MercadoLibre, Inc. (MELI) Q3 2024 Earnings Report, Transcript and Summary

Good evening, and welcome to MercadoLibre's Q3 2024 Conference Call. All participants will be in listen-only mode. [Operator Instructions] After today’s presentation, there will be an opportunity to ask questions. [Operator Instructions] Please note this event is being recorded. I would now like to turn the conference over to Richard Cathcart, Investor Relations Officer. Please go ahead.

Hello, everyone, and welcome to the MercadoLibre earnings conference call for the quarter ended September 30, 2024. Thank you for joining us. I'm Richard Cascade, MercadoLibre's Investor Relations Officer. Today, we will share our quarterly highlights on video, after which we will begin our live Q&A session with our management team. Before we go on to discuss our results for the third quarter of 2024, I remind you that management may make or refer to, and this presentation may contain forward-looking statements and non-GAAP measures. So please refer to the disclaimer on screen, which will also be available in our earnings materials on our Investor Relations website. With that, let's begin with the summary of our results. Marcos Galperín: Hello, everyone. I am pleased to share that MercadoLibre had a later quarter of outstanding growth. We had an excellent quarter across all of our businesses and geographies. So the strong performance was really all around. With these results, we reinforced our position as the leading e-commerce and fintech platform in Latin America. Every time we have invested in enhancing our customer experience, in the past, we have been rewarded with strong growth and improvements in our market position. In this quarter, we continue to make strategic investments in our platform, both again, commerce and fintech and the resulted in GMV growth of 34% year-on-year in Brazil and 27% in Mexico. This also resulted in market share gains in both countries. In Argentina, items sold grew at 16%, and we surpassed the 60 million items sold in our platform, which is the highest ever volume. We continue to see improvements of the Argentinian market. We have added a record of almost 7 million new buyers into our platform. This is a number higher than the number that we saw during the peak of the pandemic. So let me be clear on that, not even during the explosion of demand that we experienced in the pandemic, we saw a number of new users at this level. As most of you know, the penetration of e-commerce in Latin America continues to be relatively low when we compare it to other more developed regions. Only 15% of commerce is done on online. So there’s plenty of room to continue growing. As the lead player in the region, we have a crucial role in driving more volume online and increasing this penetration of e-commerce. For that purpose, we continue to invest in our strategic initiatives, such as logistics and our loyalty program. In Q3, we opened six new fulfillment senators, five in Brazil and one in Mexico, and that helps us improve or increase the fulfillment penetration by 4.5 percentage points compared to a year ago. We know that having more volume shipped out of our own fulfillment facilities results in a much better user experience for buyers and sellers in faster and more reliable delivery speeds and a much better conversion of volume, in turn, results in further growth of the ecosystem. Investment in technology also plays a critical role in bringing more business online because it removes buyer fiction (ph). A large team of developers continue to be focused on improving the buyer experience and customizing it for each individual vertical. Our advertising business represents a powerful profit engine for MercadoLibre and once again we had a strong quarter despite FX headwinds that we experienced in Mexico and Brazil. Most of the volume of advertising continue to come from product ads, so we see a tremendous opportunity going forward on display and video advertising where we're just getting started. In Fintech Services, the impressive growth of our user base and the rising engagement that we see for having with our products shows that more and more people and consumers are choosing Mercado Pago as their main financial service provider. We continue to invest behind our credit card as is a critical product to achieve principality within our users. In Q3, we issued 1.5 million new credit cards. And the credit card TPV grew by 166% compared to last year. As we see improvements in our risk models, we have accelerated the underwriting of credit, not only credit card but also consumer and merchant credits, and this resulted in growth of our portfolio of 77% compared to last year. We know that the faster growth of our credit portfolio, combined with the shift towards more credit card as a percentage of total portfolio puts some short-term pressure on 900 spreads, the broader stable NPLs that we have seen and the fact that the earlier cohorts of credit cards are already becoming profitable, gives us confidence in our investments. Turning now to our financial results. Our strong operational KPIs translated into revenue growth of 35% year-on-year despite FX headwinds that we suffer in Mexico and Brazil. A successful Q3 performance is not just a result of the great execution of our team, but also a testament of the many growth opportunities that lie ahead of us in our business and our ecosystem. Our EBIT was $557 million with a 10.5% margin. The majority of the margin compression from last year comes from the strategic initiatives to the investments that we discussed earlier. Investors can find more detail on those investments in our shareholder letter. Net income rose by 11% compared to last year, reaching $397 million in the quarter, with lower EBIT partially offset by a reduction in our FX losses. As we finish our 25th year anniversary, we continue to see tremendous growth opportunities ahead of us, both in commerce and fintech and now also in advertising. We have a clear vision of the future and are focused on the long-term growth, profitability and cash flow generation of our business. Now I want to share some examples of innovation that we have implemented during the quarter. So I'll turn it back to Richard, that will explain those to you.

MercadoLibre is a major driving force of the shift of offline to online commerce and financial inclusion in Latin America. Continuously investing in our value proposition, service levels and user experience is key to making that happen for years to come. We want to share a snapshot of just some of those investments in innovations today. In September, we have created the Meli Mas benefits package to include cash back or extra installments on marketplace purchases and new fintech benefits. This means that users are able to get back much more than they pay for subscription, making Meli Mas the smartest choice for our users. Those fintech benefits include cash back on all purchases made by the Mercado Pago credit card and a higher yield on funds held in the remunerated accounts. These benefits increase the differentiation of our credit card and digital account. Cash back is paid in the Meli dollar stable coin and is deposited into users Mercado Pago accounts where it could be held, sold or used to make purchases on the marketplace. We also introduced an additional tier in Meli Mas that offers a cheaper plan and gives consumers a choice of two plans that best suits their needs. The essential plan includes free shipping, cashback or extra instalments on marketplace purchases and all of the fintech benefits. Whilst the total plan has all the benefits of the Essential plan plus great content with access to Disney+, Deezer and discounts on other streaming services. Meli Delivery Day continues to be a core and popular piece of our loyalty program. It enables free shipping on lower value items, which drives higher purchase frequency and enables us to dilute our last mile costs. To incentivize the purchase of multiple times for delivery on the users Meli Delivery Day. We have made our free shipping of – on lower value items more visible to the user as they engage with the platform. Several other initiatives aimed at improving the online purchasing journey have gone live in Q3. In Q3, this helps us to attract a record number of new buyers even higher than the previous peak during the COVID pandemic. We continue to develop category specific UX, as a way to drive off-line retail online vertical by vertical. In Auto Parts, we launched a new feature that enables buyers to schedule an appointment with the mechanic to install the items they are purchasing, and we ship those items directly to the mechanic. In the beauty category, we introduce technology that enables users to see different makeup products apply to their face in our app. This tackles a major barrier to consumers buying makeup online. To increase cross category shopping, we launched a new tool that suggests a new high frequency category for users to shop based on that individual users profile. This personalized suggestion includes a short cut button to a category on the home page, a special coupon, a personalized landing page and push notifications. An important driver of shifting offline retail online is the shipping experience. In September, we announced plans to more than double the number of fulfillment centers in Brazil by the end of 2025. This means 11 new operations, five of which were added in Q3. These investments will enable us to increase the number of cities where we deliver on the same day in Brazil by 40%. 63% of these operations will be outside of Sao Paulo as we increasingly regionalize our network. Some of them will be metro fulfillment centers similar to the model we introduced in Rio de Janeiro last year, which are smaller operations that target individual cities. This is just a snapshot of the innovations that have gone live recently, and we have a strong pipeline of innovations that will enhance our position as the destination of choice to shop online in Latin America because as always at MercadoLibre, the best is yet to come.

We will now begin the question-and-answer session with Martin de los Santos, CFO; Ariel Szarfsztejn, Commerce President; and Osvaldo Gimenez, Fintech President. [Operator Instructions] And our first question comes from Irma Sgarz from Goldman Sachs. Please go ahead.

Yes. Hi. Thank you very much for the opportunity to ask a question. I'd like to ask about the credit card portfolio where we saw -- was a quite pronounced growth. And there's a two-fold element to my question. Firstly, what drove the incremental acceleration in the pace of quarter-over-quarter portfolio growth in the third quarter, specifically for the credit card, where you've sort of been originating at a pretty stable pace over the last couple of quarters. I know the messaging generally has been positive, and there are about acceleration. But just surprised to see that jump from one quarter to the next, and I was curious to understand sort of in which geography or customer cohorts you've been leaning into for this -- for driving this incremental acceleration and whether we should think of further acceleration from here? And then the second element to that, you emphasized that part of the NIMAL spread pressures were also from leaning into larger, longer duration loans to lower risk users in really all your geographies. And I was curious, if you could just expand a little bit more on that, what does it mean for growth ahead? And if you're generally feeling more confident about your underwriting mechanisms, why wouldn't that make you maybe go into perhaps like slightly higher risk classes where you may be initially seeding smaller loans, but then sort of have more of a ramp-up over time as you feel more confident with those new customers? Thank you.

Hi, Irma. Let me start with the second part of the question about NIMAL and then I move back to credit cards. With regards to NIMAL, yes, NIMAL spread (ph) this quarter of 15% driven by basically three key drivers. The largest one is the ramp-up I'll refer to the minute of the credit card business, whose share in the portfolio rose from 25% a year ago to 39% now. And since this is a lower spread product. This increase in product mix or in credit mix affected NIMAL. Also, we are continuing moving up market. We are very, very happy with this. We have started in the low end, and we're happy with the move up market. And the way to do that is with longer durations and also offering larger lines to lower risk customers. These are credits that are very profitable, but have lower spreads than the one we had which were smaller to lower income customers. And the third part of it is that simply by the fact that we have been accelerating the size of the portfolio whenever we grow the portfolio, initially, we have to a higher than proportion of provisions. And then the revenues from the interest comes afterwards. So whenever we're growing very rapidly, there is a hit to NIMAL. And having said that, I would say, asset quality indicators are solid. We have had broadly stable growth rates, both for the 15, 90 days NPLs, which has been stable at 7.8%. Now moving on to the first part of the question with regards to credit card. And yes, we are excited with how this business is evolving, both in Brazil and in Mexico. Brazil is a country where we have been doing this for longer. And the good news we have is that when we look at all the cohorts for -- from 2021 and 2022, all of those already have positive variable contributions and even some cohorts in early 2023 also have a positive value of contributions are already net positive. So the fact that we are growing so fast and that the recent cohort represent a significant part of our total portfolio makes -- has an impact in our returns or in our P&L in the current quarter, and we are very excited with how the product is evolving not only because all the cohorts are already profitable, but also because we see all of the drivers that the increased penetration of credit cards bring us from higher Net Promoter Scores, higher increased transactionality in users who have credit cards, both in the marketplace, but also off the marketplace, increased use of other Mercado Pago power products and in general, an increase in principality. So we are excited with how that is evolving. You mentioned too, regarding the number of cards we were issuing, I think that beyond the number of cards, which is growing, I think that we are doing two other things. One is increasing the size of the lines, and therefore, increasing the TPV, which we believe is the more relevant metrics in the TPV of the portfolio of credit cards. And all of that has been – has resulted in the increase in the penetration of cards in the portfolio, as I mentioned before.

The next question comes from Andrew Ruben from Morgan Stanley. Please go ahead.

Hi. Thanks very much for the question. Hoping to dig in a bit to the shipping investments, given the five distribution centers in Brazil in a quarter being quite sizable. So hoping you can help give us some color on how long a distribution center typically takes to ramp up and reach productivity. And you mentioned the reach outside of Sao Paulo, but perhaps if there's any incremental category or capabilities otherwise, these facilities give you? And then related, we recall last year, you had some network capacity constraints during this 4Q holiday season. So I'm curious with the network expansion so far this year, what are your views on how the capacity now compares to your demand expectations during the holiday season? Thank you.

Hey, Andrew. Ariel here. So yeah, as you said, we made a significant progress in terms of facilities for our fulfillment operations. I think we've been saying this for quite some time, fulfillment leads to higher GMV because of faster delivery promises, higher conversion rates, stronger buyer and seller NPS, and from there, higher retention rates. And I think we were clear in the past that we need to grow our fulfillment capacity because of three things. A, because of the expectation of future demand, we need to grow capacity with what we think the business will continue to grow in the future. And luckily, we see huge opportunity for growth in our marketplaces, both in Brazil and in Mexico. Second, we need higher -- we need to build capacity because of the increases in fulfillment penetration that we are consistently having in Brazil. And last but not least, there's a normalizing effect of capacity available in Mexico. So overall, we are confident that with this capacity that we are bringing into the network, we will not only have the required square meters to deal with big demands from our sellers and buyers, but more importantly, to continue growing and expanding our business in the different regions. In terms of the first part of your question, ramp-ups and productivity, we are consistently making progress in shortening the productivity ramp-up window from our representatives, but it's natural, right? So every time we open a fulfillment center, we start operating with lower utilization, which then tends to grow. So yes, to your point, in the short term, you can see some cost pressures, but this capacity is crucial for our growth, for our scale and our -- for our long-term strategy of serving our users.

The next question comes from Robert Ford from Bank of America. Please go ahead. Hi, Robert, is your line muted? Our next question comes from Marcelo Santos from JPMorgan. Please go ahead.

Hi. Good evening. Thanks for taking my question. I wanted to ask you about the differences between the Brazil credit card experience and Mexico so far. So you have these two markets where you're growing. Can you contrast like perhaps NIMAL, NPL frequency, whatever information you could bring would be interesting?

Marcelo, I'll say that probably the most relevant difference for us is that when we launched the credit card in Brazil, it was the first time we were doing this. So initially, we started relying on the models we have built for consumer loans. And then we learned and iterated on how to adjust those models. And every roughly six months, we have been able to roll out a new generation of models. And we did learn quite a lot along those lines. Therefore, when we launched the product in Mexico last year, we had more experience, and really, we're more reliable in terms of what to expect. Therefore, I would say the ramp-up in Mexico has been significantly faster than it was in Brazil. And also, there are some, I would say, significant differences in the structure of the market for credit cards in both countries. Clearly, the penetration of credit cards in Brazil is significantly higher than what it is in Mexico. And therefore, we believe the opportunity in Mexico is huge. Having said that, there is less public information about how – and credit scoring for the users in Mexico. So you have to rely a little more on internal information and less third parties information. Also, there are differences in terms of when and how the working capital works in each country because you have to sell the transactions in Mexico, the next day in Brazil is after 30 days. So the working capital requirements are higher in Mexico than they are in Brazil. And I think those are the main differences.

The next question comes from Geoffrey Elliott from Autonomous. Please go ahead.

Hello. Thanks very much for taking the question. If I could go back to the point about the investment you're making in fulfillment, you've opened quite a few fulfillment centers in Q3. You've got another, I believe, five or six planned by the end of 2025. Can you give us a sense of how long this is going to weigh on margins, weigh on profitability. Because it's clearly had a pretty significant impact relative to where you were a quarter or two ago?

Hi, Geoffrey. It’s Martin here. How are you? It's just like you said, we continue -- and I think Ariel was mentioning this before, we continue to invest in fulfillment centers in the case of Brazil to increase their fulfillment penetration and to take it to the same level of similar levels that we have in Mexico. And in Mexico, just to give up the growth of the business, which this quarter grew at 27% year-on-year. And I think the -- if you look at the margin pressure in this particular quarter, there wasn't a lot of margin compression because of fulfillment. Most of the compression came from credit cards, as Osvaldo mentioned recently. So to answer your question, we will continue to invest business as usual as we have been doing it for the past five years. And if you look back over the past five years, fulfillment penetration increased significantly. We have always put some short-term pressure on margins. But because of the incremental volume that it brings because of the better user experience and so on, then we are able to dilute those costs in addition to all the focus that we put on productivity. This quarter, we improved significantly the productivity. The actual cost per shipment is down year-on-year even though we have higher penetration of fulfillment. So I think this particular quarter, there wasn't a significant pressure for margins because of that. But we will continue to invest in fulfillment and business as usual as we have done it in the past.

Just to complement, Martin, Geoffrey, another point to highlight here is that not necessarily every new warehouse that we open is a 100 square meter facility, right? So a few quarters back, we announced that we are testing our first metro fulfillment center in Rio de Janeiro. This quarter, two of the warehouses that we opened are metro fulfilment centers potentially in the coming years, we will open more of those, which are -- have a different profile and different type of investment and, of course, different unit economics.

The next question comes from Jamie Friedman from Susquehanna. Please go ahead.

Hi. Thank you for taking my question. I wanted to ask about Slide 19. It's the margin bridge and specifically about the bad debt call out here. I'm wondering, is that -- is that like the fail first of new originations? Or is there something underlying the credit quality that's maybe different than you had modeled? Marcos Galperín: Hi, Jamie. How are you? Let me start with the last part of your question. In fact, we are very confident about the credit quality and the prediction of our models, and this is the reason, as Osvaldo explained, that we're accelerating our credit origination. So that part is not explaining this 3.4 percentage point compression that you are pointing out to. Here, what's happening is that we increased roughly $1 billion portfolio quarter-on-quarter. If you were to look at it year-on-year, we increased by 72%. And as Osvaldo mentioned, when you increase the portfolio very rapidly as we did this quarter, you provisioned the losses upfront. And that happens not only on credit card, but also on consumer credit and merchant credits, which are very profitable books. But as we accelerate, we need to provision upfront, and that generates an incremental bad debt which is generating compression margin this quarter. So it's not related to the performance of the portfolio. When you look at NPLs, which we also published (ph) this year, you can see that they are pretty much under control. And the fact that we're accelerating, I think it speaks about our confidence in the way we are originating, which by the way, we're accelerating in all of the markets and all of the products. So we feel very comfortable about the quality of our books.

Next question comes from Craig Maurer from FT Partners. Please go ahead.

Yeah. Hi. Thanks for taking the questions. I wanted to ask if you could perhaps separate out the margin compression related to one-time costs of setting up new facilities versus the ongoing cost of operating those facilities. And second, the ad penetration rate was flat or unchanged quarter-over-quarter. I was curious what your expectation is for penetration going into the fourth quarter and perhaps next year, if you can comment on that? Thanks.

Hi, Craig. How are you? It's Martin here. We don't disclose specifics of the margin compression, which part we have treated to ramp up versus overall incremental cost of fulfillment. But if you see on the disclosures and the investor letter, you can see that the impact of our investments in fulfillment was relatively low, less than 1 percentage point of compression. So that's as much as I want to say about that. I think, as I mentioned earlier, over time, we make up for those short-term pressures through incremental volume, and we feel very comfortable about the investment that we're making in fulfillment.

So, hey, Craig, this is Ariel. To your question about that, we are pleased with the performance we had this quarter. We grew 86% in FX neutral, 37% in U.S. dollars. And as you said, we reached 2% of revenues over GMB this quarter, that's an increase of 0.3 percentage points year-over-year, with Brazil and Mexico reaching record high penetration once again. So we are happy with the fact that ad revenues are growing well above our GMV. We continue to see great potential for our ads business in the future, as we develop more technology, better product stack across the full funnel of advertising. Still growth will never be linear, so we are not focused on analyzing what's happening quarter-over-quarter. There are seasonality mix effects, FX, whatever, many, many levers they are that do explain changes. But conceptually, we are excited with the progress we are making and see great results from all of our efforts. Thank you.

[Operator Instructions] Our next question comes from Deepak Mathivanan from Cantor Fitzgerald. Please go ahead.

Hey, guys. Thanks for taking the question. So on the ramp in credit cards, can you give us some color based on historical cohorts, how it's likely to drive volume growth either on e-commerce or fintech over the next few quarters? Is there anything you can share on typical customer behavior post the issuance of the credit cards just in terms of the benefits to e-commerce and then on the processing side? And then secondly, just to get some specifics on the NIMAL difference between the various products, how does it differ between credit cards and consumer credit? How much is NIMAL generally lower in credit cards. So that we can think about modeling the headwinds for the future a little more precisely. Thank you so much.

Hello, Deepak. So with regards to the impact the credit card has on our users and our ecosystem and what we see -- the first thing we see is those users who have access to a credit cards start using it. And on top of that, start using also a credit line, the users who are happy with MercadoLibre and Mercado Pago. They have the highest engagement. They use our products more often. They buy more on the marketplace. So clearly, it increases the engagement those users have with us. And then you, those are -- those users who have all of these products, then second next are those who use a credit card and then those use our credits and so on. But basically, getting a credit card significantly increases engagement. Also, another factor that is very relevant is when you look at penetration of payment methods within the marketplace, we see that we were already building a proportion of mix that was coming from our consumer loans. Mexico probably is a great case because it was the exact – the case where we had the highest penetration of consumer loans within the marketplace. And on top of that, we are adding the layer for the credit card. So we continue gaining share with credit products of all of the volume in the marketplace, and therefore, it becomes more and more profitable. Then, I'd say, those are the most relevant impacts. In terms of ramping, we'll continue increasing the cards we issue, we'll continue increasing the lines and so on. We are comfortable how this is working.

Hi. It’s Martin here. Regarding NIMAL spreads, we don’t disclose specific spreads by country or product, but I can give you directionally consumer credits and personal loans have the highest NIMAL margins throughout followed by merchant credits, which has also a relatively high margin. In both of those products, as we move up market, we expect to reach users with lower risk with whom we charge lower interest rates and they will result in lower – smaller spreads, but it will be incremental volume as Osvaldo was mentioning before. Credit card itself as it is ramping up. And remember, the credit card, it takes several months, a few months to reach profitability. So as we are scaling the NIMAL spread of the credit card is neutral or marginally negative, but we have seen the older cohorts becoming positive already. So we are confident that as we continue to increase our portfolio and clean the other cohorts who will reach profitability at a portfolio level.

The next question comes from Neha Agarwala from HSBC. Please go ahead. Hi, Neha. Is your line muted?

Hi. Can you hear me? Hello?

Yes, we can. Go ahead, please.

Okay. Perfect. Sorry, zooming in again on the credit cards. Could you just give us a breakdown as to the usage of credit cards on platform and off platform roughly, both in Brazil and Mexico? And also, any sense of what percentage of your credit card portfolio is in Brazil versus Mexico? My other question is on the software business. You mentioned that in Brazil in the acquiring space, -- you launched software to increase the monetization of the acquiring business. How the software business is in-house? Are you building software in-house or are you working with software providers and attaching it with your POS to make the usage seamless for the merchants? And what is the ambition there? If you could elaborate a bit more on that? Would it be organic, M&A, partnership? Thank you so much.

So Neha, the first part of the question was related to credit card use between the marketplace and off platform. I would say that, in general, the majority of the volume we do, both in Brazil and in Mexico is off-platform. I would say over 60% of the volume we do in both countries is off platform. Even though we do give extra incentives for people to use the card on platform, like, we believe it's great because they show that they're not just using the card for those incentives, but that we are gaining principality. And in many cases, this is the card that gets top of wallet and gets us more frequently.

With regard to software, I think it's, as you mentioned, it's an important part of the strategy of our acquiring business, and we are making good advances in the trial, we mentioned some of those in the letter, but for the most part, it will be internal developers. Typically are not -- we have not been very acquisitive in terms of buying other companies. So most likely will be on development.

Let me complement that, Martin. Thank you. I will say that -- two things. On the one hand, there are many things we want to develop. We have started with catalog, the ability to invoice and also to keep stocks and then I think that we will develop related to basically to become really the neural system of small and medium stores. There are many things that we need to do to be able to achieve that, but we are starting to see good results in terms of engagement and driving higher use of Mercado Pago, in particular with SMBs where they usually have several MPOs in front of them. Also with regards to acquiring, we have seen some – I think that this is typically very complex because you really need to integrate it with all of the payment system, and we have seen other companies struggled with this in the past. So it’s a likely – it’s always something we look into, but we usually prefer to develop it ourselves.

The next question comes from Robert Ford from Bank of America. Please go ahead.

Hi, Martin, Ariel, Osvaldo. Thanks for taking my question. Can you talk a little bit about the rate of adoption of Meli Mas in Brazil and Mexico? And more specifically, how behavior patterns change following subscription on the marketplace and also across other verticals and are there any early indications of incremental principality? And then how is Meli Mas impacting Blue funding as well? Thank you so much.

Hi, Bob. How are you? Martin. As you know, we did -- we implemented a change in our loyalty program this quarter, where we opened 2 tiers of the program: one priced at $2, which included all the benefits of free shipping that we traditionally had, including also now the fintech benefits of cash backs, both on the marketplace as well as when you use the credit card outside of MercadoLibre. And then we have the other -- the traditional one, which is called Meli Mas total, which includes all the benefits plus the content from this Meli. We don't disclose specific numbers on the loyalty program, but we have seen very positive adoptions. We have the new program, the one that lower-priced one. We have seen a lot of users engaging with that loyalty program. And then the effects of people enrolling into a program are similar to what we have seen in the past. We've seen more uses, more better conversion, better engagement with our platform. So it's all very beneficial to us. We're very excited about the relaunch of a program, but it's still too early to share our actual numbers on that. Many refer to the Blue money. As you know, we're giving cash back on purchases of MercadoLibre as well as purchases done with the Mercado Pago credit card those cash backs are done on mainly dollar. It’s still too early. I mean we’ve seen users accumulating those balances. They are accumulating more than expanding it. So actually balance their accounts is increasing, but there’s nothing really to report so far. It’s too early.

The next question comes from Joao Suarez from Citibank. Please go ahead.

Thank you. Two questions on my side. The first one, I just wanted to hear your thoughts. I mean we're looking to Brazil, the credit environment, right? It seems like October had a slight uptick in households with outstanding debt, right? So it seems like there's some -- of course, with the interest rates increasing, right, the components increase in interest rates here in Brazil to 11.25%. So it seems like you're very confident in growing keep the origination at this space, right? But how are you thinking about the macro? Could it impact somehow, right? And just going back to the question about Brazil and Mexico. I just still -- I didn't understand how much is Mexico growing versus Brazil? If you could please clarify. Thank you.

Hi. Yes. It's Martin here. Yeah. It's correct. We'll be seeing this increase in interest rate in Brazil. We have seen it in different countries where we operate. But the reality is when we look at our credit business, much more important than one or two points of increase in interest rate or the cost of funding is an ability to predict defaults. Now that's the biggest part of the cost structure of our products. Again, once again, we are accelerating because we feel very confident about the improvements that we have made on models, and we're reaching new segments of the population that are lower risk. So I think macro, overall, it does not affect this product as much as a bit to predict bad debt.

And on the second part of the question, we don't disclose precise numbers by country. What I was referring to is that if you compare the first year or the first quarter of Mexico to the first ones in Brazil, we are issuing cards in Mexico at a significantly faster rate than we did in Brazil, basically because we have had more prior experience building credit card models.

[Operator Instructions] And our next question comes from Josh Beck from Raymond James. Please go ahead.

Yes. Thanks for taking the question. I wanted to ask about the fulfillment center ambitions. You obviously said it, I think five in Brazil and one in Mexico. Is this a multiyear investment cycle. And then I'm also curious, as you have a higher density of fulfillment centers that's maybe a little bit more hub and spoke. What type of benefits are you seeing in terms of cost to serve and ability to do shorter delivery windows?

Hey, Josh, Ariel here. So to the first part of your question, there is no new multiyear cycle or change in ambitions. We've been always saying that we are -- we pretend to continue building the capacity that we need in order to make sure we can deal with the growth in demand that we are forecasting for MercadoLibre. The growth in penetration that we'll have in Brazil and other countries and normalizing the capacity required to operate in Mexico after a few years in which we were running super tight. So no new direction, no changes in terms of investment cycles, just reality on what we think we need to continue gaining market share as we have been doing over the last few years. To the second part of your question, there is definitely an angle in terms of geographic expansion. We are not only building warehouses in the main capitals of the countries as we have announced. We are also regionalizing our capacity more and that means that we'll be better equipped to serve our customers with faster delivery promises and hence, better hopefully, retention rates coming from them. Different profiles of warehouses have different unit economics, I think we have never gone that deep into disclosure, and we'll probably do not do so in the future. But as I was saying before, the story and the type of unit economics and investment modes that we're thinking are not changing from what we were thinking in the past, this is the type of investments that we need to continue growing and serving our users.

And let me just to complement from a cash flow perspective, these are investments that are business as usual. If you look at the CapEx during this quarter in which we opened six fulfillment centers, our logistics was $120 million, which was more than last quarter, of course, but as a percentage of revenue is relatively stable compared to the past several quarters. So it's something that it's not an investment cycle. It's not -- obviously, this quarter, we opened more fulfillment center in the past couple of quarters, but this is business as usual as Ariel mentioned.

Our next question is a follow-up from Geoffrey Elliott from Autonomous. Please go ahead.

Hello. Thanks very much for taking the follow-up. I just wanted to get some explanation on one point, that there's a note that G&A expenses rose 1.1 percentage points year-on-year, of which 0.8 percentage points due to a non-recurring expense that relates to certain customer refunds for prior periods. What happened there? Why refund customers for prior periods.

Hi. It's Martin here. Yeah. As you mentioned, there was a 1.1 compression on the margin because of G&A. Most of that is related to a non-recurring expense this quarter, which is a refund to fintech users for the past several years back. What happened this quarter, we made a change to our consumer and merchant credit products in Brazil, but from now on, when you user faces an installment a few days before the actual due date, we'll be refunding them the base which the advancements, sorry, in terms of the days that they prepay. So even if it's a few cents, we will refund them for that installment. So since we are already doing that going forward, we decided to go back in time and refund our users for the past several years. And those are refunds that are very, very small on average $4 per user, $0.40 per actual installment paid in advance those are few days, but we felt that it was the right thing to do to make it to – make it up for them since then, we are not making a change going forward. So this is the one-off that we registered this quarter. That’s it.

And our next question comes from Rodrigo Gastim from IBBA. Please go ahead.

Yeah. Thank you, guys. Just two questions here on the loyalty program. First, if you're seeing any kind of margin pressure coming from the loyalty program given that you just re-dampened the program this quarter. So any margin pressure on this front, how much you're investing on that, that's the first point. And the second one, what are the main unit economics difference between the essential plan, the total plan that we just launched. So just curious here to understand how you see that? What are the main KPIs and the difference in terms of economics? That’s it. Thank you, guys.

Hi. It's Martin here. I think your question is related to margin -- potential margin pressures because of the loyalty program. First of all, we -- all the benefits that I was mentioning before, the Fintech related benefits and the cashbacks, we started in this quarter, so it's still ramping up. The way we are investing that is we're relocating marketing dollars to the loyalty program. So we don't expect that to generate significant margin pressure. But also keep in mind that expectation of the loyalty program is to bring incremental volumes that should compensate that whatever investment we make on the program.

So just to complement on unit economics, while we don’t disclose the whole details on our program. Bear in mind that in the total subscription, we need to pay Disney for a big chunk of what we collect as part of our agreement with the subscription for content, while in our essential subscription, we are basically giving our users the same type of benefits within our ecosystem, but we don’t have to pay content. So basically, the revenues we get from the subscription do help us sustain or pay for the benefits that we are providing with our users.

And the next question is a follow-up from Jamie Friedman from Susquehanna. Please go ahead.

Hi. I was just hoping you could elaborate or share some observations about Argentina on the marketplace side. I see in here that in the presentation you described it as Argentina improving significantly being driven by growth initiatives and consumption trends. But could you elaborate further what the market environment looking like in Argentina? Thank you.

Hey, Jamie. Ariel here. So yes, as we mentioned in the presentation, we see an improvement in the growth rate and the health of our business in Argentina. It's, on the one hand, there is an improvement in general consumer behavior and the macro scenario that definitely play a role for us in terms of a tailwind to expand our business. But also there is something around our own performance, right? So we continue to roll out experience improvement, assortment, improvement, removing friction from our users. We've also taken measures to help our buyers and sellers while taking care and preserving our P&L to boost frequency and drive conversion rates like installment, shipping prices, etc. But in general, we remain very, very optimistic what we are seeing in the country, demand is performing well, and we are satisfied with the turnaround that we saw over the last few months.

Yes. Just to complement that, on the fintech side of the business, we're also seeing very good traction in Argentina, with interest rates coming down, NPLs under control and very good spreads on our credit product, we also accelerated credit this quarter in Argentina. So very positive signs as well from the fintech side.

There are no more questions in the queue. This concludes our question-and-answer session. I'd like to turn the conference back over to Martin de los Santos for any closing remarks.

Thanks, everybody, for joining the call tonight. We had another quarter of strong results and great growth throughout our business. As we mentioned, we invested during this quarter on credits, which is strategic for Mercado Pago as well as logistics, which is critical to continue bringing volume on line. Meli is very well positioned to capture all the investment -- the growth opportunities that we have ahead of us, and we are investing behind those with discipline, focusing on long-term and sustainable profitable growth. Let me take a minute now to remember Mario Vasquez, who passed away a couple of months ago. He was a board member for MercadoLibre for 17 years. He left a very strong mark on all of us, all of us who had a chance to work with him for those years. So we are going to miss him. And will now close the call, and thank you again for joining.

The conference has now concluded. Thank you for attending today's presentation. You may now disconnect.