However, I still decided to pass on it:
- For the stock to “work”, the core golden goose search needs to growth and accelerate. Given the cancer incident and governmental actions, it is quite likely like they miss the next quarter + continue to face skepticism on out-year growth rate. I also think that they could be over-monetizing this core business.
- While the stock implies a bid, almost oversized cash-burn for the O2O business (5-10 year peak-run-rate burn of $4 Bn / year), I find it difficult to stand behind the businesses they are plowing money into – they seem more like no-moat ventures and more of volume acquisition vehicles vs. good standalone high ROIC projects.
- The difficulty to fully understand a Chinese company and the unknown-unknown risk.
Would still like to share my thoughts. Comments welcome.
How does search look like and how should we think about it 2-3 years out?
- Congruent to generic wisdom, Baidu’s search division is dominant in china with 70-80% market share in both volume and ad dollars.
- Would caveat that for both Search and O2O, Baidu doesn’t disclose enough information for granular QoQ tracking, but sell-side models (particularly DB), press releases, and transcripts give us something to work with.
- As it stands for FY15, 52-53% of BIDU’s revenue (both search and overall) comes from Mobile. Paid clicks stand at ~63% of total search clicks, and it appears that CPC for mobile is still ~15-20% below overall average. (See below for more detailed break-down)
- Net-net assuming the 20%+ top-line CAGR continues at 50%+ EBIT margin, BIDU’s search segment can do $7.0 Bn USD of EBIT in 2017, and $9 Bn in 2018. Assuming 20% tax rate and 350 mm shares outstanding, this translates to $16-21 / share in search earnings.
- A 15-20x “fair” PE multiple gets us $240-320 / share for FY17 and $300-400 / share for FY18.
- Given how formidable the business looks, we should probably focus on what could go wrong.
- One of the trends worth highlighting is the increase of TAC as a % of search. The number went from sub 10% of search Revenue before 2013 to currently 16% of search revenue in FY15. I have not done enough work to figure out where this goes (currently assumes this cost moderates to <15 70="" approaches="" as="" mobile="" of="" revenue="" search="" span="">I recommend we dig deeper into the reasoning behind this ramp.
- I also wonder if Baidu is over-earning. Should do a benchmarking exercise vs. GOOG and YNDX. Maybe we can also check what kind of mix contributes the most to revenue vs. GOOG.
- A downside analysis on how bad the Cancer incident can be is also helpful.
- Hound seems very confident on 20%+ topline growth for many years to come with its top-down build, citing ad spend being ½ of US in terms of % of GDP, online share gain, and China search spend still being 10-20 pts below international peers. The logic seems reasonable but I think we should replicate it.
What is Baidu’s positioning in its various non-search segments?
- I’m omitting iQiyi and CTRP for now due to the relative ease for valuation.
- Given the stability and current dominance of Search, I’m also putting it on a lower priority.
- The big China on-demand report has excellent industry background. Would strongly recommend
Nuomi
- Top 3 players in China controls 95% of GroupBuy market, with Nuomi being the distant #3 at ~15-20% share (according to Robin Li at the latest conference call). Currently, catering is the biggest segment using groupbuy platforms, accounting for 63% of total transactions in 2015, followed by entertainment (17%) lifestyle (11%) and hotels (9%).
- Nuomi lost market share when it was under Renren due to lack of funding and weak execution. Baidu acquired 100% of the company in February 2014 and its market share has gradually increased.
- In October 2015, Meituan and Dianping announced they would merge to create China’s largest O2O and groupbuy services platform. However, the new company will maintain a co-CEO structure and the Meituan and Dianping brands will operate separately, including the overlapping high-frequency groupbuy coupons and restaurant instant-discount businesses. The chatter is that subsidy pressure had eased somewhat since, and there are talks about the Joint company doing an IPO by 2017 – leading to potential pressure to become more profitable.
- My current fear is that GroupBuy’s is a terrible low margin business with no moat.
- The GroupBuy sites seems like nothing but an alternate marketing flyer / billboard for local businesses: Customers’ sole purpose is saving money and have 0 switching cost (which prevents platform from charging merchants more), businesses don’t really do these promotions unless they need the traffic and are willing to try anything (which lowers stickiness and heightens churn due to adverse selection), the technological barrier is low, the data could prove valuable for alternate profit-generating purposes (which promotes incentive to compete at low to negative margins as a platform), and the “success” is somewhat self-defeating (good effect = less promotion down the road, more merchants on platform = worse advertisement benefit). None of these are foundations of a solid, high ROIC business down the road, and I suspect sell-side are too rosy on out-year margin assumptions.
- Groupon & Yelp make close to zero (if not negative) GAAP EBIT margin. Bulls argue China has a much denser city footprint – and thus GroupBuy should enjoy higher margins than its US/EU peers. I think the argument is flawed – because it also makes competitive entry that much easier / lucrative.
- In particular, being #3 as an internet company just doesn’t sound sweet. I wonder if there is anything Baidu can do to change it. This could be a good area of research.
- Net-net my current view is this segment’s SOTP value should reasonably be (a) 3-5 year cash-burn + (b) a discounted exit-value based on EV / GMV at a more matured stage.
Baidu Delivery
- Baidu’s takeout delivery service was launched in October 2014 and covers 90 cities. Baidu is currently No.3 nationwide by volume after Meituan and Ele.me, but it has become the leader among white-collar workers in half of its covered cities including Beijing, Hangzhou and five provincial cities.
- Baidu outsources its delivery service but seems to have an edge currently on order / route planning. Based on surveys, the service quality of the top-3 players are fairly similar.
- Notably, Grubhub of US and Just Eat of EU enjoy a 10-15% take-rate with 20-25%+ EBIT margin vs. Chinese companies currently at 3-5% take-rate and negative EBIT margin amidst land-grab. Notably, the annualized order count for the top 2 players in China already exceed GRUB and Just Eat by a factor of 9x, while the GMV / order being only 1/7 or 1/8. As penetration grows and inflation trends catch up to western standards, the TAM for this segment is quite exciting.
- Vs. Groupbuy, delivery’s suppliers have bigger margin to share (given incremental margin and less throughput constraint), customers can be more convenience-focused with more frequent / habitual usage (predictable, could be less inclined to switch), and the time sensitive + route density component add certain degree of economies of scale (lowers cost / delivery, Chinese city density actually does apply here) + pricing power (can charge more if consistently on-time). This is, in my opinion, a better business than GroupBuy and actually allows an of-scale player.
- Thus, BIDU might actually have a viable, solid ROIC business here. We could possibly ignore the cash-burn and value the company on a EV / Sales, EV/ GMV, or discounted EV / EBITDA basis.
Can O2O even remotely earn a margin 3-5 years out?
- The quick answer is: it seems so, if Baidu sticks to its $20 Bn initial investment plan.
- Assuming certain margin structure for Nuomi and Takeout Delivery, I estimate that BIDU spent ~4 Bn RMB (~0.6 Bn USD) in FY15 in rebates / subsidies to grow share – this is roughly 55% of its FY15 Transaction Services Revenue base and, given their announcement of the 20 Bn investment plan on June 30th, 2016, gets them about 15% there.
- If BIDU sticks to its plans and front-loads the subsidy as it implied, The $ / GMV (order) level quickly trails off after 2016: the $ / GMV intensity in FY16 will be ~80-90% of the FY15 level, and FY17 will be at 25-30% of FY15 level, with FY18 being less than 10% of FY18 level.
- Put another way, I’m projecting ~10 Bn of RMB spent in FY16, 5 Bn in FY17, and 2 Bn in FY18
- If this rapid tail-off happens (which we should intensely research) and assuming some loosely constructed cost structure, the Nuomi + Delivery segment should be able to achieve break-even exiting 2019, and reaching 5-10% operating margin by 2020 (which seems high). I personally think the rest of Transaction Service segment could still be heavily loss-making, and net-net the segment’s loss may narrow towards ~ 2-5 Bn by 2019/2020.
- A framework I think would be helpful is: Assuming a fair multiple on BIDU search, investment assets, and transaction multiple on O2O asset, what does the current market price imply on how much BIDU needs to spend on a total, per GMV $, and/or number-of-year basis?
- A key question we should also wrap our arms around is: “What should the long-term margin profile for O2O be? Is it any more than a traffic funnel / distributor / Data gatherer with low ROIC?”. I think a benchmarking exercise w/ its US/EU peers will be helpful.
- Other key questions we should ask include:
- What would it take for Baidu’s #3 O2O services to become a 25%+ share player?
- Why did Baidu dedicate its resources towards O2O – what is the real business logic?
- VC funding situation in China – are they looking to monetize? / Meituan DIanping IPO timeline.
- What does it cost / take to sign a new merchant up? How does it work?
- Leveraging the agents network for O2O, how real is that? Is it a easy cross-sell
- How should I think about the non-Nuomi + non-delivery profitability and growth trajectory?
Preliminary SOTP – implies 2018’s transaction service cost base running for 10 more years.
Other Questions Outstanding:
- Does Baidu pay taxes upon the iQiYi MBO? What is the timeline?
- What is BIDU's R&D trends? 15% of sales is a lot of bodies
Reasons for Inefficiency / Variant Perception
- Confluence of strategic initiatives obscure valuation: While BIDU’s search division in China is a behemoth with 60%+ market share and envious operating margin of 50%+, its strategic initiatives Qunar (Travel similar to TripAdvisor), iQiYi (Online Video similar to Netflix), and Online-to-Offline Initiatives (Nuomi, similar to GroupOn, and Uber China) are heavily loss-making – which in aggregate depresses headline earnings, implies an optically expensive valuation (currently 27x FY17 EPS), and can potentially deter investors. In particular, the flurry of recent deals ($1.2 Bn investment in Uber China, merger of Qunar and Ctrip, recently announced MBO of iQiYi) further obfuscates the clarity of out-year numbers.
- Our current hypothesis is that core BIDU search could be implied at a very cheap valuation assuming the rest of BIDU are not negative NPV ventures.
- The consensus bias, especially amongst many Chinese locals, against BIDU regarding its vision and execution: The common criticism include its omission of cloud infrastructure despite technical expertise (whereby Alibaba went after it in strides), missing mobile operating systems, and allocating sizable amount of capital into hypercompetitive, cash-burning O2O ventures. The company is also panned for its lack of execution of various products / projects while Alibaba and Tencent are blazing ahead.
- Our current hypothesis is that perception comes in cycles and it could still pre-matured to comment on O2O. There is more work to be done judging Robin’s vision.
Key Risks / Considerations
- China Macro / Currency Risk
- Google Entry Risk
- O2O Cash-burn and other “moon-shot” initiatives could continue cash-burn – a part of doing business.
Appendix
05/02/2016 Update:
A media attacked was launched on BIDU over the weekend. The background is that a cancer patient, in desperation, underwent unproven treatments after following the "recommended" link on Baidu's search results. The patient recently passed away and the article went through great length (1) pointing out the illegitimacy of the treatment through detective work and (2) questioning and criticizing BIDU's ethics re: its bidding system.
Notably, even People's Daily ran an article on it, and specifically quoted Xi's recent statements regarding internet / search: " ...and those who do search cannot rank solely based on how much money had been paid". Regardless, it seems like a well-orchestrated attack and, while the consequence of such false advertisements is saddening, it is the hospital / regulatory body that should be blamed (and not Baidu). In any case, we could see another wave of debate /attack on potential Baidu's pricing / bidding practice that may / may not affect the stock.
Some other notable points include:
- Medical / pharma are ~15-30% of BIDU advertisement sales depending on who you ask.
- While the Chinese government has rather standard rules regulating medical advertisements, the legal boundary on whether BIDU page-rank should be defined as advertisement is still hazy.
- Google had been involved in scandals like this and had settled with the US government, last time in 2011 they paid ~$500 mm
Source Link to People's Daily article: http://weibo.com/ttarticle/p/show?id=2309403970397409263704
Original article: http://mp.weixin.qq.com/s?__biz=MzI5NDEzMjc0OQ==&mid=3045317271&idx=1&sn=9cca5718716cb641a9a081cba60c0cc3&scene=23&srcid=0501aPeNW1CjDrsYYn0GU3Ct#rd
No comments:
Post a Comment