YY Inc Is Now In Serious Trouble by Unemon

I decided to publish my 2014 Research Notes on YY Inc. Here you can find the reasons why I was bearish on the Name. Maybe my methodology was/is flat wrong … maybe not. Do not trade on “my opinion(s)”. Do your own Due Diligence. Hopefully this Document might offer a new perspective on how to look at YY and on how to investigate it. Consider this Educational Material … aimed at fostering Effective and Unbiased Discussion around YY.

Executive Summary

  • Data collected over a period of [9 months] suggests that YY Inc has overstated their users in both gaming and music/entertainment channels. Implications of fictitious users are numerous:
    • For YY as an internet company, user base and user growth are paramount in the success of the company. The market and sell-side analysts make revenue predictions based on user growth.Slower user growth can suggest several issues such as limited market size and opportunity, sloweradoption rate, or increasing competition from similar service providers
    • An inflated reported user base also implies that the true Average Revenue per User (ARPU) is much higher than the number reported in the Company’s SEC filings. Furthermore, for YY to hitits revenue projections, they would have to grow ARPU much faster, which is not that simple as the current ARPU is already very high.
  • As an example, in the F-1/A filed on November 16, 2012 (page 131, Line 10) YY reported “9.3 million peak concurrent users on YY Client in September 2012”. Our data show that during the Month of May2014, the maximum number of concurrent users1 at any given time never exceeded 1.6 million. Once we clean our dataset for fake users (BOTs generated users), the number did never clear the 1.0 million users threshold (see Thesis 1). YY Inc claims that its user base is growing. However, our lastest data in May suggests usage is far below market expectations.
  • Furthermore, during the sample period of May 2014, our proprietary dataset suggests total users accounted for approximately 32 billion voice minutes. However, in the F-1/A filed on November 16, 2012 (page 131,Line 10), YY claimed that users were “consuming” > 43.6 billion voice minutes per month2. Given the fact that reported monthly active users (MAU) between August 2012 and March 2014 increased by 35.46%3, it would be fair to assume a similar development for voice minutes usage. But even if voice usage stayed flat yoy, our data suggests YY may have overstated voice usage by a substantial factor (43.6b/32b=44.68%) in its F-1/A filing dated November 16, 2012
    • If we take into consideration the presence of BOTs4, it follows that the voice minutes consumed by real users during the month of May 2014 amounted to approximately 16.4 billion, namely62.38% lower than the number reported in the F-1/A filing dated November 16, 2012
  • Observed real concurrent number of users online casts doubts on the ability of YY Inc to plausibly generate the reported revenue over the past year, especially in the Entertainment Sector (See Thesis 2).
    • While the data reported by YY for the Gaming Segment of the business might potentially be in line with the real achieved number, as far as the numbers concerning the Entertainment and Music categories might point to fraudulent reporting by the Company (See Thesis 2)
  • We believe that YY Inc could have overstated its bandwidth costs to keep up with the pretense of increasing users, when in fact user count is significantly lower (See Thesis 3). If the actual costs are significantly lower, this would beg the question why are they doing this?
    • They have to increase the bandwidth costs to show, lock-step, increasing users. We have shown that user count is a lot lower, and thus we would expect a lower bandwidth cost.
    • Bandwidth cost might was overstated by 300% or close to $6M+
    • So where did this money go? We can let the reader speculate, but kick-backs and fraud are very common in China
  • User count and usage on YY.com is infested with BOTs. While other authors have brought to investors’ attention the existence of such a claim has been difficult to prove so far. By looking at YY architecture, in the source code used by channel indexing, we discover two different variables to denote the number of users online on a given channel, namely users and ipUsers. It is our belief that ipUser represents the real user count, whereas users is the number that is presented on YY.com and is used to potentially inflate the true usage
    • YY may argue that many of its users use its services from internet cafes and thus users share one IP address. As we are going to illustrate in APPENDIX III, this logic does make little sense when compared with the actual data observed on channels where the presence of BOTs is minimal
    • Furthermore, internet cafes have been declining at a rapid pace (here), while YY Inc is showing dramatic increases in ipUsers, which go against common sense and market realities
  • We have also been collecting data on the total number of Concurrent Users Online for two Chinese competitors of YY, namely 6.CN and KUWO. Both competitors seem to attract more users than YY at any given time (other than during Bejing night hours which are peak times). See APPENDIX IV.
    • Interesting is the fact that the % of BOTs on YY is at its highest levels at night. If we clean YY Data for bots, then 6.CN and KUWO do have more users online also during night hours.
    • YY Inc has claimed to be one of the biggest providers of online Entainment services. However, there are private companies that seem to have larger user bases
  • For an overview of the older educational portal (http://edu.yy.com/), please refer to APPENDIX V. Simply put, it was a JOKE in terms of usage. Our data analysis suggests that education platform has minimal adoption. The new educational website is now http://www.100.com/. We have not analyzed it yet.
  • Some additional Interesting Statistics about YY Channels (See APPENDIX VI)

YY Inc Revenue Generation Process: An Overview.

Per F-1 (Oct 15, 2013 p. 1) YY platform consist of YY Client, the YY.COM and Duowan.com web portals and Mobile YY. YY Client, YY core product, was launched in China in July 2008, has attracted 400.5 million registered user accounts as of September 30, 2012. To increase the accessibility and usage of YY Client, we introduced our mobile application, Mobile YY, in September 2010.

YY Inc

“While the basic use of our platform is currently free, we monetize our user base through internet value-added services, or IVAS, and online advertising. Currently, revenues from IVAS are primarily generated through sales of virtual items and game tokens that our users may purchase for use in online activities on our platform, including online games, which are all web games, and YY Music, our music channels on YY Client.” Per F-1 (Oct 15, 2013 p. 2)

“In the years ended December 31, 2011, 2012 and 2013, we had derived our revenues from IVAS and online advertising. Our IVAS revenues are primarily comprised of revenues from the paying users of online music and entertainment, online games and, to

1, 2  - View Full Page