Xunlei Ltd (XNET) 2017 Q3 法說會逐字稿

Good morning, and thank you for standing by for Xunlei's Third Quarter 2017 Earnings Conference Call. (Operator Instructions) Today's conference call is being recorded.

I would now like to turn the conference over to your host today, Ms. Dani Zhang, IR Director of Xunlei. Ma'am, please go ahead.

Thank you, operator. Good morning, and good evening. Welcome to Xunlei's Third Quarter 2017 Earnings Call. I'm Dani Zhang, Investor Relations Director at Xunlei. With me today on the call are Mr. Lei Chen, our CEO; and Mr. Eric Zhou, our CFO. Today's conference call is being recorded, and a replay of the call will be available on our IR website following the call. Our earnings release was distributed yesterday -- today, and it is now available on our IR website.

Before I get started, please note that the discussion today will contain certain forward-looking statements made under the safe harbor provisions of the U.S. Private Securities Litigation Reform Act of 1995. These forward-looking statements are subject to risks and uncertainties that may cause actual results to differ materially from our current expectations. Please refer to our SEC filings for a more detailed description of the risk factors that may affect our results. We do not assume any obligation to update any forward-looking statements, except as required under applicable law.

During this call, we will be referring to both GAAP and non-GAAP financial measures, and the non-GAAP measures are reconciled to the comparable GAAP measures in the table attached to our earnings release, which can also be found on our IR website. Please note that all numbers are in U.S. dollars unless otherwise stated.

I will now turn the call over to Lei. Lei, please?

Good morning, and good evening, everyone. Thanks for joining us today for our earnings call for the third quarter of 2017. I'm pleased to report another quarter of progress and achievement. This is the seventh quarter in which we achieved year-over-year quarterly revenue growth. We beat the revenue guidance of between $41 million and $44 million for the third quarter of 2017. And we expect the growth momentum will continue into the fourth quarter of 2017.

During this quarter, total revenues were $47.3 million, which represented approximately 16% and 14% increase on a year-over-year on a sequential basis, respectively. The growth was mainly driven by our core cloud computing business and our new live video business. Cloud computing grew by approximately 104% on a year-over-year basis and 24% over the previous quarter, respectively. Live video grew by 1,197% on a year-over-year basis and 154% over the previous quarter, respectively. We are encouraged that our growth initiatives are working.

Now I'd like to add some color to the performance of our major products, including cloud computing, membership subscriptions and Xunlei Mobile. In early 2016, we took a unique approach to cloud computing, which we branded as crowd-sourced computing. We leveraged users' idle bandwidth, storage and CPU to offer cloud computing services to businesses. By this approach, we can help business to reduce cost and improve performance. In the past year or 2, we have seen the trend of similar and relevant technology, such as Edge Computing and Serverless PaaS, sprouting from different technology thinkers and evangelists. We believe this trend will continue for a couple of reasons. First, Internet of Things becomes more important and starts to fundamentally change how we live and how we conduct business. This leads to a variety of applications where such technologies provide better solutions. Second, the demand for bandwidth, storage and computing power grew rapidly, while the cost of CPU, flash and other computing units stopped rapid decreasing. We even saw a reverse price trend in some of these units. A solution that can fundamentally reduce cost will become more and more desirable in this market.

To strengthen our lead in crowd-sourced computing, in September, we launched a hardware device, named OneThing Cloud. This can be considered as a Google Drive as well. The online personal cloud market, which is often used to characterize market for products similar to Google Drive in China is very big. According to incomplete statistics, China has more than 400 million personal cloud users. Moreover, Chinese online personal cloud vendors often offer terabytes of free storage to users by offering a hardware that's placed at your home, but also satisfy the same needs. OneThing Cloud provides an added benefit of privacy, security and better user experience. The product is initially offered on Taobao's crowdfunding platform, then sold on JD.com. The actual sales and shipments of the product started in late September. The product has been very well received by consumers. On JD.com, customer reviews have been nearly 100% positive. And the accumulated reservations for the device just now exceeded 9 million units. We also introduced a voluntary OneThing incentive program based on OneThing Cloud. By enrolling into this program, users give permissions to us to use their idle bandwidth, storage and CPU through OneThing Cloud and receive a blockchain-based crypto token as reward. This program is also fairly popular among OneThing Cloud users.

The introduction of the new device and blockchain improves our crowd-sourced computing business. The device offers better hardware configuration, which allows us to go beyond CDN domain and offer new types of services to our customers. We already started testing with one of our top customers. Blockchain technology allows us to construct an automated, efficient and trustworthy reward mechanism. It is a crucial building block to the crowd-sourced computing ecosystem.

Let me turn brief attention to our subscription business. Revenues from our subscription business accounted for about 44% of our total revenues in the third quarter of 2017. The amount of the total revenues was relatively stable, but the percentage declined as our cloud computing business contributed more to the total revenues. The subscribers will soon officially start to offer a new product. This product is more focused on downloading and is free of such features as games and short videos. It is a clean and light downloading tool and available only to the subscribers. So far, customer feedbacks of a test version of the new product are quite positive. We will continue to focus our attention in improving downloading experiences, both to subscribers and the nonsubscribers.

Let me now briefly comment on our mobile products. Xunlei Mobile, our mobile app, allows users to search and download contents, distribute short videos and participate in live video programs. Users now, on average, spend up to 40 minutes daily using our mobile app. This is partly due to our embedded short video functionality, which started to influence their behavior. After several months of transition, more than 40% of our mobile DAUs watched short videos. As we mentioned in the second quarter, we were testing live video product on both mobile and the PC end. We're delighted to see revenue from our live video product rose by 154% in the third quarter as compared to the previous quarter. In the coming months, we will continue to manage top posts with better incentive mechanisms and try to integrate live video and short video products to enhance live video conversion rates and revenues as well. Further, building a short video and live video content ecosystem led to higher mobile advertising revenues, which grew 22.7% and 9.2% on a year-over-year and sequential basis, respectively. We hope our ongoing efforts to strengthen our mobile products will contribute to both the top and bottom lines.

With that, I'll now turn the call over to Eric to review the third quarter financials. Eric, please?

Thank you, Lei. And good morning, and good evening, everyone. And once again, welcome to Xunlei's third quarter earnings call. For the third quarter, total revenues were $47.3 million, which represented a 15.6% growth on a year-over-year basis. Increases in cloud computing and live video business were the major contributors for the year-over-year growth. Subscription revenues were $20.8 million. The number of subscribers slightly increased from 4.09 million in the second quarter to 4.18 million in the third quarter. The ARP value slightly decreased to RMB 33.2 in the third quarter from RMB 34.4 in the second quarter.

As Lei mentioned in his section, we were very focused on executing and customer care. We are taking measures to improve user experience, like releasing a new premium version accelerator product for members. For our cloud computing business, revenues from cloud computing increased 24.4% on a sequential basis primarily due to the sales of our new encouraging hardware device OneThing Cloud in September. The device is sold on JD.com and on the website of OneThing Technologies.

The accumulated reservations for the device have lately exceeded 8 million units, and the demand far exceeded available supplies. Online advertising revenues, which include mobile advertising revenue, in the third quarter were $5.7 million. Mobile advertising revenue grew 22.9% in the third quarter of 2017 compared to the corresponding period of last year.

Cost of revenues was $28.8 million compared with $24.2 million in the second quarter of 2017. The increase in cost of revenues was primarily due to increased sales and onetime accelerated depreciation of servers in amount of approximately $1.3 million.

Gross profit margin was 38.6% as compared with 41.1% in the previous quarter. The declining of the gross margin was mainly due to changes in the revenue structure, where revenues from sales of fast-growing products accounted for a larger percentage of total revenues as well as a onetime write-off expense.

Operating loss was $30.4 million compared with operating loss of $11.3 million in the previous quarter. The increase was mainly due to several onetime write-offs in an aggregate amount of $21.8 million, of which approximately $11.8 million was due to the Kankan business the company sold before, $4.8 million due to write-off of the intangible assets for a prior-period acquisition of Kuaipan Personal and $1.3 million for accelerated depreciation of servers.

Net loss for continued -- net loss from continuing operations for the quarter was $25.6 million compared with a net loss of $9.7 million in the previous quarter. The increased net loss was due to the write-offs discussed above.

Turning to our balance sheet. We continue to have a strong balance sheet. We ended the quarter with cash, cash equivalents and short-term investments of $357.7 million, which is down slightly from $364.8 million at the end of the second quarter of 2017. Cash, cash equivalents and short-term investments per ADS was approximately $5.40 at the end of the third quarter of 2017.

Let me finish by going over our guidance for the fourth quarter. We expect our revenues to be between $55 million and $60 million for the fourth quarter of 2017. The midpoint of the range represents a year-over-year increase of about 46%. We also expect significantly improved bottom line for the fourth quarter.

With that, I'd like to conclude our prepared remarks today. And I will now turn the call over to the operator for your questions. Operator, please?

(Operator Instructions) The first question comes from the line of [Jack Wu], a private investor.

Congratulations to our CEO, Mr. Chen Lei [sic], for strategic release of OneCloud, which became the 2017 hottest product in China. Congratulations to Xunlei for winning the award of China national innovative and high-technology title for shared computing and blockchain technology.

Thank you very much.

Congratulations to our CEO -- CFO, Eric Zhou. I have several questions tonight to ask. My first question is to CEO, Mr. Chen. I remember at the One Cloud release conference, you mentioned that starting from the release of One Cloud, Xunlei will be all in blockchain. My question is why blockchain is so critical to Xunlei? What kind of opportunities do you see in the application of blockchain technology. What are the advantages Xunlei has in blockchain?

Thank you, Jack, and thank you very much for your kind words. Blockchain technology is a future technology trend that everybody recognizes. It is particularly relevant to Xunlei, because Xunlei as a technology company, our technology is focused on very distributed computing. So we -- our strength starting from the P2P and P2SP technology, which is by nature distributed. Blockchain technology adds to the decentralized technical reservoir, which plays really well with the P2P technology that we have. Furthermore, blockchain technology provides mechanisms for us to involve large number of consumers to our crowd-sourced computing ecosystem. It provides a fast and small amount -- extremely small amount and efficient, trustworthy mechanism as a reward system for people who contribute their computing power. Furthermore, blockchain technology is also, we believe, relevant to our consumer business, which involves the distribution of content. So we also see a trend in the current technology development of decentralization. In an era, where privacy and security become more and more issue for consumers, decentralized technology will potentially play more and more important roles in people's life and business. Again, thank you very much, Jack. And hopefully, these answers are...

Yes, that sure answers many of my concerns. And can I still ask questions?

Yes, yes, of course.

So I noticed and you mentioned there are more than 8 million reservations for One Cloud on JD.com, and Xunlei, I noticed that it is currently running a lottery system for One Cloud. And on Alibaba people are selling One Cloud at least at triple price. My question is, what is the bottleneck for Xunlei to release more One Cloud. When will consumers be able to purchase One Cloud as they wish?

Thank you, Jack. So -- a little correction. So the accumulated reservations has exceeded 9 million as of today. Apparently, we were not prepared for customers' enthusiasm of this new product. The production -- so since September, we've been enhancing and improving our product lines, and we have more than tripled our production so far and is continuing to increase our production. So we hope that in the next few months our production can catch up with demands. But as you know, in hardware, everything is brick-and-mortar. So it really takes time. Furthermore, we're very cautious of the quality of the product I reviewed. So while we're certainly increasing our capacity, output capacity, but this has to be done without any impairment to the quality of the product that we deliver to our customers.

Okay. Thank you, CEO, Mr.Chen Lei [sic]. And my next question is to our new CFO, Eric.

Yes, go ahead, Jack, please.

Yes. So when people share their computing power through One Cloud, they are rewarded with OneThing coins or, we call, tokens. And Xunlei promised to offer future gifts and services to those owners of those coins. This sounds like a lot potential liability to me. According to U.S. generally accepted accounting principle, or U.S. GAAP, potential liability needs to be reported on financial statements if the amount can be reasonably estimated or disclosed in the notes to the financial statements. How will Xunlei deal with the liability associated with OneThing coin or the tokens on financial statement? That's my question.

Thank you, Jack. And we -- actually, we are discussing with our auditors, and also we internally -- we are also discussing how we are going to achieve this. And -- because right now we just launched the product. And so far, the impact on our financial statements are not big. But that being said, we are starting this issue, and you will see more disclosure in our annual report next year.

The next question comes from the line of Richard Chen.

Mr. Chen and Mr. Zhou, congratulations on the strong quarter, and we're also seeing your strong revenue projections for the fourth quarter as well. With regards to the OneThing Cloud, would you be able to comment on the shipment results? I understand that you mentioned that the orders reservations has reached 9 million, but just wondering where shipments would be. And perhaps also comment on the financial impact of OneThing Cloud from a sort of revenue and gross margin perspective, if you could.

Thank you, Richard. Apparently, we're not segregating -- segmenting the OneThing Cloud revenues in particular. So we're currently not disclosing the numbers you just asked.

Okay, understood. Or perhaps maybe if you could comment on the cost structures. I noticed that this current quarter bandwidth cost as a percentage of revenue has dropped slightly versus the previous 2 quarters. If you could just comment how your cost structure is evolving and whether that's related to sort of the mix -- change of the business mix that you're facing right now.

So as you know -- you may know that we started the innovation on cloud computing in 2015. Crowd-sourced computing is a very new and innovative technology. We're still spending heavily in research and development. And we're currently not making a profit on cloud computing. While cloud computing is taking up larger and larger percentage of our total revenue, it has impact on the structure -- cost structure of our total revenue.

Which, in other words, we expect the bandwidth cost -- the percentages of the bandwidth cost to total revenue will be declining in the future.

(Operator Instructions) The next question comes from the line of Leon Dean from Morgan Stanley.

Just a quick question. I guess, if I could, I have 3 questions. One is, what is the margin profile for OneThing Cloud? That's one. Two is that how do you recognize revenue for OneThing Cloud, just given that you have 9 million reservation, and you are going to speed up the production of OneThing Cloud. And I'm just curious on your fourth quarter guide and as it seems quite conservative given the ASP of OneThing Cloud and the number of units that's on back order? So that's my second question. My third question is, the long-term, I guess, product pipeline or division that you guys are thinking about, what do you guys -- what is the long-term, I guess, vision for, one is, OneThing Cloud and, two is, the entire company, what other type of products do we have in the works, whether it's hardware or software, whether it is independent from OneThing Cloud or integrated with OneThing Cloud?

Thank you, Leon. So the first question is the margin on OneThing Cloud. We are currently not disclosing that number. So unfortunately, that's not a number that we're ready to disclose at this point. So second question is about how we're recognizing revenue on OneThing Cloud. Maybe, Eric, you can talk about it.

Basically, we recognize revenue on our shipment, not reservations, because the 8 million reservations, they are just reservations, they are not the actual sales. So the actual sales -- revenue will be based on the actual sales.

So correct me if I'm wrong. So what I'm hearing is -- I'm going to paraphrase -- is that we recognize revenue after the shipment is done.

That's correct.

That's my understanding as well. So about long-term product pipeline, the current thinking of the management team is as follows. Xunlei has a downloading product and is widely recognized in Chinese market that we are the leader in this market. What we realize is that we need to continue improving our core user experience around downloading. This is very important to us and very important to our customers. And -- so the company will become more focused in efforts that improve the core user experience around our core functionality. Two, apparently, of course, that cloud computing is our main focus moving forward. The crowd-sourced computing mechanism, as I talked about earlier, I think, provide both methods to reduce -- to fundamentally and hugely reduce the cost structure of conducting business in the internet and other computational areas. It also provide better solutions to some new product -- problems that we're facing when Internet of Things as a phenomena is growing rapidly. We will be very focused and putting efforts and money on the table to not only expand our cloud business, but develop it in depth and improve both the technology and quality of service. So these are the 2 main focuses of the company today. We also have obligation to shareholders to improve our revenues and margins. Beyond the 2 main focuses that I just talked about, we will, obviously, try to monetize with both our consumer product as well as cloud computing business. So that's, I guess -- to summarize, I think the company today needs focus, needs to improve our core product and need to monetize better.

And if I could just have one follow-up regarding to OneThing Cloud. So what is the expectation for the fourth quarter shipment from -- unit wise?

Unfortunately, we're not ready to disclose that number at this time.

Got it. And have we shipped any yet? Or we have not?

No, we've shipped. We've definitely shipped. We actually had -- we actually started shipments in late September.

If I recall correctly, management -- basically, you have 9 million total. And before November 11, I think you might have 2 to 3 million that will have shipped already. Is that number -- is the numbers all right? Or...

No, 2 to 3 million, I think, is still the cumulative reservation numbers.

Got it. Got it. Got it. Okay. Good luck with everything. Hopefully, you guys can help the company transform into something much bigger and take a meaningful spot in the new Internet of Things era.

Thank you very much, Leon.

Thank you.

The next question comes from the line of Cheng Cheng from Guosen Securities.

I want to know now as you know the potential risk of the One Cloud coin and I want to know what the management will do to make money from -- obviously these 2 coin cannot trade on the website. And what will the customer do, let's say, to make money from the token coin and the potential risk continue or not? That's the first question I want to know. And second question is, in the future, how Xunlei will do in the personal cloud computation? This is my question.

Sorry, Mr. Cheng, I think I'm not 100% catching the questions that you asked.

Yes, I want to know -- yes, the first question I want to repeat is, if the customer can't trade their One Cloud coin or the token coin, how our Xunlei company to do with these coins and how to make money from this -- the customers -- how to make money?

I see. So -- okay. I see. The company has no plan to monetize over the coin at all. And as I said before, the crypto token is rewarded to users who volunteer in the OneThing reward program to encourage users to share idle bandwidth, storage and CPU computing power for us to leverage. That is a voluntary program and is set up separate from the sales of the OneThing Cloud product itself. So I'd like to just explain a little bit about the -- about this. So OneThing Cloud as a product is priced at a slightly lower the market -- the competitive market for equivalent products that are offered by other companies. And so after purchasing OneThing Cloud, the user has the option to enroll into a voluntary program and to share their computing power and be rewarded by this crypto token. From the beginning, we as a company have no plan at all to monetize on this coin, and we certainly do not in the future.

Okay. The next question I want to know, what's the difference between One Cloud and ICO? What is the initial difference between these 2?

Well, the explanation -- the definition of ICO, I think, we're, as a private company, we're not in the position to comment what is the very definition of ICO. But what I can tell our investors is that we work closely with the regulatory authorities in China, and we follow all the local regulations and laws. And we conduct business -- we have been conducting business, and we will continue to conduct business in a manner that is honest and responsible and meet all the regulations and law requirements.

The last question. Okay. Compared to China net cash and Tencent or Alibaba, what's the advantage of Xunlei in the CDN field?

Well, apparently, we use different technologies to deliver the same results or better results to our business customers. Our crowd-sourced computing leverages users' idle bandwidth, storage and CPU power. As we all know, shared economy -- the one important advantage of shared economy is low marginal cost. So with the growth of scale that our cloud computing business is continuing to obtain in the market, we will have a better cost structure. And that's one major advantage over traditional CDN offerings. I would also point out that even though we started with our CDN business, that is not the -- that definitely does not set a boundary for us. So we believe the crowd-sourced computing technology can be beneficial and applied to other domain of cloud computing. And we have initiated the effort towards that goal in this quarter. And I wish to report to you further advances in the future.

(Operator Instructions) The next question comes from the line of Ning Hu from AKQA.

I have 2 questions. The first one is for the CFO, Eric. My question is, have you ever count if the OneThing Cloud hardware totally replaced the previous version from Taobao is that could it be profit based on current CDN price? Because the CDN market has been [collapsing] for over a year. So that's my first question. The second question is, based on the company's balance sheet -- so company only has 4% debt. So does the company considering to purchase maybe other CDN business in future to grow revenue? Yes, that's it.

Thanks, Ning. And for second question and as you mentioned correctly, we have a very strong balance sheet, and we have almost like $350 million in cash and cash equivalents, which give us a lot of strength for future opportunities. And of course, we will consider everything that would be beneficial to our business. I think basically, that I can't see today and, of course, Lei may want to add a different word to that one. For the first question, the One Cloud, do you mean that if we in the future the One Cloud income replace the conventional procurement of bandwidth? Is that what's your question?

Yes, my question is because the OneThing Cloud -- I believe it will reduce your cost -- bandwidth cost. So my question is, will the company profit by actually making money based on current CDN price, which is very low? So have you ever count for that? So if you like to complete your launching the OneThing Cloud hardware?

So maybe I will try to answer your question. Currently, our cloud computing business is not making a profit. It's still in the investment phase. The reason is not -- so the reason is not that our expenses -- the Mine Crafter is certainly not the reason for not making a profit right now. Actually, the Mine Crafter and the OneThing Cloud both are ways to cross-source computing resources. And they both have an advantage of the shared economy that allow us to do business at a much lower cost. So it really isn't that OneThing Cloud or Mine Crafter that really is the issue when it comes to making a profit. It really is an issue of our investment and our technology improvement that will help us to turn the profit in the future. I'm actually prepared to say at this point that I do expect the cloud computing business will turn a profit sometime in the next year. And that will be the result of technology improvement.

(Operator Instructions) There are no further questions at this time. Over to you, sir.

Okay. Well, thank you, again, everyone, for joining us today on today's call. As always, the IR team will be available to your answer any questions you may have. Have a good day. Thanks, again.

Thank you very much.

Ladies and gentlemen, that does conclude the conference for today. Thank you for participating. You may all disconnect.