I took a small position in the last weeks in Global-e. After hearing some hard to believe comments in a recent discussion with fellow investors, I thought to delve in and look for myself. Did they really have an Apple-esque take-rate of almost 20% of GMV as one of the commentators in that discussion opined? Or did we not do our homework here? So yesterday I unpacked their numbers and came to (for me at least) some surprising conclusions.
Global e-online is a recent foreign listing. The stock opened at $25.50 on 12 May and is now at $70, for a $10bn market cap and a P/S (annualised last Q revenue) of just under 50.
The company is based in Israel and has 522 merchants customers (and counting!) from which they derive revenue. They facilitate cross-border e-commerce for these merchants - by enabling them to sell to customers outside their home target market. These 522 customer merchants are currently mainly based in the UK (49.6%) followed by the US (34.9%) and EU (13.8%).
The three co-founders are Amir Schlachet (CEO), Shahar Tamari (President) and Nir Debbi (COO). The CFO is Ofer Koren and CTO Eden Zaharoni. Each of the three co-founders has sufficient skin in the game, owning about 4% of the company worth approx $400m each at current prices, while the CFO has shares worth around $15m and the CTO around $40m. So all good on this front: good alignment of interests of the core management team and us shareholders.
Amir Schlachet, CEO (co-founder) : ±4%
Nir Debbi, President (co-founder) : ±4%
What does he do? → From 1 July Nir Debbi transitioned from Chief Marketing Officer to President responsible for sales and customer success (and in this capacity supervises the newly appointed Chief Revenue Officer). He is responsible for both business development and corporate development too. Still sounds very much like a CMO/CCO to me.
Shahar Tamari, COO (co-founder) : ±4%
Eden Zaharoni, CTO : ±0.4%
Ofer Koren, CFO : ±0.15%
What I really like about this team is that they were all in it from the start; not just the three founders, but also the CTO has been there for 8+ years. The only exception is the CFO who’s been there only a year. However he had a long stint - lastly as CFO - at Bank Hapoalim, which is where the CEO, COO, and President also worked. And the CTO worked at 888.com which is where the COO also had a stint.
These guys know each other really well and they’ve been building this together.
Amir Schlachet, the CEO, has exceptionally strong credentials. He was in the Israel Air force in tech and intelligence positions for eight years, ending that part of his career as the commander of the prestigious Talpiot Military Academy. I’ve read a little about this academy and it is exceptional. A very small group of recruits are accepted each year. Out of about ten thousand top school-leavers, a smaller group of about 200 are subjected to additional tests and from this group, about 60 recruits get to proceed with a program lasting nine years - focused on tech and leadership development. His military career was followed by two years at McKinsey and five as Exec assistant to the CEO of Bank Hapoalim before co-founding Global-e with three colleagues from the bank, which he no doubt hand-picked.
Nothing not to like on the management front.
EBITDA and cash flow
Operating cash flow $m 2018 2019 2020 -8 7 29.3 Op cash % of revenue 2018 2019 2020 -20% 10% 21%
Op cash flow is moving in the right direction and the company is cash generative already. And EBITDA shows good progress too: it was $12.8m in H1 2021 vs $2.7m in H1 2020, a big jump.
Customers & concentration
Global-e have 522 merchant customers as of the end of their Q2, up from 422 end 2020 and 283 end 2019. Decent growth. As is often the case for early stage companies with relatively few customers, they have significant, but rapidly decreasing customer concentration. Their largest customer contributed 15.9% of H1 2021 revenue, vs 22% in H1 2020.
Their top 10 merchants represented 37% of H1 2021 revenue, down from 44% in 2019, but the same as 2020. High. Very high actually…
Even though this aspect is improving customer concentration is a red flag for me.
Now we get to the meat of the matter. Revenue and revenue growth is arguably the most important metric for our companies, so I will spend some time nitpicking here.
$m 2019 12.5 12.7 16.2 24.5 2020 19.7 29.8 33.3 53.6 2021 46.2 57.2 QoQ% 2019 2% 28% 51% 2020 -20% 51% 12% 61% 2021 -14% 24% YoY% 2020 58% 135% 106% 119% 2021 135% 92%
Clearly revenue has been accelerating at a very healthy clip, and it is highly seasonal, with a much stronger H2 than H1. So quite a lot to come in the second half of the year. All good.
In 2020 59% was from UK merchants selling outside of their borders, 25% US, and 15% EU. Non UK has grown rapidly; in 2018 UK outbound still comprised 91% of revenue, so they had almost nothing outside of the UK just 3 years ago.
For Y/E 2020 geographical growth was as follows:
$m 2019 2020 UK 51.7 80.1 -> up 55% yoy US 9.5 34.1 -> up 258% yoy EU 4.3 21.2 -> up 393% yoy
Growth in the US and EU has been exploding in the last year, and this is pre the Shopify deal, which should add a firecracker to the growth rate. Still all good.
Revenue is split roughly 37% from service fees and 63% from fulfilment services. And this last part is why the GM% is so low. Not so good.
For fulfilment services, they act as principal to actually deliver the goods, but use, and pay for, DHL and other couriers/delivery agents to actually deliver the goods, with them orchestrating the whole. Because of this, they record fulfilment revenues gross and make a low-ish margin as a result. Service fees are recorded net and therefore have a much higher GM. Here, they act as agent and record only their part of the GMV which the customer sells.
→ So the service fees represent their take-rate of GMV, whereas the fulfilment fees are essentially global courier and delivery services, which they in turn sub-contract to others.
To get to what’s going on with revenues we need to dig a bit more and try to understand this split. They give the split for H1 2021 vs H1 2020:
H1 $m 2020 2021 %yoy Rev 49.5 103.4 109% Serv 17.3 38.2 121% Fulfil 32.2 65.2 102%
I guess this is good news as it shows the Service revenues are growing very fast - 121% yoy, and faster than fulfilment revenues at 102%. However given that Service revenue is simply a % of GMV, we can calculate the take-rate as 6.2% in H1 of 2020, increasing slightly to 6.4% in H1 2021. GMV shows a similar trend so I’m not including the numbers here.
So not an almost 20% take-rate; rather a much more modest, reasonable and slightly increasing take-rate of 6.5% in Q2 of 2021 / 6.4% in H1. If you divide the GMV with total revenue you get close to the 20%, but I don’t think that’s correct…
But this story is not complete without the margin picture too.
In their F-1 they give the CoS split of H1 2020 and H1 2021, so we can calculate the gross margin for the two revenue types:
H1 2020 2021 %pt up GP% 31.3% 34.7% 3% Serv 65.3% 70.4% 5% Fulfil 13.0% 13.8% 1%
And here it gets interesting. The service revenue part is the valuable part revenue-wise. It has SaaS-like margins of 70% - up a very respectable 5%pts vs the prior year - whereas fulfilment has, and will retain, mediocre 14%-ish margins as this is just a mark-up on outsourced shipping, handling and delivery fees.
Looking at it this way puts the revenue of this company into perspective and the valuation into the stratospheric range imo. The “real” SaaS revenue - the Services part - was only $21.1m in Q2 2021. Adding the full fulfilment margin to that of say 14% of $36.1m or around $5m gives $26m. Annualised that’s $104m. Lets call that annualised adjusted SaaS-like revenue - another non-GAAP measure - and we get a Price to adjusted annualised SaaS-like revenue of about 100x !!
The Shopify connection
Shopify is a major shareholder, with just under 6.7%, or 9.7m shares currently. So how did this come to pass?
Basically Global-e gave Shopify these shares in exchange for Shopify agreeing to the Services and Partnership agreement with Global-e. A big deal therefore with a big price tag for Global-e.
The mechanics are that Global-e gave Shopify a warrant to purchase 19.6m shares at $1c each - so basically for free - on 12 April 2021, and the warrant vests over time. To date Shopify has exercised the warrant to the tune of buying 9.7m shares currently worth about $700m - for $97k.
The warrant is payment to Shopify for integration of the Global-e solution into Shopify’s platform and to their merchants.
To date the cost of the partnership is ±$700m to Global-e shareholders, and at the current share price the total cost, assuming it was immediately exercised in full comes to ±$1.4bn. A pretty penny. And where will this show up in the P&L? This is where accounting gets weird. About a third of this - $470m - will be recognised as an expense over a period of four years, in equal instalments. That is because the warrants got valued when they were issued using the then-prevailing share price of around $25 and a bunch of other assumptions. So Shopify gets paid ±$1.4bn or more (if the share price appreciates further) and Global-e expenses $470m over 4 years in S&M. And quite possibly this will then get excluded when showing Non-GAAP measures.
But enough accounting weirdness. Bottom line is Global-e is paying a lot for the privilege of selling to Shopify customers.
What about the other side? What does Global-e get for this? They get to sell their service to Shopify customers which is worth reading in its entirety:
We are party to a Services and Partnership Agreement with Shopify Inc. and its affiliate, dated April 12, 2021, pursuant to which we will be an exclusive third-party provider of an end-to-end cross-border solution that includes localization, merchant of record, duty and tax calculation and remittance, and shipping services for merchants in a single solution with permissions to access and integrate into the Shopify platform checkout. Shopify however will be entitled to offer its own native, whitelabel, or branded partnership merchant of record solution to any merchants without limitation. We will pay Shopify a fee equal to a percentage of the GMV for all transactions processed through our platform for applicable Shopify merchants. Although payment of such fee may initially negatively impact margins, we believe that our expanded partnership will afford us an enhanced market position, enabling us to increase GMV, and positions us to realize efficiencies in winning and onboarding Shopify merchants as a result of our permissions to access and exclusively integrate into the Shopify platform checkout.
Ok, so let’s unpack this a bit. Global-e gets to be the exclusive partner to provide an integrated combi of localization, merchant of record, duty & tax and remittance and shipping and pays a % of their take-rate in exchange (in addition to the warrant). Shopify is free to do what they want wrt merchant of record services without limitation. They can do it themselves, partner with someone, have a branded partnership, white-label someone else’s tech. And my understanding here is that merchant of record essentially means the seller first sells to Shopify, who then sells to the end customer. Which makes this out rather wide as it can include all of the other services that are “exclusive” to Global-e, provided they become the merchant of record. I struggle to understand exactly what is then exclusive here. If Shopify insert themselves into the chain then there is no exclusivity?
On the face of it, it sounds to me like the exclusivity could be thin. But then management would have to be exceptionally inept if they were to give away $1.4bn and counting plus a share of all GMV if the exclusivity and opportunity was not, in fact, significant.
And we know management is anything but. Therefore the benefits that they are expecting must be huge and the exclusivity deemed meaningful.
However, just to put this deal into perspective, that $1.4bn is a type of marketing spend - they are paying Shopify in shares to be able to sell to Shopify customers. And even though they will only expense $470m of this in S&M, as the warrants were priced when the share price was a lot lower, that is still huge for this company. In the whole of 2020 they spent $9.8m on Sales & Marketing, up from $4.6m in 2019. The Shopify expense will dwarf this: even using the $470m over 4 years equates to $118m per annum vs the $9.8m spent in 2020. Using the economic value of $1bn+ is over $350m pa. It’s a bet the company/transformative (pick your word) type of deal imho.
I guess we will need to watch and see how this pans out. But for me this is sufficiently complex to seriously wonder whether I want to ride this out, or simply watch from the sidelines first.
After reading Starrob’s post about them and Bear and others’ comments, I listened to the conference call and came away very impressed by the team, revenue growth and the potential of the Shopify deal - even though it was not expected to contribute before next year. Management’s presentation of the progress of the business and answers to questions were really smooth and the team exuded confidence.
But I did not do the level of analysis that I did above before buying my small position. I told myself that the low-ish GM% is not such a big issue as it was steadily ticking up nicely and revenue growth was fantastic. But as I thought about it more, it started feeling to me like the expectation is that the Shopify deal will blow everything out of the water, including the poor gross margins (but, wait - didn’t they say the Shopify deal will reduce margins further because they will be paying them a part of GMV?) and customer concentration because of the much bigger base that they will now exclusively be selling to (but wait again - how water-tight is that exclusivity? And didn’t they say they will only be done with technical integration next year?).
In addition, looking more deeply at the business model presented me with the problem that their real revenue was actually much smaller than it appears to be at first. That is because of the dynamics of revenue recognition: if you’re an agent you essentially show only the margin you make on a particular deal - think estate agent selling a house. Hence, in software-land you don’t show the full GMV flowing through your platform if you’re an e-com platform, but only your small part. Like Global-e does with their services business on which they take a 6.5% cut of GMV. But in their case the bigger part of their revenue - 63% in Q2, in fact - is actually just them being on the line, and hence not an agent, for delivery of goods, which they largely subcontract. This part of their revenue they show gross, and on that part they make a measly 14% gross margin. If they were in fact deemed to not be on the line for the actual delivery but in stead DHL was - a very subtle change - then they would only show their margin part as revenue. That would cut their revenue in half.
Looking only at the “core” Services revenue and handling the fulfilment revenue as if they were an agent yields a company with very high gross margins - like our other SaaS companies, but now with a very small revenue of only $26m in Q2 - but valued at more than $10bn!!
That means either that this aspect is still poorly understood (run for the hills) or that the market price reflects a huge, huge belief (gamble?) on the future success of the Shopify deal, for which I really cannot even begin to try to generate an expectation (run for the hills). Or both.
So this morning I briskly strolled to safety and sold it all. I’ll be watching this one from a safe distance rather than being invested for now.
Of course I could be dead wrong, but at least I’ll have slept better!