SE beat revenue estimates by 6.2%. Shopee Gross Merchandise Value (GMV) growth decelerated from 28% last year to 23% this year, mainly due to ramping foreign exchange (FX) headwinds. FX neutral (FXN) growth was 26.5%, which is still almost 2 points slower than last year, but also a bit faster than two years ago. A firm dollar and rising U.S. yields pressured currencies across SE’s markets during Q3.

EBITDA margin didn’t expand despite GPM strength, strong ad revenue growth and healthy take rates. That’s because this company is choosing to reinvest all potential margin expansion back into supporting more growth. Monee credit growth is the most notable drag on that segment’s EBITDA performance.

Monetization and other margin tailwinds will keep making this business more efficient, and they’ll balance how much of that they plow back into operations and how much they allow to flow through to net income. Right now, that flow is more of a drip. Still, the drip is getting bigger where it matters most. Shopee’s adjusted EBITDA was 0.68% of GMV vs. 0.58% a year ago, so e-commerce profit per dollar of GMV is rising. A notable drag came from outside the three main segments, as the “Other Services” EBITDA loss widened to $84M from $25M Y/Y.

Shopee is still on track for about 25% GMV growth and $1B in adjusted EBITDA for 2026. Last quarter, the GMV guide was “25%+,” so this is a modest softening in wording. I’d chalk that up to the ramping FX headwinds covered above rather than anything more worrisome. That headwind isn’t going away just yet. The rupiah is still trading near record lows and has been Asia’s worst-performing major currency this year.

SE trades for about 26x forward EPS. EPS is expected to grow by 17% this year and by 39% next year. Those estimates will likely dip a bit following this EPS miss, but I don’t expect anything too drastic.

Shopee – Logistics & Fulfillment:

Similar to what we talk about with Uber and mobility, Sea Limited is building a barbell of delivery options.

On one end, it’s adding speed wherever demand and unit economics support that decision. On the other end, it’s adding cheaper options so every buyer has an affordable, convenient way to shop. Fast when you need it… cheap when you don’t. 

Starting with speedy delivery progress, its sub-one-hour (“instant”) delivery is now scaling. Like we see with Amazon and Mercado Libre, this naturally unlocks high-frequency use cases where speed is a prerequisite – mainly grocery, pharmacy and the everyday essentials umbrella. This is how you build the kind of habit and loyalty that not only earns you these orders, but many more items across other categories too. As this scaling takes place, there’s real evidence that economies of scale will drive ongoing cost per order leverage and enable long-term margin expansion. The pattern of falling costs and more instant delivery orders in Indonesia (2x Y/Y) was on full display this quarter.

For affordable delivery options, Indonesia’s slower, cheaper option is now over 50% of all orders in the market. Cost per order is 20% lower than the overall business, which gives plenty of room to pass on savings to customers by doing things like lowering free shipping minimums… again just like MELI is doing. Buyer self-collection is another one of these cheaper options. As a reminder, that’s when shoppers pick up parcels at a nearby pickup point or locker instead of getting door-to-door delivery. It now accounts for over 30% of orders in Singapore and over 15% in Malaysia. 

Fulfillment is scaling quickly too, with fulfilled order volume up 25% Q/Q. And while that’s undeniably encouraging, it’s even more so when considering improving expense dynamics as fixed cost leverage kicks in. Cost to fulfill is falling alongside delivery costs in Asia. Their Fulfilled by Shopee program continues to mature, notch better capacity utilization rates and find new sources of automation via robotics and AI. All of these trends are material and all of these remain early. They have so much more productive work left to do, creating room for more customer value and deeper relationships, as well as plenty more profit.

Improvements across all parts of fulfillment also make delivery (a key piece of fulfillment) faster. If Sea Limited gets smarter about placing inventory closer to customers or sorting goods for trucks, delivery naturally gets faster. As always, faster means higher conversion rates and happier customers. Specifically, parcels delivered the next day or sooner grew more than 60% Y/Y, more than twice as fast as overall orders.

Shopee – Engagement & Loyalty:

New user growth keeps accelerating, with 50% Y/Y new monthly active buyer (MAB) growth and 20% total MAB growth. That marks a 2 point acceleration compared to last quarter and a 5 point acceleration compared to last year – clear evidence that they should absolutely be spending to capture the large remaining runways they have ahead of them.

ShopeeVIP topped 21M members, up more than 40% Q/Q (vs. 45% Q/Q last quarter), with monthly retention holding steady compared to last quarter at about 80%. Good to see after language slipped from “above 80%” to “around 80%” for loyalty program retention between the Q1 2026 and Q2 2026 quarters. These members continue to punch well above their weight in terms of GMV contributions, driving nearly 25% of GMV despite representing a mid-single-digit % of total buyers. Just like the everyday essentials push, this is another great way to build habits. Two reasons for this. First, it’s where SE can collect the visible, high-margin subscription revenue that unlocks even more customer value. They recently introduced more tiers to strengthen that formula with new benefits like a Netflix subscription. Second, the loyalty program is also where the company can neatly display all of its customer value in one place. This is why spending curves for new members are steep. Notably, a lot more of these perks are being funded by merchants, as they see the value of this massive network effect and want their piece.

Shopee – Monetization & Macro:

Ad revenue rose more than 55% Y/Y and ad take rate (ad revenue as a % of GMV) rose by over 80 bps Y/Y. That’s a slowdown from 70%+ ad revenue growth both last quarter and last year. Part of this ongoing momentum is thanks to a strong uptake of its AI ad tools, with customers up 35% Y/Y and spending per merchant up 13% Y/Y.

There are reasons for macro fears here. Thailand’s household debt sits near 87% of GDP and its economy stalled in Q3 amid floods and rising living costs, while Indonesian consumers are dealing with a sharply weaker rupiah. Despite all of that, Shopee demand has been resilient. Why? The competitive backdrop is stable and not intensifying, which is nice to hear. Beyond that, ongoing opportunity for more market share gains helps buffer macro cyclicality. Finally, Shopee’s pricing is also among the best in class, while its GMV mix skews more towards “needs-based,” rather than “impulse buying.” That’s always more immune to fragile backdrops than more discretionary purchases.

Brazil:

Brazil remains Sea’s fastest-growing market, and it again took market share. Logistics improvements are a big reason why. Average buyer wait time fell 12% Y/Y and the percentage of parcels arriving the next day in Greater São Paulo rose from 33% to 40% Y/Y. Shopee Turbo is already under 4 hours to deliver in this part of Brazil as well. To build on this momentum elsewhere in the country, the company just added air delivery to sharply cut delivery times to Brazil’s Northern regions. 

Monee – Why The Spin-Off Now?

First, the news. Sea plans to list Monee, its financial services business, as a stand-alone public company within the next 1 to 3 years. This is coming from a point of strength and, per leadership, was always the plan. Their ability to be patient means the timing of the listing will be financial market dependent and opportunistic.

Why now? A separate listing gives it its own capital base to keep building and lets SE shed what is by far the most asset heavy, macro-tied piece of its business. That could be good for the valuation of the post-spin-off entity. Furthermore, Monee has, in a way, grown up. It needed Shopee’s scale and network effect to rapidly scale from a base of zero. More recently, the off-Shopee business in areas like BNPL and lending has grown into a much more meaningful piece of the overall segment. Off-Shopee BNPL more than doubled Y/Y to nearly 25% of the total portfolio (almost 40% in Malaysia). Monee’s stand-alone apps also more than doubled monthly transacting users Y/Y. Monee’s growth is clearly no longer dependent on the more mature e-commerce platform, which has plenty of its own investment needs. 

Leadership made one thing clear: the open data sharing and collaboration that makes Shopee and Monee each so much stronger will not go away. Monee will maintain access to this highly important information to sharpen underwriting and will remain a massive piece of checkout flows on the e-commerce platform too. While this wasn’t explicitly mentioned, I hope (and would think) the Shopee loyalty program would continue offering the Monee-based perks it already does.

Monee – More On Execution & Credit Health Update:

The goal hasn’t changed. To paraphrase, Sea wants Monee to be a full financial ecosystem built around inclusion and elite service for every need.

The star of that growth story right now is credit. Its loan book grew by 60% Y/Y compared to 62.5% growth last quarter, 71.3% growth 2 quarters ago and 69.8% growth last year. Consumer and merchant loan principal outstanding rose 60% Y/Y to $12.6B. Furthermore, average loan outstanding per user rose 23% Y/Y, and 40% of active credit users now use two or more Monee credit products. Perhaps most notably, active credit users rose about 28% Y/Y to nearly 44M, with nearly 6M first-time borrowers added this quarter. That’s a massive Q/Q addition, showing how aggressively they’re grabbing land right now. All of this means more provisioning, more front-loaded losses and more margin pressure for the segment (EBITDA margin was 18.4% vs. 26.1% Y/Y because of this).

Bold portfolio expansion is based on healthy unit economics, although they did acknowledge some deterioration in credit performance across Brazil and Thailand that prompted pullbacks in both markets. This (along with some mix shift) is why the non-performing loan (NPL) ratio worsened from 1.1% last year (and 1% last quarter) to 1.3%. Something to watch closely, but they think they’ve identified the problem cohorts and quickly made proper adjustments. Interesting that MELI didn’t cite this in their report. NU reports Thursday, so we’ll see what they say.

Sea Cloud:

The big news during the quarter was their Sea Cloud announcement. Over the last couple years, they’ve worked hard to develop AI services and tools internally for their own fraud detection, customer service and other needs. They’ve created “real capabilities across the infrastructure, model and AI agent application layers.” Now that they’ve developed these talents, they think there’s a large market to package and sell them to customers externally, better matching costs with revenue generation and giving them access to yet another secular growth story.

While Sea Cloud is very exciting, the news understandably sparked a lot of fears over hefty CapEx growth to support new project buildouts during a time when their other businesses still require a ton of attention. After hearing the actual plan for this segment, I think a lot of those fears should ease.

Furthermore, they aren’t looking to start building a ton of data centers or purchasing mountains of GPUs for bare metal rentals like a lot of neo-clouds do. Instead, they are determined to differentiate in two other areas: 

They are going after margin-rich inference and services that allow customers to use the cheapest AI asset possible. Their model doesn’t rely on selling access to compute via lofty token prices… it feeds on those token prices declining and allowing AI usage to broaden across far more companies. As leadership described, they’re building a model that thrives on token deflation because they view this ongoing deflation as inevitable (agreed). I personally love the approach, and love it even more considering the capital investment plan.

Outlays will be very limited and dependent on contract wins and direct revenue opportunities tied to the projects. Per leadership, this will be run as a “frugal start-up.” That's why this will not have any material impact on financials for a while. Great to hear from a FCF perspective.  They will not be building massive data centers either, with compute needs likely serviced via leasing models sort of like they do on the fulfillment side. They’re open to operating their own capacity down the road, but only if clear, tangible demand supports that decision. The capital outlays that are required will also be funded via customer prepayments, shrinking financial risk even further. The approach is conservative in every way, again with future demand signals informing whether or not they get more aggressive. I think there’s a great chance those signals turn out to be bright green, with this growing into another giant SE segment.

More Notes:

Great quarter. Growth looks fantastic while several trends make me confident in saying they’re right to be aggressively spending today. The runways remain massive. Competition remains fierce. They should be sprinting to take a larger piece of the pie and build more customer habits… so that’s what they’re doing. 

Separately, I selfishly love the Monee spin-off. I’ve wanted to get my hands on this name for a while but avoided doing so. I have enough exposure to international credit markets and don’t want more. With the Monee spin-off coming, that no longer looks like it will be an issue. I also love their decision to enter the AI/cloud space and how they’re approaching it in a low-risk, asset-light manner.  This quarter and the recent news all make me more interested in owning this than I already was. If I had any excess cash available at the moment, I’d seriously consider starting a position. Elite team. Elite business. Compelling multiple. Strong ongoing execution… and? A balance sheet that’s about to get a lot cleaner.