continue, but I think the improvement is definitely be better than Q3 because we have seen this traffic jam across all the main ports is improving day over day, week over week. So that's why I'm saying the shipment in -- and together, we are shipping more into China market in Q4, so that's why we don't have to suffer the logistic problems for the -- from the shipping lines. So those are the 2 factors why we say the revenue recognition in Q4, the pressure will be lighter or easier than Q3.

So regarding the guidance change, I think it's mainly because we have some delays of one of our workshops, cell workshops. That give us kind of a ceiling for this capacity expansion. So that's why we naturally adjusted our guidance accordingly.

Okay, that's helpful. And so just to be clear, you haven't seen any cancellations of shipment schedules whether for this year or into next year. Just things are changing on the schedule, but you haven't seen anything canceled outright?

We have seen pretty, let's say, several cases canceled by the shipping line companies, but we always can find some solutions because we have these multiyear shipping line contracts with all the major shipping lines. So even if they cancel the shipments or the shipping lines this week or next week, we can always try to find out a solution to squeeze them in for the next one in the, I don't know, next months or the months after, right?

Okay, fair enough. And then maybe just 2 last ones for me and I'll pass it on. Are you reiterating your CapEx guidance for 2021? I think you had said USD 1.1 billion before. And then any early thoughts on capacity increase in CapEx for 2022?

Yes, thank you. This is Pan. And for the CapEx side, we expect that in Q4, we'll have a $400 million expectation, so the total CapEx in this year will be $1.2 billion. And next year will remain the same, $1.2 billion for 2022.

Okay. So flat CapEx year-on-year. Any thoughts around the funding for $1.2 billion of CapEx in 2022?

No. We are -- our subsidiary is in the China IPO process. And we're expecting to raise at least RMB 6 billion. So that is the most reliable, the majority part. Upon that, we will make it some debt financing as well.

I will now hand the session back to you. Please go ahead, Ms. Ripple. Ms. Ripple, please go ahead.

Hello. Yes, let's end the call today. Thank you.

Ladies and gentlemen, this concludes today's conference call. Thank you for your participation. You may now disconnect.