🌊 NYSE:SE - Shopee: The King of S&M (sales & marketing exp)
Evaluating ASEAN e-commerce marketplace Shopee's performance - using GAAP accounting metrics (e.g. revenue), rather than non-GAAP metrics (e.g. adjusted EBITDA/GMV)
In this report, we shall attempt to evaluate the historical performance of SEA Ltd’s e-commerce marketplace Shopee - using GAAP accounting metrics, considering how unreliable non-GAAP metrics have been in the recent past.
Evaluating the business performance of startups (e.g. Shopee) can be performed via two measures, by asking: 1) Is There Product Market Fit? and 2) Does Operating Leverage Exist?
We can answer these two questions by observing SE’s historical GAAP accounting metrics rather than management’s promises - and from there, draw reliable conclusions with regards to their future outlook and potential valuation.
Note: All share prices listed in this report are based on the closing share price at 16 May 2022 of USD 70.33. Please adjust as necessary.
EDIT (29.05.22): One thing that I previously forgot to include in this report is that by giving out excessive free shipping promotions, Shopee could theoretically have shifted their shipping fee expenses from COGS to S&M expenses (i.e. below the Gross Profit line) - while still recognizing collections of shipping fees under Revenue. This could have had the effect of inflating their past Gross Profits.
Hence, Gross Profit growth is not the best performance trend to monitor in terms of estimating Shopee’s normalized cash-on-cash return. Given how outsized their S&M expenses have been in the past, it will be interesting to monitor whether any future correlation exists between the reduction in Shopee’s S&M expenses vs. the increase in their COGS going forward.
For ease of navigation, please click on the links below to jump to the respective chapters of this article:
(The above links which jump to the relevant chapters will only work in the desktop version of Google Chrome. If you’re using a different browser, please scroll to the relevant chapters below.)
Evaluating SE’s performance using standard GAAP accounting metrics - rather than non-GAAP metrics
SEA Ltd (NYSE:SE) probably needs no introduction to the mainstream investor - it was the darling of the stock market for quite awhile, after climbing +300% since the pandemic began, and +2,000% since its listing on the NYSE in Oct 2017. While you’d still be up +25% CAGR today if you had invested immediately before the pandemic began (or +33% CAGR from its listing), it’s still a far cry from the 250% CAGR you would have gotten if you had sold at the peak in Oct 2021 at $367.
The reason why SEA Ltd and its unicorn-like South East Asian (i.e. “SEA”) e-commerce business Shopee probably needs no introduction is because there are already thousands of similar reports like this one covering SE, amidst its meteoric share price rise over the past two years. Some of them go into extreme detail about their respective business segments - and if this is your first time stumbling across SE, these can probably do a better job of getting you up to speed with SE and Shopee than I can. Here’s an example of one such blog who I think has done a fantastic job of covering the bases of SE’s respective businesses:
Fortunately, my goal for this report isn’t to rehash the same business narrative of SE - which many of you have probably read 100x over already.
WHAT I WILL BE DOING DIFFERENTLY IN THIS REPORT will be to evaluate SE’s and Shopee’s historical performance via an accounting lens - by analyzing their GAAP accounting figures (e.g. revenues) from their three statements (i.e. P&L, B/S, CFS), rather than their pro forma or non-GAAP figures (e.g. GMV).
One nitpick that I’ve always had with all the previous SE reports that I’ve read thus far, is that most of them derive their valuation based on non-GAAP figures (e.g. GMV, EBITDA, QPU, etc) - rather than the more familiar GAAP-based line items extracted from their three statments (e.g. Revenue, Net Profit). To be fair, SE hasn’t attained profitability yet despite having been listed for nearly 5 years - hence more traditional valuation metrics like the PE ratio (or even EV/EBITDA) might be inadequate for valuation purposes.
However, that still didn’t quite scratch away the itch that these are still non-GAAP figures - which I’ve never been comfortable relying wholly and exclusively on for estimating a company’s valuation (e.g. Valeant Pharmaceutical’s infamous ‘cash earnings’). Non-GAAP figures simply haven’t withstood the test of time as robustly as the GAAP accounting line items in the three statements have - where accounting standards lay out extremely well-defined definitions for their recognition criteria. While I’m not suggesting that there has been any hanky panky going on with SE’s numbers, it still makes me uncomfortable when two non-GAAP figures are combined to justify an investment thesis - e.g. improving Adjusted EBITDA losses/GMV over time.
What I find a bit unsettling is that, up to today, I have not come across a single investment thesis for SE based exclusively on traditional GAAP accounting line items - most of them incorporate non-GAAP figures to a significant extent in order to justify the investment thesis. This is what this report aims to cure - as I will be approaching SE’s valuation from a standard GAAP accounting perspective rather than using non-GAAP figures (the way an actual PE investor analyzing SE like Tencent might).
As most of SE’s Group performance and valuation is contingent on their e-commerce business Shopee, we shall be focusing exclusively on evaluating Shopee’s performance in this report, in the interest of brevity. We will also be focusing mainly on their ASEAN operations (63% of revenues) - since most of their other regions have either been declining in revenue share (e.g. Taiwan: 14%) or still remain fledgling operations (Latin America: 19%). And with Shopee’s recent exit from both India and France, I think we should conservatively not expect too much from their non-ASEAN operations.
Most investors would also agree that it is safe to assume that the status quo maintains for the rest of their other businesses, such as their Gaming segment. While the news of India banning its hit game Free Fire is bound to impact their valuations, it should not materially affect the Gaming segment’s pre-Free Fire role of basically being an internal reinvestment channel for their E-commerce segment. Given Shopee’s outsized contribution towards SE’s valuations and in the interest of brevity, we will be assuming the status quo remains for their Gaming segment.
Check out our related articles: What The Early Amazon Investors Saw & Big Tech is Officially in Value Investing Territory
We also have a sister news blog for daily market news, Value Investing Substack NEWS. Click the box above to check it out!
The Underlying Economic Framework For Evaluating Startup Performance
Since SE’s nominal profits are getting further and further away from breakeven levels over time, we can’t use traditional earnings-based valuation metrics like the PE ratio to inform SE’s valuations. This has resulted in most sellside analysts understandably relying on arcane non-GAAP figures like Adjusted EBITDA losses/GMV or Bookings/QPU in order to arrive at their valuation estimates - with accompanying disastrous results.
However, by understanding the underlying economic framework for evaluating startup performance, we can identify SE’s key success factors from a business perspective (i.e. what a businessman would look for) - and then reverse-engineer from there to determine what kind of GAAP accounting metrics we want to use to monitor their historical performance. In this way, we avoid having to jump through hoops in order to justify an investment thesis with arbitrary non-GAAP numbers. And to be fair, these are exactly the kind of standard GAAP accounting metrics that venture capitalists (VCs) would be demanding from SE’s management if they were raising funds from PE/VC (private equity/venture capital).
Let’s be clear about one thing - SE is still effectively a startup, with startup economics to boot. While it’s true that they have a mature Gaming business segment, SE’s valuations are to this day still very much contingent on their immature E-commerce business (i.e. Shopee) - resulting in their horrendously cratering Net Losses. This is largely due to the startup mentality which has perpetuated the Big Tech sector for the past half-decade - grow market share at the expense of profits today, and then switch to monetization after you’ve commanded a monopolistic position (i.e. the Facebook model). This is why Shopee’s sales & marketing expenses (S&M) as a % of revenue has consistently hovered at legitimately mindblowing levels, and is also why I referred to Shopee as the “King of S&M” in the title - for FY21, this metric stood at 57% (a healthy level for most industries is <20%):
One of the consequences of this gargantuan S&M spend was to invoke accusations of Shopee “renting market share”, where their explosive revenue growth was simply a function of their outsized advertising & promotion budget, rather than organic customer acquisition growth - the parallel here is if the coffee place next to you gave out perpetual 50% sales discounts. If true, the implication of this allegation is that Shopee’s explosive revenue growth has not been sustainable - and that once they turn off the marketing spigot, their revenue growth will normalize to much lower levels. We shall attempt to determine if this is true or not by evaluating their historical performance in this context using standard GAAP accounting metrics below.
Consequently, the question begs: how do we attempt to derive a valuation for SE, considering that Shopee (and the group as a whole) is still enormously loss-making? This is where “startup economics” comes into play - we want to figure out whether Shopee can eventually become profitable on a sustainable basis, how long it might take to achieve that, and what those future profits might look like. In other words, we want to analyze SE’s historical performance through the lens of a VC investor - without relying on potentially distorted non-GAAP figures as supplied by management.
The first thing to understand about evaluating startup performance is that only one GAAP accounting line item on their P&L really matters - Revenue. This is because at the startup stage, whether you have sufficient Revenue growth outweighs all other cost factors - in the context of deciding whether you want to pour more good money after bad into the business. If you have enough revenue growth, it could justify being unprofitable at the beginning for awhile - while if you don’t have enough revenue growth, no amount of cost-cutting is going to be enough.
In this context of revenue growth, there are two main considerations you want answered when evaluating a startup’s historical performance:
Is There Product Market Fit?
Does Operating Leverage Exist?
1) Is There Product Market Fit? This question basically asks whether there are enough customers in the world who actually want to buy the startup’s product - i.e. does it add value to society? There are too many aspiring entrepreneurs who adopt a “build it and they will come” approach to business - without stopping to question if anyone (besides themselves and their mom) actually wants the product badly enough to pony up actual money for it. This is clearly unsustainable, as you’re going to very quickly run into a brick wall if you can’t find enough customers. We can use VC metrics like the Efficiency Score to determine this, which we’ll explore further below.
2) Does Operating Leverage Exist? Of course, it’s not enough just to have product market fit and show revenue - you need to eventually be able to earn profits as well. While being profitable at the outset is always preferable, a business can still justify being unprofitable at the outset if it can prove that it will eventually become profitable. Typically, the way that happens is via an increase in operating leverage.
Here’s an illustration: suppose that a startup has large fixed costs (e.g. depreciation of upfront CAPEX), which might result in it being unprofitable at x amount of revenue. However, since those costs are fixed in nature, there should be a theoretical breakeven point represented by a multiple of x revenue - above which the business will become profitable (e.g. if the breakeven level is 10x revenue, then any incremental revenue above 10x would represent profit).
However, the bulk of a startup’s cost tends to be represented by variable costs (e.g. COGS or staff costs) rather than fixed costs - since it rarely makes sense to commit large amounts of upfront capital into a yet unproven venture. As such, it typically doesn’t make sense to invest into a business which is unprofitable at the outset - except in those cases where most of those variable costs can somehow be transformed into fixed costs over time. This transformation is referred to in the industry as Operating Leverage - which is usually engendered via Economies of scale. The reason why Economies of scale works is because at some point, there will be enough customers to justify the large upfront CAPEX investments required to make the unit economics self-sustainable - e.g. buying a home rather than renting, since on average your monthly obligations will be lower from buying vs renting (at the expense of the large upfront purchase commitment).
This is the basic formula which all “venture” startups of the past half-decade have been following in the pursuit of future profitability - scale the loss-making business up to a level where unit economics start to become profitable; or at least be able to show investors that your unit economics are improving. This is operating leverage at work - and one of the ways we shall be evaluating SE’s historical performance.
Subscribe for just $9.90/month to access our entire catalog of Value Investing stock reports like this one! Free subscribers also get to read my latest NFLX stock report!
Is There Product Market Fit?
There is one very good way to assess whether product market fit exists for a startup or not - the Efficiency Score (i.e. Incremental Net ARR / Net Burn). Let us break down what these mean below.
Incremental Net ARR - there are 3 components to this metric: ARR, Net, and Incremental
ARR: Annual Recurring Revenue. As the name implies, recurring revenue is a way of assessing how much revenue can be reliably generated each year from recurring customers.
Net: This refers to Net ARR. Net ARR simply deducts lost recurring customers from the Gross ARR calculation. So if I gained 2 new subscribers but lost 1 throughout the year, I’ve only added one unit of Net ARR.
Incremental: This refers to incremental Net ARR amounts in excess of the previous period. You can also refer to this as Net ARR growth.
Putting all three together, Incremental Net ARR simply tries to determine how much revenue growth there is on a sustainable basis.
Net Burn - This refers to the amount of cash burn that there was during the period, after netting off any revenue gained. It’s basically FCF for the period.
By dividing Incremental Net ARR by Net Burn, we arrive at the Efficiency Score - which is a very rough cash flow approximation for Net Profit. Since the startup doesn’t tend to be actually profitable yet, what we want to see is an improving Efficiency Score ratio over time - as this means that there is sustainable revenue growth in excess of cash burn. In layman’s terms, there is enough customer demand (i.e. product market fit) to justify an eventual self-sustaining business - even if it might be loss-making at this moment in time.
There is a problem with evaluating Shopee’s historical performance according to this approach - Shopee doesn’t have recurring revenue, as it doesn’t have a SaaS-like subscription based business. However, it is still useful to serve as a barometer for what their present-day business economics look like - we can simply assign a sufficiently large margin of safety to the resulting calculation. At the very least, it serves as an economic framework whereupon we can assign GAAP accounting metrics towards, rather than being forced to rely on more subjective non-GAAP metrics.
If we treat Shopee as a startup, we can draw parallels between their S&M expenses and the standard VC industry metric ‘Customer Acquisition Costs (CAC)’ - since the outsized levels of the former implies that it is mainly used for acquiring new customers. Similarly, we can simply use their Revenues in place of ARR - since we are just trying to get a feel for their Efficiency Score. Hence, the equivalent to the Efficiency Score = Net Incremental ARR / Net Burn can be represented by their Rev / S&M ratio:
As we can see from the table and charts above, Shopee’s Rev / S&M ratio is clearly moving in the right direction. In my attempt to draw a closer parallel to Net Incremental Rev / Net Burn, I ended up coming up with Incremental Rev / “Growth” S&M. To be clear, this might as well be a non-GAAP figure - but we can clearly see that it is moving in the right direction as well.
Here’s the explanation for “Growth” S&M:
Firstly, assume that their maximum sustainable S&M/Rev = 20% over the long-term, in line with existing Retail industry standards (e.g. Samsung).
This implies that any S&M/Rev in excess of 20% can represent S&M spend in excess of the necessary levels to replenish organic churn - i.e. “Growth” S&M (or CAC).
Therefore, Incremental Rev / “Growth” S&M is supposed to represent how much incremental revenue could be attributable to CAC.
To be doubly clear, Shopee is not a SaaS business - and therefore using the Efficiency Score to evaluate whether there is product market fit might not be the best approach. But since their FY21 Rev/S&M of 1.7x is far in excess of the breakeven level of 1.0x, I would argue that a reasonable margin of error exists to justify using this metric in this context - at the very least, we can say that Product Market Fit exists for Shopee’s e-commerce business. This was probably unnecessary to prove simply based on anecdotal observations of Shopee’s active users, but at least this gives the claim some scientific credibility based on GAAP accounting metrics.
However, the observant among you will point out that this is a cash flow metric - not a profit metric. This is extremely correct - you can have positive free cash flow (FCF) without actually being profitable. Since the only disparity between long-term FCF and long-term profits is timing differences, all that does is delay the inevitable losses once those cash expenses materialize as cash outflows. Hence, there is also a need to determine whether Operating Leverage exists in Shopee’s business.
Download my Excel financial model and 1-page summary of SEA Ltd by clicking these link:
Does Operating Leverage Exist?
To answer this question, the obvious metrics to observe are SE’s profit margins. Ostensibly, if their margins are improving over time, that implies the existence of Operating Leverage. And on the surface, this appears to be true - we can see how all of their respective profit margins have improved over time, both on an annual and quarterly basis:
However, you’ll also note above how except for SE’s annual Gross Margins in the years FY20 and FY21 respectively, all of their other profit margins remain in deeply negative territory. While their Adjusted EBITDA margins are technically improving over time, I don’t think it says much when they are improving from FY18’s -319% → FY19’s -122% → FY20’s -73% - → FY21’s -57%. Those are just some ridiculously negative numbers - and that information is basically useless for historical performance evaluation purposes. I certainly don’t believe that you should be extrapolating from such trends into the future.
On top of that, consider how Operating Leverage is attained in the E-commerce sector which Shopee resides in. If you’ve read my previous stock report where I analyzed the business economics of early Amazon, you’ll know that the competitive bottleneck of the e-commerce sector lies in minimizing shipping costs - since an e-commerce marketplace is basically just a disrupted version of the commoditized B&M retail industry. This means that the future of the E-commerce sector looks eerily similar to that of the Streaming sector today - once e-commerce has completed the disruption from B&M retail, everyone is just standing on tiptoes again and e-commerce stops having any unique advantages anymore. Thus, the e-commerce sector is expected to revert to the price wars of the B&M retail sector where economies of scale to gain the lowest unit costs rules supreme - and those unit costs reductions will be focused on minimizing shipping costs.
As a result, we can observe in Amazon’s early financial statements (i.e. 2005-2010) how they sought to replicate Walmart’s hub-and-spoke logistics channel by rapidly leasing warehouses across the USA - in preparation for the eventual fight with Walmart’s inevitable entrance into the e-commerce sector. Amazon started pursuing this even from their very early days, when people still mainly viewed them as an eBay and Barnes & Nobles competitor - which also explains their consistently high PE ratios at the time. You can check out the full thesis for yourself by clicking the box below:
As such, when looking for the presence of Operating Leverage in Shopee’s business, we should be looking for signs on their balance sheet of new warehouses being built (or some sort of localized equivalent, e.g. dark stores). Now let’s refer to Shopee’s balance sheet metrics below - we can see how their Fixed Asset Turnover of 257% in FY21 hasn’t really changed much since 254% in FY18 when Shopee first started gaining traction in ASEAN. On top of that, their “regular” Asset Turnover has also not shown signs of improving over that same period. This is in stark contrast to Amazon - which despite starting from a much higher base, has seen both metrics reduce rapidly over time:
If you really think about it, Shopee’s CAPEX environment looks starkly different from the likes of Amazon’s and Alibaba’s. Amazon is located in the USA, while Alibaba is located in China - both of which are countries that are basically one gigantic contiguous piece of land. This makes it relatively simple to derive economies of scale from building out a nationwide hub-and-spoke logistics channel in your e-commerce operation - by maxing out the capacities of the 20 TEU trailers that you are trucking goods across the country in.
Now, take a look at the map of ASEAN below - and tell me what the single most defining characteristic of this map is:
If you’ve answered “SEA”, you’d be correct. The topographical difference between South East Asia and USA/China is immense - the former is mostly represented by water, while the latter two are basically completely landlocked units. This means that you can’t exactly outfit a truck-based hub-and-spoke logistics channel in ASEAN to cultivate economies of scale - since trucks can’t drive across water, and the economics of an intermodal logistics system is a lot trickier than a pure land-based one. Sure, both Amazon and Alibaba have planes in their delivery fleet - but those are for premium express shipping with premium rates, not regular economy shipping.
This might explain why Shopee hasn’t exactly been in a hurry to cultivate Operating Leverage in their e-commerce operation by building warehouses everywhere they can drop a pin on Google Maps in ASEAN - the way both Amazon and Alibaba’s 3PL partners have. It also begs the question - how exactly do they intend to replicate Amazon’s unit economics in such a starkly different topographical environment? To this day, I don’t know of any large B&M Retailer who has a unified presence in ASEAN - even the likes of McDonald’s and Tesco operate here as a separate entities in each local country, with disparate local supply chains and local branding.
This gives me pause regarding the narrative of Shopee’s improving non-GAAP performance as evidence of incremental economies of scale in Shopee’s e-commerce business - e.g. declining Adjusted EBITDA losses/order, improving Gross profit/GMV, etc. It’s very easy to generate those increased efficiencies simply by ramping up your S&M spend - since hopefully, the delta on your S&M spend per unit of revenue is positive. The question was never about how to improve their S&M spend per unit of revenue - but why are they even spending 58% of FY21 revenue on S&M expenses in the first place.
And quite frankly, I wouldn’t extrapolate a trend of S&M/Rev which starts at 93% in FY19 and has since fallen to 58% in FY21 - I’d like to see it fall to <30% before I start drawing conclusions about Shopee’s organic revenue growth. At the very least, we can say that those metrics do not neccessarily represent conclusive evidence of Shopee demonstrating improving unit economics. Anecdotally speaking, I can tell you that I only really use Shopee now whenever they have promotions, and have stopped being a daily active user (DAU) - since I was never going to use them that frequently without those promotions to begin with.
The Problem with Non-Improving Unit Economics
Well if Shopee isn’t necessarily improving its unit economics, what are the implications? The heart of the issue is best encapsulated in the following quote by this excellent article titled, “The Gross Margin Problem: Lessons for Tech-Enabled Startups”:
“Growth solves many problems at startups, unit economics is not one of them.”
The same paragraph continues, “When you’re losing money on every transaction, you can’t make it up in volume. In fact, the more revenue that a businesses with negative unit economics generates, the more money it loses.”
Herein lies the crux of the issue for Shopee. Given that they’ve only had 6 quarters of positive Gross Margins thus far, I think it’s quite a stretch to extrapolate that trend (starting from negative >-100%) and conclude that they’re going to start reporting super exciting Net Margins in the medium-term. Can they eventually become technically profitable? Yes. Can they become profitable enough to justify their immense unprofitability of the recent past? Probably not. When you look at it from the perspective of Retained Earnings (i.e. cumulative Net Profit - FY21: $ -7.2B), it’s really hard to see how they can eke out a decent ROI on that - except perhaps in the super long-term.
The same article as above has two more sterling quotes, which I’ll quote below:
“The question is what kind of money are they losing? Losing money at the corporate level is ok (all startups do); losing money at the unit level is not.”
This is really the one-two punch in the gut of the Shopee investment thesis. The general idea here is that you cannot afford to be losing money at the unit level even right from the start - there needs to be some signs of unit profitability existing, even if corporate profitability doesn’t. I know that non-GAAP figures like declining Adjusted EBITDA losses/order might imply the existence of this - but it’s also really easy to complicate the assessment of how “pure” this trend is when their FY21 S&M/Rev is at 58%. Unless they’re willing to provide further disclosure about the breakdown of their S&M expenses - and how they’re being optimized to generate incremental profits (not revenues) - I don’t think anyone would blame me for being professionally skeptical about their claims (and let’s not even get into LTV/CAC).
“Trying to re-engineer the unit economics or culture of a business that is already operating at massive scale is brutally hard. Searching for the scalable model when you’re already at scale is a contradiction in terms.”
This second quote from the above article mentions how much more difficult it is to re-engineer the unit economics of a business that is already operating at massive scale. If there’s one thing I’ve learned about Big Business, it’s that INERTIA is a powerful force - for better or for worse. In Shopee’s case, they’ve pretty much already captured at least 50% of ASEAN e-commerce market share - as shown in the charts below:
The question then begs: if they’re still posting exponentially declining losses at this scale, then at what scale will they start posting corporate profitability - nevermind unit profitability? As someone who lives in ASEAN, I can anecdotally say that the e-commerce war for new eyeballs hasn’t reduced significantly in intensity - there was a slight lull in the promotions being offered at the beginning of the Tech Crash (Dec ‘21 - Feb ‘22), but all the e-commerce competitors have since come back in full force with their sales & promotions offerings. Here are a few screenshots I’ve personally taken inside Shopee’s app over the past month alone (there are many more which I’ve omitted):
If you happen to live in another part of the world but would like to witness this phenomenon with your own eyes, just fire up a VPN and geolocate your position in any ASEAN country - then download the Shopee app and hit the big Free Shipping button on the home page. Alternatively, do the same VPN trick and just randomly watch Youtube ads - you’ll inevitably come across something like this:
The reason why I say “inevitably” is simply due to the sheer frequency of Youtube ads I’ve been seeing from Lazada (a Shopee competitor backed by Alibaba) in my ASEAN country of Malaysia - it’s not a stretch to say that 10% of all my Youtube ads are from Lazada. Now multiply that by every ASEAN country and imagine what the table stakes for that annual A&P spend looks like. If you’d like more proof, just perform a Youtube search for “<ASEAN country> Lazada Ads” and bask in the non-stop bevy of high-octane family-friendly party neon.
The point I’m trying to make is that Shopee’s wildly excessive Sales & Marketing spend ad-venture is far from over. Can they continue to decline as a % of Revenue? Sure. Is that proof of improving unit economics? Eh… I’ll wait and see. This is the digital equivalent of those Persian Rug stores with persistent “50% OFF CLOSING DOWN SALE” in your average suburban Class B mall.
One final point I’d like to make to wrap up this section: SHOPEE IS NOT A SAAS BUSINESS! As I’ve pointed out above, we are simply attempting to model their economics based on a SaaS business framework for lack of better alternatives - whereupon a Margin of Safety should be applied in order to get a general sense for their historical performance based on GAAP accounting metrics. However, if they can’t even stand up to such generalized standards in the best of circumstances, what might their future unit economics actually look like?
Recall from above how we discussed that the E-commerce sector is simply a disruption of the highly commoditized B&M Retail sector - once everyone stands on tiptoes, the entire playing field just goes back to perpetual price wars. This is quite evident in how aggressive all the respective ASEAN e-commerce players (e.g. Shopee, Lazada, Grab, GoTo) are spending on Sales & Marketing expenses - if there were even the semblance of a moat in this sector, this wouldn’t be happening. With such an unforgiving business environment as context, it stands to reason that a conservative investor would only question how much worser their actual unit economics might look like when adjusted for a regular Retail-like “SSSG” basis, rather than a “SaaS-like” recurring revenue basis.
As I’ve mentioned above, most of the assessment of a startup’s business performance can be boiled down to their Revenue line item. However, there are still a few other line items worthy of mention in SE’s GAAP-based three statements - which I shall cover below.
The elephant in the room is clearly their large Cash & Cash Equivalents (C&CE) (including Investments) of $11.2B. If we subtract their Total Debt, that brings their Net Cash balance to $7.2B. Even under the most conservative FCF assumption of roughly negative -$2B annually, that still gives them at least 3 years of cash runway.
I’d say this is most definitely a good thing - however, I don’t think investors should deduct their Net Cash from their market cap to arrive at Enterprise Value to inform their valuations, since this cash is more likely than not going to be spent over the next few years.
Why are their C&CE balances so high? They raised a total of $6B in equity and convertible bond financing via public markets on Sept 2021. This reminds me of when Facebook raised $240M from Microsoft in 2008, on the advice of investment bankers who gave the suggestion to Zuckerberg that equity markets were too hot at the time - and was arguably instrumental in allowing the then-startup to survive the 2008 GFC.
However, the other more interesting observation lies with their Cash Flows. Look at the table above: notice how their WC changes/OCF was a whopping 613% in FY21 - and a still colossal 275% in FY20? What this means is that most of SE’s positive OCF & FCF are a function of their hugely positive working capital (WC) - rather than “cash profits” from sales of goods & services.
If we take a look at the breakdown of their working capital cash flows, we can observe how two line item contribute to the bulk of their WC changes during the period - i.e. Accrued exp and other payables & Deferred revenues:
The bulk of their Accrued exp and other payables are in turn contributed by just one line item (which can be found in Note 12 of their AR21) - Escrow payables. These are just the funds they hold in escrow on behalf of buyers until they get their purchases delivered to their homes + a refund window of about 3-4 days (for potentially damaged items). The reason why these are so large in magnitude relative to their other line items, is because they are basically just escrow funds representing a % of GMV which are withheld by Shopee as buyer protection for between 1-2 weeks.
Shopee’s Deferred revenues are simply a function of the customer funds held in their digital wallets - both in the Gaming segment and E-wallet segment. The parallel to draw here is Starbucks Cards balances - which tend to be reloaded weeks in advance of the actual coffee purchase being made. This is certainly a legitimate source of float as well.
However, while both their Escrow payables and Deferred revenues provide SE with a hefty dose of cash flow (as they are basically 0% 1-week loans from customers (i.e. float)), they do not represent Shopee’s revenue - and thus should not be considered in the evaluation of their historical business performance.
Investors who try to use SE’s positive FCF as a proxy for their future profitability need to understand that this is effectively just a rolling 1-week working capital loan - borrowed at a 0% interest rate. If we were to back them out to reflect “actual” OCF, you can see how FY21’s OCF before AE+DR (i.e. OCF before accured exp + deferred revenues) is a whopping negative -$1,212M. I would argue that this is a closer proxy to their “cash profits” than their headline FCF number:
Aside from that, there is nothing particularly notable about SE’s three statements. Feel free to scroll through them yourself here:
Up to this point, there is still room to give Shopee the benefit of the doubt in terms of ramping up their unit economics going forward. I’m pretty sure that in the upcoming quarterly announcement this week, SE’s management will be announcing various cost-cutting measures in line with the capital discipline that markets are requiring from their e-commerce management teams. Given how much of their cost base is represented by S&M exp, I fully expect the sales & marketing budget to be cut significantly in forward-looking guidance.
If so - and if the ASEAN e-commerce sector as a whole is capable of maintaining competitive discipline and follows suit - then there is certainly an argument to be made for SE improving their unit economics going forward. However, there is still one final nail in the coffin for the bullish SE thesis (if there weren’t enough already) - their current Valuations.
At their latest share price of $70.33, that implies a P/S ratio of 3.8x. And since they don’t have profits yet, we’re going to attempt some valuation gymnastics in order to estimate how much Margin of Safety exists at their current share price. The way we’re going to approach it is to pull some random numbers out of the air - and if there is a sufficient Margin of Safety under these assumptions, then we can move on to the next step.
So the random numbers that we’re going to pull out of the air are: 1) Average LT net margin of 5% and 2) Average LT revenues = 200% FY21’s revenues:
The reasoning behind choosing these assumptions are as follows:
Average LT net margin of 5%: As we’ve mentioned above, the e-commerce sector is destined to revert to the commoditized state of its predecessor B&M retail sector. I understand that many SE bulls have been preaching about cultural advantages that they have like ‘hyperlocalization’ and a ‘996 work culture’ - but these are commoditized processes which can eventually be copied by competitors. The final destination of the economics of the e-commerce sector is those of its predecessor B&M retail sector - which means that we should be conservatively assigning B&M retail margins to SE’s LT net margin assumptions. These tend to hover between 2-5%; so let’s just give SE a 5% net margin assumption for the benefit of doubt that they can somehow outperform.
Average LT revenues = 200% FY21’s revenues: Here, we’re again pulling random number out of the air, simply for the purpose of estimating if any Margin of Safety exists. The number I’ve chosen for their LT normalized revenues is 200% of their latest FY21 revenues - call it $20B to be safe. This is simply a reflection of where I think their LT revenues will finally normalize - if you think this is too conservative, please feel free to adjust it as you see fit.
Under these assumptions, the LT implied PE ratio that we arrive at is 37.6x. As a reminder, we are simply pulling random numbers out of the air in order to estimate if any Margin of Safety exists - so these assumptions are not required to be pinpoint accurate.
However, we can clearly see that at a LT implied PE ratio of 37.6x there is absolutely no Margin of Safety at their current share price. Despite the high-growth backdrop of the future ASEAN economy, it would have to be closer to 20x normalized PE (i.e. $40/share) before I would consider recommending this stock as a Buy - especially considering all the issues that we’ve discussed earlier regarding their unit economics.
Hence, even if you gave SE the benefit of the doubt regarding the future state of their unit economics, I don’t think that a conservative investor can reasonably justify buying SE’s shares at its current share price. It’s not like there aren’t literally a million other stocks out there for you to consider investing your capital into.
In summary, let us wrap up everything we’ve discussed so far:
Evaluating the business performance of startups can be performed via two measures, by asking: 1) Is There Product Market Fit? and 2) Does Operating Leverage Exist? Shopee meets the first criteria but arguably not the second.
Shopee has a relatively decent proxy to the Efficiency Score - an FY21 Rev/S&M ratio of 1.7x. Even though this is only a proxy, there is a sufficiently large margin of error from the breakeven level of 1.0x to justify the existence of Product Market Fit.
While Shopee’s profit margins appear to indicate improving unit economics, most of them are still in deeply negative territory - and hence are unreliable for trend analysis (e.g. extrapolation). A comparison to early Amazon’s unit economics trajectory over time reveals that Shopee’s unit economics might possibly even be moving in the wrong direction.
In the absence of clear signs of improving unit economics, Shopee faces some extremely challenging headwinds going forward in its gradually “commoditizing” ASEAN E-commerce sector - especially considering its existing scale (>50% of ASEAN e-commerce market share) and the aggressive competitive landscape.
Can SE survive? Yes, they have ample Cash balances and healthy Cash Flows. However, SE’s positive FCF shouldn’t be taken as a proxy for their future “cash profits” - as they are mostly dominated by contributions from working capital (i.e. escrow payables + deferred revenues).
Even if we gave Shopee the benefit of the doubt on the unit economics front, SE’s current LT implied normalized PE of 37.6x is enough to turn most conservative investors off, given what they have to offer.
As such, we can quite conclusively state that SE is not undervalued now. What does that mean for existing or potential SE investors? I cannot skillfully answer that, as much of their future share price trajectory depends on macroeconomic developments - which as we can see, has become increasingly volatile. But one thing I’m pretty certain about is that they’re not currently undervalued on a purely fundamental basis.
Trust me, I started out this report with very bullish expectations - especially considering that SE’s share price had already cratered by -80% over the past 7 months. However, I was eventually forced to concede after completing my analysis that - despite looking at things from various different angles - I still wasn’t able to justify a Buy recommendation for SEA Ltd at its current share price of around $70. There are just too many headwinds to justify buying their shares today, and I cannot in good conscience recommend others to do what I myself wouldn’t.
However, I’d like to think that there is still value in this report - as it is perhaps one of the few SE investment reports out there which actually attempts to value SE and evaluate their historical performance using GAAP accounting metrics - in contrast to non-GAAP figures. Maybe the share price will fall by another 50% from here - and then we can dig this report out again for a Buy call. We’ll see.