On 10 November 2026, Sea Limited held its third-quarter results conference call.
Sea reported strong growth: revenue rose 43% year-on-year to US$8.6 billion, while adjusted EBITDA increased 29%. Its shares, however, swung sharply on the day. The stock fell more than 10% in pre-market trading, later traded as much as 9.2% above its previous close, and finished the day at US$76.80, down 1.5%.
Three of the four analysts asked about Sea Cloud, Sea’s new business offering its AI infrastructure, models and applications to external customers. They wanted to know how much capital it would require, what returns it could generate and when it might become a meaningful business.
Chairman and CEO Forrest Li, President Chris Feng and CFO Tony Hou participated in the call.
Below is the Q&A transcript, edited for readability.
Q&A Session
Q1: Good evening, management. Congratulations on the solid set of results, and thanks for taking my questions. I have two questions.
First, on Sea Cloud: with plans to invest in cloud and AI infrastructure, how should investors think about the return profile of this investment? What are the key use cases or demand drivers that give you confidence Sea Cloud can become a meaningful business over time?
Second, Monee has reached meaningful scale, with over US$12 billion in loans outstanding, about 44 million active credit users, and an increasingly diversified business beyond the Shopee ecosystem. Looking ahead, what milestones are still needed for Monee to be viewed as a standalone financial services business? Strategically, how do you think about the benefits of keeping Monee integrated within Sea versus creating a more independent business over time?
A1:
Thank you, Alicia. As I mentioned in my remarks, Sea Cloud is a very new initiative at a very early stage. We have been thinking about potential new business opportunities in the AI era.
We increasingly believe that consumers and organisations around the world, especially in our region, will go through an AI transformation. Every organisation will need to build certain AI capabilities across all the layers: compute capacity, models, how to use and deploy those models affordably, efficiently and securely, and, on top of that, what software to use and what agents to build. This will have a meaningful impact, but it is not an easy process. It will take time and effort.
Over the past few years, we have gone through this first-hand in our own businesses. Although it has been challenging and we have put in a lot of effort, we are very excited about the impact. In my remarks, I mentioned fraud detection, customer service and content generation. These benefit our businesses significantly on a daily basis.
We strongly believe that the capabilities we have built, and the know-how and knowledge we have gained through our own use cases, will be valuable to other organisations. Not every organisation has the technical capabilities and internal usage that have allowed us to build these capabilities from scratch. Our aspiration is to externalise what we have already built and use our internal capabilities and resources to serve external customers.
We also believe adoption will accelerate as models get better and agents can do more things. Token costs will continue to fall, making AI more affordable and accessible.
That is the big picture we are looking at. We think demand will be huge, and we want to enter this business early. There are still many things we do not know and need to figure out, including which segments and types of customers to focus on, and the best business model to use.
We will get a better picture as we get into the business and gain first-hand experience. I am sure we will learn very quickly once we have a real business working on this. Even over the past few months, we have gained a lot of insight into the potential opportunities.
We will keep investors updated. Because this is a new and very dynamic opportunity, we want to be very frugal with capital expenditure and cash flow spending. From the top down, we believe in the big picture ahead of us. From the bottom up, we will be disciplined in assessing each project, customer and potential deal: what return we are looking at, and what it means for us financially.
At this moment, we do not have a very clear view of the return on investment for the business as a whole. We believe it will be very good, but we also want to stay humble. There are many things we do not know. We will take financial discipline very seriously at the project level, making sure each project generates an attractive return while minimising potential financial risk.
That is our overall thinking on Sea Cloud. It is still at the formation stage. Its profile is very different from our three existing businesses, which have taken us a decade to build to their current stage. Sea Cloud is a long-term bet and one of our bets in the AI era. We are very excited about it and will keep investors and shareholders updated on major developments.
On the Monee IPO, this is also a very exciting project. It has always been our aspiration to build Monee into a standalone financial services business. We believe it has the potential to be one of the best fintech businesses in the world.
We think we are on the right track. We now feel comfortable that Monee has reached the stage and scale to sustain itself. A standalone listing would give it more room and flexibility to continue growing.
In the past, Monee benefited hugely from being part of an integrated ecosystem. That has been a critical factor in its development. Moving forward, we expect Monee to continue growing, with the flexibility and room to expand beyond that ecosystem.
At the same time, the benefits of being together with Sea and Shopee will remain. As I mentioned in my remarks, Shopee and Monee will continue to collaborate very closely. There are huge synergies, and we will definitely not give those away. From Sea’s perspective, the capabilities we build in the AI era will continue to provide technical support for Monee’s growth.
We are not in any rush. We wanted to share this direction with investors and shareholders as early as possible. We are looking at a one-to-three-year timeline. As we kick off the process, there will be a lot to learn, and we will weigh different approaches and their pros and cons. We will keep shareholders and investors updated throughout the journey.
Q2: Thank you very much. My first question is on Shopee: when does monetisation start flowing through to profitability?
Despite strong growth in take rates and advertising, margins were still flat quarter-on-quarter at about 68 basis points. Could you elaborate on what specifically absorbed these monetisation gains? Which investments should start delivering operating leverage next year? Could you also comment on the differences in the profitability runway between ASEAN and Brazil, as well as between the content ecosystem and traditional shelf-based ecommerce?
My second question is on agentic AI at Shopee. You mentioned that you are an early connector partner for Muse. If product discovery moves through an internal agent, how do the economics of a Shopee order change, particularly advertising revenue and contribution profit? Do you retain the customer relationship and data?
Could you also comment on how Migoo is performing in terms of adoption and conversion, and how you see its role at Shopee as we move towards agentic AI?
A2:
On Shopee, we did see better monetisation over the quarter, as you pointed out, and EBITDA remained relatively stable.
Overall, we still see a lot of growth potential in both Southeast Asia and Brazil. This quarter, we reported growth of 23%, but excluding foreign exchange effects, growth was more than 26%, close to 26.5%. We see huge potential in our markets and would like to grow ecommerce penetration further in both Southeast Asia and Brazil.
At the same time, as Forrest mentioned in the opening remarks, we are committed to achieving EBITDA of more than US$1 billion this year. We expect monetisation to improve, primarily through advertising, with less coming from fixed commissions. Part of that will flow through to higher EBITDA over time, and part may go towards investment to grow our markets further.
Between Brazil and Southeast Asia, I do not think there is a particular difference in trends. Both regions are relatively stable from a quarter-to-quarter perspective.
For us, content comprises both videos and livestreaming. Our content ecosystem has been growing year-on-year and quarter-on-quarter, and its profitability has also improved. On a contribution margin basis, our content business is already profitable this quarter, which is a very good milestone for us.
On the second question about agentic AI, we work with a range of partners, starting with ChatGPT and now, as you mentioned, Muse. At this point, checkout, seller relationships and buyer relationships remain within Shopee. We do not see a meaningful impact on our economics in the near future.
At the same time, we work with many platforms as affiliate partners. For example, we have been working very closely with YouTube and Meta, through Facebook and Instagram, in a similar way to how we work with platforms such as Muse. These partners generate sizeable numbers of new users and traffic for our platform and have been part of our economics over the past year. We do not see this dramatically changing the picture when we work with Muse or ChatGPT.
Q3: Hi, thanks a lot, and congratulations on the good numbers.
I have two questions. First, on Shopee: when do you think investment will peak, considering your ongoing investments in ShopeeVIP, instant delivery and fulfilment?
Did I hear correctly that the content ecosystem is already EBITDA positive? If not, how much is it weighing on profitability? Could you also share the progress on fulfilment in both Asia and Brazil?
Second, on Sea Cloud: you have explained your aspirations in detail. Have you secured any enterprise contracts already? To clarify, would you invest in data centres and GPUs, or take a capital-light approach and lease these assets? Are there any guardrails on maximum capital investment over the next two to three years?
A3:
We have been investing in ShopeeVIP, fulfilment and other initiatives such as instant delivery. It is hard to say when investment will peak, as we are rolling out these programmes across different countries and markets.
We roll out these programmes in different countries and markets at different times. Some programmes in the countries where we started earlier have passed peak investment, while other countries started later. The country mix and pace of rollout have a significant impact on the pace of investment.
As I shared in the previous answers, we expect EBITDA per GMV to continue improving year-on-year, including the investments in these programmes. We manage this by balancing how much take rate we can achieve with how much we can invest while growing our ecosystem and user base.
For content, including short-form video and livestreaming, it is hard to isolate a particular segment’s EBITDA because there are many shared costs. We look at variable costs and contribution margins to our platform. Our content business has a positive contribution margin, so growing it adds to our profits.
On fulfilment penetration, overall fulfilment has been growing significantly year-on-year. We started in Southeast Asia earlier than in Brazil, so penetration as a percentage of the overall platform is higher in Southeast Asia. We believe penetration will continue to grow in both Brazil and Southeast Asia.
We also spend time optimising the cost of fulfilment, which is very important to us. We are seeing meaningful reductions in both Southeast Asia and Brazil. Part of this comes from the scale we have built over time, and part comes from better automation, which is an important tool for reducing costs.
Part of it also comes from optimising how we work with sellers, including inbound processes and how long we keep items in fulfilment. These contribute to fulfilment growth while improving the economics over time.
Sea Cloud is at the very beginning of its business. At this moment, we have not secured any enterprise contracts.
On investment in data centres and GPUs, we will probably invest in GPUs primarily for our own use, regardless of whether there is a Sea Cloud business or how it develops. We will need them for internal use cases.
On data centres, we will see what opportunities look like. In general, we are not interested in investing in data centres, but we may spend money to build them for our own use cases. For example, for Monee, regulatory requirements mean a lot of data needs to stay in Singapore. In those cases, we could evaluate building our own data centres, with a clear use case, versus leasing data centre capacity.
We will keep an open mind, but this is separate from the Sea Cloud business. From Sea’s perspective, we will keep those options open.
On capital investment guardrails, we will be very cautious. This is different from a consumer business, where you may need to invest a huge amount upfront, reach critical mass, and then see unit economics turn from negative to positive as density and volume increase. We have been through several rounds of that and know it very well.
The nature of Sea Cloud’s business is different. At least initially, it will be driven deal by deal, with specific customers, revenue projections, timelines and investment requirements.
As mentioned in my remarks, we will only pursue opportunities with clear customer commitments and a clear return profile. We will judge case by case whether the return is good enough for us. We do not have to pursue an opportunity if we do not like its return profile.
At the same time, we will look for ways to minimise the impact on our financials, especially cash flow and the balance sheet. There are ways to work with potential partners and through customer prepayments. The aim is to minimise the financial impact on us.
Q4: Hi, good evening. Thanks, management, for taking my questions. I have two questions.
The first is on Sea Cloud. Given that you are approaching it project by project and looking at each project’s return profile, could you share management’s strategic perspective? Looking three to five years ahead, how should we think about Sea Cloud in the overall business or profit mix? Could it become another meaningful business within that period?
My second question is on Shopee. With a dynamic macro environment, fuel prices and other uncertainties, how should we think about GMV growth in local currency going forward? Are you seeing a meaningful or sequential slowdown given these headwinds?
A4:
Thank you, Thomas. Strategically, and in terms of the big picture, we are very bullish and optimistic about Sea Cloud.
First, we believe AI is a fantastic technology that will have a huge impact on every organisation and business. Every organisation will need to figure out its own strategy, action plan and execution: how to make that transformation and stay relevant in the AI era. From that perspective, demand is huge.
Second, we believe AI adoption across industries will accelerate. Models are getting better, and agents can do more things and make a real impact on workflows and day-to-day use cases. At the same time, token costs and prices are falling rapidly. These factors give us conviction that the opportunity could be huge.
Whether this happens in three or five years depends. I do not have a clear view of the exact timeline, and things move very quickly in this industry.
We are building this business for more than three or five years. We want it to become a sustainable, permanent part of our businesses. That is our aspiration. Of course, we would only do this if we believed it could generate meaningful, sizeable revenue and profit over time.
Management, on Shopee
We do not see a big slowdown in Shopee’s GMV growth rate. As we shared earlier, in Q3 we saw meaningful growth, at a similar year-on-year rate to the previous quarter when excluding foreign exchange effects.
You are right that many macro factors are affecting consumption. Although we have noticed more value-seeking behaviour in our markets due to the macro environment, overall demand has been relatively resilient.
I would reiterate that our competitive advantages differentiate us in the current macro environment. We have built our platform around long-tail products, and our pricing is much more competitive than on other online platforms or offline alternatives.
On top of that, a large part of our platform’s GMV is driven by people who need to make purchases rather than in-house buying. These factors put us in a more resilient position.
For ecommerce growth, we will continue to monitor our markets very closely for any changes.
Given the things we have mentioned, and with ecommerce penetration still relatively low compared with many other markets, we believe there is a long runway to grow ecommerce in the region.
[THE END]
Disclaimer:
This translated transcript was compiled by Momentum Works from Sea Limited’s Q3 earnings call recording. It is intended for informational and analytical purposes only. The recording has not been confirmed by the speakers, and the translation may contain unintentional errors or omissions.
All original spoken content belongs to the respective speakers. Please refer to any official announcements or statements for authoritative reference.
Any analysis or commentary by Momentum Works is independent and does not represent the views of Sea Limited or any other organisations mentioned.
