*Please see our disclaimers at the bottom of the page and our full disclaimer here. By using our website, you acknowledge you have read our disclaimers and agree to our terms of use.

Thank you so much to all of our subscribers. This is our first piece that is behind the paywall and given how gargantuan of a company Alibaba is, it is rather long! We know you will find high value in reading it though and additionally we will release a “PM Summary” this weekend that will be included for free in your subscription.

Given the prominence of Jack Ma and Alibaba, we have spent more time than usual on the businesses history, with sections on “Founding History” and “Business Background”. Feel free to skip right to “The Business” if you’re familiar with their story!

Founding History.

Jack Ma, Alibaba’s founder, was born in Hangzhou, China in 1964, when private enterprises were forbidden from operating. However, by time he was a young teen, there were signs of China moving towards a new direction. Deng Xiaoping launched China’s Open Door Policy in 1978, which coupled with President Nixon’s visit to Ma’s hometown a few years earlier, led to tourism flourishing. Jack took this opportunity to practice English by offering tours of the city. After failing the math portion of the GaoKao (China’s college entrance exam) twice, he became an English teacher and started his first company: The Hangzhou Haibo Translation Agency. Because he was fluent in English, the municipal government sent him to the US to investigate the status of a high construction project (it turned out to be fraudulent) and he was reportedly kidnapped (Jack is sparse on details). It was in the US that Jack first learned of the Internet. He went back to China with one of the first computers in the region.

Noticing China’s non-existent footprint on the internet, he started a Yellow Pages for China, dubbed, to help people find Chinese businesses globally. The internet was too nascent in 1995 though, and Ma had trouble explaining what it was and why they it was important: people treated him as if he was a con man. After a rough period of trying to win clients, they entered into an agreement with an SOE who took a majority ownership in exchange for a much-needed cash infusion. While they were supposed to work on the project together, the SOE was quick to ignore all input from Jack and he essentially became a glorified public servant in a stultified government organization. After a few years, he started plotting his next venture with 17 colleagues who would later become his cofounders in a business that would focus on the “shrimp” of B2B (a reference to American B2B sites being whales but 85% of the fish in the sea being shrimp-sized). In the cramped Lakeside Garden apartment in Hangzhou, Alibaba was born.

Open Sesame. 

Lakeside Gardens Apartment in Hangzhou where Alibaba was founded.


Alibaba was launched in 1999 to help facilitate B2B transactions. Initially, it was a collection of bulletins and message boards that helped various businesses source products, with them completing the transaction offline. A few months after launching, Joe Tsai, a Taiwanese-born, Yale Law School graduate became so captivated by Jack Ma that he left his lucrative private equity job to help Alibaba raise money. At the time, while they had already gotten good traction with over 28,000 members, Tsai was instrumental in helping legitimize the business (it wasn’t incorporated prior) and shore up investors. Through Tsai’s network they were able to secure a $5mn investment from Goldman Sachs for half the company (Goldman would later sell their remaining stake in 2004 for $22mn). This was critical to getting billionaire Mayoshi Son’s attention, who decided on the spot of meeting Jack to invest $20mn in Alibaba for 30%, which had now grown members to 100,000+. This investment was crucial to keeping Alibaba afloat post-dot com bubble burst as US internet companies Yahoo and eBay entered China, as Alibaba still wasn’t generating revenue (nor quite figured out how they would yet).

Alibaba webpage from 2000

eBay entering the Chinese market via acquisition of auction house EachNet was a pivotal event for Alibaba. While eBay focused more on C2C and B2C, the concern was that large sellers on EachNet could eventually broach into the B2B space. To obviate that fate, Alibaba moved downstream by focusing on smaller sellers and consumers with Taobao, which launched in 2003. Sellers on Taobao enjoyed free listings and no commissions, in contrast to “Greedbay’s” ever increasing fees. Shortly after launching Taobao, Alibaba launched Alipay, a simple escrow service to help facilitate payments on Taobao. eBay’s bureaucratic management paralyzed platform development, and after moving where the platform was hosted from China to San Jose, web page loading became prohibitively slow and users churned in mass. This effectively ended eBay’s commerce ambitions in China. This was a huge feat for Alibaba as eBay was one of the world’s premier technology companies at the time. Yahoo had similar troubles fending off local competitors in China and instead decided to make a strategic investment for 40% of Alibaba for $1bn in 2005 and let them run Yahoo China (which was in a downward spiral after achieving muted success). This investment helped Alibaba continue to focus on building platform activity rather than monetization, but also likely played a role in Ant Financial ultimately being transferred from Alibaba (more to come).

The Taobao site from 2003 picture above, which was more geared towards consumers and smaller merchants.

By 2007, had turned profitable thanks to the membership fees they began charging sellers to join the platform. But since Taobao was still burning cash flow, the decision was made to list only in a 2007 Hong Kong IPO, which raised ~$1.7bn. The fortuitous timing of these funds helped insulate them from the worst of the financial crisis and let them continue to build the platform. Notably, as global trade was hit hard, Jack Ma decided that there is more opportunity focusing within China and launched Tmall, an ecommerce platform that helps branded businesses sell to the Chinese consumer. A year later in 2009 they launched Alibaba Cloud.  

In 2010, the PBOC issued vague rules on domestic third-party payment platforms which required all payment companies to receive one of the limited licenses or risk having to immediately cease operations. It was not clear whether majority foreign owned entities could control payment platforms or not, but Jack Ma maintains that in order to avoid risking having to shut down Alipay, he transferred the asset to himself and another cofounder for ~$51mn. For additional context, Yahoo and Jack Ma were publicly clashing at the time due to Yahoo China’s continued languishing under their stewardship (likely an inevitability anyway), but Ma’s concern was that Yahoo’s large stake could let them replace him. We don’t think it will ever be clear if Jack Ma was acting out of an abundance of caution, if moving Alipay under his control was an insurance policy to threaten Yahoo, or if something else was happening in the shadows. It could of course also be that the situation is simply as bad as it looks and he was only vying for a higher economic interest but given his character and history of readily sharing equity, this seems the most unlikely. However, AliPay did receive license #1 and eventually Alibaba received a 1/3rd interest in Ant Financial, the new company that housed AliPay. Nevertheless, this instance highlights the major risks of the VIE structure with loose governance.

Early Alibaba Timeline

In 2013 Alibaba pushed farther into reducing ecommerce friction by partnering with 4 of the largest logistics providers in China to start Cainiao. Since over half of all packages were originating from Alibaba sites, it made sense to add a software layer that would integrate all of the different logistics providers, which streamlined operations for everyone and eliminated the need for each logistics company to individually build out their own technology. This also helped them compete against, who operates their own delivery network which is highly regarded by consumers. As Alibaba was becoming increasingly ambitious with their businesses aims, they IPOed in 2014 on the NYSE to bolster their investment capabilities and allow early shareholders to cash out. In the years that followed, they launched several other businesses (Freshippo, Tmall Genie, Taobao Deals) as well as acquired wholly or took stakes in a ton of others including: ChinaVision Media (renamed to Alibaba Pictures), Youku (top 3 long form video hosting sites), (local delivery), Lazada, and many others.

Jack Ma stepped down as CEO in 2013 and Chairman in 2019 to focus on philanthropy. Daniel Zhang, previously the president of Taobao & Tmall, has been the CEO since 2015 with Joseph Tsai still involved as executive vice chairman.

The Business.

Alibaba operates many disparate businesses, but reports them through four segments: 1) Core Commerce, 2) Cloud Computing, 3) Digital Media & Entertainment, and 4) Innovation Initiatives & Others. We will start with Core Commerce, which is ~86% of revenues and the most important piece of their business.

Commerce is the largest segment, but the other segments, especially cloud, will be important for future growth.

Core Commerce.

Core commerce is further broken up into seven segments, splitting it up between business in China vs. international, retail vs. wholesale, and other commerce services including logistics and local (which is mostly on demand food and CPG delivery, similar to Door Dash). BABA generated over RMB 600bn, or almost USD 100bn from their core commerce businesses in 2021, growing revenues an incredible ~3.5x since just 2017. (Note that their fiscal year ends March 31st).

Despite being fairly mature, commerce grew 42% last year, buoyed by Covid tailwinds. While growth will likely slow in their domestic commerce businesses, they still have many international and ancillary commerce businesses that have long growth runways.

As seen below China Commerce Retail is >3/4th of their commerce revenues so we will mainly focus on this piece of the business, which includes the immensely popular Taobao and Tmall platforms.

While Core Commerce generates RMB ~170bn in EBITA, this is net of several money losing operations, absent of which EBITA would be even higher.

Core Commerce: China Retail.

Taobao and Tmall are two of the largest ecommerce platforms globally and together they facilitated RMB ~7.5 trillion of GMV last year (or ~$1.1tn) with that split roughly 50/50 between each platform. For a sense of scale, Amazon is estimated to support ~$500bn of GMV, or less than half of what Alibaba did last year. Just to hammer this point home, while GDP and GMV are not perfect comparisons, there are only 17 countries whose GDP is over $1tn. BABA also has a large buyer base with 800mn+ annual customers just in China and MAUs of 900mn+. As shown below, buyers of Alibaba have not only grown every year, but also stepped up purchasing on BABA’s platforms to hit just shy of RMB ~10,000 per buyer. This is important to keep in mind as we talk about competition later on.

First, we dive deep into the dynamics of the platforms and why customers and merchants incessantly use their platforms.

Tmall or “Tianmao” is pictured on the left, while Taobao is on the right.

Taobao and Tmall.

As mentioned prior, Taobao was started as a sort of eBay competitor, but quickly took on a life of its own, and is now markedly different than anything you will see elsewhere. This is partly because of the underdeveloped retail environment in early 2000’s China, which accelerated the usage of ecommerce penetration. Readers may recall a similar dynamic in our Sea Limited piece, whereby the relative infancy of physical retail was met with a jump straight to ecommerce, especially mobile-native ecommerce. 

From the Alibaba S1 which shows developed countries have much more retail per capita than China. This was more exaggerated when Taobao was launched in 2003.

While eBay was focusing on selling used goods, Taobao knew that that wouldn’t work well in China where the average person had fewer possessions and most things they owned weren’t of high enough quality to be of value to a 2nd owner. Thus, Taobao focused on becoming a platform for small merchants and sellers, and today hosts over 10 million of them. While Taobao does have a reputation for being lower quality, its usage is high even among higher income consumers in Tier 1 cities because of their wide selection and great value with incredibly cheap products.

Swipe through to get a feel for the main tabs on the Taobao app.

Taobao offers consumers a comprehensive suite of products and services, personalized based on consumer preferences and interactive features.

Taobao does not charge a listing or selling fee, which sets it apart from virtually every other ecommerce platform. Instead, they monetize through sponsored listings and other seller services. The no fee structure helped seed the marketplace with more sellers early on and today it has more selection than any other ecommerce platform at 1 billion+ products. Compare that to Amazon, which has an estimated ~350mn products. To fill the gap of higher end products though, BABA has Tmall.  

Swipe to get an idea of what it’s like to interact with the Tmall app.

Tmall Hey Box is a dedicated platform for brands and retailers to launch new products.

With Tmall, BABA tried to move more upmarket with higher quality and branded goods—an area JD was already focused on. More Chinese consumers were becoming familiar with global brands, but there were limited ways to buy them, and most global retailers weren’t sure how to approach the Chinese market. Tmall aimed to solve both of these issues by making it simple for a multi-national company to set up a Tmall storefront and leverage their payment and logistics infrastructure for Tmall customers. Over time, many Chinese brands also started selling on Tmall and today it is regarded as higher end than Taobao and almost as good as, if not on par with JD, which is the gold standard for authentic products given its heritage as 100% 1P. However, JD is often more expensive and is much more limited in selection than Tmall. Most of the consumers we talked to have used Tmall before for higher priced items, but for particularly large ticket items (especially high-end electronics) they prefer JD. Of course, this also varies by income levels. In the competition section below we will get more into when and why a user prefers one platform over another. Tmall’s importance to Alibaba has been gaining in recent years with GMV growing from RMB 500bn in 2014 to RMB 3.2tn last year, almost on par with Taobao.

Tmall and Taobao now generate almost equal amounts of GMV.

Unlike Taobao, Tmall charges a seller fee that ranges from 0.3 – 5.0% and comprises a sizeable portion of revenue. They stopped disclosing revenue from commissions last year, but it was about 40% of what they generated from advertising (in total it was ~21% of China Retail revenues, but that is understating its contribution because of 1P sales from their New Retail initiatives). Tmall has a standalone app, but a lot of traffic is also directed from Taobao. When searching for items in Taobao you can receive results that are listed in Tmall (but not the other way around). This helps Alibaba cross-sell to their large and highly engaged Taobao user base. For instance, a user could search for cooking ware in the Taobao app, but than also be shown cooking ware from the branded Tmall store in the results.

This is Alibaba’s very popular annual 11/11 Single’s Day sales event, think Black Friday on steroids. Many Chinese consumers plan for weeks what they want to buy. Last year, GMV on Single’s Day alone reached RMB ~500bn.

There have been many developments in ecommerce shopping since they launched these platforms, most notably: 1) live streaming, 2) group buying, 3) gamification, 4) short videos, and 5) AR-enabled experiences.

1) Live streaming: Taobao Live-streaming is very popular and has generated over RMB 500bn in GMV in 2021. Live-streaming can often be similar to a home network shopping channel, but is tailored to a specific users interest with a wider variety of different streamers. KOLs (Key Opinion Leaders) often will live-stream products they like and receive a commission on sales (similar to how Instagram influencers will post products they like for a fee). For instance, a KOL may give a tutorial of how to apply a beauty product or show how to make the best stew with a pressure cooker all while promoting the items at a discounted price for viewers.  Alibaba has an affiliates program too whereby they help merchants find KOLs to showcase their products and the merchant gets to set the commission fee (with BABA taking a small portion). The other aspect of live-streaming is more “personal” shopping experience where a farmer or craftsman can showcase their creation. Huang Wensheng or “Uncle Farmer” for example is a popular tea farmer from rural China whose live-streams regularly receives millions of views, dishing out tea expertise and life advice while exhibiting the minutia of farming life (rural farming live streaming has become very popular recently). This format crosses commerce and entertainment, helping make Taobao a higher engagement app.

Taobao Live – Merchants and KOLs stream directly to their fans and potential customers

2) Short-form videos: Similar to live-streaming, this is another format for KOLs and brands to showcase their products. It differs in that it is much “punchier” content that tend to be more produced, although in some cases it is edited down live-streaming clips. Essentially it is a more commerce focused version of Douyin, built right into the Taobao app. The videos are used to help consumers find products and also see them demonstrated. Video has been a big push for Alibaba and they even required sellers to upload a minimum amount of short-form content to be promoted during the big Single‘s Day event.

3) Group buying: This has been given a lot of attention as of late with the rise of PinDuoDuo (PDD). While there are a few different ways to implement this, the general idea is that the platform helps aggregate purchase orders and sends them directly to the manufacturer, allowing them to circumvent a distributor who traditionally played this roll. The value prop to the manufacturer is clarity in demand forecasting coupled with scale, allowing them to offer customers very cheap prices that a distributor would usually buy at. The consumer receives cheaper goods, but they can often arrive slower as they are not on demand. PDD also has group leaders that can start a group purchase order and enlists friends often via WeChat. The group leader (depending on the model) can receive a commission from the order and is also responsible for collecting and distributing the items, which is a very elegant way to solve last mile delivery. However, as PDD has gained scale though, the groups of buyers have dropped from 10+ to often just 2 (and regular home delivery is often used). Once you have a scaled user base, like PDD’s 824mn buyers, the need for groups drops on popular products because the demand is always coming in, however the social functionality that comes with group buying continues to make it more interactive than traditional ecommerce, especially when you add a gamification layer. Clearly PDD found a niche here, but Taobao is ready to compete in this space as well with Taobao Deals launched a little over a year ago. This standalone app has already garnered 150mn+ buyers and they are expanding their product offerings while leveraging their massive merchant and consumer base to grow. Last year 70% of new buyers were from lower tier cities. It’s also worth mentioning, BABA has had another business in the “Money for Value” space for long time, called Juhuasuan, which is housed in the Taobao app for flash sales and BABA has recently revamped it. The shift of focus to the Money for Value space comes amid maturing ecommerce for middle and higher income consumers versus lower tier cities still having low ecommerce penetration. More to come in the competition section below, but this is a growing area and focus for BABA as they’ve committed to reinvesting to fend off competition.

4) Gamification:  Similar to group buying, this has also been a popular buzz word as of late. While PDD may have popularized this by executing particularly well with their gamification features, some aspect of this was present in Chinese commerce for a while (Ant Forest was a game, launched in 2016 by Ant Financial, where users would receive “green energy” whenever they used public transit or bought e-tickets to grow a virtual tree. Real trees would then be planted once the virtual tree matured… sound familiar to a certain recently popularized ecommerce game?). The idea of gamification is simple: make the app look more like a game and reward the consumer behaviors that you want to incentivize within the game. In one of the original PDD games, users receive “water droplets” and “fertilizer” by sharing with friends, buying stuff, visiting merchant pages, and checking in daily, which they can use to grow a fruit tree of their choice, receiving a package of actual fruit at maturation. Features like this help increase user engagement, stickiness and virality. As such, Taobao has been focused on further gamifying their app, with little hesitation to emulate the more popular features of competitor products (see below).

Pinduoduo Orchards is on the left and a Taobao game showcased during their 2020 investor day is on the right

5) AR-Enabled Experiences:  While still nascent, Alibaba has already demonstrated several new shopping experiences that are powered by AR in their apps. For example, virtual showrooms that are interactive or trying on clothing & makeup virtually. These efforts are still early innings and, as the AR space develops, there is going to be many other novel experiences that can be offered. All of these features though are technically hard to implement, and they can potentially raise the barriers to compete in ecommerce over time.

All these features together have helped Taobao remain relevant while also giving competitors an opportunity to gain share. Ecommerce is ~25% penetrated, but will continue to grow as China incomes rise and physical retail becomes more meshed with online (a lot of BABA initiatives here in their New Retail segment that we get into below). As you can see below, despite competitive intensity increasing, Alibaba has grown China retail revenues and active annual buyers every year. If you look at total spend per user (GMV per buyer) across Alibaba platforms, that figure has increased every year. While we don’t have granular cohort data, we know that Alibaba is higher penetrated among higher income users and spend per user increases every year. Thus, we can ascertain that aged cohorts increase spend per user more than the average spend per buyer increases. This implies the spend per buyer figures we calculate in the exhibit below are dragged down by new buyers and the buyer base simply “aging” will result in GMV per buyer growth. Said simply, we show GMV per buyer growing ~6% since 2014, but the (hard to read) investor day slide below implies the 2016 cohort was growing spend ~40%+. New buyers are likely lower income so will not likely spend at those rates ever, but this is good GMV growth driver to be aware of and points to the large value consumers see in Alibaba.

If ARPU grows 3x in 4 years then consumer spend per buyer is growing much faster than taking a simple average of GMV/buyer would suggest.

The table below shows core China commerce metrics for Taobao and Tmall. GMV per buyer grows from RMB ~6,500 in 2014 to RMB ~10,000 last year.

However, while spend per user is still growing, increasing the monetization rate (net revenue as a % of GMV) has been the more powerful lever in topline growth as of late. In the chart below we show what portion of total revenue growth can be ascribed to GMV growth versus increasing their monetization rate of that GMV. While GMV growth used to drive ~40%+ of revenue growth, it has been a smaller revenue driver (ex-covid bump). In other words, their ability to further monetize their existing GMV is more important to incremental revenue growth than growing GMV further (but of course they can do both). 

To answer how can they increase their monetization rate, we turn to the lesser known, but quiet driver of almost all of Alibaba’s revenues: Alimama. 

China Retail Monetization: Alimama.

Alimama is a marketing ad platform, and the principal way BABA monetizes their Taobao and Tmall sites. BABA breaks down their China Commerce Retail revenues into 1) Customer management, 2) Commissions and 3) Other. Commissions are mostly from the Tmall marketplace, and the “Other” category is also known as New Retail, which we will touch on later. Customer Management revenues consistent of several products for merchants to bid on keywords for search rankings, buy display advertising and buy sponsored placements on a CPC basis. There are also programs like Taobao Ke, an affiliated marketing program whereby Alibaba helps match merchants to individuals to promote their products. These individuals are KOLs (key opinion leaders) and receive a fixed commission (set by the merchant) for making sales through their channels, which can include external platforms–this is how many live streamers can monetize their channels.

While it is not important to grasp the nuances of all these products, we wanted to explore the main monetization tools of Alibaba, which sheds light on why they can be considered to be closer to a Google or Facebook, rather than Amazon. In fact, BABA describes themselves as a “media” company rather than a commerce company, and that is because at the end of the day, they are largely in the business of monetizing time spent on their platforms. Compare that to Amazon, who despite their budding advertising business, is designed to get you off their platform as quick as possible by reducing all friction in the buying process. This was evident as early as 1999 when they patented the “1 click” buy it now button. While Alibaba of course does not want to create unnecessary friction in the purchase process, the platform is geared more towards engagement with a target recommendation feed, live streamers showcasing their favorite products, short product videos, and gamification. Even the name seems to suggest the service is meant to invite engagement: Taobao means “searching for treasure” (although treasure is too generous of a description of the products you’ll find on there). Without this context, it would seem weird that Alibaba reports MAUs and DAUs, metrics usually only seen from social media companies. Understanding this is important because the competitor set and product expectations change when a company is also in the “time spent” business.

BABA also enjoys a huge data advantage compared to peers. They not only benefit from owning customer purchase data, which is very high value as it actually shows what a consumer spends money on, from Taobao and Tmall, but they also have consumer data from Ant Financial/ Alipay, in addition to their other ecommerce platforms and media properties like YouKu. They even have their own unique user identity product—Uni ID– that ties all this data together and allows BABA to effectively track users across disparate platforms. This data set allows them to better serve ads and promotions to relevant consumers and is why they also have an advantage in their 3rd party ad network business.

BABA’s Uni ID. Each BABA user has a Unified ID that tracks them across the entire BABA ecosystem.

While seldom mentioned by management, Alibaba has an ad exchange (Taobao Ad Network and Exchange or TANX) that serves ads off platform. This is similar to Google’s DoubleClick business and is likely low margin, but it does give them an interesting competitive advantage. A Tmall merchant, for example, can buy ads through Alimama that are displayed on Weibo and then send the user back to Tmall if they click on it. Alimama not only gets paid for serving the ad, but also the benefits from the user traffic as well as 3rd party data. Similar to Google and Facebook, the algorithms that suggest ads will continue to get better overtime which 1) leads to more sales and 2) opens up more inventory as an ad can be shown fewer times to be effective. However, following this line of think of Alibaba as a “media matrix” (their words) monetizing time spent with ROI driven sellers will quickly make an investor see some potential competitive threats they might not of otherwise. What we layout below is somewhat speculative, but worth being aware of as a potential investment risk.

Alibaba, Just a Media Company?

Alimama essentially monetizes time spent by selling ads and thus it makes sense to look at the health of their advertising ecosystem. A lot of their advertising can generally be considered more “discovery” based commerce as a lot of it is geared towards showing a consumer something they didn’t know they wanted. However, the bidding on keywords is clearly more intent based. This following line of reasoning is more applicable to the “discovery” portion of commerce, however it still effects intent-based commerce as it relates to merchant ROAS (return on ads spend). We don’t know how much of Alibaba’s GMV is intent based versus recommendation generated, but it could be fairly material as in our conversations with dozens of consumers, many noted that they had purchased Taobao recommended products and management has touted Feed Recommendation improvements as driving incremental purchases before.

Above you can see theme-based recommendations and “curated” recommendations on the Taobao platform – both are powered by Alimama and the data collected from the wider Alibaba ecosystem. .

For a long time, Taobao was one of the few places to get high quality recommendations as direct response advertising is very nascent in China with few scaled platforms. Merchants thus have very limited options to drive traffic beyond BABA’s “Media Matrix”. While there are other ad platforms, very few had the advertiser tools, user data, and sufficiently advanced targeting algorithms that could offer merchants a good ad product that drives traffic. This is why Weibo and even Bytedance’s Toutiao and Douyin have ads served on their platforms from Alimama. However, this is quickly changing. Douyin, Kuaishou, Bilibili and others are all building out their advertising capabilities which have created new channels for retailers to advertise through and they are already formidable in ecommerce today with Douyin and Kuaishou driving over RMB 500bn and RMB 300bn of GMV last year, respectively. While a sizeable portion of that activity gets redirected back to Taobao or Tmall, more of it is likely to be completed on platform, starving Alibaba from the traffic.

The real issue comes though when we look at ROAS of Taobao versus Douyin. As expected, ROAS range widely, but a high number of Taobao merchants and a former senior Alimama executive purport ROAS (or ROI) of ~1.2x versus Douyin advertisers of >10x. Several merchants we talked to also didn’t even know their ROAS and instead felt like you had to run the promotions just to drive traffic (without any idea if that was profitable or not).

Frankly, it is not clear how a 1.2x ROAS even makes financial sense with the low gross margin of most merchants, but one contact suggested that is because most Taobao ad buyers are factory bosses who are more focused on consistent factory throughput than profits. It is also possible that these factory bosses only buy traffic when they have an unsold portion of a batch order, or merchants are hoping to make money on a 2nd sale after establishing a customer relationship, but either way these are clearly unsettling comments that suggest usage is supported by uneconomic buyers and it draws to question the sustainability of Taobao’s monetization rate. Even if this magnitude of difference is only partially correct, then merchants will be better off shifting ad budgets elsewhere to generate traffic. As it seems today, most merchants do not know their return on ad spend as it is often (seemingly) not an objective, but other self-serve platforms could train them to focus on this, which could result in a pullback of uneconomic ad spend. If Alibaba is just a “media” company monetizing time spent, then ROAS is king.

The exhibit below walks through the math of why a 1.2x ROAS doesn’t make sense. Informing the merchant assumptions below is an example merchant P&L from BABA’s 2020 investor day. We show what the minimum break even is for most merchants with a ~25% gross margin and what gross margin a merchant would require to break-even on a 1.2x ROAS. The math suggests that it is possible a lot of merchants are losing money advertising.

If a merchant is really getting a 1.2x ROAS, then they are very likely losing money.

While we think this is an important point to surface and a real risk, given the opaqueness of seller ROAS and the wide breadth of answers we received, we are not going to swallow hole the suggestion that Taobao is built on unsavvy merchants’ consistent uneconomic ad spend, suggesting Taobao is over-monetizing and at risk of other ad platforms cannibalizing their ad budgets. In response, we would like to make several points:

1) ROAS can vary for many reasons.  Stating the obvious, but ROAS are not a uniform metric and vary by advertiser, product, brand, audience, ad spend, ad price, and content, which all make it hard to extrapolate out generalized statements (especially with only a dozen data points). While a low ROAS was more common than we would have liked to have seen, it is also true that most ad buying still goes through agencies and operating companies who are more closed-mouth about client ROAS and more probable to be higher. Nevertheless, there still were several merchants who disclosed as high as 8x ROIs. There are ~4mn advertising merchants on Taobao and while it is likely a good portion of them do not have consistent profitable ad campaigns, they will fall out of bidding and new merchants will take their place in the auction. 

2) Ad spend is less fungible than you may think. Many products sold on Taobao are fairly substitutable and it is hard to generate high ROAS anywhere when your product is similar to the rest. Said another way, there is a selection bias between those merchants that can advertise on Douyin and those that only advertise on Taobao. Douyin advertisers are more likely to put more effort and time into content, which if successful garners a high ROAS, but the small hair scrunchie or dining placemat merchant is probably not going to be able to create compelling content.

3) There will be many winners as the digital ad ecosystem evolves. When looking at potential advertising TAMs for companies like Facebook and Google a decade ago, analyst made the mistake of thinking these companies would simply just take market share from the existing advertising players. Instead, what happened was Facebook and Google create new use cases and grew the advertising market substantially. Similarly it is likely that Douyin and other new platforms create new commerce opportunities and do not just cannibalize pre-existing TAMs.

4) Data advantage. As mentioned, Alibaba has a trove of data on users not just on Taobao and Tmall, but also all of their Alipay history, behavior across their media properties and data from their 3rd party ad network: more than just knowing what you like, they know what you actually buy. This puts them in a great position to continue to refine their algorithms to increase ad success rates. A more refined algorithm means higher ROAS which then get bid down again. This process could go on for a long time. Furthermore, we wonder even if Douyin, Kuaishou and Bilibili build out their self-serve platforms, if BABA can better monetize those impressions given their better consumer data. One industry insider claimed BABA was still better at monetizing Douyin inventory than Bytedance.

5) Taobao keeps improving and they are best in class for their ad product types. Remember, advertisers today use Alimama for impressions on 3rd party properties like Toutiao and that is because Alimama can monetize the inventory better than competitors. A different industry contact noted that recent improvements in Alimama’s news feed ad product has resulted in ROAS on Toutiao of 2.5-3x what other DSPs (demand side platforms) are able to achieve (which in itself is still a lower monetization rate than Taobao Feed).

6) Global peers suggest Taobao is under monetizing. Amazon is estimated to have advertising revenues of ~5% of GMV, with 1P estimated to be anywhere from 45-60% of that. For conservatism, assume 3P is 55%, then that implies that their ad revenues are 9% of GMV in addition (!) to seller fees. Other platforms like Mercardo Libre and Coupang are also estimated to be able to generate at least 3-5% of their GMV in advertising revenues (again, in addition to seller fees). In total, across both platforms, Alibaba monetizes GMV <4%. You can construe this as evidence of Alibaba ecommerce environment being more competitive, which is partly true, but we believe that it is more indicative of their monstrous growth and still early innings of monetizing of their platforms. We will revisit this point later, but we think this supports the thesis that they are more likely under monetizing than over monetizing.

7) ROAS are being driven down by high demand. Given their ability to drive traffic directly to merchants store better than peers, the (potentially) low ROAS are being driven by high demand on the merchant side. Every time Alimama roles out ad improvements that increase click through rates (and reduce cost per click), the ads get bid up again. This strongly suggest that the market (for whatever reason) is at an equilibrium and pricing is sustainable at this rate.

8) You are not paying for an explosive ad monetization story. To rationalize Alibaba’s valuation (explored later), it is unnecessary to pencil out a long runway of high growth in ad monetization. Simply gaining confidence that the current monetization rate is sustainable could suffice. While we will cede it would have been more comforting to see higher ROAS for merchants, we have to remember how cutthroat ecommerce is in China. Even on Alibaba’s investor day they showed only a 4% profit margin for a representative merchant (after ad spend).

9) ROAS are not capturing the true value of advertising. In talks with multiple merchants, they suggested that they are not buying ads or promotions on ROAS metrics, but rather because if you don’t Alibaba will throttle your traffic. We have no way of verifying if this is true, but one possibility is that successful advertising results in a higher relevancy score in the search engine result page which then leads to more sales. Perhaps there is a factor in the relevancy score that weighs more recent purchases higher than older purchases, so too long without advertising and slowing sales means you need to “reset” the flywheel by topping up ad spend. This means the true value of the ad spend is the direct sales, plus the knock-on effects of being ranked higher in organic search. The second piece to this is that there are other promotion channels that you become eligible for when you buy ads that may not directly be accounting for in the ROAS figure, so there could be some under attribution.

While we had many points to make, often the best rebuttals are simple, so we will leave it at this: Despite low ROAS and new advertising options opening up the past few years, you actually saw more ad spend on Taobao, not less.


China is notorious for how brutal their start-up competition can be with hundreds of companies going after the same market: Meituan’s success, for instance, was dubbed the War of 1,000 Groupons (in an ode to how many copycat’s they had to beat out). Given this backdrop, it should be no surprise that competitors are not disheartened by Alibaba’s size and continue to try to gain a chunk of China’s ecommerce and supplant BABA as best they can.

Bytedance. Following from our comments in the prior section, Bytedance with Douyin/ Tiktok has been one of the newest disruptors to ecommerce with a highly addictive app that users often spend over an hour on daily and enabling over RMB 500bn of GMV. The large amount of user time spent with full screen videos makes Douyin a natural venue to serve ads and facilitate in-app commerce. While the nature of these purchases tends to be discovery-driven and more impulsive, they are building out their search capabilities that could serve more intentionality-driven purchases in the future. Their large base of videos could serve as a great way for users to learn more about specific products and see them in use. There is a very clear use case for verticals like makeup where a user could search “dark lipstick” and get a wide range of high-quality videos that showcase different products and in the future Douyin will allow you to purchase it in app (even if it is not an advertisement with a link). However most users do not think “Douyin” when they want to buy something and changing how a user interacts with a product is hard. Also, today many merchants drive traffic back to Taobao or Tmall store, but Bytedance is clamping down on that as they are trying to get merchants to set up stores on-platform, but even when they do though, it is often for a limited amount of their SKUs. Generally, we see Douyin as a threat to Taobao’s discovery feed and live-streaming GMV which serves more of the impulsive/discovery purchases, as well as a potential threat to stealing more of the ad budget (as mentioned above). However, BABA still has much better data and users visit Taobao with clear commercial intent, with most of the consumers we talked to having positive opinions on the Taobao Feed recommendations. As long as BABA keeps users on their apps, they will get their share of recommendation commerce.

Although Douyin is wading into the ecommerce space, their core focus is still video. They compete with Alibaba for eyeballs and advertising dollars. For about a decade was Alibaba’s most formidable (and really only) competitor. has a unique value prop versus Alibaba as they take everything inhouse: they own their inventory so there are no concerns of fakes, they operate their own logistics network which is very reliable and fast, and they have a good consumer centric culture. It is for these reasons that tends to be the ecommerce provider of choice among the wealthiest, despite moderately higher pricing (although ecommerce platform usage is far from mutually exclusive). JD’s no hassle, free returns policy is also valued by consumers versus Taobao/ Tmall purchasers who have to deal directly with the merchant (some merchants are wonderful, others not so much, and a “Xiaoer” mediates the dispute). As mentioned, Tmall helped BABA gain consumer mind share as a higher end platform and they continue to lean in on that with Tmall Luxury and Tmall Flagship Store 2.0, but today JD still is generally seen as higher-end. Our conversations with dozens of consumers showed most would recommend a friend go to JD for high end items followed by Tmall and then Xiaohongshu (similar to Pinterest), so BABA still has work to gain more mind share here.

JD mainly competes with Tmall for higher ticket, branded products.

To address JD’s superior logistics, BABA is investing more in Cainaio, their inhouse logistics platform that connects with 3PL providers, to decrease delivery times and errors. However, for quick and reliable delivery, clearly wins today, with most of the consumers we talked to (in tier 1 cities) getting one day shipping with JD and a mix of one, two, and three day shipping with Cainiao (two day seemed to be the most common). However, Alibaba is still investing in Cainiao and has a long-term goal of delivering anywhere in China in 24 hours (and internationally in 72 hours). Remember though, most of the “shipping time sensitive” purchases are already on JD and JD getting to same day shipping is much harder than BABA simply catching up to where JD is now, so JD is at risk of losing some purchases to BABA as their logistics advantage closes. In our talks with consumers, it was very seldom that a user cancelled or didn’t buy an item because it didn’t arrive quick enough, so we believe that logistics is increasingly less of a differentiating factor once you get to 2/3 day shipping. (Of course consumers may always want quicker shipping, we just do not see them comparing shipping times across different platforms after they have found an item they want).

Meituan. Although, the truly time sensitive purchases are now going to Meituan and which originally focused on food delivery, but have expanded to delivering anything locally available, usually in under 40 minutes. Local selection is of course more limited, but can still cover all of a consumers higher frequency purchases like grocery, household products, medicine, and personal care items. High frequency items drive consumer habit and leave an open window every order for the provider to move more purchases into the consumers basket. While the cross-sell is tough with only sourcing locally, BABA cannot afford to capitulate these high frequency purchases. BABA invested in in 2016 and acquired them in 2018. While has a distant 2nd place market share with Meituan commanding ~65% of the local delivery market, it is a strategic asset that they can leverage with some of their other New Retail initiatives, especially their grocery push. In our conversations with consumers, they tend to use Meituan for food delivery, grocery, and some other goods like paper towels and notebooks. In order for Meituan to become a real threat to Alibaba’s Taobao and Tmall businesses, they would have to aggressively expand selection, which will naturally lead to trade-offs in delivery efficiency, not to mention the price of a personal carrier limiting use cases. In some sense these high frequency purchases are worth more to Alibaba than Meituan as getting consumers in their Alibaba Ecosystem presents more selling opportunities given their larger variety of goods, but of course Meituan’s whole business are these high-frequency purchases, so they will not cede them. Critically though, we think it will be very hard for Meituan to move past food delivery, grocery and FMCG without their delivery value prop deteriorating, a move that will quickly lead to an undifferentiated service.

Meituan mainly competes with Alibaba’s Local Services ( and Koubei) for restaurant recommendations and food delivery.

PinDuoDuo. Entertainment, social functions, and even cheaper goods were how PDD differentiated from Taobao. PDD levered group buying as a customer acquisition strategy, while also keeping high engagement with gamification features. This was an incredible combination to attack user growth & user retention and is how they were able to surpass BABA as the ecommerce platform with the most buyers in China, at 824mn versus BABA’s 811mn. However, in the context of BABA’s existing consumers spend, there is relatively muted overlap. This is because PDD’s target market tends to be lower income, rural China and the purchases are more spontaneous. Careful to generalize all consumers behaviors, we see PDD similar to an outlet mall where goods are cheaper and usually of lower quality, but most importantly the channel itself will preclude many consumers from wanted to shop there for certain items. Furthermore, more purchases start on PDD through friends sharing or their discovery feed than search, whereas most people shop on Taobao/ Tmall with intentionality, but, admittedly, discovery has become a more important means of incremental sales and thus this is where we see PDD and Taobao overlapping. (Taobao Deals aside, which is a platform launched directly at PDD). Though, Taobao benefits here from having better consumer data and more merchant products (of low and high quality) which translates to better targeting. While we don’t want to dismiss PDD as they have accomplished something incredible, we just see the nature of these purchases as different and not often cannibalistic. In other words, it is not a zero-sum game and spending across both platforms is likely to continue to increase. We will acknowledge that incremental sales though are more likely to go to new channels (PDD, Douyin, KuaiShou, Xiaohongshu), but this is because they are growing the market with new use cases. This does imply that BABA market share will continue to fall but spend per consumer is a far more relevant statistic and it has continued to go up every year, despite averaging down with lower income consumers—70% of new users have been coming from lower tier cities. (BABA market share was likely unsustainably high anyway at ~60% versus Amazon is around ~40% in the US.  In the context of anti-trust scrutiny, more competition can be a good thing).

Pinduoduo mostly targets the price-sensitive consumer – a userbase Taobao initially controlled, then shed in order to gain legitimacy with US investors. They have recently forayed back into the space with Taobao Deals.

Shown below is spend per user and how it comps to competitors. GMV per annual buyer has grown from RMB ~6,500 to almost RMB ~10,000 despiteoverlapping with the period that PDD gained 800mn+ users that spends RMB ~2,000. The concern we hear most from investors on PDD though is that they have built up their business on low margins so they will be better adapted to move up market and BABA will not be willing to undercut themselves: this argument harkens back to the Innovators Dilemma framework. This framework though does not apply here for two simple reasons: 1) Alibaba has already committed to invest in the money-for-value space and has already rolled out a competitor product—Taobao Deals—with 150mn+ MAUs, as well as refreshed their flash sale product, Juhuasuan. 2) There adjusted EBITDA margins have dropped from 57% in 2014 to 27% last fiscal year—clearly, they are not running their business to a margin metric and there was no dilemma moving down market. BABA’s willingness to forgo margin is similar to a sentiment Bezos expressed in early shareholder letters: “Our pricing strategy does not attempt to maximize percentages, but instead seeks to drive maximum value for the customers and thereby create a much larger bottom line—in the long term”. 

It also shouldn’t be understated how hard of a task it is to go up market either. Consumers associate certain distribution channels with specific expectations and a company’s willingness to attack higher margin segments is in no way an advantage. That’s not to say PDD won’t have some success breaking into new categories, but BABA had to launch a whole separate platform from Taobao to move up market (Tmall) and then invest prodigiously in support and anti-counterfeit measures to ensure high quality. In fact, a lot of the counterfeit goods Alibaba stomped out ended up on PDD, which PDD will have to contend with. Additionally, moving up market risks secluding their current consumer base and retailers risk brand dilution coming onto the platform. Furthermore, most purchases on PDD are price-driven with no recognizable brands and it will be hard to win them as the point of brand building is usually to demand higher prices, at odds with PDD’s current value prop. Lastly, they will also have to catch up to BABA who is well ahead of them on shipping times with Cainaio. While PDD has some initiatives around getting more brands and building out logistics, these are not easy problems to solve and the trade-offs give us confidence that PDD is not an existential threat to BABA, especially when you consider the ecosystem BABA has built.

Alibaba spend per user is ~5x PDD’s and almost twice JD’s.

Alibaba. Alibaba benefits from having ~10mn merchant sellers with the widest selection across both low-end products on Taobao and high-end products on Tmall, with the ability to cross-seed traffic from Taobao without making brands feel their image is being diluted. Merchants on Taobao are known for being very responsive to consumer questions with 40% of all transactions involving a merchant query prior to buying and many consumers receive responses in mere minutes. The rating and review system on Taobao keeps merchants very invested in making their consumers happy, with merchants generally very accommodating for product issues and often dropping an extra small knick-knack or sample into purchases (the large team of Xiaoers act as a back stop should a customer still be unhappy). This effectively turns merchants into their own customer service, fielding product questions and addressing issues. Merchants also devote significant effort to curating their store content for Taobao, including videos, live-streams, affiliated content, all of which helps educate consumers, but also is hard to port over to other platforms in full. Hosting a lot of this content is Alibaba’s Weitao which is sort of like a commerce-oriented Instagram that is integrated directly into Taobao. Weitao users can follow businesses or KOLs to see educational and entertaining content about brands or products. Many Weitao KOLs, who get paid commissions for sales, find it to be such a compelling platform that they do not build up much of an audience elsewhere. This is all activity that is unlikely to leave Taobao. Additionally, as mentioned, Alibaba has virtually all of a consumer’s purchase history (among other data) so they are well positioned to have the most personalized product suggestions. Furthermore, Taobao has multiple selling formats allowing merchants the most options in how to sell their products from live-streaming and short-form videos to branded stores or promotional flash sales, all of which is supported by Alipay integration and Cainaio’s logistics. Lastly, if you want to reach over 800mn+ consumers who spend ~10,000 RMB annually, you have to be on Taobao.

There is no doubt that BABA would be a much better business financially if JD, PDD, Meituan, and Bytedance never existed. Generally speaking, fear of competition has driven BABA’s margins down overtime as they continue to invest heavily in making the consumer experience better. This is actually exactly what you would want to see though as the alternative would be very concerning.

If you remember our eBay conversations earlier on in this report, one of their biggest mistakes was not taking over more of the transaction: they didn’t help with shipping, fulfilment, consumer service, or merchant support. The result of this was two-fold: 1) poor consumer experience when buyers were inevitable let down, and 2) competitive advantages that dwindled overtime as more places to connect buyers and sellers popped up. eBay wanted to be a pure marketplace and keep the high ROIC and margins that entailed, and even today they have some of the highest gross margins of any marketplace. Amazon, of course, took the opposite approach and built out all the infrastructure needed to provide the consumer with the best experience, which is of course costly, but that’s why Bezos touts “Your margin is my opportunity”. Fast forward a few decades its so clear which approach was better, we don’t even need to say it. We bring this up to make the point that Alibaba has transitioned overtime from essentially eBay’s modus operandi with minimal enduring competitive advantages to a hybrid between the two and this was happening long before the new set of competition was an issue. Every year that goes by BABA reinvests to make each transaction a better experience with more efficient warehousing picking, slightly faster shipping, better product recommendations, new selling formats, and deepening the value of their buyer & seller network with products like live-streaming, Weitao, and short-form videos. We won’t make the claim that BABA is as well insulated as Amazon is with Prime lock-in and merchant’s inventory literally stored in their warehouses, but there is simply no ecommerce player that could replace what BABA does today at the scale they do it. 

We understand the fear of competition: PDD is gaining share on lower end commerce, is relentlessly attacking the high-end, Meituan is going after high frequency local and Kuaishou, Douyin, and others threaten share in discovery commerce. But BABA has a strong hand to remain relevant for a long time with the deepest product catalog, the most selling formats, a unique product description platform, the most purchase data with a 3rd party ad network to drive traffic and over 10mn merchants who are highly invested in BABA’s platforms with it often being their most important sales channel. Add in 800mn+ buyers who buy an average of twice a week, seamless integrations with the country’s largest payment method–which itself is a large funnel of traffic–and Cainiao’s logistics network and you can see why we are not that worried for BABA. This is before mentioning all of the other strategic initiatives and assets in New Retail, Money-for-value, augmented reality, and many more that can further create a unique value prop.

Open Sesame. 

Alibaba warehouse.

New Retail.

Alibaba has been focusing on what they call “New Retail” which is ironically physical retail (but with a software layer). There have been three big developments here so far and most of them relate to grocery because of its high frequency, room for consumer improvement and large TAM (estimated over >$1.5tn and growing). As mentioned prior, since physical retail was less developed in China, more of an emphasis was placed on ecommerce earlier on. However digital commerce could not fully fill the need for more grocery stores as many consumers like to browse markets and see & smell the food before buying. Plus, most shoppers buy food multiple times a week, which means the products must be close to the consumer. The grocery category is also tougher to penetrate because many foods need cold storage and building out cold-chain logistics is capex heavy. The easiest solution to getting customers cold food quickly is simply to build out a grocery store footprint where the store also serves as a mini-storage facility for their inventory. This, tied with the strategic rational of why grocery is important mentioned in the Meituan section above (high frequency purchases increase user tie-in and drive increases in total spending across services), is why BABA has opted to focus on physical grocery stores and try to link them to their digital platforms.

Taoxianda. This is effectively a “store OS” that helps integrate online and offline channels with inventory management systems. This was initially rolled out and implemented at a portion of Sun Art’s stores, a Chinese supermarket chain with ~500 stores that Alibaba invested in. This software makes all of a retail store’s inventory shoppable online and plugs into Alibaba’s different demand channels and logistics platform for delivery. Taoxianda enables inventory in Sun Art’s stores to become available to users on and Tmall Supermarket, which expands grocery selection and range, while speeding up delivery times. After successful trials with Sun Art, Alibaba took a controlling position in Sun Art and now owns ~72%. It’s worth mentioning that Taoxianda, has a natural synergy with AliCloud as adopting this Store OS will often lead to new compute requirements. 

One of the first campus supermarkets powered by Taoxianda.

Tmall Supermarket. This is a platform for users to order mainly groceries, but also other small CPG across multiple different supermarkets. Tmall Supermarket has also been tied to their food delivery app,, which cross-seeds traffic and utilizes the same driver network. The strategy is to use the Tmall and traffic to bolster their grocery delivery business, while they simultaneously onboard supply through Taoxianda.

Freshippo (or Hema in China). In addition to aggregating existing supermarkets onto their platform, they are creating their own. Freshippo is Alibaba’s new chain of high-end grocery stores that imbeds digital capabilities directly into the store with several different store formats. New features include 1) scan & go which displays info about an item as well as offers delivery options and the ability to pay for it right on your phone, 2) digital price tags that update in real time for dynamic pricing, 3) stores as fulfillment centers with employees picking in store and prepping orders for back of store delivery, 4) facial recognition payment, 5) delivery under 30 minutes within 3km, and 6) experiential retail with automated robot servers. Listed below are several store formats they are currently experimenting with. Across all of their formats, their Freshippo stores become profitable within 6-12 months.

1) Freshippo Farmers Market are in residential districts, mostly in outskirts of cities and offer a wider variety of fresh produce and meats alongside regular market items. 

2) Freshippo Mini is a smaller neighborhood store format that offers high quality supermarket items and can deliver to anyone in a 1.5km distance.

Freshippo Mini Store

3) Freshippo F2 stands for fast and fresh and they sell fresh fruits, drinks, and other small packaged food items that you can self-check out from a kiosk as well as order warm food that is held in meal lockers.

4) Freshippo Pick n Go is located at subway stops and busy intersections. A user orders ahead of time on their phone different warm meal options and then picks them up from a locker with a QR code.

Freshippo Pick n Go

5) Freshippo Station which fulfills produce orders for local residents to pick up or have shipped within a 1.5km distance.

6) Freshippo Membership Store which is similar to Costco where a customer buys a membership for the right to shop there and is given discounts on bulk items in return. They charge RMB 258 versus RMB 299 for Costco and they have over 500,000 paying members in Shanghai (only membership stat disclosed).

7) Freshippo Mall is a full shopping center with family attractions, retail, and food. They only have one of these currently, but it is an interesting look into how the mall of the future can turn out with all of a retail stores’ inventory purchasable on your mobile phone.

While it is hard to know where the future of retail ends up, we believe that success with Alibaba’s omnichannel approach could position them at a significant edge over competitors. Many consumers still enjoy shopping in-person, but are put off by the annoyance of check-out lines, crowds, or out-of-stock items, which are problems Alibaba can solve, while further entrenching their relationship with the consumer. This will create another unique value prop that is hard to replicate as it’s in the world of atoms and not bits. Furthermore, crossing the online/offline boundary with Taoxianda and initiatives like Freshippo Mall, could result in Alibaba having the most shoppable inventory that exists simultaneously in real stores and online. This would effectively make every retail store on their systems a mini-warehouse for their ecommerce platforms while simultaneously offering the retail stores an opportunity to drive in-person traffic from their mobile apps. Success here could look like a consumer searching for a sweater online and deciding to try it on in-person from a shop they know with certainty has it in stock and then purchasing it on their phone to have it shipped to them later, so they don’t have to carry it with them to their next destination. This is all speculative, but we think retail will move in this direction at some point and Alibaba is well positioned.  

China Wholesale.

The remainder of Alibaba’s core commerce operations are a much smaller portion of revenue. China Wholesale operations are just ~2% of revenues, and consist of the and Lingshoutong businesses. is B2B platform entirely in Chinese for domestic manufacturers and suppliers to connect with wholesalers and distributors. In order to sell on the platform, you must be a member which costs RMB 6,688 annually and you need a government licenses to sell. The gives buyers some confidence in the legitimacy of the supplier. Banking restrictions also tend to bar any foreign companies from buying or selling on the platform given how complicated it would be to remit payment (but this is what is for). Today they have over 40mn active buyers and ~1mn suppliers who pay for the right to sell on the platform in addition to receiving premium analytics and customer management services. This is the largest B2B commerce platform in China and cover a large swath of goods. As of last disclosures, of their 40mn active buyers, an average of 7mn are daily active users with 76% of buying done in-app.  Renewal rates were 70%, but number of buyers still were growing ~20% with daily average buyers increasing +50% and paying members growing ~40%.

Lingshoutong helps SMB businesses digitize by offering them a slew of different tools including access to a larger variety of goods they can offer their customers, online/offline integration, delivery, promotions and even store footprint layout.

These are both relatively small pools of revenue for Alibaba, but they help support the rest of the Alibaba ecosystem and are good businesses with less competition focused here. Last year this segment grew 15%, partly adversely affected by Covid delays on domestic manufacturing. Alibaba should have plenty of pricing power with membership given the demand generation for suppliers and few alternatives, but we don’t expect much more than low double digits growth over the long term.

International Retail.

There are several different platforms Alibaba puts in this segment, most of which they acquired. This segment comprises only 6% of revenue, but is considered a strategic priority longer-term, especially for them to reach their 2036 of 2 billion Alibaba Ecosystem consumers globally. Several of these platforms are tiny and do not have any disclosures, but we walk through what all of them are below.

AliExpress is global marketplace, available in 17 different languages and 200+ countries, that connects buyers and merchants globally. However, despite their stated ambitions for a global platform, most of the supply is still sourced from China and the buyers tend to only shop for cheap stuff. Additionally, this has also become a popular channel for “part-time sellers” to buy in bulk from China and resale on other platforms, like Amazon, because it’s cheaper since they skipped the traditional supply chain with manufacturers selling directly to consumers. The last breakout was in 2019 which showed USD ~$10bn GMV with 80mn annual buyers that tilted young (60% under 35 years old). They also broke out the most popular items by country, which shows how “different” consumer demand is globally: Brazil—quartz wrist watches, Spain—vacuum cleaners, Netherlands—diamond painting cross stiches, UK—laced wigs, and the US—hair bundles with closure. 

Tmall Taobao World is basically their domestic platform offerings ported overseas just for the Chinese consumer outside of China. This platform is aimed at the tens of millions of overseas Chinese consumers and basically lets them keep using Taobao / Tmall even though they are abroad.

Lazada was mentioned in depth in our Sea Limited piece, but it is a South East Asian ecommerce platform that Alibaba took a controlling stake in in 2016. At that that time they were arguably the #1 ecommerce platform in SEA, as Shopee was barely a year old. However, today it looks like they have lost the lead and are distant #2 or #3 in all geographies they operate in. There have been issues with Lazada management being dropped in from China and not properly localizing their offerings to the region and instead just trying to push what worked in China. Even though they are losing ground, they have not capitulated SEA in any way, and we should expect an increased investment here, although their chance to succeed in SEA seems fleeting with Shopee moving very fast ahead of them.  Having said that, they do still have the benefits of a huge Chinese seller and consumer base, so perhaps they can leverage that better to onboard merchants and attract users. Last disclosed they had 70mn unique customers, with orders growing ~100% y/y and 75% of parcels going through their own facilities. These statistics are pre-covid though and they have lost ground since, but are likely still growing nicely since the market is so under penetrated. Frankly though, there isn’t much evidence to support that they are doing well here against Shopee. Interestingly, they also have a large stake in Tokopedia (~28% before the last private funding round), another Shopee competitor that is merging with Gojek and is closer in catching up to Shopee in Indonesia.

Trendyol is the leading ecommerce platform in Turkey that Alibaba acquired a majority stake in 2018 and further increased their stake to 86.5% with a $330mn investment. It has a rumored valuation of $11bn currently with GMV on track to be $10bn for this year.

Daraz, started by Rocket Internet who also incubated Lazada, is the leading ecommerce platform in Pakistan. Alibaba bought them in 2018 for a rumored ~$200mn. 

Tmall Global is a platform that helps overseas businesses sell to mainland Chinese consumers without the need for any physical locations (note that most large, multi-national brands that sell to Chinese consumers have Chinese offices and are counted in China Retail). Tmall Global GMV is not broken out, but it was disclosed they were growing 40%+ in 2020. This helps open up the Chinese consumer to more international sellers, growing their offerings, a potential advantage versus the competition.

Kaola was acquired for $2bn in 2019 and is a membership-focused service that is geared towards higher-end, overseas, import products offered at a fair price. Essentially, Kaola places large orders to get discounts on a variety of goods and then passes along the price savings to members. While you can use the platform without becoming a member, those that do pay the $40 fee receive premium discounts, lower delivery fees and get assigned a permanent customer service rep to help them shop. Kaola active buyers are growing +20%.

These businesses have some natural advantages riding on Alibaba’s commerce, payment, and shipping infrastructure, but those do not tend to extend well outside of China. Lack of familiarity with local customs has been a barrier to knockout success internationally. The import business focused on the Chinese consumer seem to be better positioned as they more naturally fall in Alibaba’s circle of competence: serving Chinese consumers. These international efforts are still important focuses for management, but frankly their efforts here seem to lack focus. Of course, this could change and the opportunity remains as SEA ecommerce is still in early innings.

International Wholesale.

This segment consist of just, the very first business Alibaba launched in 1999. was originally created to help connect Chinese and international wholesalers. Today it is the largest wholesale marketplace with 26mn buyers from 190 countries. Suppliers do not have to pay to list their items, but if they do they get vetted by a 3rd party agency and receive an official logo which helps signal some legitimacy. The fee to become a member (all members are dubbed Gold members) is either $3,500 annually for just the official logo or $11,000 annually, which includes more ad spend for keywords and they receive help to create product posts. As of last disclosure in 2020 they had 190,000 paying members with 76% renewal rates and 1.6mn DAUs and DAU buyers growing +100%. Renewal rates seem somewhat low, but it is likely that the free offering is compelling enough for suppliers that if they only see a muted increase in sales, they downgrade their membership. Alibaba could devalue the free service to push the memberships more, but we don’t see them doing that as it would weaken the ecosystem. has a lot of the same listings as, but tends to be cheaper, with more selection and better customer service, partly owing to being in the same time zone and no language barrier. has machine translation messaging to help bridge the language gap as well as virtual factory tours to instill more buyer confidence. Additionally, they have some other services like omnichannel data visualization, intelligent marketing, and financial services.

Local Consumer Services.

This consists namely of the, Koubei, and, Fliggy. and Koubei were both acquired in 2018 for $9.5bn and $8bn respectively, and the combined company became Local Consumer Services with some other assets. They separately raised ~$4bn of capital in 2018 at a $30bn valuation from Softbank and others. Alibaba currently retains ~73% of this business., which translates to “Are you hungry?” is a food delivery service that has expanded to delivering any local goods. As mentioned in the Meituan Competition section, they are the #2 market share leader in most markets with ~35% share versus Meituan’s 65%. In our conversations with consumers, most preferred Meituan but a few have use(d), and were satisfied with the experience, but felt Meituan had better selection and delivery speed. Meituan had a restrictive policy that made restaurants make Meituan their exclusive delivery provider if they used them and so that limited’s ability to onboard restaurants given they had a smaller user base (whereas usually a restaurant would just consider the incremental demand as worth it). Since this practice of “two choose one” is no longer allowed, we could see gain more restaurants and perhaps relevance. Alipay’s super app also serves as a funnel for traffic for where users can order food in the Alipay app (40% of new users in 2020 were from Alipay). In the New Retail section we talked about Alibaba’s initiatives in grocery and having the best grocery selection could also help boost’s market share. Commanding a higher market share is important in local delivery as the local density effects can draw more delivery carriers and users, which kicks off a virtuous cycle: more orders allow better route matching to more carriers who can deliver more orders quicker, reducing the cost of delivery, while increasing the carriers total pay and decreasing user delivery costs, thus bringing in more users. still has a smaller footprint with presence only in Tier 1 and Tier 2 cities, versus Meituan who has started to expand out, however it is questionable how well the economics pencil out when you go to lower tier cities which likely have smaller AOVs (average order value) and worse density. Alibaba is likely still losing money in this business and will continue to for some time. It is somewhat strategic as they tie it to their grocery initiatives which tie into their core marketplaces, but its hard to see them becoming the leader anytime here (and owing to the local density and network effects mentioned above, the winner takes a disproportionate share of the industries profit).’s standalone offerings are not quite as diverse nor built out as Meituan’s, but it is still a force to be reckoned with.

Koubei is the leading restaurants & local services platform, similar to Yelp. However, Meituan owns Dianping, which is a similar service and more popular. There are natural synergies between the a platform that helps in local search and’s local delivery network that can deliver those items to the consumer, but they still are distant 2nd to Meituan. In a consumer survey, ~70% of respondents viewed Meituan Dianping more favorably than Koubei. Given Meituan relentless focus and sharp execution, we wouldn’t expect this to change, however it is possible Alibaba’s New Retail initiatives do breathe life into their local consumer businesses, but that is not our base assumption.

Fliggy is an online travel platform for purchasing plane & train tickets, booking hotel rooms and finding local events. Fliggy also uses AliPay traffic to drive usage, but nevertheless is still a distant 2nd to market leader Ctrip. Consumers we spoke with viewed their experience on the app poorly.

All of these businesses are interesting and do have some strategic value, but as of now they are likely destroying shareholder value as investment and management focus is poured into areas they are unlikely to win outright, if at all. Investors may wonder why they don’t simply partner with the best-in-class, which could achieve much of the same benefits, but at a cheaper cost. That does not seem possible as Tencent is involved in most of these competitor companies and would not likely welcome a partnership. To management’s credit though, even if these businesses are losing money now, ensuring access to a fast delivery service and local services platform (even if a #2 service) could turn out to be a wise strategic move, as we can never know how the future of commerce unfolds. Today, we assign very little value to this segment (detailed in the valuation section), so we consider it largely a call option should things turn out much more successful.


As touched on, Cainiao logistics is essentially their software integration with various 3rd party logistics providers that makes shipping items for merchants very simple, with the merchant able to use Cainiao to clearly see who the best logistics provider is for each order. They started this effort in partnership with several large logistics providers and thus didn’t have a majority stake, however since then they have been increasing their ownership stake (they have a majority now it is consolidated on their P&L). Their last investment of $3.3bn in 2019 increased their stake ~12% to ~63%, implying a ~$28bn valuation (and later upped it to 66%).

Cainiao directly helps disperse inventory with warehouses, but their software integration enables the whole logistics process.

Since Alibaba’s platforms source around ~half of all packages delivered (and as high as 70% by some estimates), having direct integration with the logistics providers helped increase efficiencies and speed delivery times. Their smart routing and sorting services reduce logistic partners delivery errors by >40%, which is key to reduce customer churn from poor experiences. Improvements on shipping times can be exemplified by their progress on Single’s Day: it used to take them 9 days to deliver the first 100mn packages in 2013, but last disclosed in 2018 that took just 2.6 days. They are continuing to invest in speeding up deliveries for their 5mn+ daily packages, including investing in robotics and autonomous trucks.

Autonomous delivery robot in testing.

In addition to Cainaio, Alibaba also made direct investments in the logistics providers and now holds stakes in several, shown below. (These stakes are accounted for using the equity method and are not recorded as revenues in Cainaio). 

Additionally, Cainiaio has built out a smart warehouse platform which enables fast product picking and packaging, leaning heavily on robotics: products are packed in an average of just 3 minutes.  They also have a network of package pick-up locations numbering >60,000 at last disclosure.

Cainiao has a network of pick-up locations to give consumers more delivery options.

While we mostly think of Cainiao as a enabler for Alibaba’s ecommerce businesses as it traditionally has been a drag on cash flow, however they announced that Cainiao just became operating cash flow positive. Nevertheless though, logistics will continue to be a drain of cash as they move towards 1 day domestic shipping and 72 hour anywhere global shipping. Compared to JD Logistics’ ~9% gross margins though, Cainiao should be a much higher margin business as they do not own all of the physical equipment and warehouses involved, but we have no segment financial disclosures to dig deeper. You can see revenue growth reaccelerated last year to +68% y/y, helped by Covid increasing ecommerce utilization and delivery demand. Look at Cainiao as % of GMV, their monetization rate increases faster than GMV growth, showing the ability to take some pricing as they continue to drive cost savings and efficiencies for delivery partners. We would expect future efficiency gains to lead to incremental pricing power, however, as a strategic enabler of their ecommerce business they are unlikely to be too aggressive here.  

Other: Alihealth.

Other core commerce is ~2% of total revenues.

Alihealth is consolidated in their core commerce revenues and is the majority contributor in the “other” category. The beginnings of the company were after acquiring a majority stake in Citic, an integrated information provider that provided the tech foundation to build off of. The result is Alihealth, Alibaba’s healthcare company that works to provide fair and affordable healthcare to 1bn people. It has several businesses: cloud-based pharmacy, medical & healthcare services, and tracking & digital health. Within these businesses there is an ecommerce platform for buying medical and pharmaceuticals (through Tmall Pharma). Their app, Doctor Deer, lets users search medical information, make appointments, and conduct virtual doctor consolations. Revenues are RMB ~15.5bn, but profit is just RMB 0.4bn for fiscal 2021. Alihealth is currently listed as a standalone company on the Hong Kong Exchange (HK: 0241) and BABA holds a 60% ownership stake. We will not go into more details here, but we may do a separate in-depth write-up at another time.

Cloud Computing.

Moving beyond Alibaba’s commerce businesses, we look at their 2nd most important segment, Aliyun, also known as Alibaba Cloud.

Alibaba was early to cloud computing, launching this business in 2009. They are China’s market leader with over 3mn+ customers, >40% market share in China and ~30% market share across all of Asia. Today Alibaba’s cloud segment is just 8% of revenues, but is growing rapidly at a ~50% clip (that includes the hit from losing the Bytedance contract for non-technical reasons, further touched on below). While China’s cloud penetration is still in early innings, Aliyun just recorded its first profit in 3Q21.

BABA management estimates that IT spending for internet companies is around $80bn USD and the total spend including public sectors is closer to USD 400bn (compared to BABA cloud revenues of ~$9bn). Additionally, this TAM will continue to grow as businesses rely on more data and compute. However, today the cloud market is around just ~$25bn USD which is 1/8th of the US market’s size and industry estimates point to that figure growing to >$80bn by 2024. Our industry checks noted that 90%+ of Chinese companies are still paper-based, using physical documentation for critical business functions, which also will become digitized overtime. Cloud services are a cost savings to the organization as they do not have to maintain their own servers, IT department and tech spend to keep everything updated. However, lack of vertical software vendors means relying on bespoke software is still more common, which makes the immediate cost savings of moving to the cloud less convincing. However, there are still strong arguments for moving to the cloud for flexibility, speed to iterate products and better utilization of data. Nevertheless though, it is probably an easier sell in the US when the CFO can pencil out direct cost savings in the near term, versus trying to sell non-tech savvy business managers on the value creation of moving to the cloud, but the market will still move here overtime. To create more of a Saas ecosystem to help accelerate the transition to the cloud, BABA is partnering with various software developers to help foster this ecosystem.

Taking a step back, there are several other cloud players in China with a split between public and private cloud. Private cloud is more popular with SOEs, especially in the financial services and public sector verticals. The public cloud demand is more broad-based but largely driven by the newer tech companies for now. Alibaba only has a small private cloud business and the largest player here is Huawei, with a ~80% share supported by a close government relationship. As Aliyun is China’s first cloud provider, they’ve had the most time to build up their platform. With them recently moving all of their commerce activity onto it, they have proven that their cloud is capable of functioning with gargantuan amounts of activity. Several industry experts noted that being able to handle this massive amount of traffic was only made possible by them having a very high performing instantaneous QPS platform, which shows just how far ahead of the competition they are technically. Alibaba’s lack of profit was also in part a byproduct of them continuously investing in their cloud platform. In addition to investing in their cloud infrastructure, they have been further developing, distributing, and integrating their services into their other business, most notably by developing AliOS, a mobile operating system designed for smart cars and IoT devices, as well as reinforcing services like DingTalk (explained more in the Innovations Initiatives and Others section).

1) Amazon is of course the original cloud player in the US with top tier tech, reliability, and performance, but regulation and the complexity of the Chinese market has hampered their cloud efforts here. It is estimated they only have a mid-single digit market share. Many companies are likely to prefer a domestic service to avoid any potential conflicts with the government on data sensitivity.

2) Tencent is the distant second market leader with approximately ~12% market share and their technological abilities are far behind Alibaba’s.

3) Huawei has the biggest private cloud business, but their public cloud business is estimated to command only a high single digit market share. They benefit from close government ties and our expectation is that they win virtually all of the China’s governments business.  

4) Kingsoft Cloud is smaller player currently with only a mid-single digit market-share, but they are still considered a top player behind Alibaba and Tencent. This is in part because they are the only pure cloud provider that doesn’t have other businesses that could compete with customers.

5) Baidu Cloud is another small player with a tiny market share despite being considered one of the internet giants. We have heard underwhelming support for their products.

Alibaba is well positioned to continue to keep their ~40% public cloud share against the competition, as BABA is known as the most reliable, as well as the easiest to use and maintain. This is why ~60% of A-listed companies picked Alibaba as their cloud provider. Overtime, it is possible we even see BABA gain share as smaller competitors drop out. Foreign competitors’ disadvantage is likely to only increase going forward as the Chinese government is likely to support keeping data with domestic companies. The Bytdance debacle, where the US forced Tiktok to be run on a US cloud provider with servers located in the US is only one instance of “data localization”. The EU also is implementing similar rules. Losing Bytedance international business was material drag to their 2021 growth rate, but isn’t a big deal relative to the remaining Chinese opportunity. There is little customer concentration risk now with the top 10 customers representing ~8% of total revenues. Given how underpenetrated China’s TAM is and the large value add of the service, Alibaba could grow cloud ~50% for many years to come. AWS operating margin at ~30% and could be ~40% at maturity and management notes that over time there is no reason their margins cannot match Western peers. However, they are likely to continue to heavily invest for some time, weighing on margins. It is worth mentioning that as far back as 2013 we know AWS was turning a profit, just 6 years after starting their cloud business versus Alibaba who took twice as long to become “profitable”, which happened in 3Q21 and only after adding back SBC. We think this is largely a byproduct of the quicker adoption Amazon experienced to the cloud in the US and supports the idea that the shift to the cloud is much more nascent in China. While it’s not the cleanest comparison due to PPP (purchasing power parity) differences, in 2013 AWS was generating $3.1bn of sales, an amount BABA didn’t match until a little over 2 years ago.

As seen above Cloud Revenues are up ~50x just since 2015 and is still growing 50% before adjusting for the lost Bytedance contract. Growth vectors are not limited to just adding new customers either. As data utilization increases and new value add services are added, spend per customer increases. In 2Q21 BABA disclosed that among their A-listed companies, spend is growing 45% per customer still. While its hard to be precise about future growth rates, we are comfortable with the assumption that China’s ~$400bn of IT spend will only go up overtime and the 3% penetration of IT budgets spent on the public cloud will increase too. The US currently spends only 10% of their IT budget on the cloud, which will also go up overtime. Our unscientific conservative estimate, helped formed by various industry experts, is that ~30% penetration of IT spend on the public cloud is very feasible over the long-term with upside should there be breakthroughs in secure multi-party computation, which eliminates privacy concerns, making more computation cloud eligible.  Without the illusion of false precision, if you assume China reaches parity with US IT spend in 30 years and US IT grows just 4% annually (grew 8% in 2021) with 30% public cloud penetration, then that implies cloud growth of in the low-teens for 3 decades—a much more impressive feat than it may sound. In the revenue build below though we will revisit these estimates and apply them to AliCloud.

While this slide is a few years dated, it is in the same general ballpark despite public cloud having gained higher incremental share of IT spend.

Digital Media and Entertainment.

This segment houses a variety of different assets that help support the Alibaba Ecosystem. Frankly though, all financial data given here is of limited use given re-segmentations and consolidations of assets without full disclosure. We will go through each of the businesses in this segment but spend limited time on the financial contributions of this business given that it is still loss making with an unclear path to profitability. There is a slew of different initiatives housed in this segment, but most of them aim to support the Alibaba Ecosystem, more so then to generate profit in the interim.

Alipictures is an integrated, internet platform for promotion & distribution, intellectual property licensing, as well as ticketing & data management. They invest in and help distribute films that end up in actual movie theatres and a few have won awards like “Green Book” which won 3 Oscars. There big value ad is in bring foreign movies to Chinese audiences by licensing them. 8 of the top 10 grossing films in China were coproduced and distributed by Alipictures. However, it looks like they could be taking a more active role in content creation with the launch of a new inhouse studio last September 2020, called Surprise Works. This is a shift in them taking over more of the creative process rather than just financier and distributing partner. They are originally going to focus on “Current Youth”, “Interesting Lives”, and “Near future Sci-Fi” and have a roster of 20 theatrical films already.

Alipictures is a standalone company that is listed on the Hong Kong exchange (HKG: 1060) and currently has a market cap of ~$4bn. BABA currently holds a majority ~51% stake.

YouKu is a long-form video distribution platform that is the 3rd most popular in China and hosts variety of different self-produced content as well as high quality licensed content. Additionally, YouKu hosts the very popular 11.11 shopping event live streams. Separate from Alipictures, YouKu also licenses content and has their own production company to create content that is exclusive to their paying users. Over the past few years, they have deemphasized licensing with more than 50% of their content now owned. They have dozens of original TV shows that have garnered millions of views each, as well as some blockbuster movies.

Alibaba is leveraging YouKu as a way to mix commerce with entertainment.

While Youku frankly doesn’t have a great reputation among younger Chinese as the content can seem stale, they are continuing to add more shows to attract a younger audience like “Street Dance of China” which went viral. Their main competition here is Tencent Video, iQiyi, and increasingly Bilibili. Their ecommerce initiatives with short videos, discovery and livestreaming content featuring KOLs can help bring in a younger audience, while also supporting their commerce businesses. There is also an opportunity to cross sell products when a tv show happens to showcase a product or food dish. Similar to Queens Gambit leading to a huge increase in chess sets in the US, showing a certain dish on a tv show can lead to more food orders for it on or showcasing a location in a tv show leads to more tourism there, which can be captured by Fliggy pushing promotions. Furthermore, there is product merchandising and events opportunities that can be monetized from successful shows. For instance, with “Street Dance of China”, a live dance competition show, they were able to use their ticketing platform to sell tickets for live viewing, license the brand for in-person dance classes & to sell branded merchandise on Taobao, and then then signed the most popular contestants to their talent agency. While they do not break out user info, they are estimated to have around 85mn paid subscribers (we range it between 50-100mn). Lastly, they also have a lot of free content that is monetized through ads. With Alibaba having a ton of user purchase data they are able to very effectively personalize ads for users and drive traffic to Taobao and Tmall.

Street Dance of China performance.

UC Web is a mobile internet company that operates a browser, a search engine and a news product. They were one of the most popular browsers in India with over 500mn MAUs at peak, however recent security measures in India resulted in the app being banned and a scandal in China where promoted search results commonly showed false medical advertising led to it being removed from several app stores. As a result, the user base today is much smaller and this business may be shuttered at some point.

Alisports is trying to modernize and grow the Chinese sports industry. They are getting broadcasting rights for various sports events that they stream on YouKu and there is new interactivity features embedded into the platform. The data they collect can also help them retarget merchandise. For instance if you watch a lot of one basketball team then they can cross sale you their star athletes basketball shoes. Additionally, they are also trying to encourage people to play more sports to spawn more viewership for the professional leagues, which also creates tie-ins for Alihealth (discounted health services for more athletic people). Currently, these efforts are still experimental with them focusing on marathons to build out the technology with the ambition to expand it out to all other sports.

There are slew of other businesses (most of which sit in AliPictures) including Alifish (copy right trading platform), Beacon (helps film makers understand their audience), Cool Young (artists agency), Damai (online ticketing platform for live events), Funghuangyuanzhi (theatre management platform), Maizuo (platform for live venues), Shuqi (ebook app), Taopiaopaio (movie ticket platform), Xiami (music platform that has been shuttered), and Yulebao (film and entertainment crowd funding platform).

88VIP (8 sounds like “ba” in Chinese) is a loyalty program, that offers benefits across Alibaba’s retail businesses, media and local services. Members receive a free subscription to Youku, promotions on various services (such as Fliggy, and Taopiaopiao), and discounts on Tmall, Taobao and others Alibaba properties. They also secure discounts for other stores and international brands like Ray-Bans, Beats by Dr. Dre, GoPro, L’oreal, Marriot (automatically get gold elite status) and others that are available year round. The price is 888 RMB annually, however that drops to just 88 RMB based on engagement on the platform like answering product questions, rating items, buying items, among other factors. 88VIP members spend 9x as much as other users and 38% of people signed up for Youku for the first time through the program. A “lighter” version of 88VIP, Taobao pass, only offers the discounts on Taobao and Tmall (and ARPU increases 30% after users sign up). Between the two, last disclosed Sep. 2020, there are currently 35mn members. Frankly, we were surprised this figure wasn’t higher as Amazon Prime has much higher penetration among their userbase in the US. However, the prime comparison may not be the best analog as RMB 888 annually (or a lot of engagement for the discount) is a very high bar for the average user and this offering truly is tailored to the most loyal & fervent users of Alibaba. Nevertheless, we would like to see a broader offering that has a value prop that appeals to the average Alibaba buyer to increase consumer lock-in and better connect the Alibaba Ecosystem between retail and media.  

All in all, there are a plethora of interesting businesses here, but none of them are likely to move the needle for Alibaba anytime soon in terms of profit contribution. As you can see, they still are generating a loss of RMB ~10bn and we think this will continue to be a money losing operation as they continue to invest in more content and technology. However, even without contributing direct profits, these businesses can help drive the ecosystem with Alipay app integrations, advertising driving platform traffic, data leading to better consumer understanding, and tying Youku membership benefits to the 88VIP loyalty program.

Innovation Initiatives and Other.

This is similar to Google’s “Other Bets” segment and houses a variety of different assets, but there are only three main ones we will call out. Unexpectedly, most of these businesses are not highly monetized and this segment represents just ~1% of revenues.

AMAP is a digital map service like Google Maps where you can get directions and search for local retail and restaurants. There are nice synergies here with as you can find a local store or restaurant on the map and then place an order for it to be delivered. Additionally, they provide access to Ant Financials bike sharing service, Hello Bike. AMAP also provides traffic data to 3rd party in-vehicle navigation systems and the data is also utilized in Cainiao’s logistics operations. Baidu Maps is the biggest player here though with DAUs over 300mn versus AMAPs of ~100mn DAUs.

DingTalk is a digital workplace collaboration tool for business communication, similar to Slack or Microsoft Teams but with more features for employers to track employees. They have 300mn+ DAUs across over 15mn different companies and is also common among students. This is China’s largest intra-business communications platform, but it is fairly hated by most employees because of poor user experience and powerful monitoring tools for employers. DingTalk has a feature which allows an alert called a “ding” to be sent to a fellow employee which cannot be silenced, and messages show read receipts that cannot be turned off. Clocking into work can be done by GPS, but employees complain of it being buggy and erroneously marking them late. There is a 2nd option to clocking in, which is to send a photo, however in which case your smile will be ranked against others employees. All of these features are unsurprisingly unpopular with most employees. There are some competitors entering this space like Bytedance’s Lark and WeChat’s work with positive reception, but Dingtalk is still the biggest. Alibaba is adding more features to Dingtalk like high quality video conferencing with up to 300 video participants and looking for ways to tie it to their cloud business to push more adoption.

Tmall Genie is Alibaba’s smart, AI-powered speaker similar to Alexa. They not only sell hardware with Tmall Genie but also are expanding it to in car applications. The idea here is to reduce frictions to purchasing items, enable more smart in-home features, and to have a funnel into consumer recommendations. Baidu and Xiaomi also have competing products, and all seem to be similarly regarded by users, but Baidu’s Xiaodi was #1 by market share. It is unclear when (or if) consumers will rely more on voice in the future, but BABA is well positioned if they should.

DAMO Academy stands for Discovery, Adventure, Momentum, Outlook and is Alibaba’s “Google X”, responsible for all cutting-edge research. They intend to marry the forefront of scientific research and industry application. They are focused in areas like autonomous driving, quantum computing, 5G network applications, robotics, blockchain, and biometrics. This Academy also serves as a funnel for talent acquisition with various competitions and scientific fellowships.

We generally think of this segment as a “call option” with the belief that, despite substantial losses, at the very least BABA is not destroying value in aggregate across all of their projects. However, we acknowledge that is a possibility and a very conservative valuation may instead of just zeroing this segment out, consider it to be of negative value. We think that would be too punitive though as there is clearly some value in these businesses, albeit ambiguous.


Alibaba has an investment portfolio of many companies (public and private), which they last estimated to be worth at least $45bn in 2020 (excluding Ant Financial). Below we will only touch on a few of their larger and more notably investments which can have strategic importance to the Alibaba Ecosystem.

Ant Financial. We will do a separate write-up on Ant Financial when they IPO, but we want to mention them in the context of the Alibaba Ecosystem and also as a source of material value given their 33% ownership stake in them.

Illustration from Ant Group IPO prospectus detailing its different use cases.

Ant Financial is one of the world largest financial institutions with over 1.3bn users (including their ewallet partners) with 700mn+ MAUs, over 80mn merchants and supporting over RMB 120tn in payment volume. They first became popular on the back of Alipay which was one of the only ways to digitally buy something and the only means for many people to purchase goods online given the low penetration of debit and credit cards in China. Alipay was crucial to Taobao’s early success, but the ease of the payment mechanism was quickly adopted elsewhere, even offline. Today Alipay still has a ~55% market share with WePay by Tencent being the 2nd most popular payment mechanism. Alipay became a “super app” with many functionalities beyond simple payment embedded in their mobile app. Users can do anything from order food, ride hail, book hotels, shop, and much more. All of these functionalities are usually supported by Alibaba businesses, so the AliPay app helps funnel traffic for a variety of BABA’s consumer businesses. Of course, they also innovated beyond payment on financial services side in wealth management, micro financing and insurance. These financial services contributed over half of Ant’s revenues (with payment processing being the remainder). For many of the financial products they partner with various financial institutions to bring products to consumers in a relatively capital light way with for instance ~98% of their loans being syndicated to 3rd parties, but it looks like regulation could be changing this. A couple of their main products include Yuebao (online money market account), Huabei (buy now pay later), Jiebei (consumer lending product), Zhima Credit (credit scoring), and MyBank (online bank). We will dive into all of this much deeper when we cover Ant Financial in full.

Ant Financial attempted to IPO back in November 2020, but changes in China’s regulatory environment that would later cause a restructuring of Ant delayed it. We will not go into all of the details of it here, but the byproduct of the intervention is that growth expectations have been significantly cut and it is likely they will have to hold more capital given their lending activities. The rumored IPO valuation at the time was ~$320bn, but it has come down since. We will touch on valuation later when we go through our SOTP valuation below.

Weibo is sort of like a Chinese Twitter. It has 500mn+ MAUs, 225mn+ DAUs and generated over $1.7bn last year. Alibaba is one of the largest advertisers on here (through their TANX) and they own a ~30% stake in the company. They trade on the Nasdaq under WB with a ~$11bn market cap. It was rumored they may have to shed their media assets as a result of anti-trust, including Weibo, but the original article that purported that has since been removed and we haven’t heard anything since. BABA CEO Daniel Zhang is a director of Weibo.

Weibo home page, which shares great resemblance with its US counterpart, Twitter.

Baozun provides integrated ecommerce solutions for brands, similar to Shopify. They publicly trade under BZUN on the Nasdaq with a market cap of ~$5.5bn. Today, BABA owns 11% of the shares outstanding. We will cover them in depth in another write-up.

Bilibili in an online long form video platform. See our previous report for more details. BABA owns a ~7% share in them compared to Tencent’s ~12%.

Didi is China’s largest ride hailing service with almost ~500mn active users. It is not disclosed how much BABA owns, but it must be less than 5%. BABA CEO Daniel Zhang is also a director. 

Xiaohongshu is similar to a Pinterest with more social activity, and has also been compared to a social-commerce-native Instagram. They last raised capital at a $6bn valuation and are rumored to IPO around $10bn. They have 100mn MAUs and are known as B2K2C, whereby the K is KOL. We do not know BABA’s ownership but they co-led a $1.1bn round with Tencent at a $3bn valuation and then a subsequent round for ~$400-500mn at $6bn. We will cover them at IPO. 

Tokopedia was mentioned at length in the competition piece of our last Sea Limited piece and after the Gojek Tokopedia merger Alibaba will own ~13% of the combined company.

We only wanted to call out some of the larger investments that have clear tie-ins with the Alibaba Ecosystem, but below is a list of their most notable investment stakes.

Sum of the Parts.

In the SOTP (sum of the parts) analysis below we assign fair market value to all of Alibaba’s largest assets and see what the market implied valued of the core commerce business is. Remember, Alibaba is a company with 4 main segments, 3 of which are unprofitable (on GAAP) so any aggregated earnings amounts will be combining loss pools with earnings streams and make BABA look more expensive when looking at multiples. This is one of the reasons why BABA may be more attractive than it appears prima facie.

Cloud. We will size up Cloud’s valuation in two ways: 1) Looking at the long-term TAM and applying steady-state business and market share assumptions, discounting back to today and 2) looking 3 year forward revenues and applying market multiples to steady-state earnings. 

Revisiting our TAM analysis assumptions (US IT spend grows 4% from ~1.25tn, China reaches US parity and 30% public cloud penetration in 3 decades) implies BABA could be generating over RMB 3tn in sales from their cloud business over the super long term, more than 30x higher than what next year’s revenues estimates are. Taking these future earnings and applying a 30% mature margin and 15x mature multiple implies a ~$300bn valuation today at a 7% discount rate (nothing scientific about this rate, it’s just the approximate implied discount rate we have observed throughout our coverage and a fair required return in this interest rate environment). This seems to be on the higher side of BABA’s cloud valuation, but that is mostly because we would want a higher implied return for such a long-term holding. For an implied 10% return you would need to pay ~$125bn.

The second way we can look at sizing up cloud is just taking our 3 year-out estimates of RMB ~180bn and applying a ~6x sales multiple to this, which maps to a market multiple (~22x) on mature NOPAT at a 30% margins in 3 years. This would yield a valuation around ~$150bn. The idea here is that you are prepaying for 3 years of growth, which is fair as long as you have high confidence in BABA getting to that level of sales. The range of these valuation methods points to cloud being worth somewhere between $125-$300bn. We will pick $150bn for our SOTP analysis, but of course all excels will be available to members plus subscribers to change assumptions as they wish.

Digital Media & Entertainment. We will only count two main assets in valuing this segment, 1) Alibaba’s 51% interest in the publicly listed Alipictures and 2) Youku. To value Alipictures we will take a short-cut and simply use their market price which prices their stake at ~$2bn. Given the relative tiny size in the context of Alibaba Group’s market cap and the lack of disclosures it is not worth spending the effort to try to dive in deeper here—no result will be material. (If there is interest though in Alipictures as a stand-alone investment, we can dive deeper in a separate write-up). Youku also suffers from the same lack of disclosures, but we can piece together some information that gives us confidence in a valuation floor. 3rd party data places paid Youku subs around ~85mn, but if you follow managements commentary on sub growth–which is admittedly ambiguous–from the disclosed 1mn paid subscribers in early 2015, it implies closer to ~50mn paid subscribers. We know that at some point Youku MAUs were 500mn and while it has become a generally less popular service, especially with young Chinese, there still is a lot of traffic driven from their ecommerce tie-ins and certain hit shows, so we tend to think the MAU figure is still roughly correct with the important caveat that time spent on the platform likely fell. We cannot be precise given the spotty information, but it is typical for a lot of free services to have 10% paid penetration which comports with our lower estimate of paid subscribers. The exhibit below runs through our Youku valuation build, which gets us revenues of $2.7-4bn for Youku. Note that this implies Youku is ~60% to potentially 90%+ of all non-Alipictures digital media revenues. We would think Youku would be a majority, but whether 90%+ is too high is too hard for us to know.

We will pick a middle of the range valuation of $15bn for Youku and add the ~$2bn Alipictures stake to that for the SOTP.

Ant Financial. As noted, the original IPO valuation for Ant Financial was over $300bn, but then new estimates have ranged as low as $30bn to $150bn. While we acknowledge that the regulatory changes will weigh on Ant’s ROIC as they are going to be required to carry more capital and their growth will be hampered by the new regulatory changes. However, if we just look at Ants payments business, we see that it processes more than double the payment volume of Visa. While it doesn’t monetize this TPV as lucratively as VISA, putting any small take-rate on the ~$18tn of payments and assuming ~40% margins, you will be able to rationalize at least a $120bn value, but likely higher. We are being a little vague here on purpose given the minutiae of Ant’s different businesses and how they monetize each service, but given the sheer size of the businesses and its instrumentality to daily Chinese commerce, we are comfortable assuming a total valuation of $150bn for Ant. We will dig in deeper here in our full Ant write-up, but feel free to adjust our valuation for the conservatism you feel it requires.

Innovation Initiatives and Others. Since we are considering this section a free call option, we will not be assigning any value to it.

Investments. BABA has a slew of investments, and the market has been very rewarding for virtually anything in the tech space as of late. However, we will not attempt to adjust the value of each public and private stake to current fair value and instead will use BABA’s disclosed investment portfolio value of $45bn from September 2020. Valuations have by and large gone up since then, but for conservatism we will use that figure for the investment portfolio.

Cainiao. If you recall, both this segment and local services sit in the core commerce segment, however Cainiao is barely breaking even and Local Services is very likely still running at a loss. In order to value the mature core commerce businesses, we will back both of these out in our SOTP. BABA last invested in Cainiao at a ~$28bn valuation in 2019 Its hard to figure out what the value of Cainiao would be as a separate business given its deep integration with BABA and that it is likely under monetizing. (We cannot look to JD logistics to get a great picture either since they own their delivery fleet versus Cainiao that is mostly software). Last fiscal year revenues were ~$5.8bn and if you think that steady-state margins would be 25% (conservative) then the price BABA paid is only at a slight premium to a market multiple today. You could argue that a mostly software company with deep integrations to their customers with the ability to directly attribute cost savings should be worth much more, but we will just use $30bn for conservatism.

Local Services. In theme with other segments, Local Services is hard to value given the limited disclosures. We know that and Koubei were each valued around ~$8-9bn when they were purchased and $4bn of capital was raised for the Local Consumer Services division at a $30bn valuation from Softbank and other investors (leaving BABA with a 73% stake). reached positive unit economics in June 2020, but they could still be generating a loss at the segment level. Nevertheless, we have seen that these sorts of business can ultimately become profitable, albeit with fairly low margins (Meituan currently is at ~4% in food delivery). As they add more advertising servings and other value-added services this could likely go up. They generated just shy of $5bn last year, but they are likely under-monetizing the platform as they vie for adoption so we cannot apply a steady-state margin to that figure as it would under report the true earnings power of the platform. Having said that, it appears to us that the valuation is more predicated on the value to the Alibaba Ecosystem, rather than the direct profits that could potentially be extracted from Consumer Services. The math is tough to pencil out without believing that order frequency will go up a lot, which is very possible (especially with grocery), but if you doubled their current revenues (grew just 24% last year) and applied a 5% margin, applying a 30x multiple only yields ~$13bn. A banker would want to say that Meituan trades at ~7x sale and so a valuation of up to $42bn could be fair based on forward revenues, but we will use $13bn for the SOTP. This could be too conservative and high level, but it is unlikely to be very material. When we do our deep dive on Meituan we can revisit our valuation assumptions.

Alihealth. For Alihealth we simply apply their current market value, which is ~$30bn or 18bn for Alibaba’s 60% stake. We may revisit Alihealth more in depth in a future write-up.

Rolling it all up. Putting it all together we get Core Commerce, which includes the China Retail and Wholesale businesses (Taobao, Tmall, New Retail, 1688, Lingshoutong) and the International Businesses (Lazada, Aliexpress,, and others) valued at $225bn after backing out the business stakes mentioned above.

The table below shows that all of these businesses are worth ~$310bn compared to Alibaba’s total enterprise value today (after accounting for ~$50bn of cash) of ~535bn. Taking just last years actual Core Commerce EBIT of $25bn (RMB 159bn) and adding back amortization and taxing it at 18% gets us NOPAT of $22bn. (We choose to add back amortization as it does not represent an economic cost, but we do not back out stock-based comp). It is very important to acknowledge that there are still imbedded losses in this $22bn NOPAT figure from businesses that we backed out from the EV (Cainiao and Consumer Local Services) and also losses from businesses that are not backed out of the EV (New Retail, International Retail). This means that the implied ~10x NOPAT multiple for the Core Commerce businesses would be lower after backing out the loss pools. There is simply not enough disclosure to do this though. We also want to emphasize that these figures are on LTM revenues, and that multiple would come down further as their businesses continue to grow. (BABA’s next year guidance calls for 30% y/y growth across their entire company which would be hard to achieve without the core commerce growing at least double digits, but more likely ~30% especially with the 1P dynamic for New Retail.)

BABA’s Core Commerce businesses are trading at below 10x NOPAT after backing out newer loss-making initiatives. Based off of BABA’s 6/18/2021 close of $212/share.


We have just shown that the market seems to think Alibaba’s core commerce is a ~10x business, a multiple usually only applied to businesses with fickle earnings and zero competitive advantages: is this a fair characterization of Alibaba?  

Since they IPOed in 2014, Alibaba has grown GMV, revenues, buyers and GMV per buyer every year. Last year alone, when there was more competition than ever before, Alibaba generated >$200bn more GMV than the year prior–that is adding almost half of an Amazon in a single year. They also grew revenues 24% y/y, which means BABA’s one year revenue increase was the equivalent of PDD’s entire 2020 revenues. It’s true though that as they get bigger, their growth will naturally slow. Our growth expectations for BABA in the mid-teens (explained in the next section) may look lackluster in the context of PDD’s 200%+ revenue growth, but you are not paying for a high growth stock being priced on an EV/ Sales multiple, rather you are buying a mature company with a durable runway of growth being priced like dying one.

It’s true that competition is coming from many angles, but competition is nothing new for BABA. This is not to say competition isn’t “scary”, but it was scary when when eBay moved into China with vastly superior resources and technical talent, it was scary when JD launched a fully integrated logistics operations at a time when BABA had virtually no logistics support, and today it’s scary that Pinduoduo has more annual buyers than BABA and Bytedance has ~2bn users, high technical talent, and large ambitions in ecommerce. But despite all this, we actually view most of these new competitor products as by and large orthogonal to what the core Alibaba customer uses Taobao and Tmall for. As we have said before, there is going to be more than one “winner” in ecommerce, and we are making the case that BABA isn’t about to be “decrowned” as one of those winners, as the Alibaba Ecosystem is more formidable than appreciated. There is simply no ecommerce player today that can replace what Alibaba does today at the scale that they do it.

Alibaba has over >800mn users who spend an average of RMB, ~10,000 annually, with very high retention of 98% among their 140mn highest spenders. Almost ~40% of users are on the app daily and order an average of twice a week. The longer a consumer uses Alibaba, the more they spend, with spend per user going up every year.

Their 10mn merchants offer consumers the widest product catalog with the most information about those products that you can get anywhere. Their in-platform commerce-oriented, social network, Weitao, provides users with informative and entertaining content, with a lot of it produced by KOLs they look to for guidance and are largely absent from other platforms. If a consumer still can’t find the information they are looking for, they may directly message the merchant, who often responds in mere minutes. In fact, 40% of transactions have a messaging element to them, a feature not natural to any other commerce platform. With consumer ratings dictating where Taobao and Tmall merchants will show up in search rankings and Taobao garnering the most search-driven commerce, merchants have become highly attune to the needs of Taobao customers and often go out of their way to delight them with sample products and rapid responses. Another platform could copy all of their features but would not be able to duplicate the merchant loyalty to Taobao that Alibaba has built as it is many merchants’ most important sales channel.

Adding to the ecosystem, Alibaba is deeply integrated with China’s most popular payment mechanism, which imbeds lending capabilities, and has an extremely efficient logistics platform that can deliver anywhere in ~2/3 days. They also have the most consumer data and one of the largest 3rd party ad networks that drives traffic back to their platforms. All of their new initiatives from New Retail and local delivery to AR will also only help better entrench them.

In a world of capricious consumer tastes, ever changing shopping formats, and increasing expectations all around, BABA has innovated out ahead of paradigm shifts in the consumer landscape whether it be from desktop to mobile, B2B to B2C, or “hands-off” marketplace to fully integrated ecommerce provide. Peculiar enough, the other most common investor concern apart from competition, is that they will get regulated as a monopoly. Truthfully, we would be naïve to think we know what form regulation will take, but there is no evidence today to suggest a break-up or any draconian restrictions on BABA’s ability to operate will be implemented (we touch on this more in the regulatory section below).

So, rolling all of that up together, we have the worlds largest ecommerce platform that is highly entrenched in user behaviors with spend going up every year, more merchants who are devoted to making sales on this platform than anywhere else, and an ecosystem supported by China’s largest payment method, the most consumer data, a highly scaled logistics platform, and a huge 3rd party ad network, all being priced like the business is in secular decline. We will explore more on business expectations and valuation below, but our general thoughts on Alibaba are simple: we don’t think their future competitive position will be that different from their current competitive position. 

Commerce Revenue Build.

We have just explored how much the market is valuing BABA’s core commerce businesses and gone over our qualitative assessment, but below we talk about our expectations for the business. As we have already explored valuing the bigger pieces of Alibaba Group we will focus just on their Core Commerce business here. Additionally, we will only concentrate on the China business here as international is less than 10% of revenues.

As BABA is the largest ecommerce player and a sizeable portion of the Chinese economy, future growth will converge more with macro-trends, so we will touch on how we think about that. Very simply, we see three key drivers to revenue growth 1) buyers increasing on population growth and penetration, 2) disposable income growth increasing total consumer spending, and 3) monetization rates increasing.

1) The first piece of growth is BABA increasing their number of buyers from population growth and increasing penetration. Population growth is minimal right now at ~50bps annually, but there are some signs that this could be changing with new policies. The bigger piece of this will be more Chinese getting online: today about ~400mn people do not have access to the internet. Buyers grew 12% last year as they added 85mn new buyers and they currently count ~85% of internet users as buyers. Continuing this level of penetration means they will add 340mn (85% of 400mn) new buyers when everyone in China has access to internet. If you assume these users come on over 15 years, that equates to ~2.5% annual growth in buyers.  

2) On our second point, we size up how much discretionary income (money after essentials) will grow ahead of disposable income (total income after taxes).  According to a McKinsey Study, roughly 40% of average Chinese income is discretionary. China’s GDP is growing ~6% and we believe that is a fair proxy for discretionary income growth. So, taking 6% discretionary income growth and assuming incremental income is ~100% disposable, equates to 15% disposable income growth (6% / 40%). This means the pool of money that can be spent on BABA’s services is expanding at a mid-teens rate, though it should be noted BABA also has offerings that cater to the “essential” spending piece, like grocery and medicine, where they will also benefit.

The next piece of this is how does 15% disposable income growth translate to BABA’s GMV? We believe that incremental growth in spending will slightly fall on BABA’s platforms as new places to spend money (PDD, Douyin, Kuaishou, ect.) continue to proliferate. However, we still expect BABA to continue to grow. Recall there is also a tailwind from buyer spend increasing the longer they are on the platform. While GMV growth going forward will be less than historical (5-year CAGR of 21%), we think a high single digit growth figure over the long term is reasonable.

3) The last piece to BABA’s China Commerce growth is monetization rate: the % of GMV they turn into revenues. Today that stands at 3.8% and has been increasing fairly consistently the last 5 years at a ~8% CAGR. We believe that this is going to continue to be a powerful driver of revenues in the future and do not believe they will hit a ceiling anytime soon. As mentioned before, other ecommerce platforms like Amazon can take a direct cut of the sales in the mid-teens and then receive advertising fees on top of that (which is similar to Mercado Libre). While we do not think their monetization rate will necessarily reach parity with global peers, there shouldn’t be a natural limit anytime soon given their relative low cut of the transaction and importance in demand generation and helping facilitate the entire chain of commerce. Furthermore, simply improving efficacy of advertising with more data and refined algorithms will result in more advertising revenues; an incremental innovation process that can play out for decades. Lastly, new innovations like AR can create new opportunities to charge for value added services, expanding the growth runway over the long term. 

If you roll these three drivers up, and err on the side of conservatism, you can reach a long-term growth rate of ~15%, as shown below. However, even hair-cutting these numbers substantially still yields a solid growth rate in the context of this business trading <10x TTM earnings. As a general rule of thumb, businesses trading under 10x tend to be either very cyclical or melting ice cubes, neither of which we believe describes BABA’s Core China Commerce business. Below we will dive into the cost-side of the commerce business. The concern that margins will go lower is well founded and correct in our view, but does not change this conclusion.

Margin & EPS.

There are a lot of dynamics that play into their margins with consolidations, new 1P businesses, acquisitions and increasing investments all convoluting estimates. Instead of trying to granularly account for all of this (and fooling ourselves into thinking we can know all of these variables), we rather focus on what we believe with high confidence will be a reasonable bad case scenario to allow an investor to see if they can find that an acceptable outcome.

For starters, core commerce EBITA margins have dropped every year from 57% in 2017 to 27% last year. There are several pieces to this, one of the larger reasons being acquiring or launching several money losing businesses and second mix-shift towards lower margin 1P businesses which comprise ~23% of revenues last year.

Management noted that all incremental profits will be reinvested as they see investment opportunities and want to fend off competitors. Gross Margins were 41% last year, so we exclude the impact of the 1P business (China Retail Other) and assume gross margin contracts 2 points on the remaining revenue and then layer back in 1P cost of revenues at an assumed 25% gross margin. This yields RMB ~600bn of expenses (36% total gross margin), however contributions from cloud and other loss pools shrinking should offset some of this, but we will ignore that to make the point that despite adding RMB 175bn in cost of revenues, gross profits still increase RMB ~35bn.

We have operating expenses growing ~20%, which comes out to total flat operating income y/y or RMB ~110bn in total. The following years though we do not assume further gross margin contraction after this investment step-up, but we do not assume margins increase on lower investment levels either. We expect continued operating leverage on the G&A and R&D lines, as we’ve observed every year (excluding the monopoly fine). If you zero out the equity investments income and other investment gains to get a core EPS, then we are arriving at ~$5.35 per ADS for 2022E, which grows at a ~18% CAGR to $12.50 by 2026. This means BABA trades at 36x forward earnings, which does not sound too attractive, however, it is critical to remember that these GAAP figures are not representative of the true earnings power of the businesses given the consolidation of loss-generating businesses. We wanted to make the point though that despite increases in investment, earnings will grow and over time we expect that for reported earnings to converge with their true earnings power as growth investment ameliorates and loss pools turn to profitable.

Summary Model.

Our summary model above is presented to give readers a sense of how different line items flow through the P&L. As always, Members Plus Subscribers will have access to the underlying excels and can change any assumptions they wish. We do not usually have an opinion on how each specific line item will appear in each specific years, but the model format forces us to estimate granular details that we cannot possibly know. We zeroed out investment income and equity method income for lack of confidence in estimating given disclosures, but this will prove to be too conservative.

Valuation Conclusion.

If you put the two pieces of our analysis together, the SOTP valuation and our earnings expectations, there are two takeaways we want to emphasis: 1) Consolidated figures are misleading and the core business trades at <10x LTM NOPAT, 2) Earnings are still growing despite slowing revenues and increased costs. If you simply back out the loss-making stakes from the EV, but not their associated losses from earnings, it trades at ~16x 2022E, which drops to ~6x by 2026. With market multiples sitting >20x, it is hard to rationalize why a very dominant business that is growing faster than the market should be priced at a below market multiple. Current valuation gap aside, it our belief that BABA will continue to grow their intrinsic value overtime. However, it is an investor’s decision to judge where we were being potentially too dismissive of competition, over-estimating managements’ competence, or not giving proper credence to a slew of other risk factors, some of which we will dive into below. There is no investment without some risks of course, and it is the investor’s judgement of those risks and whether they are properly being priced that matters.


Below we identify several potential key risks.

1) Competition. Competition is able to gain a higher share of consumer spending and consumers spend less on Alibaba’s properties as competitors offer better experiences with similar selection, price and “good enough” shipping. New users grow up on different platforms and shop differently, using Taobao and Tmall less. BABA consumers switch to PDD or JD for more intentional purchases and Douyin, Kuaishou and other social platforms will take over a majority of a consumer’s discovery-based commerce through direct response ads or KOL recommendations. Furthermore, as users become used to purchasing on Douyin/ Kuaishou they can start their purchase journeys there with search (a recent focus of Douyin). Lastly, Meituan keeps expanding selection on local commerce and more purchasing shifts to them. BABA loses not just market share, but GMV falls for the first time.

2) Brands go D2C. As more brands become mature in China, they can push D2C efforts which cut BABA out of the sales process entirely. Perhaps brands start leaving Tmall to focus on owning the consumer relationship and emphasis direct response advertising on apps like Douyin and Bilibili to find customers. Tmall’s value prop as an aggregator of merchants with wide selection falls, which results in fewer customer visits and even more merchants leaving. Tmall loses revenue from fewer sales to take a commission on and also ad inventory dropping.

3) Better advertising elsewhere. Monetization Rates do not increase as the price of ads are already very expensive and other advertising options present a much better return. This will weigh on revenue growth and at the extreme could have merchants shift ad purchasing elsewhere. 

4) Share Data. Alibaba is forced to share data with competitors, making their advertising offerings less attractive as their first party data gives them an advantage in personalizing ads.

5) Ant Financial stops favoring Alibaba. Ant Financial stops supporting the Alibaba ecosystem and no longer helps send traffic to Alibaba businesses. Ant Financial decides to charge Alibaba a higher fee for their services and BABA loses their favorable treatment.

6) Losses never stop. Loss generating businesses never turn profitable. Management continues to invest in these money-losing businesses despite no path to shareholder value creation.

7) Poor capital allocation. Management makes bad acquisitions that destroy value and spend on investments with bad ROI. Due to the complicated financial statements, it would be hard to judge if an investment was poor.

8) Regulation. More Anti-trust charges or other regulation from China hampers BABA’s ability to compete. At the extreme, BABA is forced to divest assets at prices that are considered unfair.

9) Trade restrictions or tariffs. Restrictions on international trade can hurt BABA’s international businesses and domestic businesses that sell international products. Another trade war could result in higher prices and thus dent consumer demand.

10) Accounting Fraud. International companies have less stringent auditing processes than US ones, which when combined with very opaque financial statements and corporate structure, gives life to the opportunity to defraud investors.

11) VIE / Corporate Governance. As evidenced by the Ant Financial/ Alipay debacle whereby shareholders lost their full interest in the business, VIE’s tenuous claims on the underlying assets. There is ample opportunity to move assets at will and shareholders have little or no recourse. Investors do not have the same voice in corporate matters either.

Stock-based Compensation.

It is worth touching on their SBC program. From Alibaba’s founding with Jack Ma making all 17 of his coworkers founders with equity stakes, BABA has always been very generous with equity shares. Today, that takes the form of a $7.6bn (RMB 50bn) expense, which is large but not egregious in the context of their market cap. This equates to 1-2% annual dilution. Remember, in all of our earnings analyses we do not back out stock-comp and we treat it as if it was a cash expense. However, it is a significant portion of free cash flow in the ~25-30% range.

Free Cash Flow.

Below we show a condensed cash flow statement. Stock-based comp is a material source of cash flow, however dilution in total has run <2% on average. Its also worth mentioning that free cash flow conversion is often over 100%, giving BABA high amounts of internally generated cash to grow.

On a FCF per share basis, BABA is trading at ~21x, but we prefer valuing them on the metrics we dove into above. However, it is good to see cash flow-based valuation support.

Regulation and Anti-Trust.

Alibaba was fined RMB 18.2bn ($2.7bn) last quarter by China’s State Administration for Market Regulation for violating Anti-monopoly Law. This fine had to do with the “2 choose 1 policy” of forcing merchants to restrict selling activity to just their platform or risk being removed. While perhaps this policy could have helped BABA earlier on, today they have enough platform advantages that such a policy would be almost superfluous. However, it is worth being mindful that very harsh competitive tactics will likely no longer be accepted by the Chinese government. This can actually be a positive for BABA in certain areas.

Meituan was hit with a fine for similar seller restrictions, whereby restaurants had to choose between them or Given Meituan’s larger user base and thus ability to generate more traffic, most restaurants choose Meituan. However, with this restriction now deemed illegal, more restaurants are likely to join given that restaurants are generally agnostic to where consumer orders are coming from with the exception that they would probably prefer less single-channel reliance. Both of which bodes well for onboarding more restaurants, thus increasing selection. 

In an unexpected turn of events, Alibaba also will benefit from Tencent’s WeChat no longer restricting mini programs from companies they consider competitors, however they are still blocking Alibaba links. Taobao Deals launched its mini-program, at the beginning of the year, but it was removed shortly after, so we think Tencent is still playing games to delay having to host BABA apps, given their investments in PDD and JD. Given that PDD user growth was essentially grown off of sharing on WeChat, BABA’s Taobao Deals can eventually run the same playbook and leverage WeChat traffic to cheaply acquire users. Interestingly, Taobao Deals will likely accept WePay, but we haven’t heard any push for WePay to be added onto Taobao, with Alipay still by far the most popular payment method on Taobao/ Tmall (over 70% of GMV). There could be downside for Alipay if WeChat is allowed across all of BABA’s platforms, however Alipay ties to many services and financial products which is likely to make churn minimal. Nevertheless, it does reduce the value prop and most users would prefer one wallet and P2P payments is a highly valued function that works much better with WeChat. This is a risk worth monitoring.

When attempting to send a Taobao link on WeChat, the link is converted to a text message with emojis and code that needs to be manually inputted into the Taobao App.

Community Group Buying is also becoming an area of focus for consumers. While CGB is good for consumers and manufacturers, cutting out the middleman means millions of Chinese resellers will lose their livelihood. For now, it seems only excessive subsidies are under fire, but this could be an area of further action. PDD and Meituan are more at risk here than Alibaba.

As mentioned in the Ant Financial section, Ant has had to restructure, and they are going to be more regulated going forward. Lending is going to be curbed, which on margin could dampen some consumer spending, but we do not view it as material. It has also been reported that their credit scoring business might be put in a JV with a SOE and they may no longer control the proprietary scoring data. The credit scores, if broadly provided to competitors could theoretically allow others to compete in lending more effectively, but Alipay’s captive user base and integrations are a strong competitive advantage. This would clearly not be a positive development for them though. The most direct effect of these actions on Alibaba for now is that their stake is worth less, but we do not see these regulatory changes as being a critical threat to the Alibaba ecosystem yet.

While to be frank, it is always hard to judge how regulation or future legislative action will adversely impact a business, we generally take confidence in the fact that Alibaba supports a lot of merchant’s livelihoods, generates a significant amount of economic activity, and is well liked by the consumer. Generally speaking, such businesses provide too much value to society for there to be an appetite to unfairly cripple them.

Management and Culture.

Today Alibaba’s management team is large, which is not surprising given they started with an unheard of amount of cofounders. Many of the early cofounders and executives brought in in the early days are still there. Jack Ma was of course the visionary founder that could inspire the team to execute technically difficult things despite his lack of domain expertise. While he stepped down as CEO in 2013 and as Chairman in 2019, his unbridled ambition continues to permeate BABA.

Daniel Zhang is the current CEO and has not only overseen phenomenal growth since 2015, but also led BABA in creating and acquiring several new businesses. Some investors have compared the Jack Ma/ Daniel Zhang dynamic to Steve Jobs/ Tim Cook. In our conversations with employees, he seemed generally well liked, but some of the unhinged determination from Jack Ma was missed. Joe Tsai still remains somewhat involved and is a good consigliere as he is known for having vetoed several of Jack Ma’s ideas (not known which) to focus the company on their current opportunities. The general consensus on management from our conversations is that management is fairly vanilla, but very dedicated to Alibaba and focused on executing growing the businesses. Since most of them have been with the company for a long time, there is a sense of camaraderie, but also a lack of innovation with perhaps managers becoming a little too comfortable.

As far as employee talent goes, Alibaba is still considered a top tier employer, however, it is no longer the most exciting place to work (that is Bytedance). Alibaba employees are talented though, evidenced by a disproportionate amount of VC funding going to ex-Alibaba employees, which can be considered a positive or negative. As far as employee experience goes, we have heard that Alibaba has an “extreme” culture that demands a lot of buy-in and can verge on being cult-like. Its normal to be sitting in a meeting with employees chanting the Alibaba slogan, “Make it easy to do business anywhere” and on Ali Day (Alibaba’s “Thanksgiving”), Jack Ma would preside of employee weddings. Even long-time ex-employees are Alibaba evangelists for all of their products and refuse to use competitors. Most employees tend to stick around for many years and opportunities for upward mobility internally are abundant for talented employees.


Thank you to all of our subscribers for supporting us! Believe it or not there is still a lot more on Alibaba that we have left out! We will periodically be releasing more content in our Details pieces and future write-ups will dive into some more pieces of their businesses, but let us know if there is anything you wish to see more analysis on. Thanks again!

*Disclaimer: I have a position in Alibaba and people who have helped craft this analysis may individually (or through funds they work at) have a position in the company. Absolutely nothing in this report is investment advice nor should it be construed as such. We make no claims to the veracity of all facts and figures presented in this report. Certain figures could be stated erroneously and certain analysis could be incorrect. Please see the full disclaimer linked above.

DJY Research, its author(s), any of its contributors, and any funds they may advise at may own securities in the companies, or related companies, written about on this website.