Seeking Alpha
Short only, independent research, litigation
Profile| Send Message| ()  

Note to Readers: The intent of this research is not to cause short term volatility in the trading of the stock. Citron believes the following research is worthy to be placed in the public domain so all investors can make investment decisions based on reading and considering carefully all available information.

Citron has developed a tremendous volume of forensic research on Qihoo 360 (Nasdaq:QIHU), and has concluded that the company has either not been forthright about their revenue model to Wall St., or there is a financial fraud unfolding. Citron has received numerous comments on its previous writings on Qihoo, mostly to the effect that "Citron doesn't understand, " "Citron doesn't understand China," "Citron doesn't understand the internet," "Citron doesn't understand the future," etc.

One thing Citron does understand is math. Math is universal: 1 + 1 = 2, in the US, Antarctica, Mars, and yes, China. And Qihu's numbers don't even come close to adding up … in any universe.

Can Qihoo's Topline Revenue Withstand any Analytical Scrutiny?

QIHU reported 47.5 million USD in topline revenues and guides next quarter to 55 million USD. What do we know about these revenues?

The company states that 75% of its advertising revenue is generated on its homepage hao.360.cn. It further states that appx $4 million was generated from referring search to Google (GOOG).

Qihoo Revenue Analysis for its Most Recent Quarter

Qihoo Topline Revenue

All figures USD in millions for Most Recent Quarter

47.5

Comments

Search Referral

7.3

Net search revenue was 4.6m, (plus 2.7m back payment from Baidu after settlement of 2010 dispute and claims)

IVAS (mostly games)

12.1

Includes Games and other internet services

Third party antivirus

0.2

Business being phased out

Advertising Revenue

27.9

Home Page

75%

"Last quarter, we were at like, I think, 75%." – conference call by Zuoli Xu, CFO.

Advertising from home page

20.9

Revenue generated by home page (if these numbers are to be relied upon).

$20.9 million and the missing number

What we know: Homepage advertising link revenue is the largest component of Qihoo's revenue. The Company sells links on this page on a fixed fee per month basis. The company states that the fees vary based on the customer's business type. Yet it was extremely challenging to find the single most important metric : The average advertising revenue per paid link.

We read everything we could find on Qihu, to no avail. But finally, in a single analyst report by Macquarie (the only one that seems to reflect any detailed knowledge of the company's actual business model) we read:

The average advertising price for a link on Hao.360.cn’s front page is RMB 200-250 K per month.

This was published on August 29th, and notes it was after "the company raised its ad prices by 20% – 25% sequentially in Q1 2011". Of course an average is an average — some links will be be priced higher, and others lower. To make the math easy we will run the estimates on $225,000 RMB per month, a little more than $100,000 USD per paid advertiser per quarter.

It is easy to see what sites are paid because they have tracking URL's attached to them; thereby the advertiser can judge the effectiveness of the campaign. Links such as Hotmail, Vogue, CNTV sports, are displayed, as in all the Chinese directory home pages, for free, for user convenience. Citron has made a heat map so investors can see how many links on the HAO.360.CN page are tracked, untracked, or lead to another 360 company page. Here is the current hao.360.cn homepage, which generates for 75% of Qihoo's ad revenue.

GREEN: Tracked Link to a 3rd party website.

RED: Link to another 360.cn or qihoo.com page – a Qihoo company page

YELLOW: Untracked link, presumed free.

With this key, let's look at QIHU's homepage again [click to enlarge images]:

There are about 92 green links on this page; however many are duplicates and dispatch clicks to the same domain. Eliminating links to Google (paid separately) and Baidu (BIDU), there remain about 60 – 65 links with tracking IDs to the websites of 3rd party entities. (There are about 9 links to Taobao.com, which Qihoo claims is now about 10% of Qihoo's revenue; presumably these are at the high end of the scale of revenue per link.)

This is what we assume are the paid links. UBS verifies this by publishing a list of Qihoo's paid links in its Nov 17th note, totaling about 65. UBS includes several untracked links in their list, but for now let's not quibble about details. UBS's list very closely matches the spreadsheet we compiled with all the links from the hao.360.cn homepage here.

An interview with competitor 2345.com confirmed that around 60 links on their homepage are sold and they believe the same to be with their competition.

This is a very similar pattern to Baidu's directory page hao123.com:

So now we try to construct a picture of Qihu's homepage revenue:

Min

Max

Paid Links

60

70

Average Revenue per paid link in RMB

200,000 RMB/mo

240,000 RMB/mo

Average Revenue per paid link in USD

30,769 USD/Mo

36,923 USD/Mo

Estimated monthly advertising revenue USD

1,846,154 USD/Mo

2,584,615 USD/Mo

Estimated quarterly advertising revenue USD

5,538,462 USD/Qtr

7,753,846 USD/Qtr

Exactly where is the $20 million in advertising revenue from the home page, or anything even close to it?

We expect the company will now claim that many of the non-tracked links are paid. This would be very challenging to believe, as anyone paying real money for a link would want to track the traffic generated. And we would need an additional 130 paid advertisers on the home page. Who are they? (We wanted to ask this pivotal question on the conference call but the company would only take questions from their “cheerleading” analysts.)

Or calculating in the reverse direction:

Homepage Advertising

Min

Max

Revenue in millions USD for most recent quarter

20.9

20.9

Paid Links

60

70

Average Revenue per link per qtr in USD

348,333

298,571

Average Revenue per link per month in USD

116,111

99,523

Average Revenue per link per month in RMB

736,144

630,981

Neither the company nor any of its analysts, nor any of the numerous mainline ad agencies in China nor any of Qihoo's competitors, have ever quoted an advertising rate for a web traffic link in China within an order of magnitude of this astronomical number.

And now the 7 million in ad revenue that doesn't come from the homepage

Now we try to construct a picture of Qihu's non-homepage and non-game revenue (the other 25% of advertising revenue): Citron was able to procure a contract for a non-home page link. We were quoted a rate of $20,000 RMB per month, which seemed reasonably within expectations, as rates fall off rapidly because traffic is much lower by any measure on pages only linked from the home page.

Non-homepage advertising revenue per qtr in USD

7,000,000

Non-homepage advertising revenue per month in USD

/ 3

2,333,333

Non-homepage advertising revenue per month in RMB

x 6.34 RMB/USD

14,793,331

Cost per link per month, avg in RMB (from company source)

Est per contract

20,000

Number of customers required to generate this much revenue

740

Who are these customers? Citron has searched all pages on 360.cn and QIHOO.com and cannot come anywhere close to a number approaching 700. Some of the 360.cn linked pages have only one customer…or none.

But the "Games" Do Not Stop There!

Qihoo runs a game directory page, which is extremely similar in appearance to those run by many major competitors, for example Sohu (SOHU) and RenRen (RENN):

What isn't similar is the astonishing revenue per customer Qihoo claims in their last quarterly report.

Within IVAS, web games revenue were $9.7 million, up 187% from same period last year and 47% from the prior quarter. The strong performance was driven by a large user base on our web game platform.

At the end of the second quarter, we have about 62,000 paid game players. At the end of the third quarter, we have about 87,000 paid game users.

Qihoo Games Revenue 2011

Games Gross Revenue

Qtr over qtr increase

Paid Game Players

Qtr over qtr increase

Revenue per paying player per qtr USD

Revenue per paying player per month USD

Gross Revenue paid per paying customer per month in USD– (Qihoo and /developers split 70:30)

Qtr 3

9,700,000

47%

87,000

40%

111.49

37.16

53.09 USD

Qtr 2

6,600,000

62,000

84.48

28.16

40.23 USD

These revenue numbers are simply absurd. Nobody in the games business, in China or the US or anywhere, generates $53.00 per month per paying customer – not NetEase (NTES), not Changyou (CYOU) and not Shanda Interactive (SNDA). Other Chinese games vendors report ARPUs about $3 to $8 per month USD. Zynga generates about $1.00 to $4.00 per month per paying customer, depending on which number you compute. Even the "gold standard" Blizzard International, whose World of Warcraft is in a league of its own for online gaming, charges US and European gamers, who typically pay far more than Chinese gamers, $15/month for subscription.

And nobody in China is growing games revenues at 47% quarter over quarter. For comparison, RenRen, one of the hottest properties on the internet, grew games revenue at 26.3% year over year, and Shanda, one of China's largest game companies, increased revenue just 3.7% quarter over quarter.

Like the advertising revenue numbers, these numbers are simply not believable either. Yet the analysts simply nod and offer the perfunctory "great quarter, guys!"

Citron has no idea how the company will even attempt to explain this glaringly unbelievable revenue per player number. Too bad not one analyst on the call even asked about it!

Why does Qihoo make it impossible to verify their web traffic claims?

Citron has presented a compelling case as to why the current business of QIHU can be questioned. Yet, the multiples afforded to the stock by analysts are obviously not based on their current business but rather their future business, which is all predicated on the web traffic generated from their homepage — both in its generic and "personalized" forms.

Even though the analysts don't agree on how Qihoo will generate money in the future, the one thing they all agree on is that its current revenue depends solely on internet traffic generated through its website. Therefore the company should be taking extra steps to clarify and make transparent 3rd party verification of the reach, penetration, and page views they claim. But instead, the company does the opposite.

Citron first challenged Qihoo's hao.360.cn homepage traffic claims because the industry standard measurements of internet traffic – including Alexa, Google, CNZZ, and Comscore — all report ranges of lower penetration, reach and traffic rankings than the company's paid-for statistics from iResearch. The company's defense was that "Citron doesn't understand" the internet – because the company's anti-virus software prevents all analytic tools from being installed on a user's computer.

All tools except of course those from iResearch and their 200,000 sample size.

Wait a minute! What kind of excuse is this? How is it that Qihoo's software coexists with iResearch's monitoring tool, but can't coexist with Alexa? Does anyone seriously consider Alexa's voluntarily installed toolbar to be malware? Like all anti-virus software, Qihoo's anti-virus updates the desktop without user intervention, so the company could easily change its software to make it Alexa compatible any time it wanted. The fact that the company actively prevents any independent verification of web traffic metrics remains a huge red flag for this company. This is not Citron's issue, it is Qihoo's issue. In the real world, this limitation would impose a huge obstacle for a company trying to sell web advertising.

This should make any interested investor question the depth of the relationship between iResearch and Qihoo.

What business is Qihoo in?….a terminal one

Each of the analysts covering Qihoo seems to present a different idea of what business Qihoo is in and where the company is headed. One says its "mobile" (which won't be monetized before 2014), one says it's "search" as though they operate in a fantasy world where they could turn on a switch and compete with Baidu. Another says it's going to operate an app-store with an interface copied from iPhone. The only thing these three strategies have in common is that all of them currently generate zero revenue for Qihoo.

Citron will not address search, mobile, or desktop platform because the company has generated no revenues nor presented any notable proprietary technology. The analysts involved in the name do not even have a cohesive story as to where this company will be in 3 years — every analyst describes the company's future differently.

The truth is Qihoo's revenues come nearly 100% from what can only be called "Web Directory Services". They devised a clever low-cost strategy to give away anti-virus software as an indirect method to entice internet users to their home page. But they generate no revenue from the anti-virus software, and no revenue from their browser.

But this model — "get in front of the eyeballs" — is hardly new. We can observe the same in the early days of the internet in the US, where the gold rush was on for eyeballs and clicks, and of course Wall Street's dollars.

AltaVista, Lycos, Excite, anyone? Does anyone even remember these web directory businesses? Billions of investors' dollars went into them, all of which was vaporized. Why? Because they provided neither durable content nor services, they quickly ebbed into the background. They were mostly a short-run and shallow land rush to "get in front" of the user's internet experience. Look at what happened to the US Web Directories — they are a virtual internet graveyard. Even Qihoo's biggest cheerleader Mirae says:

Our assumption is that PSP will run out of momentum by the end of 2012.

So where does that leave Qihoo?

Analyst Malpractice

Citron has commented on the incredibly shoddy analyst work on this stock, but in our opinion, the work of Mirae goes even farther, to professional malpractice. For a firm that writes glossies on QIHU nearly every week, we find the most disturbing pattern of promotional claims without foundation that we have ever seen from an investment banking firm.

Mirae will probably want to comment on this piece as well, so we thought we'd offer them some topics to address. Mirae has raised its 2013 topline revenue estimates for Qihu twice just in November.

This topline number which includes no breakdown or financial basis looks like this:

Mirae report date

Oct 13 2011

Nov 15 2011

Nov 18 2011

% over QIHU's 2011 rev est

2011 Revenue Est

148.1

153.0

167.0

2012 Revenue Est

307.4

324.9

330.8

105.5%

2013 Revenue Est

472.7

563.3

573.3

355.4%

Target

37.50

40.00

40.00

Current Mkt cap at date of report

2.1 Billion

2.4 Billion

2.2 Billion

Mkt cap at this target

4.4 Billion

4.7 Billion

4.7 Billion

The Mirae analyst's work reveals the following gross deficiencies:

  • Provides no breakdown or foundation of the prediction of massive revenue gains for Qihoo in 2012 or 2013.
  • Completely ignores the red flag inconsistencies in Qihoo's current reported revenue, the lack of transparency in how the revenue is generated, the gross discrepancies in Qihoo's claims of reach and penetration vs industry standard metrics from Comscore, Alexa, Doubleclick and CNZZ, and ignores the crazy ARPU in Qihoo's games revenue
  • Completely ignores that China e-commerce was grossly disappointing in 2011; investments have collapsed due to gross revenue shortfalls, massive losses, lack of fulfillment infrastructure, low demand of Chinese consumers wary of poorly made counterfeit products that they feel the need to inspect before buying, and other obstacles. See the charts of DANG and MCOX – both down 80% year to date.
  • Displays a clear and fundamental misunderstanding of Qihoo's business. Unfathomably, in the November conference call, despite management having made perfectly clear that their revenue is derived from selling links on its homepage on a negotiated fixed-fee-per-month basis, Ms. Nancy Yang, Mirae's internet / media analyst, asks:

Firstly, can you help us get a sense how much of your fourth quarter revenue guidance will come from growth in users and growth in per click per user? It seems that growth in per click per user is much faster than growth in monthly active users, should we expect this trend to continue? (She got her answer in Chinese)

Instead of analysis, Mirae publishes screen shots of a mockup of a yet-unreleased Google product and claim that because Qihoo has hired an unidentified former senior Google employee, that Qihoo will be able to monetize this unreleased technology in 2012. Note that the product sketch simply clutters up the user's screen with more ads – exactly the opposite strategy of Qihoo's homepage. The mockup's main feature is a product price comparison, which would itself be an anathema to Qihoo's main e-commerce advertising customers.

Don't believe us? Have a look here: Mirae Nov 15th.

The US Securities and Exchange Commission has a long history of regulatory enforcement actions against analysts and other 3rd party enablers of stock promotion who issue "baseless or unfounded revenue projections". Literally dozens of regulatory links against firms and individuals for this practice have been filed over the years. Here are just a few examples:

http://www.sec.gov/litigation/litreleases/2008/lr20644.htm

http://www.sec.gov/litigation/complaints/comp17674.htm

http://www.sec.gov/litigation/complaints/2007/comp20375.pdf

http://www.sec.gov/litigation/litreleases/lr16456.htm

http://www.sec.gov/litigation/litreleases/lr17050.htm

http://www.sec.gov/litigation/litreleases/2010/lr21565.htm

http://www.sec.gov/litigation/complaints/complr17300.htm

http://www.sec.gov/litigation/litreleases/2009/lr20940.htm

http://www.sec.gov/litigation/litreleases/lr16924.htm

http://www.sec.gov/litigation/complaints/comp18652.htm

Citron believes Mirae Research is publishing thinly disguised reports on Qihoo that are in fact promotional materials masquerading behind the appearance of "analyst coverage", and is worthy of immediate SEC enforcement scrutiny for stock promotion activities.

Conclusion

Based on the data presented above, and the interviews we conducted to gather it, it is Citron's belief that Qihoo's current web directory business is generating far less in revenue than the company is reporting.

Its presence in China's web market is, by the company's own choice, opaque to third party metrics. Its advertising links are sold only by a small sales staff, not any of China's many agencies that would represent major advertisers. There is no pricing transparency. There is no media kit, typical of major online advertising providers. We are told that their rates are soaring, but e-commerce revenue and investment in China is currently being crushed due to brutal price competition and logistics pressures. We are told there is an "auction process" driving ad rates higher, but there is no auction functionality. In short, nothing about Qihoo's main line of revenue is independently verifiable. That was a common problem among several dozen US-traded China stocks halted and delisted in the last year.

But even if Qihoo's revenues were reported accurate to the penny, the company is still overvalued by a factor of 4 or 5 x.

Analyst claims that Qihoo could be in search, or could be in mobile, or could be a leading app store, are patently ridiculous. Qihoo has no more claims on those highly competitive spaces than Citron. It is pure fiction. Qihoo is no more likely to be able to compete with Baidu in search than Microsoft's (MSFT) money-losing attempt to compete with Google.

What Qihoo does have is a web directory, which competes with other web directories for advertising dollars. The company's topline would have to double and double again just to grow into its current valuation.

You could say that the last standing monetizable "directory" in the US is Yahoo. Of course Yahoo offers a lot more than a menu to web services, it has tons more content than Qihoo. But even so, when you strip out its other equity positions, the multiple afforded by the market for its advertising revenue is about 1:1. For Qihoo, we're talking 14:1 or higher (even if you believe their highly questionable numbers).

If Qihoo makes sufficient metrics available to allow independent verification of its claimed web penetration and reach, and if Qihoo were to provide transparency to its advertising revenue sources so that a credible revenue model could be built, Citron would generously afford a 2:1 multiple on its current revenue run rate to the company. This plus the company's cash yields a price target of about $5.00 per ADR.

Cautious Investing to All.

Note to all readers : This report is not to be misunderstood as Citron holding any antipathy toward China or China investments. Citron believes the China investing space (especially China-domiciled companies trading in US markets) is subject to extreme risks due to the absence of independent 3rd party verification and regulatory remedies for corporate misbehavior at all levels. This condition is similar to the US market before the Securities Acts of 1933 and 1934, which set the standard for investor protection that underpins the US equity market. There will be successes and failures in this space, and as the market matures, the worst ones will tend to be washed out.

Disclosure: As of the date of publication, the editor of this report is short QIHU for the reasons stated above, and long SINA and SOHU.

Source: Qihoo: Fraudulent Financials, Terminal Business, Or Both