Sea Limited

Edit: Please check out our most recent deep dive, a 24,000 word report on Alibaba!

Hi! Thanks for checking us out at DJY Research. Every month, we will do a deep dive on an Asian equity. If you like our research, please subscribe. This is our second report and final free sample. Next month, only those who become Members will have access. Those who join Members Plus will also be able to download our excel worksheets.

*If you are not ready to subscribe now, drop your email below to stay in touch and get free updates.

This month, we will be doing a deep dive on Sea Limited.

*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.

Sea Limited

TLDR:

Sea Limited is the premier internet company of the greater South East Asian region. Its founding is rooted in gaming services with Garena, but has since expanded into digital financial services and eCommerce. First, we briefly go through Sea’s background and provide an overview of each of Sea’s businesses. We then dive into how we think about analyzing and valuing each of its segments.

*Throughout this report, we will use “Sea” to refer to the company and “SEA” to refer to the South East Asia region. Typically, South East Asia refers to Indonesia, Malaysia, The Philippines, Singapore, Thailand, and Vietnam. However, Sea also includes Taiwan as part of SEA.

**In preparing our report we have talked to various industry experts, investors, and perhaps most importantly, ~200 users of various Sea products across every South East Asian country. 

Enthusiastic eSports fans at a Garena sponsored event.

Background:

Singapore-based Garena, the successor company to a shaky eSports social gaming network dubbed “GG”, was founded in 2009 by Forrest Li. Their first product, Garena+, was a revamped platform for online game discovery, distribution, and play with a social overlay that connected the gaming community with friend lists, discussion boards and messaging. Initially, they focused on internet cafes, where their high penetration was hard to match, giving them a clear distribution advantage for game publishers on Garena. Its success with gamers in SEA quickly drew the recognition of Chinese gaming giant Tencent, who identified Garena as a key partner to help expand their reach beyond China given their wide existing distribution. In 2010, Tencent made a significant investment in Garena and gave them preferential access to their portfolio of games. Distribution rights to wildly popular games like League of Legends helped cement Garena’s position as the largest gaming distribution platform in SEA at a time when the PC gaming market was by far the largest end market. 

The home screen of Garena’s desktop app, where mostly third party games are showcased. Scroll down to see some screenshots from the mobile version.

With the popularization of the smartphone and cheaper data plans, the opportunity with mobile gaming was starting to ripen. However, as the number of potential gamers increased to everyone with a smartphone, the amount of people who could pay for gaming lagged as most transactions were still done in the physical world, with cash. In 2014, Garena launched AirPay to help facilitate payments for their gaming services in these underbanked markets. AirPay was unique as it closely tied the ability to digitally transact with a network of tens of thousands of physical counters that allow the swapping of hard cash for digital credits, giving users access to the internet economy. For Garena, while this initially meant a larger portion of people could pay for their services and circumvent payment facilitator fees, AirPay quickly became an ecosystem on its own (more on this later).

Partnering with Tencent meant that Garena had access to games like Arena of Valor, which solidified their foothold in mobile gaming in addition to PC. However, Garena wanted to own their IP. In 2017, they acquired 111 Dots Studios, a Vietnamese gaming studio that had Free Fire, a battle royale style game in beta, and brought it to the world. In designing the game, it was purposefully made to work well on lower quality smartphones, and the app size was kept small with data limitations and costs in mind. This, coupled with Garena’s gamer mindshare and hyperlocalization with skins, led to Free Fire quickly becoming one of the most popular games on the Google Play Store and a significant source of cash flow for Garena.

Garena earned the rights to distribute the popular League of Legends game early on. Gameplay shown above.

Rewinding back a few years, similar to how AirPay was created to solve a clear market need, eCommerce (and retail in general) was very undeveloped compared to other countries. Given Garena’s desire to be one of the greatest internet companies in SEA, they were undeterred by the general lack of overlap with their current businesses and created a mobile-first eCommerce platform in 2015 dubbed Shopee, launching in each of their seven markets simultaneously (!). While there were several competitors already operating in those markets, most notably the Alibaba majority-owned Lazada and several home-grown local operators, Garena was betting that laser-focused product localization, frictionless seller onboarding, and payment integration would lead to the best customer offering. Their bet paid off as they quickly climbed into the Top 3 eCommerce provider in all of their markets.

In April 2017, Garena changed their name to Sea Limited as they now had two separate businesses beyond Garena. At the end of the year, they IPOed on the NYSE, raising over $880mn at a ~$7bn valuation. Over the next few years they continued to make strides in eCommerce, Airpay (segment later dubbed SeaMoney) and gaming. With a sense of Sea’s background and history in mind, we will go more in-depth into each of Sea’s businesses and how they monetize starting with gaming.

Monetization:

We split up Sea into three segments: 1) Digital Entertainment which is the Garena gaming distribution platform and their developed games (namely Free Fire), 2) eCommerce which is the Shopee App and 3) Digital Financial Services which includes AirPay and other financial services that sit under the SeaMoney umbrella.

1) Digital Entertainment

Here you can get a sense of what it’s like to use the Garena mobile app. These screenshots are from the Thai version. In order, they are: 1) Home Feed, 2) Home Forum, 3) Home News, 4) Games, 5) Streaming, 6) Profile.

This is a slide taken from an old investor presentation, explaining to investors their feedback loop of exclusive content driving a captive user base, whose spending and loyalty drives further developer partnerships.

Sea primarily profits from taking a cut of all revenues generated through distributing publishers’ games on their platform, which ranges from 65-80% of gross billings* (see accounting note below for definition). On games they developed, like Free Fire, they keep 100% of the economics. Publishers are willing to pay so much because Garena has aggregated the largest user base and increased their engagement with tied-in social aspects like friends lists and group chats, as well as eSports events (their events get more than 250k people attending them with tens of millions more viewing online) which together makes it hard for a standalone publisher be outside the Garena ecosystem. Additionally, given the infancy of financial service penetration, these publishers would have a hard time monetizing these users anyway (although this has changed in the past few years). As SEA starts to mature, the economics for a standalone publisher of a franchise to circumvent Garena might make more sense, especially if they have strong IP. However, it should be noted that the 65-80% of gross billings is optically higher than what Garena’s true take rate is as it includes app store fees (30% of billings) and payment fees, which a publisher would have had to incur in distribution anyway. In 2019, Tencent subsidiary Riot Games, the publisher of the League of Legends, announced that Garena would no longer have publishing rights to their new and upcoming games, as they would start self-publishing through their own channels. Over time, if more publishers go direct, Garena could become more irrelevant. Compounding their issues is a proliferation of messaging apps and game discovery channels on mobile, which has led to fewer gamers using Garena chatting functions (almost no one we spoke to used, or even knew of, Garena). The desire to develop proprietary IP is firmly rooted in this fear of increased irrelevance. But as it stands, this will not be that material to Sea financially. For one, Sea has retained the rights to still carry the largest game, League of Legends, on their platform. And secondly, the majority of Sea’s digital entertainment revenues have been coming a single game, Free Fire (which they fully own).

Photo from a Garena eSports event.

Free Fire puts Sea’s gaming franchise in an immensely stronger position as they now have proprietary IP which makes their gaming revenues immune from fickle partnerships. As shown below, Free Fire reinvigorated gaming growth and has been a significant driver of revenues the past two years. In 2020, Free Fire was the most downloaded mobile app globally and is the highest grossing app in both South East Asia and Latin America, a position it has maintained for the past 6 quarters.

*Accounting Note: “Billings” are used throughout their press releases and is basically revenue plus deferred revenue or can be thought of as the cash received for a service which may not have been fully provided yet. When buying digital items, like clothes for a character, Sea must amortize the amount received for that item over the expected useful life. In reality, it is hard to accurately figure out how long each user will consider each item “useful.” However in aggregate, their assumption of 9-12 months is probably close. This also means that billings will be a leading indicator for revenues.

While Free Fire is clearly a knockout success, having the majority of revenues predicated on a single game seems like a precarious position. However, a popular gaming franchise is much more durable than one might think. Take Activision’s Call of Duty, for example, which is on its 18th (?) iteration of the game, not to mention the many other versions, mobile editions, and add-ons released as well. Perhaps lasting franchises are harder to create on mobile given the more casual nature of play, but Candy Crush is almost a decade old and still attracts an estimated 250mn+ MAUs. It’s true that few games will reach that level of permanence, but Free Fire is on the right track with 72 billion (yes, billions; no, not a typo) views on YouTube for Free Fire related-content (most watched game globally) and Free Fire eSports tournaments accumulating over 170mn views to date.

Strong franchises can yield many years of popular games. Pictured above are the many iterations of Call of Duty, with the original having been launched in 2003 (pictured in the top left).

Interestingly enough, as the Garena platform becomes more mature with less relevance in a mobile focused world, Sea has a new social gaming opportunity with “BOOYAH!”. This is an app that was originally created to host short-form Free Fire content, but has since expanded to other games, memes and now even hosts live content. The maturing of this app could help them recreate a “mobile-first” version of Garena with social integration, video sharing, and live game streaming. BOOYAH! is surprisingly popular with 1.1 million ratings on the Google Play store and an average rating of 4.6 stars. On its own, this could not only be monetized over time, but will also help strengthen the Free Fire community.

Above, we show how QAUs have reached >600mn users and that paying users have meaningfully increased, almost 10x in 4 years, although average revenue per paying user has dropped almost in half. This makes sense given that Free Fire is a free to play game that upsells users with virtual item sales (this style is called freemium). Over time, if paid penetration continues to increase, it is likely that the average revenue per paying user drops as the most ardent gamers tend to adopt first, but the number of new paid users more than makes up for it.

2) eCommerce

Here you can get a sense of what it’s like to use the Shopee mobile app. These screenshots are from the Indonesian version. In order, they are: 1) Home Page, 2) Shopee Feed, 3) Flash Sales, 4) Live Streaming, 5) Notifications, 6) Live Chat. More on these under “Shopee Product Experience”

Shopee is Sea’s eCommerce platform that connects merchants and buyers. Originally, they were modeled as other pure marketplaces like eBay: in other words, they hold zero inventory and do not own logistics or fulfillment. Rather, their initial go-to market strategy was to partner with local logistics infrastructure and integrate their payment mechanism into the platform to make selling relatively seamless for merchants. This capital light approach to building out the network allowed them to move quickly. Furthermore, they initially focused on onboarding merchants with clothing and other high margin products that are light (cheaper shipping), low urgency (few people care if it takes till the end of the week for their blouse to arrive), and where “fakes” are not important (an unbranded red blouse can’t really be “fake”).

Slide from an old investor relations deck that shows network effects and where they will focus to differentiate themselves for buyers and sellers (although many of these ideas are commoditized, what matters is execution).

Enabling merchants to sell their products on the Shopee platform is referred to as 3rd party merchants or “3P” and Sea primarily monetizes by taking a % of the sale, dubbed a “take rate”. Take rates initially have been kept low or at zero to incentivize merchants to add inventory to their platform. Currently, take rates vary by country and seller category but the aggregated average across all verticals and countries is ~5%. This compares to more mature marketplaces like Amazon that charge up to ~15% depending on the product. Additionally, Sea can also monetize with other seller services over time, the most material of which will be advertising. For instance, when a buyer searches on Amazon for a product, merchants can bid on keywords to be displayed as a “sponsored listing” higher up on the search result page, which increases product visibility. For Amazon, this is a non-trivial source of revenues and is estimated to be around 3-5% of total GMV (gross merchandise value). Importantly, advertising income is virtually 100% incremental margin and necessary for these marketplaces to show real profitability (more details below). Shopee has already started offering advertising under seller services, but it is still very nascent.

As Shopee matures, they have been moving into more asset heavy services like fulfillment, logistics and first party sales, or “1P”. Sea has been ramping up “1P” offerings, which in contrast with “3P” is capital intensive and requires Shopee become the “merchant”, which denotes inventory risk and accounting implications. This is a critical distinction as being a “merchant” in 1P greatly distorts the P&L and eCommerce growth. Other analyses that do not delineate between the two are highly misleading and hide the underlying profitability of the core 3P marketplace.

*Accounting Note: The exhibit above shows the different accounting impact of 1P vs 3P. In a 1P model Sea acts as a “Principal” versus an “Agent” in a 3P model. Note that the absolute revenue recorded is higher under a 1P model than a 3P model, but the gross margin is much lower. In 1P Sea records the full cost of the item as revenue and the associated inventory costs, versus in a 3P model where they just record net commissions as revenues. Also note that we are showing that dollars of gross profit is higher under a 1P model than 3P. While generally true (this is effectively the return on the higher asset intensity), 1P gross profit is dependent on the strategy.

There are two lines of thinking that we’ll dub the Mercado Libre Model and the Amazon Model. Mercado Libre is a LATAM based eCommerce company and they use 1P primarily to fill out their inventory selection for items that merchants do not want to carry because the profit margins and dollars are too small to deal with the hassle. Mercado uses 1P thus not to make more gross profit necessarily, but rather to make their platform a more compelling service, driving user frequency and blocking the possibility of the buyer going elsewhere. The Amazon Model uses 1P sales as a gross profit driver (higher $, lower margin), so Amazon will sell unbranded goods or create their own label of goods that compete with other offerings in order to receive a higher portion of the economics. Note that with Amazon 1P, even though they will make more gross profit dollars than selling a 3P equivalent, it is still often at a lower price to the end consumer as Amazon will be able to flex its bulk purchasing power to receive better pricing at the risk of inventory becoming obsolete and incurring inventory holding costs in the process. While in reality, both Amazon and Mercado do both strategies to an extent, we wanted to make this distinction to exhibit the different strategies. Given the miserly gross margin reported on “Sales of Goods” (1P), it looks like they are taking the Mercado approach for now, which investor relations confirmed in our correspondence.

There are other cash flow impacts to be aware of: under a 3P model, Sea receives cash up front and then distributes it to the seller afterwards on a delayed basis, whereas under a 1P model, there is an initial cash outlay to buy the inventory (of course they could finance this, but that bears costs). However, it should be noted that cash owed to sellers is often restricted and thus not a true source of working capital.

1P has been growing as a % of total revenues to a quarter of revenue, but still represents only a tiny portion of GMV.

As shown above, GMV reached $35bn last year, +101% y/y, and is run-rating at $40bn+. 3P revenues and 1P sales almost tripled.  Gross orders have similarly been growing at a rapid pace, to over 2.8bn in 2020, +133% y/y. Looking at the consolidated financials, they report that the “eCommerce & other” segment reported a ~2% gross margin, which is hiding the profitability of their 3P marketplace.

In our analysis below, we peg the profitability of the Shopee marketplace (3P) closer to ~10% with incremental margins of ~40%, an improvement from running at a loss the previous year. While marketplaces are inherently high gross margin business with little incremental costs as they scale up, Shopee was purposely under-earning with 0% take-rates in some geographies to incentivize new sellers to onboard, including subsidizing shipping. Peer pure 3P marketplace businesses can run gross margins in the vicinity of 50%-80% depending on logistics and fulfillment involvement. Now, blended take-rates average around 5%, but will likely go up more over time. However, it is also possible that given the higher competitive intensity of SEA that take-rates never reach Amazon’s blended ~15%. In order to outflank competition, Sea has been ramping up spending on marketing to attract buyers and sellers as quick as possible, spending $1.2bn last year. Because of the network effects inherent in a marketplace—the one with the most buyers will draw the most sellers, which in turn draws more buyers—becoming vastly larger than any competitor is the only way to cement their leading position (we caveat this later).

Backing out 1P and disclosed operating items, we can see that the core 3P marketplace is more profitable than the reported eCommerce segment shows. Shopee should enjoy a lot of operating leverage over time as take-rates increase and incremental sales incur very low costs.

Shopee Product Experience.

Shopee has increased engagement with users by gamifying its platform and introducing social functions through 1) games, 2) live streaming, 3) Shopee feed, 4) Live chat, and 5) group buying).

1) Games are in-app mini games that earn you ShopeeCoins used in platform for discounts. For example, ShopeeFarm (think Farmville or Animal Crossing) allows you to tend to friends’ farms and earn rewards together, driving frequency as it requires continuous app log ins and rewards that must be spent on Shopee.

2) Live streaming which is content creators showing off products live (similar to a QVC) is a decent idea in light of success of live streaming in China (check out our Bilibili piece if you’re interested in this), but has had poor usage with our channel checks and most streams show <100 viewers.

This is the live streaming piece of the Shopee app. In the first photo, you can see the various channels to pick from. In the second and third photos, you can get a sense of what it’s like to watch/participate in a live auction.

3) Shopee Feed is essentially a commerce-focused Instagram that allows merchants to showcase products, provide vouchers, get insights on their posts, and engage with potential customers through posts, stories, and comments. Shoppers are able to follow businesses, share what they’re buying & liking with friends, discover new items, and see & leave product reviews.

This is the Shopee Feed. As you can see, it is designed very similarly to Instagram.

4) Live chat enables buyers to talk to merchants (in 2019, over a million messages were exchanged each hour) with their own translation engine ranked better than leading 3rd party APIs. This helps buyers build confidence in their purchases and decreases the chance of negative surprises (which leads to churn).

5) Group buying is similar to Pinduoduo where lower prices are offered to larger groups. All of these functions help better entrench customer usage on Shopee, making it harder for a competitor to displace.

These screenshots teach Shopee users how to take advantage of group buying deals. Scroll through them to get a sense of how the process works.

*If there is interest, we may do individual product deep dives or walkthroughs.

eCommerce progress and competition.

2020 was a big year for Shopee, as they jumped ahead of the competition and took the #1 or #2 market position in all seven of the major geographies that they operate in. Most notably, Lazada, the Alibaba majority-controlled eCommerce business that was hailed as the most likely to “win” eCommerce in the South East Asia region just a few years ago, has been significantly tripped up and outcompeted by Sea. Bloated management and offerings that were not tailored to the local markets (used the China playbook with management dropped in from China) were to blame. Contrast that with Sea, whose hyperlocalized offerings were a big selling point with buyers. For example, Tokopedia was out promoting their eCommerce offerings with BTS, a Korean boy band in Indonesia, whereas Shopee outflanked them by renting the adjacent billboard and plastering it with local Indonesian celebrities. Apart from hyperlocalizing, Shopee also simply out-executed their competitors. Channel checks revealed that before Tokopedia launched livestream shopping, they studied the market for an entire year. On the other hand, Shopee immediately launched the product and began learning. This “move fast, break things” approach has allowed them to separate themselves from the crowd. It should be noted Alibaba has shifted focus from expanding their empire, with ventures like Lazada, to defending their empire from domestic competitors like Pinduoduo and Meituan who are increasingly moving onto their turf (perhaps a topic for next month…).

Shopee has a separate website for each country that it operates in, and this exhibit showcases what the front page menu looks like – notice how while the formatting and number of menu items are consistent, no two are exactly the same in order or in categories.

Sea’s real competition at this point are the super app companies like GoJek and Grab, who can use their dominant position in delivery to cross sell their captive audience, effectively acquiring their customers for zero CAC. While their efforts in eCommerce are nascent (especially Grab), the opportunity is there, especially to partner or acquire. Additionally, there is a slew of local grown tech companies that offer an experience tailored to their native audience like Blibli, Bukalapak, Tiki, Qoo10, Sendo, Zalora, and many others. The best of breed in this group is Tokopedia, a decade old Indonesian based eCommerce company that is 2nd only to Shopee in their home country. Interestingly, GoJek is reported to be completing a merger with Tokopedia, which would create a very formidable competitor. Coupang, the leading eCommerce company in Korea and an incredibly well-run organization announced they plan to expand internationally into Singapore, which could be a foothold into a much broader SEA strategy. Additionally, JD.com is also making a concentrated SEA push with JD Central focused in Thailand now.

This chart has been widely circulated online as “proof” that Shopee has already surpassed its LATAM incumbents and that winning the region is a foregone conclusion.

As long as we’re talking about geographic expansion, there’s been a lot of fervor around Sea’s progress outside of SEA. The graphic above, which has been widely circulated online, seems to suggest that Sea is a much more material player in the Latin America market than anyone thought. We have seen pronouncements that Sea is already a global eCommerce player, pointing to app downloads and monthly users in Brazil. However, an astute chart reader would notice how parabolic monthly user growth has been compared to other players that have operated in the region for decades. We suspect that virtually all of this app usage has been driven by promotions within the popular Free Fire game, offering free items for opening the Shopee app to receive a code to punch into Free Fire. Importantly, these promotions do not require the making of any purchases, but simply the registration of an email. On one hand, this speaks to the power of having a captive audience and Sea’s ability to direct them to other services, but low CAC is not a business model and their eCommerce initiatives are not competitive compared to incumbents. Latin America has few common carriers and those they do have, have sub-par delivery speeds with limited geographical coverage—this is why Mercado Libre launched their own shipping service, with a network of drop off locations and warehouses offering fulfillment. Density is critical to improving logistics and so overtime the big tend to be able to offer increasingly better service at a lower cost. Furthermore, all of the network effects between sellers and buyers that make a marketplace vibrant work against them here as they do not have any material amount of local sellers and likely very few buyers. However, it is possible that Sea has some success with selling some more unique and cheaper goods from Asia to Latin America, but it is unlikely to ever be a big business without significant investment, which frankly would be a misallocation of capital given their somewhat tenuous lead in eCommerce in their core SEA markets, where there total addressable market remains largely unaddressed.

Shopee x Garena promotions in Brazil. Simply logging into Shopee (and doing nothing else) from 2/14-2/20 would get you the jacket and logging in from 2/21-3/1 would get you the gun.

As far as consumer experience goes, there is no clear winner. Shopee tends to have the lowest prices and great shipping promotions with high consumer trust, but is lacking selection in many offerings with mixed shipping experiences. They are improving rapidly, however. Tokopedia has more selection, but prices tend to be inline or higher than Shopee as they do not have as many sellers from China or the greater SEA area onboarded. Consumer trust is also very high with Tokopedia though, especially in the electronics category. Users appreciate Lazada’s authorized resellers of branded products, but there is also more fear of scams and the general experience is sub-par. Sea launched Shopee Mall to fill this need, which hosts brands and large retailers that can prominently feature their brands. On the payment side, it’s worth noting Sea has ShopeePay/AirPay integration with Shopee (~half of orders now use ShopeePay in Indonesia) which helps remove purchase friction, but they also limited other payment options that frustrate some users. While this is only a small subset of users’ opinions, we think it supports the thesis that Lazada is falling behind and Tokopedia is becoming their biggest threat of the current players, with Shopee still in the lead. Furthermore, Shopee is expanding commerce categories, entering the grocery and food delivery space, which if successful can help build consumer loyalty by habituating them to one app for more services. This is a growing market with ~$12bn GMV today, but there are many more mature and entrenched players like Grab and GoJek here. We touch on competition more in the business analysis section below.

A ShopeeFood delivery courier in Indonesia.

3) Digital Financial Services

Here you can get a sense of what it’s like to use Sea’s digital financial services. The first photo is of the home page of the Indonesian version of ShopeePay, which is integrated into the Shopee app. The second photo is of home page of the Thai version of ShopeePay, which is a standalone app (rebranded from AirPay). The third photo is of the home page of the Vietnamese version of AirPay, where it has retained its original branding as a standalone app.

To understand the importance of digital financial services, it’s critical to know that the general SEA population is very underserved in banking with the majority of people not having a bank account or a credit card, which makes participating in the digital economy impossible. This was the initial impetus to create AirPay, since more people owned smartphones than bank accounts and it allowed users to trade cash for digital credits. While this was originally a tool to garner higher paid penetration with Garena, it quickly became a business of its own. The network of top-up counters that accept physical cash was a huge advantage to gain early adoption and today remains a big selling point with over 200k counters.

Source: SmartKarma

As shown below, today Sea operates under the AirPay and ShopeePay brands. In most markets, Airpay was rebranded as ShopeePay with more limited functionality (touched on later). Functionality varies by country due to different payment and banking regulations. Notably, AirPay was closed in Indonesia because of 2017 regulation that barred top-up payments without licensing, effectively destroying their service overnight. However, they relaunched as ShopeePay only accepting on platform payments, but since regaining their license, the partner network accepting ShopeePay has expanded. 

The AirPay brand was kept where it was most popular and well known (and didn’t face regulatory pressure). ShopeePay benefits from marketing synergies with their eCommerce platform.
This diagram shows the use cases and flow of transactions with AirPay. Below you can also see the UI of the Thai version of the AirPay app.
Slide taken from an old investor presentation detailing the UI of the Thai version of the AirPay app.

AirPay was intended to be its own mini version of a super app, with a lot of functionality beyond payment, however most of this other functionality either was removed in the ShopeePay rebrand or is not used by many users—the eWallet and payment mechanism remains the main use case. An eWallet is essentially a digital version of a physical wallet: instead of keeping physical cash in your wallet you can keep digital cash in your eWallet, which can be spent online or through various mPOS systems via QR codes. On the eWallet front they are marginally ahead with many competitors like GoPay (GoJek), OVO and GrabPay very close behind. Given how underbanked a lot of SEA is there is still ample opportunity to gain share (or lose it). It also isn’t a winner-take-all market with opportunity for a few players to be leaders in the space and eWallet usage not being mutual exclusive.

Below we show users have increased to over 23mn in the last quarter (versus a total SEA population of ~600mn). You can also see that TPV (total payment volume) has been increasing to run-rating ~$12bn annually.

While TPV has grown substantially, revenues remain small as they are more focused on adoption than monetization at this point. The real revenue opportunity will be from expanding on adjacent financial services, offering products like asset management, insurance, credit cards, mortgage, and other loans. We try to size up what this could look like in our revenue build below. Notably, they just earned a Singaporean banking license which should allow them to start offering bank accounts and other services. Below we show the fee schedule of various products. Note that typical credit card fees are 2-3%, so there is latent pricing power, but we do not expect them to exercise it.

Source: SmartKarma

COVID Impact

As alluded to earlier, COVID had a big impact on Sea’s business. Despite how hackneyed commenting on Covid-related business acceleration in the tech space has become, we still don’t think it can be overstated how transformative of year this was for Sea. Several things happened that made their business materially more attractive. 1) General eCommerce penetration accelerated which disproportionately accrued to the Shopee, who took share from Lazada and the smaller platforms. In 2018 frequency was 3.6 orders per month and has ballooned to 5.7 in the last quarter. 2) Adoption of eWallets grew alongside eCommerce growth and a preference for touchless transactions with Airpay/ ShopeePay taking lead in most markets (albeit only marginally) 3) Stay at home set the gaming business on fire with Free Fire now the highest revenue generating game in South East Asia, Latin America, and India.  4) Sea’s exploding stock price (up ~4x) and exuberant capital markets environment enabled them to raise a $2.6bn secondary offering, cash which will help them keep their lead. Similarly, hiring engineers with large stock comp becomes less dilutive to existing shareholders as the market cap swells.

Business Analysis and The Thesis.

Sea is an amalgamation of one great business and two maturing bets that seem to be paying off in large and growing TAMs, but their success is far from assured. Garena, specifically Free Fire, is a highly scalable, high margin business with an engaged and active user base that spends prodigiously on the game. As mentioned, Free Fire is so popular that we actually think it is going to allow them to reposition the Garena social network, tying in more gamers to Free Fire and their on-platform 3rd party games through BOOYAH! with all the stickiness that messaging with your friends entails. While Sea is still building up their own ability to create more games (remember Free Fire was acquired), we assume that the Free Fire franchise is relatively stable overtime and continues to throw off billions in free cash flow that will continue to fund their eCommerce and fintech initiatives.

Without question, eCommerce is very underpenetrated as a percent of retail sales (which itself is small compared to more developed economies) and SEA has many of the fastest growing economies in the world. Banking services are very under penetrated and will be instrumental to facilitate further economic growth as well as their budding digital economy. If you had to pick two “TAMs” to go after, it would be hard to pick two bigger (or more exciting) ones. Expectantly, large opportunities draw many competitors and while Amazon may have never had a credible challenger (eBay never moved beyond an auction marketplace until very late), that is not the case in South East Asia. Sea themselves were a challenger that took on Lazada and a slew of local competitors to take the lead (by GMV, app downloads, and web visits) within 5 years.

Shopee is the #1 app in both the AppStore and PlayStore and the #1 most visited eCommerce website in the 6 major markets in which they operate.

Drawing a direct analog to Amazon virtually ensures a misconception of competitive intensity in SEA eCommerce. Along that vein, there is no reason to think in a decade from now there will only be one big, scaled eCommerce player. As we are seeing play out in China, eCommerce can be fractionated into many different areas beyond brand, quality, price and shipping, like group buying and city tier. Alibaba was the clear leader in eCommerce only to cede share to Pinduoduo and Meituan with others like Bytedance and Tencent looking to get bigger in the space. This is important when thinking about what share of eCommerce Sea can ultimately achieve. Amazon is estimated to have 40-50% share of eCommerce despite having two distinct advantages that insulate itself from competition: Prime and Fulfilment. Prime creates customer lock-in, increases order frequency and total spend (~$1,400 vs $600 for non-prime), and reduces churn. Fulfilment creates seller exclusivity (can only have inventory listed on one site) and drastically improves shipping speeds. Sea is starting to build out fulfillment but it could be a decade before that is a significant portion of GMV and user loyalty is very fickle now. In our conversations with ~200 South East Asian consumers, we found that shipping promotions and discounts are the overwhelming reason as to why they choose to shop on Shopee. To be fair, this may not be the condemnation that it appears to be, as the point of the promotion is to draw buyers in. But the point is, we don’t know where user preferences will lie when the promotions disappear. Spreading the ~$1.2bn they spent on eCommerce marketing across their ~50mn buyers would equate to their promotions being >10% of the average monthly income of a SEA resident, a very substantial incentive to buy something on their platform. Promotions are also running on the seller side where fees are kept artificially low or at zero.

These promotions are all about driving platform activity to benefit from network effects: more buyers draws more sellers, which in turn draw more buyers. However, network effects are not durable advantages without some level of exclusivity (Think Amazon Prime on the buyer side and fulfillment on the seller side) or superior consumer experience. This is because sellers can list on multiple sites and Shopee may just end up plowing resources into teaching merchants to sell online for them to apply that knowledge to other sites (somewhat similar to how after Instagram stole the story format from Snapchat and popularize it, it benefited Snapchat because Facebook made advertisers adapt to shooting in the vertical format). Furthermore, aggregating all sellers in one place matters less as a means to help users find items (because general internet search can do that) and is more about driving a consistent experience with buyer trust. When you buy something on Amazon you know that you will either get your item as expected or Amazon will make it right, which took decades of relentless efforts to achieve. Today, the Shopee experience is mixed with an inconsistent shipping experience with many partners including the local postal service who several of the buyers we talked to “dread”.

Of the 63% of merchants we surveyed who sold on both platforms, most believe that while Shopee is more “complete” and has better features, it receives far less visitor traffic and selling is slow unless you buy ads.

The most important thing to building enduring value for a company is that your customers do not leave. In marketing this is referred to as a “leaky bucket”, where you pay handsomely for a customer just to have them leave because of a bad experience or poor value. This is why capital is not a moat, it is a bridge – capital can help you get somewhere but does not insulate you from competition. Their substantial investment in acquiring users will be entirely wasted if customers do not come back and they cannot compete indefinitely on underpricing their services. All of this marketing spend is to buy them time until they can get to the point where the consumer experience is so superior that it can stand on its own. However, getting more sellers on the platform is an instrumental piece of driving more activity and, thus build higher density so shipping speeds can improve. Their marketing spend is hopefully what can help them eventually get there. However, in our survey work, it was apparent that Sea is at risk of a leaky bucket. Focusing on Indonesia, their largest market, concerning user behavior includes: 1) ~70% of users search at least two sites (usually Tokopedia and Shopee) and ~25% search three sites (the third is usually Lazada and occasionally Bukalapak) before buying anything, namely focused solely on price and discount offers. This is an issue as it signals that these sites are substitutable in consumers’ minds with Shopee not offering a superior enough consumer experience that it’s worth paying a higher price. 2) Users believe different platforms serve different purposes. Generally speaking, Shopee is associated with clothes, fashion accessories, & beauty items and Tokopedia is used for electronics and kitchen equipment, but there is some overlap. The concern here is Tokopedia could “specialize” in certain verticals, making it so a buyer always goes to Tokopedia for certain items first. Over time, if Tokopedia consistently offers the best or similar prices to other sites, they will stop checking and only shop on Tokopedia for these items. However, critically, our survey work has shown that of the 40% of customers across all of SEA who do not check more than one platform before buying, virtually all are shopping exclusively on Shopee. This is the strongest piece of data that we have found that supports Shopee is “winning” SEA eCommerce.

*Note: This chart is made up of responses from ~150 of the South East Asians we surveyed, whereas the statistic mentioned above highlighted just the Indonesian cohort, as Indonesia is their largest market.
Price is the main concern for most buyers and Shopee doesn’t have a competitive advantage in getting cheaper prices. They will either have to uniquely source with cheaper prices competitors can’t offer or build other features that customers value more to be sticky (we suspect the latter happens).

Additionally, when pressed on which site had the best shipping experience and UI Shopee did win with 2/3rds of respondents. We interpret all this to mean that there is still substantial price differences for certain items between Tokopedia, Shopee and Lazada that warrant the extra hassle of price checking. Over time, we can expect these differences to disappear, and once price becomes more consistent, buyers will start focusing on other factors when picking where to buy, which should bode well for Shopee provided they can continue to improve their shipping experience (which while it is still far behind Baba or Amazon, it is currently better than competitors). This will be crucial to them achieving the >200mn buyers and >$30bn of eCommerce revenues we have modeled below.

When it comes to shipping, Shopee is the clear leader in almost all of the countries in which it operates. Interestingly, although 60% of respondents listed Shopee as their favorite eCommerce platform, only 4% said Shopee had the best shipping experience (although this makes sense, as price takes precedence over shipping experience.)

There is no doubt that of the current competitors, Sea is best positioned to capture the largest share of the eCommerce market, but this is a highly dynamic market with many players who are willing to plow cash into the market to gain share. GoJek’s merger with Tokopedia means that Indonesia is only going to get more competitive, Coupang is just launching in Sea’s home Singapore market, JD is moving into Thailand with JD Central, and while Lazada may have temporarily pulled back, would it really be that surprising if Alibaba recommits to the region with aggressive promotions that would still only represent a fraction of their cash flows? Sea would be in a better position if they made progress on their own logistics network and fulfillment, creating something that is hard for competitors to quickly emulate. Shopee Pay helps adoption, but the proliferation of other eWallets and fintech’s means that it is a rather trivial friction reduction. Having said that, as long as they can continue to outrun the competition, they will stay in the lead.

Like eCommerce, bringing financial services to the underbanked and digitizing banking services is a huge opportunity that Sea is addressing with SeaMoney. However, regulation and sharper competition is curtailing their addressable market. As mentioned before AirPay had to be rolled back in Indonesia and was relaunched as Shopee Pay with much more limited features. In fact, the original Airpay branding was only retained in a single country, Vietnam, which in part is a decision to better realize marketing synergies between the Shopee brand, but mostly is a byproduct of the limited progress Sea has made with their financial services and AirPay brand. They retained the brand in Vietnam because it has high adoption there, but even in Vietnam our conversations with Vietnamese users showed a slight preference overall for other eWallets like Momo and Zalopay. Airpay as a super app was never a reality beyond an investor deck slide (that hasn’t been reprinted in a couple years). In Thailand, where AirPay was rebranded as ShopeePay TH but retained its “super-app”-like features, over 90% of respondents indicated their favorite or only eWallet is True Wallet.

Even though Thailand was one of only two countries where AirPay remained a standalone app (it has since rebranded to ShopeePay but has not been rolled into Shopee as it has been in the other countries), our surveys showed that almost no one regarded it as their favorite.

ShopeePay is showing more success though. Tying usage to their eCommerce platform (and nascent food delivery business) with promotions for using ShopeePay has led to increased adoption. In a 3rd party Indonesian survey (MarkPlus), they pegged ShopeePay eWallet usage at 26% versus OVO at 24% and GoPay at 23%, which nicely displays how fierce of a battle financial services will be. However, these are not winner-take-all markets and a user can have more than one eWallet. The real prize is not the eWallet business which is hard to directly make money off of without huge volumes, but rather the cross-selling of other financial services, especially the extending of credit. If ShopeePay could significantly penetrate the consumer and business lending market, which they have recently gained a foothold in, could more than double their Digital Financial Services revenues. Other services like insurance, mortgage and brokerage could also be very lucrative, but could prove to be hard to get into.

Before getting too ambitious though, in our conversations with ~200 users, most keep the bare minimum needed in their eWallet and use it mostly because of the discounts offered and have multiple wallets with no preference for ShopeePay. Tying ShopeePay to Shopee did help drive adoption, but also frustrated some users who wanted to use other eWallets. There does not seem to be another service within ShopeePay that draws users back and they have limited acceptance with 3rd party merchants. Furthermore, as touched on prior, the QR code standardization implemented in Indonesia potentially lowers the bar for new fintech players to enter the space and similar laws could roll out elsewhere. In short, there is no single payment/eWallet leader and it does not look like any of the current competitors are well positioned to be the single full service payment provider. In other words, we doubt there will be an “Ant Financial” of South East Asia ever, however Sea could still have a lot of success with their offerings coming well short of that. We see ~175mn users and $140bn of TPV as plausible in a decade if all goes well, noting that if they do become a dominant wallet with higher than modeled eCommerce penetration and other efforts like food delivery becoming successful, TPV could be much larger.

One QR code to rule them all. Indonesia, the largest and most competitive SEA market, has already rolled out a standard QR code which lowers the barriers to entry (and exit) for fintechs. Other countries are said to be in the process of developing or rolling out their own standards.

While we have touched on competitive risks, it is also possible that the TAM is not what it appears to be. We bring this up to emphasize the risks of extrapolating Sea’s current success and growth far out into the future. Back to eCommerce for example, the next leg of growth could be much harder to achieve than the last with the immediately addressable TAM much smaller than it appears. In Indonesia for instance, per capita GDP is $5,000, but the average Shopee buyer is currently spending ~$800 annually (Amazon prime members spend only ~2x as much despite GDP per Capita being 13x higher). Not to conflate GDP and earnings, but this is clearly indicative that most people cannot currently afford to spend on Shopee at the current rate and new users will be less valuable. Put another way, they have higher penetration in tier 1 cities. High density and scale are critical to achieving strong logistics networks that can quickly deliver packages with low variable costs, absent of which the eCommerce model will never be as profitable or competitively strong as Alibaba or Amazon.

Wrapping all of this up, as you’ll see below that there is a path to Sea being a $250bn+ business with some upside to that within their preexisting business lines and several call options on nascent internal initiatives. However, their growth and success in these markets are hardly guaranteed, and investors will have to judge for themselves whether or not they believe they are being adequately compensated for these risks.

Management.

Sea’s executive team on the day of their IPO.

There is limited information on Forest Li and he tends to not speak much to the investing community, but there is no doubt that he has a large vision that he has executed well on so far. We know that the values important to him (as they are the company’s official values) are to “serve”, “adapt”, “run”, “commit” and “stay humble”. And you can see them permeate Sea’s approach to business. Sea today is materially different than the GG platform it was a decade ago and Forrest Li deserves a lot of credit for that transition, including launching multiple new businesses and becoming the biggest public company in SEA. There is more info on Chris Feng who is the head of Shopee and SeaMoney. Internal employees have talked positively about Chris’s execution with them launching big new initiatives every years since they took over, first with Shopee University, then Shopee Mall then the China Marketplace portal. He also helped lead the charge of hyperlocalizing Shopee and is more of the Zuckerberg “move fast and break things” school of thought — as mentioned before, Tokopedia studied livestreaming for a year before launching a product, whereas they quickly rolled out a MVP and iterated. Among investors, Chris Feng is noted as a key man risk. As far as management philosophy goes, they tend to preference building over partnering or buying for things within their competence, saying “we are a tech company, let’s just build it”. However, they will partner for regulatory or scale reasons. Ultimately, this is still a founder-led company with a large, exciting vision, which tends to attract quality management and employees.

If you’re interested in reading a deeper dive into management, please be sure to check out our Details post: Sea Limited: Management, Culture, and Execution.

Revenue Build.

Below we have our revenue builds for eCommerce, digital financial services and our segment build which combines all of Sea’s segments. Note we do not have a separate revenue build for digital entertainment and instead input our assumptions directly into the combined segment build exhibit as there is no great way to model a hit-driven business and we have no opinion on games in the development pipeline. Of course, premium subscribers will get access to all of our excel work and may change any assumptions they wish!

eCommerce Build.

Below is our eCommerce build which forecasts what the eCommerce business could look like 10 years from now. We start our build with the SEA population and layer in internet penetration and then % of internet users who utilize Shopee. Interestingly, order frequency is up from 3.7x monthly in 2018 (the last time it was disclosed) to 5.7x monthly in the last quarter. We assume this could be slightly elevated due to covid, but overtime could increase a bit. Peers in China, like Alibaba, have buyers that order an average of ~twice a week, or 8 times a month, so our assumption could prove to be too low. It’s important to note that the physical retail environment is much less developed in SEA than it is in the US. There is roughly 45x more physical retail space in the US and 5x more in China than compared to SEA on a per capita basis (!). Similar to how in more underdeveloped economies, they skipped the intermediary step of credit cards and went directly to eWallets, QR codes and, mPOS, retail will largely go directly to online and skip the stage of many large specialty stores and chain department stores. There will still of course be some physical retail growth as certain formats are symbiotic with online sales, but that is all to say there is reason to believe ultimate eCommerce penetration as a % of retail sales could be even higher than in China (estimated around ~40%), which is already more than ~2x what it is in the US. Put more simply, the opportunity is big. Having said that, there is no reason to think there will only be one winner. Our build below implies that Sea will have a ~40% share of eCommerce in a decade. Perhaps if Sea executes exceptionally well and all competitors stumble, then they can reach up to ~60%, but the eCommerce environment is still very nascent in SEA so it would be hard to have high confidence in that assumption.

We model Shopee buyers going from ~55mn currently to over ~200mn with GMV surpassing $140bn. With careful delineation between 3P and 1P sales, that sums up to ~30bn of revenues, roughly a ~30% growth rate for a decade, a feat few companies have accomplished (taking 1p into account, gross profit still grows at a high 30% CAGR – we will discuss this later in our SEA segment build).

Digital Financial Services Build

There are two builds below for DFS. The first is projecting out Sea’s current revenue streams from DFS which are basically just taking a fee on a variety of different transactions or transfers and whatever they earn on interest from holding deposits in eWallets. Our DFS build has quarterly frequency modestly climbing overtime to a little less than once per day and spend per user grows ~5% (spend increasing with existing users offset by higher penetration of lower income users). More ambitiously, we have paying users growing to ~175mn, which would mean around a quarter of everyone in the SEA region uses SeaMoney. This is a rather optimistic assumption and would require that ShopeePay expands their use cases, but ultimately there will not be more than a handful of different ways to pay for a service or product across SEA. Once there are a few good options (AirPay/ShopeePay, GoPay, OVO, GrabPay) there is less of a desire to set up accepting payment from a new provider (although in Indonesia QRIS could mean merchants accept payment from anyone). From a consumer perspective, it is a hassle to have many wallets as the point of a wallet is to have everything in one place and additional wallets have very incremental utility for a payer, but they still are nevertheless not mutually exclusive.

The second financial build below shows what cross selling into other financial services could look like. We focus only on the lending and asset management opportunity, but there are many other services they could sell or act as lead generators for, such as insurance and mortgage. Below we show that lending could be a ~$1.5bn pool of revenues by 15% of payment volume being facilitated by credit (compared to roughly 1/3rd of purchases made with credit in the US, although interest is not charged on the majority of that). This figure could be conservative, especially If you include merchant lending. Moving into asset management with money market funds and other financial products could allow them to earn a ~60bps fee on funds invested. We assume 15% of users save an average of $800 which implies a very high portion of total users who are capable of saving since most people in South East Asia cannot save in size, but this will change over time and once you are locked into an eWallet there is a good conversion rate to other services. Also as the economy continues to grow, more people will be able to save.

Combined Segments Revenue Build.

Below is our combined segment revenue and gross profit build.

P&L.

Below we continue from the segment build below and layer in our operating cost assumptions. We see a path to $8bn of EBIT and ~$6.5bn of NOPAT in a decade.

Valuation.

Below we apply different P/E multiples to ~2030 earnings. The margins we are sensitized around below are considered to be in “steady state” so you may apply a multiple that does not need to factor further margin expansion. In our opinion a 35x multiple could be justified if Sea is still growing earnings >20%, implying a duration of growth very few companies have accomplished. A more moderate 20-25x would require a less demanding low double digit to mid-teens earnings growth to support in our opinion. We pegged our base case of margins in the low 20’s with a better chance it is lower than higher especially with increased investment in logistics, but nevertheless if they can really scale efficiently 30% is plausible. Importantly, we are not modeling the most ambitious SeaMoney initiatives, which if wildly successful could be worth another $100bn+. Other upside could come from another hit game, monetizing BOOYAH!, higher ecommerce penetration and perhaps even new businesses that have yet to be launched. An investor will have to make their own judgement of the probability of Sea accomplishing all of this and whether the return below appropriately compensates them for execution risk and the competitive response.

Other far-flung call options could include: 1) an external cloud service (they are working on building an internal cloud infrastructure to handle their own traffic, similar to what Amazon did before launching AWS), 2) becoming a leader in food delivery (they just launched ShopeeFood in Indonesia), 3) Sea Capital investments (formed after acqui-hiring David Ma’s Composite Capital Management) which could copy Tencent’s investment strategy of investing in promising start-ups and could strike it big with some investments, 4) AI Labs, their artificial intelligence R&D initiatives could produce something of value, 5) New businesses could still be launched that play into the their greater ecosystem.

Risks.

1) Competition. Even benign competition can materially stymie profitability of SEA, requiring them to offer their services for less to be competitive. Competitors can leverage other assets they have (super app, messaging apps with high usage, social media, ect) to cross sell similar services and may be able to offer overall better customer and merchant experiences.

2) Churn. Many users were enticed to join the platform through limited time promotions where the ultimate customer life is unknown. Shopee and SeaMoney may only seem to be good values to customers at these discounted prices and usage could drop off materially when subsidies are rolled back. Competition may also entice users to leave with minimal lock in.

3) Steady state margin structure not achieved. We value Sea at margins of a more mature business but it is possible the competitive environment does not allow this margin structure to be realized or there is less operating leverage in the model than anticipated. Furthermore, perhaps promotions rather than only being a strategy to collect users become necessary to retain them, depressing steady state margins.

4) Free Fire popularity fades. Video games like other media can come in and out of vogue. As this is a material source of cash flow for the company’s investments, losing this cash flow would be detrimental for their other businesses and could require a large capital raise in order to continue growing at the same speed.

5) New financial and possibly gaming regulation. There were financial regulations in the past that materially altered SeaMoney and there could be more changes in the future. Most of Sea has a very young financial system that will likely vary from what it looks like when the region is more developed. While the governments are different, we saw game restriction in China on Tencent and perhaps some governments could take similar action or Chinese regulations could affect gamers in SEA.

Summary Model.

The Sea model summary below is not meant to match perfectly with the other exhibit assumptions throughout the write-up. Rather it is presented so you can get an approximate picture of how different line items flow through. Of course, Members Plus subscribers will have access to the underlying excels and can change any assumptions. Thanks for reading! If you are not ready to subscribe today to receive the next in-depth report, you can still receive periodic updates for free, just drop your email below!

Special thanks to Dennis, Muji, Akshay, Byrne, and Frederik for reading early versions of this report and providing invaluable feedback. An additional round of thanks goes to those took the time out of their day to speak with us and share their thoughts on and experiences with each of the products.

If you enjoyed this report, please consider becoming a Member! Every month, we will publish a deep dive report on an Asian company similar to this one. From now on, only Members will have access to these reports. Those who join Members Plus will also be able to download our excel worksheets.

*Disclaimer: I have a position in Sea 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.

All rights reserved. No content on the DJY Research website can be redistributed without the express written consent of DJY Research.

3 Replies to “Sea Limited”

Leave a Reply