Tuesday, September 21, 2010

The internet travel company that sells far more tickets than it has page views: another trip down the rabbit hole with Universal Travel Group

Numerous people have suggested that it doesn’t really matter that you can’t buy tickets on the Universal Travel Group’s website. After all, all you need to do is give a valid Chinese mobile phone number and they will ring you back to conclude the payment. In that view it hardly matters that the company advised the US capital markets that they had an online payment ability.* People even reference this by pointing to Alexa traffic numbers. So I will do so before revealing the fault in that analysis.

Here is the Alexa page for CNUTG. It reveals that CNUTG is the 17,563rd site in the Alexa database and the 2,492nd site in China. If I compare it to www.check-in.com.au (a real, albeit minor, travel company in Australia) it checks out quite nicely. Check-in is the 1,893rd site in Australia. In other words – the bulls argue – www.cnutg.com is a real travel company albeit with a clunky website and lack of a payment mechanism.

There are fallacies in this argument. The main one is about the relevance of the Alexa database. Alexa provides data on global internet use by people using Alexa toolbars (or similar). This makes CNUTG the 2,492nd site in China of people whose Alexa toolbars are reporting data back. Alas Alexa data is notoriously inaccurate in China as the Chinese internet police block data from Alexa toolbars.  Alexa toolbars thus bias Chinese sites to Chinese sites visited by non-residents of China. After all it is the Chinese authorities who monitor website usage in China – and they don’t much like other people doing it.

And that gives the game away… we are not tracking Chinese travel visitors. And it is not surprising that, of China sites, CNUTG is popular amongst non-Chinese. The stock is – after all – listed on the NYSE.

The Chinese Alexa

We can do much better – there is a Chinese version of Alexa which produces far more accurate data for China. That site is China Rank. It is in Chinese only – but it allows us to check out the web traffic per million internet users in China for the various travel companies.

Here is a comparison for Elong, CTrip, MangoCity, Qunar and the two main Universal Travel sites CBA-Hotel and CNUTG. Here is the link .

And here is a snapshot of the page in the original Chinese

image

And after the Google translator -

image

 

You are reading this right – the pages get 1 to 2 users per million internet users in China.

There are 420 million internet users in China – so these sites get about a thousand users in total (and presumably far less users daily).

The only problem is that the company has said in conference calls that it gets about 7 thousand to 10 thousand bookings per day. (They have implied but not confirmed that these are primarily through the website.  They might be able to do this by not using the internet - but that is not how it sold the stock and is not consistent with the traffic numbers that the company claims for its website.)

The company claims to have taken 2.4 million flight bookings last year and sold 2.3 million hotel room nights. (Those claims were made by the former CFO at the March 2010 China Rising conference.) Presuming this was on the internet (as the company has implied many times) this is an unusual travel business – in that it sells approximately 1000 users 4.5 million individual bookings per year. That – if true – would be a lot of revenue per unique user!

[For reference: the company has stated in SEC filings that in 2006 the CBA-hotel.com site booked 500 thousand rooms on the internet. They have also stated several times that internet bookings have grown substantially since then. They are thus categorical that a large number of these hotel rooms are booked on the internet.]

Alternative explanations

I am racking my brain about alternative explanations. Again I asked the management if they could explain (but without response). The obvious explanation is that the bookings never happened (and hence the revenue of Universal Travel Group is fake).

That would of course be consistent with the other problems. For instance very limited sales is consistent with the dysfunctional website (making it hard to book), the lack of reasonable cookies on the website (which is not supportive of return business), the lack of customer acquisition cost (typically a major cost for such a business) and the lack of interest earned (which is supportive of the notion that the cash balances the company claims are not existent and hence the earnings are not existent).

A top 100 travel companies in China

China Rank also – conveniently – gives us the top 100 travel companies in China. The source is linked and the list is below. None of the Universal Travel Groups sites rank in the top 100.

1 www.ctrip.com
2 www.17u.com
3 www.flights.ctrip.com
4 www.elong.com
5 www.huochepiao.com
6 www.travelsky.com
7 www.ipiao.com.cn
8 www.huoche.com.cn
9 www.piaojia.cn
10 www.tuniu.com
11 www.feeyo.com
12 www.china-sss.com
13 www.mangocity.com
14 www.9tour.cn
15 www.hrs.com
16 www.zhuna.cn
17 www.etpass.com
18 www.homeinns.com
19 www.piaowutong.com
20 www.airasia.com
21 www.china747.com
22 www.piao.com.cn
23 www.szair.cn
24 www.228.com.cn
25 www.goldenholiday.com
26 www.9588.com
27 www.book-hotel.cn
28 www.shangri-la.com
29 www.128uu.com
30 www.ctqcp.com
31 www.china-holiday.com
32 www.cct.cn
33 www.shal63.com
34 www.sh-holiday.com
35 www.go3.icpcn.com
36 www.yuding8.com
37 www.chinahotel.com
38 www.hzwy.com
39 www.oklx.com
40 www.cityhome365.com
41 www.easy-linkholiday.com
42 www.caissa.com.cn
43 www.053500411.com
44 www.51piao.com
45 www.hostelcn.com
46 www.piao.com
47 www.cntour365.com
48 www.cthy.com
49 www.114huoche.com
50 www.chinaonly.cn
51 www.junli.com
52 www.urthebest.com
53 www.quhappy.com
54 www.egpiao.com
55 www.chinapiao.com.cn
56 www.airprice.com
57 www.huilv.com
58 www.hkmohotel.com
59 www.cytsonline.com
60 www.jsj.com.cn
61 www.chinaticket.com
62 www.ehome365.cn
63 www.hxshuncheng.com
64 www.pekfly.com
65 www.cctsol.com
66 www.dalu.com
67 www.piao123.com
68 www.bjthjp.com
69 www.meini99.com
70 www.trip.com
71 www.hotel.com.hk
72 www.shasm.com
73 www.xz123.com
74 www.youcq.com
75 www.hotelhk.com
76 www.fridayol.com
77 www.zdpw.com
78 www.hottickets.cn
79 www.hotelinhongkong.net
80 www.xinyour.com
81 www.89838888.com
82 www.hohoho.com.cn
83 www.babycity.com.cn
84 www.fjp88.com.cn
85 www.kunmingguoji.com
86 www.actrip.com
87 www.010jp.cn
88 www.hn766.com
89 www.365sunny.com
90 www.95160.com
91 www.guilinyangshuo.com
92 www.yanchupiao.com
93 www.365666.cn
94 www.ginhoo.com
95 www.guilinhome.com
96 www.87667777.com
97 www.0532.com
98 www.51order.com
99 www.paio.cn
100 www.suzhoutravel.com

Stating the obvious: if you can’t rank in the top-100 sites you are not a leading online travel company as per the stock promotion.

 

 

 

John

 

 

 

Postscript: The company has irregularly given different traffic numbers in SEC filings. Here is an example (from the last annual filing).

In August 8, 2007, we acquired Shanghai Lanbao Travel Service Company Limited ("SLB") in exchange for 200,000 shares of our Common Stock and interest-free promissory notes in the aggregate principal amount of $2,828,000, payable no later than August 8, 2008. The note has been repaid in full.

SLB was established in 2002 and its core business focus is a centralized real-time booking system providing consumers and travel related businesses with hotel bookings, air ticket and tourism information via the internet and mobile phone text-messaging technology. It owns and manages the award winning China Booking Association website, http://www.cba-hotel.com/, which receives approximately 200,000 visitors daily.

These numbers are about 100 times the Alexa traffic estimate and maybe 200 to 400 times the China Rank estimate of total users for the site.

 

 

*I should note that the company has provided some online payment ability since I wrote my initial post. This ability is limited and clunky. Other major problems with the sites remain – for instance the lack of cookies to appropriately manage return customers.

 

 

First postscript quantification:  China Rank as far as I understand measures users.  Alexa measures viewers in a particular day.  In both cases however the usage is way too low to sell that many tickets.  The distinction here is second order – but if I have glossed over it then I apologize.  Whatever – I noted the visitors per day claimed by UTA for CBA-HOTEL (200 thousand) are 100 times Alexa estimates and 200-400 times China Rank estimates.  The nuances in definitions of visitor numbers are small compared to these multiples.

12 comments:

  1. Good post again John. I've been looking into the interest thing myself and made some comparisons. It seems that your assumption that anyone with a few millions in cash should be able to get a 2% return on it was an overstatement. The returns I checked in a few cash rich companies (SORL, CNTF, CAST) were far lower than that, but not quite as low as 0,16%). I have no position in UTA or any of these companies, I might add. Perhaps an update on the previous post would not be amiss? Anyway, UTA looks a stinker to me and it's good that fraud is exposed. Should it turn out to be what they claim to be, than at the very least UTA has made some serious lapses of judgement in their shareholders communications.

    ReplyDelete
  2. Your numbers are wrong, uniques aren't adjusted, they are absolute numbers. So 1-2M uniques are 1-2M unique users. Uniques are used to distinguish from page views.

    And I don't know much about the company, just thought I'd point it out. And even with the correct numbers, I would never, ever use them for anything important, they are fairly easy to manipulate and nontransparent, different sites use different sources (toolbars, isp logs, cookies) and different algorithms to extrapolate and come up with their guesstimates.

    ReplyDelete
  3. I am aware of the problem. Alexa measures visits - China Rank measures Uniques.

    I have a counter on this blog - and the Alexa data underestimates visits to this blog - but not by a whole lot.

    The uniques and the visits are pretty close.

    --

    Whatever - on both Alexa and CHINARANK the numbers are NOTHING LIKE ENOUGH to sell that many hotel rooms.

    John

    ReplyDelete
  4. I was wondering what happened to the stock today. And then it dawned on me: another whack at the pinata! I liked the first article. And while I think you are right it is slowly becoming a mud fight. I know you are playing with other peoples money and it is good
    that you are taking this fight seriously. But as a spectator sport this has become a bit tedious. Waiting for the Sherlock Holmes effect. Will the SEC come to the rescue or is this going to drag on for years? What are the most likely scenarios here?

    ReplyDelete
  5. John,

    While I typically enjoy your writing, I think you are completely off-base on this blog post.

    Do you know the difference between a unique visitor and a page view? Your post, while usually coherent, is not articulate or clear this time.

    You also failed to explain how Alexa ranks it's website.

    From my observation of the stats from Alexa and this "so-called" ChinaRank website, there doesn't seem to be anything fishy going on.

    You failed to point out that CNUTG.COM is ranked 17,270 on Alexa in the world and 2,442 in China.

    Those rankings are extremely impressive considering there are millions of websites in the world. Do you know the kind of sites that are ranked in the 1000's or in China? Let's just say that a company like elong.com is ranked 2,818 on Alexa.

    I recommend that you do a little more research on the internet, page views vs visitors, etc. before you write your next article.

    From Alexa:

    "About Cnutg.com (cnutg.com):

    Cnutg.com is ranked #17,270 in the world according to the three-month Alexa traffic rankings. Almost no visits to the site consist of only one pageview (i.e., are bounces). We estimate that 82% of visitors to this site come from China, where it has attained a traffic rank of 2,442. Visitors to Cnutg.com spend about six minutes per visit to the site and 52 seconds per pageview, and roughly 51% of visits to the site are referred by search engines."

    ReplyDelete
  6. That Top 100 list is damning evidence.

    ReplyDelete
  7. The irony is the Cnutg.com rankings will probably now go through the roof as Western investors hit their website testing out John's thesis.

    May be they can create a new business selling investment tips to westerners?

    ReplyDelete
  8. John,

    Please, if you are going to advise people on their investments about UTA and about websites in general, please, oh please, BUY a subscription to COMSCORE or HITWISE.

    You're basing your investment advice on ALEXA?! You've GOT to be JOKING right?!?

    John, go to HITWISE.COM OR COMSCORE.COM and find the traffic stats for UTA's website there.

    Your analysis will be far more convincing and legitimate.

    By quoting ALEXA and this "chinarank" site as the "authority" on web traffic, you are truly showing how much of a NOVICE you are when it comes to web site traffic knowledge.

    Please John, get a COMSCORE subscription. Even the Interns at no-name investment banks have access to this data -- and so should you. Particularly when you are trying to advise people on investment decisions on UTA.

    At least get your facts in order.

    p.s. DO NOT TRUST ALEXA STATS. ALEXA IS NOT AUTHORITATIVE -- IT IS HIGHLY INACCURATE AND SO YOU SHOULD NEVER BASE YOUR INVESTMENT DECISIONS ON ALEXA. GO WITH A PROVEN/LEGITIMATE SOURCE LIKE FORRESTER RESEARCH, HITWISE, COMSCORE, ETC.

    Jake

    ReplyDelete
  9. Reposted:

    It saddens me to see that people these days are using Alexa to aid them in making investment decisions. Or worst yet, John Hempton is advising people on investment decisions based on Alexa stats. Sad sad sad...

    Alexa's stats are inaccurate and highly variable and thus, should never be used to make investment decisions. If you want a trusted or at least, somewhat trusted source of web analytics, you should really buy a subscription to Comscore (www.comscore.com) or Hitwise (www.hitwise.com).

    Most stock analysts covering Internet stocks have subscriptions to these paid-subscription websites. Why? Because they are far more accurate and thus, a much better gauge of traffic stats.

    Alexa is NOT to be trusted and their sample size is often so small (because they measure stats based on their toolbar, which you must have installed for them to "count" you), it's highly variable and inaccurate.

    So please everyone, don't base your investment decisions on Alexa. Base your decisions on REAL stats that are verified, accurate, and trusted in the industry such as Comscore or Hitwise.

    Don't listen to that idiot John Hempton -- he doesn't even have a subscription to Hitwise or Comscore -- what makes him an expert in web traffic when he doesn't even have the most basic tools that even an intern at any investment bank would have?

    Just for the record, I am not saying Alexa is completely useless. It's just not useful for making investment decisions.

    People who use Alexa -- do so to "eyeball" trends. For example, you can see whether a website is getting more or less traffic. But in terms of overall raw # of users, visitors, etc. Nobody really uses Alexa to actually measure specifics.

    Which leads me to this point -- do you really want to invest your money based on someone who "eyeballs" a company and tells you whether you should buy or sell?

    Or do you want someone who actually did his research and homework and came up with a firm, unbiased conclusion of whether you should buy or sell a company?

    John Hempton didn't even bother buying a subscription to Hitwise or Comscore. He didn't even bother to spend a little bit of money to do the proper research on a company before passing judgment on it (obviously, because he shorted the company and will dig up every bit of dirt on it and tell you to sell -- because that helps him make more money)

    ReplyDelete
  10. This is an article published at a Chinese newspaper 21 Century Business Herald on 08/31/2010.

    http://nf.nfdaily.cn/epaper/21cn/content/20100831/ArticelJ01003FM.htm

    It talked about fraud controversy. Have your Chinese friend translate the article for you, you will find more information of how some Chinese companies listed in oversea market through reverse takeover, including Universal Travel Group.

    ReplyDelete
  11. Wow!! you think its impossible but every post of this man is worse...

    everyday he have a poor argument....

    not a clue of how alexa works...o how to evaluate traffic in a web page...

    the information in this post is awful, missleading and incorrect...

    what else you could expect of a man that doesnt even have a decent web page??? just a simple blog....

    ReplyDelete
  12. Well let me disclose my positions first: I own shares in UTA.

    I'm a native chinese speaker and I can tell you that UTA's chinese website: http://www.cnutg.com/ works absolutely fine.

    Sure enough it doesn't offer the complete online booking experience as we are normally used to here in the US or Europe, all it does is letting you search for some hotel and flight fares and then take you to an "reservation" page where you have to leave your contact details (phone numbers etc), and they will contact you back. This transaction method doesn't particularly strike me as odd and in fact telephone booking is very common in China -- most people in China probably have never paid for holidays online.

    What's more the company seems to be based mainly around shenzhen, a booming southern east coast city with relatively high standard of living in China, I suspect it primarily services holiday packages around that area but I could be wrong. If we do a search on Baidu.com (not google, you probably know who's the big guy on searching in China) for shenzhen flights, ("深圳机票" in chinese), then it comes up on the 1st page in Baidu (ranked 8th on the 1st page to be precise on 22/Sep/2010), if it can get such a high ranking then I doubt its website only receives 1 or 2 visits per day as suggested previously by the author, (this may well be the case for its English version website, but let's be honest, its customer base is not international in nature).

    While I'd certainly like to see UTA appoint one of the big 4s as its auditors to restore some investor confidence, but to give the impression that this company is phantom based on its english website is a bit of over-kill and unfair. But at the end of the day, a good trade is a good trade, if someone made money by riding this news and shorting the stock, then it is fair play as long as it is legal.

    Kevin

    ReplyDelete