Description

Summary:  XNET is a Chinese internet technology company listed on Nasdaq. The stock is trading around $3.60 with cash per ADS at $5.70. The company has a profitable but flat/declining core business. Management has chosen to make significant investments in a new cloud business division, which is growing rapidly (80% in latest Q) but incurring heavy losses. My bet is this new initiative won’t last forever – either the business turns profitable or mgmt sells the technology/shuts it down. In a downside scenario, I assume two more years of EBIT loss similar to 2016 level despite ramping revenue, and I still get to cash/share of almost $4.8 at end of 2018, 30% above current level. In an upside scenario, a SOTP of cash, the core business and the new cloud business could lead to a double.

It has to be a fraud, right? Since this is a Chinese company trading below cash, the natural question is whether this is a real company or a fraud. One could never be sure, but I am reasonably confident in the company and its financials. First, Xunlei’s legacy download acceleration program has been around and popular for a long time. Essentially all the Chinese websites that offer download service of large files (games, movies, etc) would have Xunlei’s program as one of the default download choices. Secondly, most of the cash on the balance sheet was from IPO proceeds and investments made by Xiaomi at much higher valuation. Thirdly, XNET’s new cloud business is led by the former head of cloud computing at Tencent, and they have won business from some well-known industry players including IQiyi (BIDU) and Panda. Still, given the stigma and low liquidity, please feel free to stop reading and/or give it a low rating. Aside from the usual baggage associated with Chinese companies, the operating loss and lack of analyst coverage and mgmt guidance may have also contributed to the languishing stock price.

Core business: The core/legacy Xunlei Download program was developed by the two co-founders (both were Duke CS graduate students). The program enables users to download and manage digital media content. I would venture a guess that 90%+ of the files downloaded are pirated TV/movies and porn – my “DD” included downloading the latest Billions episode in 3 minutes (15 minutes without the program). The platform claims to have 152 million monthly unique visitors in Dec 2016. Only a small % of those users are paying subscribers (enjoying faster download speed and other features), and the subscriber base has been flattish/declining from 5.02m at end of 2015 to 4.97m at end of 2016, with a growing mobile subscriber base offsetting a declining PC desktop base. Notably ARPU at 30.1 RMB grew 19% YoY. XNET has not disclosed segment margins, but mgmt has indicated in meetings that this segment consistently generates 15%+ EBIT margins, which can also be partially corroborated by looking at reported financials prior to 2015 when the company started investing in the cloud business.

This is a “traditional and mature” business – revenue in last four years were $87, $98, $82 and $90m There is some risk of subscriber loss and copyright lawsuits (less today after they sold the online video streaming business two years ago). Applying a 4x EBIT multiple @ 15% op margin on $90m revenue base, I get to a $54m valuation or 80c/ADS.   

Cloud business: Mgmt has long realized the need to diversify away from the legacy business. Still, their initial foray into online video streaming business was a dud that ended in a fire sale of Xunlei Kankan for a mere $130m RMB. XNET launched Project Crystal, which is essentially a content delivery network (CDN, think Akamai) that operates a network of servers and rents/sells capacity on these servers to customers (online video streaming players) who want their websites to work faster by distributing content from locations closer to end users. The CDN idea is not new, and indeed the space already has quite a few players, but XNET’s angle is that they use/extend their core technology to crowd source idle uplink capacity from their user base, which would allow XNET to spend less on costly bandwidth/servers and undercut competitors on price.

The jury is still out on whether XNET would succeed in this new venture. I have spoken to a couple of industry contacts and received mixed reviews of the product. Disclosure is quite poor as the cloud business is buried under “Other internet value-add services”, which also includes an online gaming division that is declining. My best guess is the new cloud business did $25m revenue in 2016 against a $50m+ cost base (the majority in R&D and bandwidth costs) – XNET started booking revenue in 2015 and the cloud revenue grew 230% in 2016 and 80% in Q4.  Putting a 2x sales multiple on 2018 revenue of $60m (50% CAGR) gets me $120m. Akamai does 20%+ op margin and trades at 4x sales.

Ownership: Management owns 20% of shares. Xiaomi, a fallen star in the competitive smartphone industry and a strategic partner with XNET owns 30%. King Venture, the venture holding of Kingsoft (3888 HK) owns another 11%. There has long been speculation that Xiaomi (once valued at $50B) would take over XNET just for its technology. Given how far Xiaomi has fallen from grace, I am not holding my breath, but the cash/asset value and concentrated ownership does make XNET an interesting target for privatization and/or PE fund, and lessens the risk of indefinite cash burn, which is the biggest risk on the stock.

Catalyst

1) continued ramp in cloud business

2) privitization