QIHOO 360 Technology (NYSE:QIHU)
If a person of sound mind tried to tell you that he had a three-foot crowbar inside a one foot square box, you wouldn't need to (1) look inside that box, (2) get a subpoena from a judge, or (3) debate the man's honesty. You know that (1) this is mathematically impossible, (2) additional information would be nice, but it wouldn't be necessary, and you would also know instantly that (3) he's lying to you. That's what's so nice about deduction. It's neat; it's clean; it's done.
Qihoo 360 recently changed its reported ARPU quotes from Monthly to Quarterly figures. (ARPU stands for Average Revenue per Paying User.) What was once quoted as somewhere between the impossible 380 and 440 monthly ARPU is now reported at 400 … for the quarter. Qihoo also tells us that this is gross ARPU.
In Tuesday night's Conference Call:
400 RMB converts to $62.96 USD. (here)
Unfortunately, those numbers belong on the left side of the revenue equation and cannot possibly equal what Qihoo claims for the right side: total revenue. They said,
Is this revenue net of sharing with developers?
"We concluded that game developers are the principals based on the fact that the games are primarily hosted on the game developers' servers and such developers are responsible for the maintenance of the games and determination of the prices of the virtual items used in the games. Therefore, we report such revenue on a net basis." ~ QIHU SEC 2011 20-F
But of course, in the real world none of this matters. Even if filed with the SEC, words and definitions are constantly changing in this business ‒ with apparently little consequence. (EG, here and here.) And then we have the press at large, who resist the hardest facts in order to keep the broadest audience. So our most serious debates are left to message boards. As I write this I can already hear the "refutation" ‒ "They meant gross-this, but you said net-that" ‒ or vice versa. So I'd like to pin this squirming little bug down to the Styrofoam. Let's forget about the revenue split with developers and net ARPU. We will take the gross ARPU of 400 RMB as claimed in the conference call and multiply it by the maximum ‒ the ending user base for June. We have 148,000 users in June, up from 139,000 users last March, which means that the average user base for the quarter would be somewhere in between, but we're going to forget that. Let's count the whole 148,000 users for the entire quarter. This gives Qihoo maximum room for the math to work.
148,000 x $62.96 = $9.3 million USD … versus the total $19.1 million that Qihoo claims they made. (And as a reminder, this is before revenue share with developers.)
As you can see, the reported numbers are so bad that even if we gave Qihoo maximum elasticity to reinterpret "net" and "gross," the numbers are not even close. If someone wanted to say that the $19.1 million represents revenue before revenue sharing (forget the other implications for now), then the bar chart shown here would accurately show the disproportion. The point here is not to illustrate revenue as it is, but to demonstrate the impossibility of using Qihoo's numbers and arriving at Qihoo's totals ‒ while providing Qihoo with the best possible set of numbers.
Let's cover every possibility … or, impossibility rather. If we refer to the $19.1 million (left bar) as total revenue net of sharing with developers, then our calculated $9.3 million (right bar) does not correspond and is much too high, since for argument's sake it has been calculated with gross ARPU. In actuality, revenue sharing at a ratio of 70:30 with June's user base brings us to $6.5 million. The disproportion is even greater than that shown in the bar chart.
As you can see, even if the revenue were reported on a net basis and the ratio of revenue share with developers were said to be 99:1, Qihoo's numbers would still be impossible. As good as it gets is real bad.
To grow faster than mathematical possibility … that's really humbling for the rest of us who were born within the limits of the finite world. In fact, these guys are soooo fast that the only thing faster than Qihoo's web game growth is Qihoo's story-telling.
And Qihoo needs stories. Short sellers and others have made strong cases that …
- The advertising revenue doesn't add up. (here)
- The web game revenue doesn't add up.(here)
- The web traffic doesn't add up. (here)
- And so far the new phones appear to be a flop. (here)
This is what I meant by "Qihoo needs stories." It's "web search" makes front page news, and any questions of the authenticity of Qihoo's current revenue are rarely even mentioned. Apparently, people need to look forward and not to the very real problems in the here and now; if they did they just might sell.
As it is, Qihoo's price has been going wild, as if to say, "Tomorrow's happy story is more comfortable than today's impossible revenue." …. Well, with a big enough crowd you can shout down good math with bad, but a person needs to be sure that the situation is inconsequential; for example, things just might not work out for those who try to drive across a 100 foot canyon on a 20 foot bridge. Bad math won't save them. With the stock market of course it's not that simple. Trying to calculate how high the irrational will carry the price is not a rational exercise. Nonetheless, I'll take my chances with the math.