Pinned Tweet
I am thrilled to announce the launch of Vision Capital Fund, which has been in the making for over seven years.
We are focused on growing wealth for generations by delivering outstanding long-term investment returns from investing in exceptional companies that best reflect our vision for our future.
Website: visioncapitalfund.co/
Press Release: tinyurl.com/VCFpressrelease
Day One Investor Letter: tinyurl.com/VCFDayOne
NVIDIA $NVDA 3Q27 Earnings
- Rev $115.8b +103% ⤴️🟢
- GP $86.7b +107% ⤴️🟢 margin 74.8% +141 bps ✅
- NG EBIT $77.9b +116% ⤴️🟢 margin 67.2% +391 bps ✅
- EBIT $77.5b +115% ⤴️🟢 margin 66.9% +375 bps ✅
- NG Net Inc $64.9b +109% ⤴️🟢 margin 56.0% +165 bps ✅
- Net Inc $65.3b +105% ⤴️🟢 margin 56.4% +38 bps ✅
- OCF $48.3b +103% ⤴️🟢 margin 41.7% -1 bps ↘️🔴
- FCF $43.8b +98% ⤴️🟢 margin 37.8% -96 bps ↘️🔴
Revenue by Segment
- Compute & Networking $106b +108% ⤴️🟢
- Graphics $9.8b +61% ↗️🟢
- Data Center $107.6b +110% ⤴️🟢
- Data Center (Hyperscale) $60.1b +71% ⤴️🟢
- Data Center (AI Clouds, Industrial, & Enterprise) $47.5b +195% ⤴️🟢
- Edge Computing $8.2b +42% ↗️🟢
4Q27 Mgmt Guide
- Rev $130b +91% ⤴️🟢
- GPM 72%
1 | Founder, President, and CEO Jensen Huang resigned and the board unanimously accepted his resignation, and does not intend to reinstate Jensen as CEO.
On November 12, our Board of Directors announced that after disclosure of personal behavior inconsistent with the company's code of conduct, standards and cultural norms. Jensen Huang resigned his position as President, CEO and Board member. The NVIDIA board unanimously accepted his resignation and does not intend to reinstate Jensen as CEO.
2 | CFO Colette Kress is the interim CEO, and the Board is searching both internally and externally for a new CEO.
The Board appointed Colette Kress, NVIDIA's long time and well-respected CFO as interim CEO. In addition, the Board announced it was initiating a search for a permanent Chief Executive, and we have since retained Heidrick & Struggles to lead the search.
We have an incredibly talented leadership team and we have every confidence in Colette's leadership, integrity and ability to deliver as interim CEO. Colette has been at the center of this company's extraordinary growth for the last 13 years, instrumental in executing strategy and with a deep understanding of our business, customers, operations and financial position.
3 | While the board wants to find the new CEO rapidly, the right one matters.
Our new CEO will have to navigate those waters very carefully. So the search will be thorough, and we want to install the CEO, the permanent CEO as quickly as possible, but we won't expedite the CEO before finding exactly the right person.
4 | The board has established a national security and export control oversight committee.
This morning, the company announced that our Board has established a national security and export control oversight committee that will include a select group of our independent Board of Directors, senior external advisers and legal counsel. This committee will enhance our engagement with the U.S. government, including all relevant agencies. We will not be addressing or taking questions on today's call regarding the circumstances of Jensen's resignation and the ongoing investigation. Our Board will communicate further when appropriate.
On behalf of our shareholders, the Board is proactively dealing with immediately -- the immediate consequences facing the company as well as enhancing our oversight capacity going forward.
5 | NVIDIA has the best technologists with a world-class multi-year road map. The question is what happens after the current line-up.
Remember, we have the best technologists on the planet with a world-class multiyear road map, and I think that's important. From a business perspective, it's business as usual. We've engaged with our customers. We've engaged with our suppliers, our partners. They're all 100% up behind us. Customers have expressed their support to the business outlook and helping us through that.
6 | Vera Rubin is off to a strong start in its first quarter of production shipments.
Vera Rubin is off to an extraordinary start. In its first quarter of production shipments, the Rubin platform generated nearly $20 billion in revenue, making it the fastest product ramp in NVIDIA's history. We're addressing strong demand for compute across a diverse and extensive set of end customers, including hyperscalers, frontier labs, AI natives, enterprises and governments.
7 | Raising FY 2028 revenue guide to grow higher 75% instead of 70%.
This, combined with increased support from our supply base, we are raising our FY 2028 outlook. We expect revenue to grow approximately 75% YoY, up from 70% and off what is now a higher FY2027 base. Our outlook considers data center readiness, the strength and durability of end customer demand and our customers' access to capital…But we feel confident in terms of both the demand that we have in front of us as well as the supply and additional supply that we can do the 75%.
8 | Data Center revenue grew rapidly supported by hyperscalers, AI clouds, and industrial and enterprise segments.
Q3 data center revenue increased 21% QoQ to $108 billion with balanced growth across hyperscale and AI cloud, industrial and enterprise subsegments. Hyperscale revenue of $60 billion grew 23% sequentially, driven by the Vera Rubin ramp and sustained strength in Grace Blackwell. Our hyperscale customers cloud revenue growth accelerated for the sixth consecutive quarter.
9 | Cloud industry backlog now sits at $2.5T up from $2T in 2Q27. Expect major frontier labs to account for 15% of data centre
Cloud industry backlog now sits at $2.5 trillion, up from $2 trillion just 3 months ago, underscoring the signi cant multiyear opportunity ahead of our partners and of NVIDIA. Proprietary models from OpenAI and Anthropic continue to advance the frontier, with OpenAI launching its latest GPT 6 models and Anthropic releasing its Claude 5.5 models. As frontier models become more capable, adoption and revenue are scaling rapidly with Anthropic's annualized revenue reaching $81 billion at the end of October. We expect the major Frontier Labs to contribute approximately 15% of our data center revenue this year.
10 | Acquiring Hugging Face to scale it and make intelligence more open, capable, and more intelligent.
In September, we entered into a definitive agreement to acquire Hugging Face, the home of open model community, serving more than 118 million developers and 200,000 companies. Together, we will scale its platform and make intelligence more open, more capable and more accessible. nemotron is now the most widely used U.S. open model on OpenRouter and is where developers start to build their own intelligence. That foundation is empowering the ecosystem.
11 | Many leading enterprise software are turning open models into them own, using their own data, tools, and workflows.
This past quarter, leading enterprise software and security companies, including Microsoft, Salesforce, Palantir, CrowdStrike, Cadence and Synopsys announced products that use post trained open models tailored to domain-speci c use cases. Training partners such as Scale, Mercor, Prime Intellect and Applied Compute are helping enterprises turn open models into their own, using their data, tools and work ows. -- and start-ups across specialized domains, including [ Leland ] Science, Harvey, Therma and Abridge are using post-trained open models to deliver results in science, law and health care.
12 | Token prices have fallen 60%, and cheaper tokens are accelerating token usage and are spurring more demand. GPU rental prices have continued to rise and strong strongly.
Token prices have fallen nearly 60% since May, accelerating token usage and demand for compute. As a result, GPU rental prices are rising across generations. Hopper GPUs have recontracted at up to twice their original price under the multiyear agreements, while the B-300 rates have risen more than 60% since their introduction. NVIDIA compute remains fully utilized across every hyperscaler and neocloud we serve, underscoring the productivity, durability and fungibility of our infrastructure.
13 | Coreweave, Meta, NScale and OpenAI, AWS, OCI and SpaceXSI have all now deployed Vera Rubin, expect CPU revenues to 2X in FY28.
Meta extends to expand its deployment of Vera Rubin hardware with the goal of maximizing performance and efficiency for its next generation of AI agents. Agentic AI is driving extraordinary demand for CPUs far outstripping supply. NVIDIA Vera CPU meets the growing demand, delivering leading performance at data center processing and scientific simulation workloads. Coreweave, Meta, NScale and OpenAI have now adopted Vera, joining AWS, OCI and SpaceXSI, the leading partners we announced last quarter. We continue to see very strong demand for Vera and expect total CPU revenue to more than double in FY28.
14 | Physical AI revenue nearly 3X in Q2 and expect to cross $25b for FY27.
Physical AI revenue nearly tripled YoY in Q3 and is now on track to exceed $25 billion for the full year. At [ TTC Berlin ], we expanded our NVIDIA Isaac robotics platform with a new humanoid reference design built on Isaac GR00T and Jetson Thor.
15 | Q3 revenues of Hopper to China was $2.4b, no expectating any for Q4.
Q3 sales of licensed Hopper data center products to China was $2.4 billion. Our guidance does not have any licensed Hopper data center products included and we are no longer shipping licensed Hopper data center products to China.
16 | Purchase commitments increased to $390b, data center to $100b, neoclouds to $32b, and total guarantees $118b.
In Q3, our purchase commitments increased to $390 billion reflecting our confidence in the long-term sustainability of demand for NVIDIA infrastructure.
Together, our data center lease and cloud service commitments increased to a little over $100 billion, driven primarily by our long-term leases.
Commitments associated with our neocloud agreements totaled $32 billion at the end of Q3. Under these agreements, we earn revenue from the upfront sale of our infrastructure and participate in upside in rental revenue above a minimum commitment.
Total guarantees were $118 billion at the end of Q3, including the $105 billion in credit support for land power and shell capacity for the PORTS-Pike campus we discussed last quarter. The site will exclusively host NVIDIA infrastructure under a 20-year leases to OpenAI, SoftBank Energy plans are on track with capacity expected to come online in phases beginning in calendar 2028.
17 | NVIDIA’s neocloud commitments only take effect via conditional guarantees that only take effect if the neoclouds secure their own financing.
In the CFO commentary, it looks to be a new sentence, new paragraph on the neocloud. It says your commitments to take e ect only if the neoclouds secure nancing to purchase the data center infrastructure products and related compute capacity is deployed, maximum commitment, blah, blah, blah and it says certain commitments may be reduced or eliminated if they don't do this.
What we are indicating in our statement there is that is when our guarantee starts. So if all of that begins and unfortunately, they are delayed or they have a challenge in terms of building that, our guarantee doesn't exist anymore. Our guarantee for most of these are really about when they are ready to go,
Is it new? Did we change anything? No. What this is only saying is that we wanted to make sure that the clarity, there was other communication in another part that would have already communicated this, but we just wanted to make sure it was very clear as we've been receiving some of the questions on how that exactly want. .
18 | Vera Rubin delivered 10x more tokens per watt than Grace Blackwell, transition thus far has been exceptionally smooth, production timelines halved.
On the MLPerf endpoints benchmark, Vera Rubin delivered 10x more tokens per watt over GB00. With Vera Rubin still early in its life cycle, we expect continuous software optimization to deliver even greater performance over time. The transition from Grace Blackwell to Vera Rubin has been exceptionally smooth. Our hyperscale partners brought Vera Rubin into production in about 1 month and half the time required with Grace Blackwell.
19 | Networking revenues crossed $21b, and CPO roadmap remains on track. Vera Rubin Ultra in 2027 and Feynman in 2028 remains on track.
Our networking franchise continues to scale, with Q3 revenue exceeding $21 billion. Our co-packaged optics roadmap remains on track with CPO scale-out and volume production. At Supercomputing '26, Lambda announced the first production deployment of NVIDIA photonics, CPO switches, connecting more than 10,000 GB300 GPUs. Compared with pluggable optics, NVIDIA photonic switches delivered 5x better power efficiency, 15x higher reliability and 10% lower latency. We remain on track to launch Vera Rubin Ultra next year and Feynman in 2028.
20 | Q4 GPM guide is the bottom, due to NVIDIA honoring previous PO despite higher COGS pricing with higher memory prices, etc. Expect GPM to improve as they re-price them higher.
What we provided here in Q4 is our bottom. If you recall that was our bottom as we honored many of the different POs that we had already received from our customers, and we just wanted to make sure that, that was held solid as we move to a new pricing as we moved into the new year. your statement regarding the changes that we are making. Those changes regarding FY28 are in line with what we communicated last quarter that we see in FY28. We see no changes at this time.
21 | NVIDIA is focusing on allocating more to customers that are ready to go with the land, power, shell, to deploy the GPUs
So when we think about our allocation, we do think about very important parts that we need to understand to make sure that in the quarter that we need to ship, they will be able to use and use very efficiently in the very beginning. What that means is that land, power, shell, whether or not it's structured, ready to go, is already set up and completed to do so. That's one of the most important things. The second thing, though, is also to think about their capital needs. Their capital, it needs to be set up, ready to go for us to work in terms of that allocation.
22 | Non-IG customers will need to have solid offtake that they can run before they get GPU allocations.
Many of our customers outside of the IG will want to provide and think through in terms of their offtake as we will be needing an understanding of do they have that solid offtake that they will be able to run as well. So our allocation runs from we're here for everyone, particularly when there is already a PO available, and we feel those means in terms of capital will happen, and we will continue to be allocating as quickly as possible when we know the amount of supply that we will think we will have. So I believe this is a process that will continue to grow as more supply comes in. We'll continue working in terms of more allocation.
23 | Demand continues to exceed supply, and NVIDIA is trying to get even more supply to meet the overwhelming demand.
Where we stand right now, there is so much demand with such the largest companies continuing to work on so many agentic solutions, new models that are also available. And that's been our focus to assist. I'd say even in the last week, the continued ask, how are we going to receive more and more supply to help them with the demand that they have in front of them. They're all getting ready for that.
➡️ Final Takeaways on NVIDIA $NVDA:
NVIDIA continues to ride the long-term AI structural tailwind by being one of the key innovators and suppliers of the AI architecture supporting strong visibility with multi-year long-term capex. While NVIDIA dominates with strong unified and backward compatible architecture, software CUDA, growing networking moat, and its ability to keep innovating faster than its competitors at a faster cadence and at an ever larger scale. Vera CPUs in particular look attractive for NVIDIA to gain share as AI workloads increasingly shift towards more agentic and execution.
While the loss of the visionary leadership of Jensen Huang is unlikely to impact the next 1-2 years as the multi-year product cycle has already been previously laid out. It remains to be seen if the longer-term innovation will still occur without a visionary founder-led leader. While growth remains strong, we are concerned about NVIDIA's growth especially from the stronger hyperscalers who prefer to have a mix of NVIDIA’s and their own custom chips, and a higher mix from the more vulnerable neoclouds.
Watch if demand for AI compute persists and hyperscalers will reaccelerate capex spend growth significantly into 4Q27 and FY28, if they do not, the quality of NVIDIA’s long-term growth could be on a less firm footing that would warrant more caution on our long-term positioning.
Wise $WISE 1H27 Earnings
- Rev $1.4b +23% ↗️🟢
- EBT $303m +6% ↗️🟡 margin 20.8% -323 bps ↘️🔴
- Net Inc $249m +6% ↗️🟡 margin 17.1% -268 bps ↘️🔴
Revenue by Segment
- Txn Rev $1.1b +25% ↗️🟢
- Net Int Inc $351m +17% ↗️🟢
- Cross-Border Rev $705m +19% ↗️🟢
- Card Rev $252m +43% ↗️🟢
- Other Rev $145m +27% ↗️🟢
Revenue by Geography
- Europe (ex-UK) $394m +16% ↗️🟢
- APAC $326m +38% ↗️🟢
- US $225m +34% ↗️🟢
- UK $317m +9% ↗️🟡
- ROW $191m +27% ↗️🟢
Biz Metrics
- Active Cust 16.2m +21% ↗️🟢
- Personal 15.5m +21% ↗️🟢
- Business 769k +25% ↗️🟢
- Cross-Border Volume $141b +24% ↗️🟢
- Personal $97.5b +19% ↗️🟢
- Business $43.6b +35% ↗️🟢
- Cust Balances $32.1b +22% ↗️🟢
- Personal $20.2b +21% ↗️🟢
- Business $11.9b +23% ↗️🟢
- Asset Under Custody $11.4b +52% ↗️🟢
- Cust Holdings $43.5b +28% ↗️🟢
- Cross-Border Take-Rate 0.50% -2bps ↘️🟡
- Personal 0.55% -2bps ↘️🟡
- Business 0.38% -1bps ↘️🟡
1 | Wise now operates 80 regulatory licenses and continues to add more, adding Chile and Thailand recently.
We're already operating 80 regulatory licenses and adding more, recently started operating in Chile and Thailand. We built technical connections to local interbank clearing systems. In this period, we went live as a member in the Malaysian real-time DuitNow system. That follows Japan and Brazil from the previous period.
We launched transfers for people in Thailand earlier this year. And now in September, we also gave our Thai customers access to the multi-currency Wise account along with the Wise Card. In Chile, we're rolling out services starting with transfers.
2 | Went live with Malaysia’ DuitNow and able to better service the customer base.
I share your excitement about Malaysia, I love the country and our -- and we've actually built up a nice, substantial customer base in Malaysia over the years, our direct customer base. We have Wise Platform partners, banks in Malaysia. So it is within Asia, Malaysia is a good -- quite meaningful substantial country that we serve. So connecting to their direct payment system, their QR code system is very good and interesting.
3 | 78% of transactions arrived <20 seconds, and average price is 0.49%, 3bps lower YoY.
Investing in this infrastructure means transfers are getting faster and cheaper for our customers and for our Wise Platform clients alike. In the last quarter, 78% of transactions arrived instantly in under 20 seconds in the recipient's bank account on the other side of the world. Such speed was unimaginable even a few years ago. Thanks to the efficiency and scale of our infrastructure, we brought our average price down to just 0.49% last quarter. That's down 3 basis points compared to a year earlier.
4 | Wise Platform (B2B) continues to grow the fastest, +50% YoY and now is 6% of volumes vs 4% 18 months ago.
Wise Platform was also a highlight that it continued to be our fastest growing offering. In H1 '27, Wise Platform represents circa 6% of crossborder volume and delivered growth of nearly 50% YoY.
5 | Wise added new Wise platform partners and 2 large undisclosed North American banks in Q2, in addition to Morgan Stanley.
That's both adding new Wise platform partners, including Raiffeisen Bank, Standard Chartered, and Capitec on top of large partners increasing in their scale, likes of Morgan Stanley and Itau.
Further, in this period, we've added 2 large North American banks, including Morgan Stanley. This brings our coverage to 3 out of the top 20 North American banks live on platform.
For the 2 specific banks that you are mentioning, this is legacy banks that are joining the multitude of partners that we're having today.
6 | Added FX Forward hedging (up to 6 months) for business customers in the UK, approval tools for German businesses and accounting tools for Dutch and Belgian customers.
We are on the journey of making Wise Business more and more useful for larger, small businesses. This means supporting more sophisticated use cases, more accounting and approval flows. For example, we started rolling out the FX forwards feature in the U.K., which lets business customers lock an exchange rate up to 6 months ahead. An important threat for our business customers is how their international business banking on Wise is integrated to their local tooling. Accounting tools, for example. This quarter, we enabled syncing with DATEV for our German businesses and the local accounting tools for our Dutch and Belgian customers.
7 | Enabling FX forward hedging for UK business customers because they are getting bigger, and have more complicated cash flows and want to hedge their FX exposures, and because Wise already actively hedges for themselves, enabling this for their clients is easy.
I think I'd say you could try out the FX forward if you're a U.K. business customer. It is, like all of our features, we felt really simple. But it is indicative that our small business customers that we're serving are all getting bigger themselves, but also we're onboarding bigger and bigger small and medium businesses. And their cash flows are slightly more complicated, and they want to do slightly more sophisticated things with their -- with managing their cash flows. One is just having certainty on exchange rates so that they don't expose themselves into volatility and have more certainty in their business. So this is something that is easy for us to support them with. If we have probably one of the more sophisticated treasury operations of the world regarding to FX and the risk management features that we're using for ourselves, for our Wise Platform and our own client flows are very sophisticated already. So taking on the FX forwards thoughtfully is a small -- in some ways, a small step for our treasury team, but a big step for our customers.
8 | When Wise customers use it for cross-border payments, they tend to send and convert more, and become stickier.
our customers are using Wise for so much more. In H1 '27, for each $1 of cross-border volume sent or convert by Personal customers, they use Wise twice as much for all other activities, such as domestic transfers, cash payments, holding of assets or balances. And this is even more relevant for Business customers where from each $1 of cross-border volume, they use Wise 5x as much for all the activities, showing a strong adoption of our products.
9 | Wise incurred higher OPEX due to larger regulatory and compliance operations, and automation.
In H1 '27, servicing spent $251 million, up 15% versus H2 '26. A significant portion of this investment was related to regulatory and compliance operations as we build trusted compliance systems. We also focus on delivering an excellent experience to our customers. For this, we follow a hybrid approach supported by automation. In H1, circa 50% of customer service chatbot replies were automated. Our colleagues also leverage AI tools to increase the efficiency in their interactions with customers. In H1, 75% of inquiries received initial response within 60 seconds on live channels and 60 minutes on e-mail. And we resolved 95% of supported case within 12 hours, a 25% YoY increase.
10 | Wise Belgium subsidiary has agreed to a non-material settlement with the Brussel prosecutors with no admission of fault or wrongdoing.
Here, we announced today that we've been engaging -- Wise Belgium subsidiary have been engaging with our Brussels prosecutors in response to inquiries about our business. And an agreement has been reached to fully resolve and settle this matter with no admission of wrongdoing. So that's what we announced today.
Actually, the settlement, we agree with both parties that we will not disclose, but it was not material because, as you know, otherwise, we will have to disclose it to the market. So it's not material compared to the rest of our results today. The fact is we settled and we achieved an agreement with no admission of fault, which of wrongdoing, which is very important to us.
11 | Wise sees OPEX as an investment and want to be very efficient around it.
And on OpEx there, I really like the question because basically, it's also part of our philosophy and culture. We want to be e cient. We want to be disciplined. And OpEx is for us an investment. And as such, we really look at the return that we generate. And we think that in H1, we've been investing in a very ROI manner, and we want to continue to do so.
12 | Wise sees the elasticity on price reduction as long term, and don’t think they can do this in a sustainable manner, will consider cutting prices when they can.
The price reduction, as you know, the elasticity on price reduction is usually very long term. So I don't think that -- we always do this in a sustainable manner. So once we get the efficiency gain, when we are a clear view that these price adjustments are justified because of our cost structure and our discipline in terms of margin -- cost plus margin. So I will not put this directly in relation, but we will continue to consider price reduction where we -- when we see that we have the room to do this. I think this discipline is through the entire structure of our P&L. This is true for price adjustments. This is true for OpEx. So I don't -- I will not see this too critical.
13 | Wise invested more in R&D, with over 1,300 engineers, and deployments grew 57%. Added a second cloud hosting location to increase redundancy.
In H1 '27, we invested $292 million in tech development. We now have over 1,300 engineers deploying code. We are not just only growing our tech team, we are also implementing AI tools to increase the efficiency. In H1, our teams complete 45,000 deployments, an increase of 57% YoY. We also increased the redundancy of our infrastructure in H1 by adding a second cloud hosting location. We achieved this while still delivering efficiency in our transactional unit cost.
14 | Wise focused much more making its products safer with Wise Call, improving onboarding flow and costs
Our teams work to manage our risk profile, building tools that allow us to improve the resilience of our products and make our products even safer for customers. A couple of examples include financial crime detection with more targeted personal and business scams models and Wise Call Status, an update which allow customers to know if it is Wise who's speaking to them on the phone with the aim to reduce scams. Our teams also work really hard to drive scale and efficiency. For example, we recently deployed products to improve the onboarding flow for customers in Japan and in the U.S., significantly reducing the time it takes them to join Wise and the onboarding costs associated to -- with this. And very importantly, the aim of some of these is to drive growth. Kristo has already covered some examples in this section, but these would include the work needed to launch in Chile, our direct integrations in Malaysia, the launch of our products in Thailand, our [ Asian ] offering and many others. Lastly, we are continuing to invest in our core functions.
15 | Expect 2027 revenue growth to be 20%, and EBT margins at the high end of the 2025 range.
Before we conclude, I'd like to cover our margins and outlook. Income before tax was $335 million in H1 '27 with a margin of 22% as we seek to deliver margins in line with our medium-term targets. This margin demonstrates that it is possible to offer customers the best prices while investing into building a high pro table business. Our investment framework is evident in our financial projections. For 2027, we expect net revenue growth to be at the upper half of our medium-term guidance of 15% to 20%, and we continue to aim for full year margins around the high end of the 2025 range.
➡️ Key Takeaways on Wise $WISE
Wise continues to grow its business of B2C with personal and business strongly across customers, volumes, spend, balances. Lower prices and take-rates are strategic to win more new, drive higher retention, and usage. Higher near-term reinvestment re-normalizes profitability back to long-term levels that are highly strategic to strengthening its growing cross-border payments infrastructure moat of over 70 licenses and soon to be 8 direct instant payment connections that is increasingly becoming difficult to replicate and compete against, with the strength in on/off ramps. B2B platforms are growing faster and Wise becomes the lowest cost provider powering the 2nd/3rd tier FIs and gradually extending that to the more dominant ones. Wise is almost the “Costco/Shopify” of Cross-Border Payments” building the cross-border payments infrastructure moat of a great enduring generational business.
Beautiful article on good, quality, enduring, durable, controlled, restrained growth by @mastersinvest.
mastersinvest.com/newblog/20…
3
22
2,928
Nu Holdings $NU 3Q26 Earnings
- Rev $6.3b +45% ↗️🟢
- GP $2.6b +43% ↗️🟢 margin 41.5% -45 bps ↘️🔴
- EBT $1.7b +41% ↗️🟢 margin 27.9% -70 bps ↘️🔴
- Net Inc $1.1b +39% ↗️🟢 margin 17.4% -69 bps ↘️🔴
Revenue by Geography
- Brazil $4.0b +40% ↗️🟢
- Mexico $349m +58% ↗️🟢
- Colombia $143m +146% ⤴️🟢
Revenue by Type
- Rev $6.3b +45% ↗️🟢
- Credit Income $3.8b +52% ↗️🟢
- Float Income $1.6b +27% ↗️🟢
- Fee Income $0.9b +51% ↗️🟢
Biz Metrics
- Customers 142.6m +12% ↗️🟡
- Brazil Cust 119.9m +9% ↗️🟡
- BR Individual $113m +8% ↗️🟡
- BR SME $6.9m +17% ↗️🟢
- Mexico Cust 16.6m +27% ↗️🟢
- Colombia Cust 6.1m +61% ↗️🟢
- Total Active Cust 119.3m +13% ↗️🟡
- Credit Card Cust 59.8m +5% ↗️🟡
- Unsecured Lending Cust 15.2m +7% ↗️🟡
- Payment Volume $44.2b +21% ↗️🟢
- Credit PV $30.4b +26% ↗️🟢
- Debit PV $13.8b +12% ↗️🟡
- Total Loan Portfolio $41.6b +37% ↗️🟢
- Credit Card $27.2b +36% ↗️🟢
- Unsecured $11b +43% ↗️🟢
- Secured $3.4b +26% ↗️🟢
- Total IEP $25.2b +42% ↗️🟢
- Credit Card non-IEP $16.5b +30% ↗️🟢
- Credit Card IEP $10.7b +46% ↗️🟢
- Cust Loans $14.4b +39% ↗️🟢
- Total Deposits $47.5b +22% ↗️🟢
- Brazil $38.6b +27% ↗️🟢
- Mexico $5.3b -13% ↘️🔴
- Colombia $3.5b +52% ↗️🟢
- Cost of Deposits 87% of I/B rate ↘️🟢
- MARPAC (rev) $17.7 +21% ↗️🟢
- MACPAC (cost) $1.1 +10% ↗️🟡
- Efficiency ratio 19.3% -100bps ↘️🟢
- NII $3.9b +56% ↗️🟢
- NIM $2.1b +45% ↗️🟢 22.5% margin +290bps ✅
- Risk-adj NIM 12.1% +130bps ↗️🟢
- ROE 32% +50bps ↗️🟢
- NPL 15-90 4.9% +60bps ↗️🟡
- NPL 90+ 7.3% +60bps ↗️🟠
- NPL 15-90 over IEP 8.0% ↘️🟢
- NPL 90+ over IEP 12.1% ➡️🟢
1 | Continued strong growth across Brazil, Mexico, and Colombia.
In Brazil, we are approaching 120 million customers, and our activity rate expanded for the seventh consecutive quarter to a record 86.3%. That is remarkable at our stage in the country where we already serve more than 65% of the adult population.
Mexico is where we are now building that same scale. We reached 16.6 million customers, growing 27% YoY and now serving 17% of the adult population. This quarter also marked a major milestone in our transition to a full bank. We added 870,000 customers in the quarter, and our Nu banking status makes our proposition even more compelling.
Colombia is becoming another clear proof that our model travels. We closed the quarter with 6.1 million customers, already reaching 15% of the country's adult population and growing 62% YoY.
2 | Two one-off items that lowered risk-adjusted NIM, Desenrola had 55bps in Q2, little in Q3, and one-time receivables charge from BCB that reduced risk-adj NIM by 63bps. Ex these, recurring risk-adj NIM expanded from 11.9% to 12.7% QoQ.
I also want to acknowledge 2 one-off items that affected reported risk-adjusted NIM in the second and third quarters. In Q2, Desenrola contributed 55 basis points to risk-adjusted NIM and only had an immaterial benefit in Q3. In the third quarter, we recorded a onetime charge on receivables acquired from payment institutions that were subsequently placed under extrajudicial liquidation by Brazil Central Bank, reducing riskadjusted NIM by 63 basis points. This was unrelated to the credit performance of our customer loan portfolio. So excluding these events, our recurring risk-adjusted NIM expanded from 11.9% in Q2 to 12.7% in Q3.
3 | Not giving loan growth guidance for 2027 as it can be volatile QoQ in the short-term due to opportunistic lending initiatives. Not expecting a dramatic growth deceleration.
First, in terms of loan growth, we're not giving guidance for next year. But as you do point out, we have had very strong loan growth this past quarter and just in general, for the past couple of quarters, if you look at a YoY basis. What I'd ask this community to do is not to focus too much on individual quarter loan growth or quarter-over-quarter because we sometimes will expand our lending opportunistically. For example, when we have a new generation of credit models that are launched, that provides an opportunity to, for example, offer credit line increases or new loans for existing customer segments. With that being said, as David and I both discussed earlier, we're not seeing anything at this point that would cause us to expect a dramatic decrease in the near future in terms of our loan growth. We understand that there is a -- it's a fraught economic environment, but we are -- our underwriting approach is built for that.
4 | Nu is well prepared for BCB’s new rules on underwriting, and remains high disciplined with credit underwriting.
We assume that there is going to be a downturn in the next day and that losses will double. And we still have to remain profitable even under those circumstances. And that's been one of the core underwriting philosophies we've had since day 1. As far as the actual macro prudential measures go that the Central Bank has just announced, we believe that these rules reward responsible underwriting and transparency, and that's always guided how we o er credit. We are waiting for these rules to go into e ect, of course. But at this point, we are very well prepared for them, again, both individually as well as holistically in terms of the impact on our business.
5 | Expect risk-adj NIM to remain strong into Q4, don’t expect a significant decline. Nu continues to take disciplined risks. Q1 has a seasonally higher cost of risk.
And we expect risk-adjusted NIM to keep its underlying economic strength into Q4. We see no reason why that would be shifting. Keep in mind that when we take risks, it's deliberate. We expand into segments when the risk-adjusted returns justify it and where we can o er market-leading pricing. That's our approach. Now for next year, we're not giving guidance. And I would just remind the audience that in Q1, there is a higher seasonal cost of risk.
6 | NU partnered with >35 merchants, Amazon, Shopee, Samsung, and Uber.
In the third quarter, we partnered with more than 35 merchants, including Amazon, Shopee, Samsung and Uber to o er additional bene ts to our customers. This growing ecosystem helps us reward engagement and deepen customer relationships. This is what compounding at scale looks like, continuing to grow our credit business, expanding monetization opportunities and deepening relationships with an already massive customer base.
7 | Nu US is leveraging on the proven model in Brazil that is being replicated in Mexico, and accelerating in Colombia.
Brazil is our proven model. Mexico is replicating it with strong economics, and Colombia is now accelerating. Meanwhile, Nu U.S. and Nu Global are laying the foundations for our next phase of growth. What connects them is the same technology platform, operating model and customer-centric approach. We're not building separate businesses from scratch, but taking a proven model into new markets, leveraging what we learn along the way. This quarter marked an important step in taking our model beyond Latin America.
8 | Nu US rolled out accounts offering 3.5% APY interest and transfers to BR/MX/CO. Seeing early signs of word-of-mouth virality similar to early days.
On September 10, we launched Nu U.S., beginning the rollout of an account offering 3.5% APY with fee-free transfers to Brazil, Mexico and Colombia and a credit card with no annual fee and a unlimited 1.5% cash back. In just our rst few weeks, we already reached customers across all 50 states. In only the rst few weeks, we have begun seeing the early signs of wordof-mouth growth and virality that reminds us of our early days of Brazil, Mexico and Colombia growth. We expect word of mouth to be an important growth engine from the start, particularly among customers with ties to Latin America. This is exactly the strategy we're pursuing, reaching communities connected to our core markets and offering them a natural bridge between their nancial lives in the U.S. and their families and friends back home. Fee free transfer makes that bridge real. Customers can send money directly to Pix accounts in Brazil, to bank accounts in Mexico through CLABE and to Bre-B-enabled accounts in Colombia.
9 | Nu US is not competing in the very high-income, well-served competitive segment that require physical banking branches.
First, the U.S. is a gigantic market. And when you kind of dig in and double-click in some of the areas, what you realize is that there is a segment of the market that is incredibly well served, very competitive. Lots of different banks compete for these customers, very high income. You need banking branches to serve them well. You have millions of dollars.
10 | Nu is focusing on the mass segment in the US that has less financial access and lesser competition, where banks don’t pay interest on deposits, similar playbook to LATAM.
And then you have the largest percentage of the market that increasingly actually have less access to financial services, and there is actually less competition. And when you start realizing what that lack of competition means, you start realizing that consumers in the U.S. are actually paying a lot of money for banking services in nonobvious ways. So for example, over 60% to 70% of all deposits in the U.S., banks are paying 0. That's actually a very similar metric that what we saw in Brazil and Mexico, where big banks paid 0 to something like 70% of deposits before we began. The fact that we were able to build a consumer brand and use the significant cost structure advantage that we have to offer high yield and pass the yield to our customers is one way to use our cost structure advantage to e ectively shift the profit pool that the big incumbent banks have that today is going to pay banking branches all the way to the consumer. And so that's one opportunity.
11 | The US banks charge a lot of fees to the mass market consumer due to the lack of competition, similarly see opportunity for Nu to come in and flatten and take market share.
There is a lot of fees. The amount of fees when you look at fee revenue as a percent of total revenue is similar to what we saw in Latin America 15 years ago. It's partly is also a function of lack of competition, frankly, especially in the mass market consumer. You're paying very high fees for checking accounts, savings accounts, overdraft and remittances, which is an area where we obviously launch 5% to 10% to send money. So there is a lot of -- once you add all the di erent fees, when you add the lack of a yield, when you add some of the interest rates in certain segments, this is a very expensive banking system. And so from that perspective, if you think about it from rst principles, it's a bit of the same of what we saw in Brazil and Mexico.
12 | Nu is leveraging on their credit underwriting, massive hiring from top US lenders, and the willingness to build a great brand and be the primary accounts, to be the fintech that eventually really cracks the US market.
So from first principles, for some subsegments and niches in the U.S., it's the same in some ways, a very similar playbook. And we have a number of di erent advantages that also are not obvious, for example, on credit capabilities. We think we are, as I said earlier, perhaps one of the best, if not the best in the world and now really knowing how to underwrite credit on behalf of our consumers. And a lot of that was because we brought a lot of DNA from some of the top lenders in the U.S. since the beginning. So this credit as a superpower, which would be huge in a market like the U.S. that requires credit. And a lot of people ask, why haven't fintechs cracked the U.S.? Well, some of them sort of have, but partly it's also because none of them really were able to do or willing to do credit well, while at the same time, building a great brand and consumer obsession and becoming a primary bank account and kind of playing all the elements of our business model. So anyway, we think it's a very big opportunity.
13 | Nu will be able to move faster in the US versus LATAM because they have a strong positive credit bureau and FICO scores for new fintech entrants, whereas that was non-existent in Brazil, allowing them to iterate faster on credit.
So there are some differences between Brazil and the U.S. Obviously, in the U.S., you end up having a positive credit bureau and a FICO score readily available for anybody that comes in as a newbie in the market. In Brazil, we didn't really have that. So that took a long time for us in Brazil. It also took us several years to -- and we continue to do so to optimize our -- basically our team, our credit infrastructure, the way we do our policies, testing a number of different algorithms. And so that entire credit muscle took a learning curve that by the time we got to Mexico and Colombia was much more developed.
And now that we get to the U.S. is a very well-tuned system that can iterate very fast in developing new models and allow us to test and learn much faster.
14 | That said, it is credit and it will take time to scale in a disciplined manner over time, treat credit losses as capex/R&D as long-term investments to build the credit scoring model.
That being said, obviously, this is credit. You do -- it does take some time. You need losses so that your model learns from the losses. In a way, losses tend to be -- is the CapEx, is the R&D investment. So these models need to have some losses, you need to iterate. You need to do foundational models. So it's effectively provided different type of credits and di erent type of modalities to different type of segments so that your models start learning. And that will take some time to learn and get it for a new market.
It's not your traditional consumer internet product that you can scale to millions of customers on day 1. You need a lot of testing. You need some foundational testing, but it will denitely be much faster and better than any other country we've done in the past.
15 | Nu remains committed to scale investment only if they see product-market-fit and unit economics trending towards hurdle rates, using 100bps of efficiency ratio for 2026 and 2027.
Importantly, we're approaching both initiatives with discipline. We will scale investment only as we validate product-market fit, see unit economics trending toward our hurdle rates and clear key regulatory and risk checkpoints. Under our current plans, Nu U.S. and Nu Global together remain within our commitment of up to 100bps of efficiency ratio impact in each of 2026 and 2027.
16 | Headwind from FGTS loans was sufficiently offset by both public and private payroll lending.
Improvements in both public and private payroll lending allowed us to offset the headwind from FGTS-backed loans and overall originations are now running at a higher pace than before that decline. Private payroll originations continue to accelerate month-over-month as we gain condence in scaling the product supported by improving underwriting models and increasingly mature processes across the market that help reduce operational risk.
17 | Nu is getting more comfortable with private payrolls as they have more performance data and is growing rapidly off a small base. Expect to be one of the leading players next year.
And the short answer is yes; we are getting very comfortable with private payroll at this point. As you know, we didn't launch at the same time as other banks last year because we were concerned both about the credit risk as well as the operational risk elements within the private payroll space. We now have enough performance data as well as the operational processes in the industry and the environment have improved enough where we can accelerate in that space. It's growing very quickly right now, doubling quarter-over-quarter o of a small base, just to be clear. And we see no reason why we shouldn't be one of the leading, if not the leading private payroll providers at some point in the next year…So we feel quite condent both in terms of going after the private payroll space, but also about the overall development of that market, not having a material impact or negative second order e ect on our existing portfolio.
18 | Mexico’s deposit base has been shrinking because Nu is deliberately optimising it towards lower cost funding.
In Mexico, we continue to deliberately optimize our deposit base towards lower cost funding while maintaining ample liquidity.
19 | Optimistic about Mexico digitalizing more of financial services, where it is currently still very physically cash heavy. Directionally will be a secular tailwind to Nu.
It's a super positive development in Mexico…And I think Mexico, after a few years of testing at different approaches, we see in President Sheinbaum, really real conviction of going in this direction. Digitalized in Mexico means more access to financial services for more people. Mexico still has 12% credit card penetration. So it's tiny really, a lot of cash heavy. That also means decreasing informality. That also means decreasing crime. That means increasing access to credit because once you digitalize, then you generate a lot of data that can be used by rms like us to provide credit. So anyway, it kind of goes very much in the direction of what the government has been saying about financial inclusion.
20 | See opportunity for Colombia to change how the interest rate cap is calculated protecting consumers from borrowing at high rates from loan sharks.
On Colombia, there is -- I think there is a good chance that there will be some kind of change in the way the interest rate cap is calculated. It's maybe less likely that it will be removed completely. But the way it's currently calculated makes no sense because it's an average of consumer and corporate credit, and they both have very different risk profiles. So -- and it is -- I think it's becoming clearer and clearer that having such a low cap basically means the creation of a very large black market where customers are -- majority of customers in Colombia are getting lending and credit at 1,000% interest rate from loan sharks. It's one of those laws that have good intentions about protecting consumers from high interest rates. But in reality, it just throws everybody to -- or the majority of the population to the black market and then you don't have any protection, any regulation and you pay way higher rates. So I think it's high likelihood that there is a rethinking how at least as a rst step on how this rate is calculated to liberalize a little bit. And depending on what that means, there could be a significant improvement in access to credit.
21 | NPLs should be better assessed over IEP than the total portfolio as IEP as % of total portfolio grows.
In order to interpret these numbers, one structural shift matters. As we optimize for risk-adjusted returns, a growing share of our portfolio earns interest. Interest-earning balances reached 61% of the total portfolio this quarter, up from 59% last quarter and 44% 3 years ago. This matters because delinquency is concentrated in interest-earning balances rather than in credit card purchases that customers pay in full every month. As their share grows, delinquency measured over the total portfolio naturally increases even without a deterioration in underlying credit performance or an increase in exposure.
22 | Nu does not take macro positions, they simply assume that the future will be worse than the past, and that buffer will still make it a good decision.
So we don't necessarily take a macro position about the economy. We just simply assume that the future is going to be much worse than the past because unfortunately, that's what we've seen since we began in 2013. The macro environment we have seen in the past 13 years has been not great to use a great word. So the only thing we have seen, unfortunately, is bad macro. And so when we make a credit decision, we assume the future will be worse than the past. And therefore, we have to use a significant buffer on underwriting decisions. And even then, that decision would have to be positive NPV…So from that perspective, all of our credit decisions right now and the way we continue are covered with a significant buffer for them to be good decisions.
23 | While Nu has done numerous small investments, and don’t rule out a full M&A, their preference is for organic growth.
And so that means that we will look at potential M&A. We have in the past. We've announced, for example, an investment in Tyme in South Africa and the Philippines. We've done a small investment in India. We've done a number of investments like that. And we will consider -- we can consider M&As. But I would say our strong preference is for organic growth. And I don't say -- never say never. The bar would have to be extremely high for us to do a full M&A. I wouldn't say it's impossible, but it's a very, very, very high bar.
24 | Nu is still small in the overall Brazil market, they have cherry picked the best credit and risk-adjusted opportunities that are short-term duration and information rich. Expect to grow faster and keep taking share from competitors.
The second point is we are to consider is we are still very small in the overall Brazilian market. And this means that we have effectively grown and continue to grow cherry-picking some of the best credit and riskadjusted opportunities in the market. We are e ectively operating in products that have very short-term duration that are very information-rich and that allow us to accelerate or decelerate relatively quickly as we see di erent opportunities. And therefore, being small, we can continue growing mainly by taking market share from our competitors in the market, and that opportunity will continue in 2027 and beyond as we are -- everything we have done in challenging scenario has been e ectively taking market share, and we expect that we'll continue taking market share given the signi cant operating advantages and competitive advantages that our model has.
25 | Nu focuses on being the primary banking of their customers, because they get richer data that allows them to assess their credit risk better leading to better underwriting, and results in higher loan repayment priority.
A key part of this resilience comes from our primary banking relationships. The more customers concentrate their financial lives with Nu, the richer the data we generate, the better we can assess their credit risk and the more important our position becomes in their payment priorities. This creates a powerful reinforcing cycle. Deeper customer relationships support better underwriting, which in turn allows us to keep expanding credit while optimizing risk-adjusted returns. Once again, this is why we view customer primacy as a fundamental component of our credit superpower.
26 | Being the primary bank account results in positive selection, and Nu can become the highest in the credit stack for customers, that they will be the first to be paid off, and that becomes a very strategic cost advantage.
And finally, I think it's important to just highlight one point that Rob made, which is we have reached a position that is very unique in the Brazilian market and very soon also in Mexico and Colombia, which is we are the largest primary bank account in the country, 2x largest, we think, than the other -- the second player. And being the primary bank account means effectively you get positive selection, not negative selection, you get positive selection. And you are, in a way, senior in the credit stack of the customer. You are the card that consumers will pay first. And so that gives us a huge advantage in underwriting credit and managing delinquency in a tougher scenario. We're not the marginal lender. We're not the one that's giving the fth card or the sixth card or the sixth loan. We're the primary relationship for the majority of our customers, for 60% of our customers. And so that's a very strategic position because we get that positive selection.
27| That brings a lot of proprietary data which can then be fed into Nu’s credit models, which is one of the most sophisticated.
And obviously, that also brings a lot of proprietary data that then is fed into perhaps the most sophisticated credit stack in the world, which we think we're increasingly building that here in the markets that we operate, signi cant advantage on credit underwriting. So anyway, all those data points kind of tells you -- I won't give you specific guidance on growth rates, but we think as we have done in all cycles, if there is a cycle coming in, we continue to think that our growth is more secular. It's more in a way, a little uncorrelated with the market because people continue to transfer markets -- transfer market share away from some of the big incumbent banks towards Nubank, and that's the trend that will continue irrespective of the macro scenario.
➡️ Final Takeaways on Nu Holdings $NU:
NU is the “GEICO” of LATAM (and potentially US) digital banking. Strong business traction supported by 📶 customers, more/faster cross-sell, leading ↗️ revenues per customer, combined with the significant competitive cost advantage of a lower cost to acquire & serve ➡️, strong risk management, and lower cost of funding (retail) that really drives its long-term highly durable profitable growth. Continue to watch its expansion into Mexico and Colombia, and leveraging its strong base with Brazil. Strategic about entering the US with limited downside and strong upside if they can execute strongly. Not worried about higher near-term (4-6 quarters) expenses of higher efficiency ratios weighing on profitability. If markets give us better buying opportunities should that happens, we will gladly take it.
Very interesting insights on Netflix:
- Netflix does not count views.
- Netflix does not make money per show, it makes money per subscriber.
- Netflix worries most about popular expensive shows that do not change user’s subscription behaviours
- Netflix far prefers users to finish, and finish early, then starting, and not finishing it.
- Netflix wants shows that keep people from leaving.
- Netflix focuses on cost per completed viewer.
threads.com/share/BABS5YBsPX
Sea Limited $SE 3Q26 Earnings
- Rev $8.6b +43% ↗️🟢
- GP $3.9b +51% ↗️🟢 margin 45.9% +243 bps ✅
- Adj EBITDA $931m +29% ↗️🟢 margin 10.9% -122 bps ↘️🔴
- EBIT $664m +40% ↗️🟢 margin 7.7% -21 bps
- Net Inc $502m +34% ↗️🟢 margin 5.9% -41 bps ↘️🔴
Shopee
- Orders 4.5b +26% ↗️🟢
- GMV $39.7b +24% ↗️🟢
- Take-Rate 15.3% +195 bps ↗️🟢
- Rev $6.1b +41% ↗️🟢
- Core Marketplace Rev $4.6b +51% ↗️🟢
- VAS Rev $724m +0% ➡️🟠
- Goods Rev $749m +45% ↗️🟢
- GP $1.9b +49% ↗️🟢 margin 31.2% +163 bps ✅
- Adj EBITDA $267m +44% ↗️🟢 margin 4.4% +7 bps ✅
- EBIT $155m +41% ↗️🟢 margin 2.6% 0 bps ↘️🔴
Monee
- Consumer & SME Loans $12.6b +60% ↗️🟢
- Active credit users 44m +28% ↗️🟢
- Avg loan o/s +23% ↗️🟢
- NPL >90days 1.3% +20bps ↗️🟡
- Rev $1.6b +64% ↗️🟢
- GP $1.4b +66% ↗️🟢 margin 88.9% +155 bps ✅
- Adj EBITDA $289m +12% ↗️🟡 margin 17.8% -826 bps ↘️🔴
- EBIT $288m +16% ↗️🟢 margin 17.8% -729 bps ↘️🔴
Garena
- Bookings $890m +6% ↗️🟡
- QAU 671.7m +0% ➡️🟡
- QPU 68.6m +4% ↗️🟡
- QPU/QAU 10.2% +40bps ↗️🟡
- ABPU $1.33 +6% ↗️🟡
- Rev $840m +29% ↗️🟢
- GP $600m +40.0% ↗️🟢 margin 71.4% +577 bps ↘️🔴
- Adj EBITDA $458m +46% ↗️🟢 margin 54.5% +645 bps ✅
- EBIT $453m +46% ↗️🟢 margin 53.9% +650 bps ✅
Service
- Rev $7.8b +43% ↗️🟢
- GP $3.9b +51% ↗️🟢 margin 49.7% +269 bps ✅
Goods
- Rev $749m +45% ↗️🟢
- GP $38m +48% ↗️🟢 margin 5.1% +13 bps ✅
1 | Shopee delivered another quarter of strong growth, continue to see long growth runway and expect unit economics to keep improving. Expect FY26 GMV growth of 25% and Adj EBITDA of $1bn.
Shopee delivered another quarter of strong growth while generating the highest quarterly profit we have ever achieved. Our strategies have proven effective at strengthening our fundamentals, giving us a growing buyer base, stronger monetization and improving unit economics. We still see a long runway of opportunity ahead, and we will keep executing with the discipline to capture it. We remain on track to deliver GMV growth of around 25% and an adjusted EBITDA of $1 billion for the full year of 2026.
2 | Shopee saw strong acceleration in new user growth, driving strong growth in active buyers.
The acceleration in new user growth we saw in the second quarter has continued into the third. We engaged and reengaged a broad range of user through both online and offline campaigns, brought our content to more social and video platforms and widened our service options, especially in delivery to serve different user needs. With these efforts, new average monthly active buyers grew more than 50% and the total average monthly active buyers increased by 20% YoY.
3 | Ecommerce monetization also doing well, driven by ads, where ad-paying sellers grew 35%, ad revenue grew 55% and ad-take rates rose 80bps.
Monetization has also continued to do well, driven by ads. Sellers are responding well to the greater value they see in our AI-driven ad solutions. Ad paying sellers rose more than 35% while average asset per seller increased around 13% YoY, and revenue grew more than 55% and our ad take-rate improved by over 80bps YoY.
4 | Expect Shopee monetization to improve, primarily coming from higher-margin advertising and to flow to growing EBITDA.
We expect going forward, our monetization will also be better, primarily from the ad side, less from the xed commission side. And part of that will ow to the bottom line EBITDA improvement over time. And part of that might go to the investment to grow our market further as well.
5 | Continued logistics improvements is allowing Shopee to serve a wider range of buyer needs from instant, same day to low cost services.
Our operational priorities to enhance service quality, price competitiveness and the discovery remain consistent and continue to be driven by our efforts in Logistics, ShopeeVIP and our content ecosystem. Each made good progress this quarter. First, Logistics. We expected our delivery options to serve a wider range of buyer needs from instant and same-day delivery to low-cost services. Our instant delivery service can now deliver as quickly as 1 hour. This has expanded our addressable market by unlocking new spending scenarios where speed matters most. These tend to be high-frequency categories such as the growth rate and the pharmaceutical items, boosting our overall purchase frequency.
6 | Strong traction in Indonesia and Thailand, orders in Indonesia 2X, same day delivery now 15% of orders in Bangkok.
In Indonesia, orders using the services more than doubled YoY, with cost per order declining year-on-year. Retailer share have increased the service. We now have more than 12,000 offline stores on our instant platform and growing.
Adoption is also growing in Thailand, where instant and same-day delivery rate now account for nearly 15% of orders in Greater Bangkok. We will continue to grow this service in areas where demand is high and the unit economics make sense.
7 | Buyer self-collection reached 30% in Singapore and 15% in Malaysia, which is a much cheap shipping option.
In the third quarter, buyer self-collection order penetration reached more than 30% in Singapore and more than 15% in Malaysia. In Indonesia, our economical shipping option has gained strong adoption and now accounts for more than half of total orders. This cost of order is more than 20% lower than the platform average in the market. These initiatives allowed us to offer free shipping at a lower minimum spend, giving our buyers the exibility to choose between cost and speed and letting us serve a broader range of buyers' needs.
8 | Shopee remains focused on scaling fulfilment quickly, orders grew 25% QoQ, unit economics improving, parcels delivered next day or sooner grew twice as fast as overall growth.
We have also continued to scale fulfillment quickly. Sellers have responded well to our fulfillment offerings as we help them take on the complexity of storing, packing and shipping. This represents significant efficiency gains for them, especially during peak campaign period where orders may be several times higher than normal. With our services, sellers can build their businesses more easily without having to increase their own warehouse space or staff size. In the third quarter, fulflled order volumes accelerated to around 25% QoQ. Unit economics for our fulfilled by Shopee service improved sequentially in Asia, and we see further room to gain efficiency through higher warehouse utilization and automation. Fulfillment also drives synergies with delivery, enabling us to deliver more packages more quickly. The number of parcels delivered by the next day or sooner grew over 16% YoY, more than twice as fast as our overall order growth.
9 | ShopeeVIP continues to grow rapidly, hitting 21m +40% QoQ, and members now account for 25% of Shopee’s overall GMV, and have higher monthly retention of 80%, and spend more.
ShopeeVIP is growing fast. Membership across markets exceeded 21 million at the end of September, up more than 40% quarter-on-quarter and members contributed close to 25% of Shopee's overall GMV. The program is demonstrating stickiness with monthly retention around 80%. Members show a consistent double-digit percentage uplift in spend on average after subscribing. And with more sellers and partners investing alongside us in membership benefits, the program's unit economics improved sequentially.
10 | Orders from live streaming and short-form video grew 45% and now account for 25% of physical good orders in SEA.
Orders from live streaming and short-form video grew more than 45% YoY and accounted for more than 25% of physical goods orders in Southeast Asia. Unit economics also improved sequentially. Beyond our app, our a liate program remains an effective and growing channel to bring new buyers to Shopee from other platforms.
11 | Content business is profit margin accretive rather than dilutive.
For content ecosystem, which includes video live stream, I think it's a bit hard to split a particular segment to the EBITDA level because there are many shared costs below that. I think for us, we look at sort of variable cost or contribution margins to our platform, the content platform that has a positive contribution margin. Essentially, what we're saying is if we grow content business more in our platforms, we'll have more profit rather than has less profit for our businesses.
12 | Social media orders from YouTube, Facebook, and Instagram grew rapidly from affiliate creators.
In the third quarter, orders driven by YouTube content more than doubled YoY and Shopee affiliate orders generated by Facebook linked creators grew more than 90% QoQ. On Instagram, more than 4 million reels have been tagged with Shopee products since we launched the partnership in April with the number growing more than 4x QoQ.
13 | Shopee is one of the few ecommerce platforms that is open to agents with Meta Muse, while still early, expect it to become a solid channel for buyer acquisition and GMV growth.
Shopee is also one of the early connector partners for Meta's Muse AI personal assistant that has just launched in Singapore. Users will be able to research and compare products on Shopee through Muse, helping them make purchase decisions tailored to their preferences. The partnership is still at an early stage, but we see potential for it to become a solid additional channel for buyer acquisition and incremental GMV growth over time.
14 | Open to working with agents, because checkout, seller and buyer relationships still remain with Shopee.
Coming to the second question on the agentic AI, we actually partnered with a range of partners starting from ChatGPT and now as you mentioned, the Muse. At this point in time, all the checkout and seller relationships and buyer relationships happen within the Shopee platform. And we didn't see a meaningful impact to our economics for the near future.
15 | Brazil remains the fastest-growing for Shopee, and grew faster than the broader market, gaining share, and continue to improve buyer waiting time by 12%.
Brazil remains our fastest-growing market, and we again grew faster than the broader market in the third quarter. Logistics improvements underpinned this growth. We continue to widen the range of faster delivery options we offer across the country with average buyer waiting time improving 12% YoY. We have branched into air delivery, cutting delivery times to the North and the Northeast regions significantly.
16 | In São Paulo, ~40% of parcels now arrive next days do 60% within 2 days, stronger delivery capabilities is adding more sellers to Shopee in Brazil
In Greater Sao Paulo, around 40% of parcels now arrive at the next day and nearly 60% within 2 days. Our new instant delivery service there, Shopee Turbo already delivers in under 4 hours from more than 7,000 partner stores. Stronger delivery capabilities have enabled us to onboard more sellers to fulfillment.
17 | This is allowing Shopee to expand beyond smaller-item ecommerce to more expensive and larger items into more upmarket buyer segments and higher value categories, and with Shopee Mall, which is where MELI dominates.
For example, we rolled out an in-house delivery service for bulky items such as the large appliances and furniture, combined with our strong service reputation and reliable cross nation coverage. This has made more large item sellers willing to ful ll with us.
Across the country, fulfillment order volume more than doubled YoY. These service quality improvements have helped drive stronger growth in higher-value categories. More upmarket buyers are joining Shopee and existing buyers are shopping across more categories with us. Automotive and home improvement categories grew more than 80% yearon-year and more brand sellers have come on board.
Shopee Mall orders carried an average basket size, almost 3x the platform average. Looking ahead, we see a lot of headroom to grow in this exciting market.
18 | Don’t see a meaningful growth deceleration for Shopee GMV yet, but noticed more value-seeking ecommerce buying behaviour.
On the Shopee growth, first of all, we don't see a big Shopee GMV growth rate slowdown yet. As we shared earlier, in Q3, we do see a meaningful growth, similar growth rate compared to last quarter from a year-to-year perspective, if we take away the ForEx impact. You are right that there are many macro factors like fuel price, et cetera, impacting the consumption.
Overall, we see that although we have noticed some more value-seeking behaviors in our market due to all this macro environment, but the general overall demand is relatively resistant.
19 | Shopee is well positioned as they are focused on long-tail products with competitive pricing that positions them better with people who need to buy rather than impulse buying. Still see long growth runway in e-commerce.
But again, I would like to reiterate that our competitive advantage will continue to differentiate us from the peers even in the macro situation that we are in right now, given that we are -- we built our platform around long-tail products and also the pricing in our platform is a lot more competitive compared to other platforms online or compared to the alternative offline as well. On top of that, many of our -- large part of our platform's GMV are driven by the people who have the need to purchase rather than people who have impulse buying. All those make us -- put us in a more resistance position on the growth for our e-commerce in our market. I think we will continue to monitor our market very closely for any changes. But again, given all the things we mentioned above, and also the e-commerce penetration are still relatively low compared to many other markets, we still believe that there is a big runway for us to grow our e-commerce in the regions.
20 | Monee has grown beyond its original origins of payments into mobile wallet, credit, and digitial banking, and is growing fast and very profitable.
Monee has grown well beyond its origins as a payment infrastructure serving our internal business needs. Today, it has evolved into a financial services business across mobile wallet, credit, digital banking and more. Monee has shown consistent strong growth and has been profitable for the last 3 fiscal years.
21 | Monee has reached the scale and maturity to benefit from becoming independent, and a standalone public listing in the coming 1-3 years will make sense.
We believe Monee has reached the scale and maturity to extend itself and would benefit from becoming more independent…Our ambition is to keep growing Monee into a full financial ecosystem with inclusion at its heart, serving users across their life cycle needs from unsecured loans to secured lending, insurance, wealth management and more. We believe a stand-alone listing would give Monee the dedicated capital base it needs to build this ecosystem and let it invest for growth while managing risk with discipline.
We hope to list the Monee at the right time in the next 1 to 3 years, taking into consideration factors such as market conditions, macro environment and regulatory requirements.
22 | Monee and Shopee will continue to work together, and remain a key payment and credit partner for Shopee, and insights from Shopee used to drive Monee.
I want to be clear about what does not change. Monee and Shopee will continue to work closely together as they do today. Monee will remain a key payment and credit partner for Shopee and Shopee will remain a key platform for Monee to reach users. Insights from our businesses will continue to empower Monee's growth.
23 | BNPL is growing rapidly, driving faster growth in Shopee’s SPayLater portfolio >2X, and is now 25% overall and nearly 40% in Malaysia.
As pay later is becoming an everyday way to pay both online and offine in many of our markets, our off-Shopee SPayLater portfolio more than doubled year-on-year to nearly 25% of our entire SPayLater portfolio at the end of September. In Malaysia, its share is nearly 40%. In Thailand, the ShopeePay unlimited cards allows SPayLater users to accept their approved SPayLater credit limit wherever Mastercard is accepted.
24 | Risk management remains top priority for Monee, 90 day NPL ratio inched up small to 1.3%. Continue to proactively adjust underwriting criteria depending on macro.
Risk management remains our top priority as we continue to grow our loan book. Our 90-day NPL ratio was 1.3% at the end of September. The modest uptick primarily reects an evolving credit environment, especially in Thailand and Brazil as well as market and product mix, including faster growth in Brazil in o-Shopee SPayLater. Our underwriting is highly responsive to market conditions. We proactively adjusted our underwriting criteria and credit limit management for selected customer segments at early signs of softness to maintain disciplined growth and optimize risk-adjusted returns. We are staying close to the data and we'll keep adjusting as needed. So our loan book growth stays healthy and sustainable.
25 | Garena grew bookings LSD despite strong comps with improving profitability.
Garena delivered a solid third quarter with bookings continuing to grow from the high base set last year. Profitability also improved. Free Fire anchored this performance with a strong bookings growth and Arena of Valor continued its strong momentum from its 10th year anniversary celebrations.
26 | Setting up Sea Cloud as a frugal start-up, not seeking to own large amounts of physical capacity, only for customers with dedicated computing capacity, and pursue with clear customer commitments and attractive returns.
We are setting up Sea Cloud as a frugal start-up so that it can stay nimble and respond quickly to market demand. Our focus is the software and customer relationships, not owning large amounts of physical capacity, where customers need dedicated computing capacity, we will only pursue opportunities with clear customer commitments and attractive returns. We will mainly look to partners and customer prepayments to fund the physical infrastructure.
27 | Still early, do not expect Sea Cloud to have a significant impact on financials, they still figuring out. Probably work better for their SMB merchants than the AI labs or the enterprises which is what the hyperscalers focus on.
As a result, we do not expect Sea Cloud to have a significant impact on our financials for now. It is still early, and we will update investors on any meaningful developments.
So that is kind of the big picture we are looking at, and we think the demand will be huge, and this is just at a very, very early stage. So we want to get into this business early. There's a lot of things we don't know yet. And again, this is very early. This is kind of a start-up. So there's a lot of things we need to gure out, for example, what exactly the segment we should focus on, what exactly type of customers we should focus on and what is the best business model we should use. We think we'll get a better picture when we really get into the -- get our own feet into the game and we start to get the rst-hand experience.
28 | Sea Cloud has not secured any enterprise contracts yet, will primarily buy GPUs (as don’t have own chips). Expect the business to be more deal driven than its current businesses.
Sea Cloud, for your question, as you know, this is just at the very, very beginning of the business. At this moment, we don't have any -- we haven't secured any enterprise contracts. And for the investment in the data center and the GPUs, we -- let me put it in this way, we denitely probably will invest in the GPUs primarily for our own use. So regardless if there is a Sea Cloud business or not, like how the business progress down the road, we are going to invest into GPUs and primarily for all the internal use cases.
But the business nature is a bit di erent from the Sea Cloud. I think it will be very much deal by deal driven, at least at the beginning stage and with very, very speci c target in terms of the custom
29 | Another strong quarter across all three businesses with solid growth momentum, operational efficiency, and improving profitability.
In conclusion, we have delivered another strong quarter. Our 3 businesses have all expanded their addressable opportunity and strengthened their competitive position. Unit economics continue to improve across many of our initiatives, a testament to our financial discipline and operational efficiency. Our positive momentum across the rst 3 quarters give us greater condence for the rest of the year.
➡️ Key takeaways on Sea Limited: SEA’s most dominant ecommerce platform with Shopee and increasingly its logistics fulfilment platform SPX and its credit lending business Monee riding on Shopee. Garena continues to recover and provides the reinvestment ability to double down. Continue to see a long runway of growth opportunity, and not concerned over the near-term higher investments weighing on near-term profitability, which are beneficial for longer term growth and profitability. Valuations remain undemanding and attractive.
A black swan and her two cygnets at Hangzhou's Qingshan Lake Water Forest 杭州青山湖水上森林, also known as Matcha Lake.
"Black swan" is used to describe a rare, unexpected event with big consequences. The term goes back to when Europeans believed all swans were white. In 1697, Dutch explorers found black swans native to Australia and that belief fell apart.
Watching this family swim by earlier today, I was reminded that rare depends on who is looking. To Europe, black swans were unthinkable. To Australia, they were just swans.
Perspective matters. Rarity says more about the observer than the swan itself.
10
1,779






























