Earnings Call Transcript
Nu Holdings Ltd. (NU)
Earnings Call Transcript - NU Q2 2022
Operator, Operator
Good evening ladies and gentlemen. Welcome to Nu Holdings Conference Call to Discuss the Results for the Second Quarter 2022. A slide presentation accompanies today's webcast, which is available in Nu's Investor Relations website. This conference is being recorded and a replay can also be accessed on the company's IR website. Please be advised that all participants will be listening only mode. I would now like to turn the call over to Mr. Jorg Friedemann Investor Relations Officer at Nu Holdings. You may proceed.
Jorg Friedemann, Investor Relations Officer
Thank you very much operator. Good evening everyone and thank you for joining our earnings call today. If you have not seen our earnings release, a copy is posted in the Results Center section of our Investor Relations website. With me on today's call are David Vélez, our Founder, Chief Executive Officer, and Chairman; Youssef Lahrech, our President and Chief Operating Officer; and Guilherme Lago, our Chief Financial Officer. Additionally, Jag Duggal, our Chief Product Officer will join us for the Q&A session of the call. Throughout this conference call, the company will be presenting non-IFRS financial information including adjusted net income. These are important financial measures for the company, but are not financial measures as defined by IFRS. Reconciliations of the company's non-IFRS financial information to the IFRS financial information are available in our earnings press release. Unless noted otherwise, all growth rates are on a year-over-year and FX-neutral basis. I would also like to remind everyone that today's discussion might include forward-looking statements which are not guarantees of future performance and therefore, you should not put undue reliance on them. These statements are subject to numerous risks and uncertainties that could cause actual results to differ materially from the company's expectations. Please refer to the forward-looking statements disclosure in the company's earnings press release. Today our Founder and CEO, David Vélez will discuss the main highlights of our second quarter 2022 results and some of the opportunities ahead. Subsequently, Guilherme Lago, our CFO; and Youssef Lahrech, our President and COO will take you through our financial performance for the quarter. After which time we will be happy to take your questions. I would now like to turn the call over to David. David, please go ahead. Thank you.
David Vélez, Founder & CEO
Thank you, Jorg. Hello everyone. Thank you for being with us today. Happy to announce that we had another very strong quarter combining growth, efficiency, and significant progress towards building the largest digital finance platform in Latin America. On the customer acquisition front, we attracted another 5.7 million customers, reaching more than 65 million customers at the end of the quarter. This growth continued to be primarily fueled by organic channels with very low customer acquisition costs. Notably, our monthly activity rate improved even further to 80%, a new historical high mark where customers now account for 36% of the adult population of Brazil. The growing number of active customers coupled with the continuous maturation of our earlier cohorts resulted in new historical highs of purchase volume, credit portfolio, and revenue. Our purchase volumes reached $20 billion, up 94% year-on-year. We became the number four cards player in Brazil this quarter, surpassing century-old incumbent institutions and we continue to gain market share quarter-after-quarter. Our consumer finance credit portfolio grew significantly above market, up 107% year-on-year to a total of $9.1 billion, while maintaining very strong unit economics. And our credit portfolio remains concentrated in credit cards and unsecured personal loans. We have not even started our journey into secured credit products. Our revenue more than tripled year-on-year to about $1.2 billion this quarter, driven by both product upsell and cross-sell and resulting in a historical high average revenue per active customer or ARPAC. While we have discussed our ecosystem in the past in today's presentation, we wanted to spend a bit more time highlighting the significant progress we have made to-date in cross-sell, a key piece of our strategy. Going from a mono-product to a multi-product platform and from a mono-country to a multi-country platform, while improving profitability as we scale. First, moving from a mono-product to a multi-product platform. We started in 2013 with the strategy of building the future of financial services in Latin America with a credit card as our first product. We chose credit card to address the large consumer pain in Brazil had very attractive unit economics and enabled us to start building the credit underwriting capabilities from day one. We were very successful in using credit cards to build a large consumer base, and one of the strongest consumer brands in the region. In about nine years, we became one of the largest credit card issuers and one of the top 10 most loved brands in Brazil. And we continue to expand this core product, at a very significant pace. It remains one of the highest credit cards in the Brazilian market, while already generating hundreds of millions of reals in earnings, which we then are able to use to finance the growth of new verticals. Most fintechs globally have had a very hard time going from being a successful monoliner, to them being able to build a multi-product platform. And here, is where we think Nu is a significant exception, in having been able to successfully introduce products across many verticals in financial services. Our second product or savings account NuConta, has now surpassed our first product credit cards, and today comes with over 45 million active customers. It has displayed exceptional levels of engagement and deposits growth. NuConta has allowed us to become one of the largest payment platforms in Brazil today, with around 23% of all the big transactions in the country processed through our pipes. NuConta has also allowed us to develop a unique deposit franchise. Today, our retail deposits exceed our interest-earning portfolio by more than 4x. We do not need to rely on portfolio securitization or wholesale funding. We own our own destiny with respect to funding, a rare feat for our consumer fintech. Our third product, personal loans, currently has 4 million customers and is generating meaningful earnings for our business. Here our growth bottleneck is our own willingness to underwrite credit and enable this feature for our customers, as we tend to go slowly and conservatively with any credit product. Our growth bottleneck is certainly not distribution. As of April 2022, our almost 59 million Brazilian customers account for approximately 32% of the entire personal loans market in Brazil. So we do not need to go outside our existing ecosystem to tap into approximately one-third of the Brazilian consumer finance industry. Our fourth product or savings account for small and medium enterprises, has brought us access to the SME segment, whose profit pool is estimated at approximately 25% of the consumer profit pool. In less than three years, and with a particular focus on micro entrepreneurs and small enterprises, we became one of the top five SME players in Brazil, in terms of number of active customers, with 12% market share of corporate checking accounts in Brazil. As of the end of this first quarter, we had two million SME customers up 150% year-on-year. Beyond these core four products, we launched insurance last year where our first product Nu Vida, a life insurance product became the fastest-growing insurance product in the history of Brazil. Today Nu Vida has over 700,000 policies and maintains the highest Net Promoter Score of the sector. Then, we entered the investment space through the acquisition of the largest independent direct-to-consumer platform in Brazil, Easyinvest rebranded NuInvest, which was a key part in allowing us to reach now over five million active customers in our investment business, already likely the market leader in the Latin American, direct-to-consumer investment space. And even more recently, we announced that we reached more than one million customers in our crypto product NuCripto, three weeks after launching it to our entire customer base. So in just a few years, we were able to go from a credit card monoliner to achieving leadership positions in the verticals we enter: deposit accounts, personal loans, SMEs, life insurance and investments while reaching over one-third of the adult population in Brazil. This to us represents significant progress in advancing our product diversification and cross-sell strategy. Second, moving from a mono-country to a multi-country platform. On the international expansion front, we reached 2.7 million customers in Mexico consolidating our position as the number one issuer of new cards in this country. In Colombia, we grew our customer base to over 300,000 customers and have received over 1.5 million applications to date, demonstrating a combination of the significant pent-up demand, as well as the selectivity we're having in this country. Likely, we are the number one credit card issuer in Colombia as well by now. All of this was achieved while sustaining a Net Promoter Score in the 90s, and high levels of customer virality in both geographies. It's worth highlighting that, one week ago Nu Colombia received approval for its requested license to establish a financing company in the country. With this approval, we now have banking licenses in the three countries we operate, transposing an important barrier to entry and enabling us in the future to intermediate financial products and offer deposit products to our customers. Together, Mexico and Colombia are largely the size of Brazil in terms of both GDP and population. We're excited with the prospects of these two countries and confident that they can be as relevant to us in the future as Brazil has begun. Finally, all of this growth has required tremendous investment and we realized that it might be hard for investors to assess whether all of this growth is profitable. Therefore, we thought it would be useful to share a few KPIs of our operations in Brazil our core market, which we do here for the first time. In the first half of 2022, NuBrazil generated an account profit of $13 million, up from a loss of $19.7 million in the full year 2021, while continuing to provision significantly as a result of the continued fast growth of our credit portfolios, as well as expensing large investments of new products and features. Our core products credit cards personal loans and NuConta are by now very profitable. And therefore, we're able to reinvest all of those profits into the expansion and improvement of our operations. And even with all of that investment, Brazil is now profitable and generates profits which we now expect to reinvest into the growth of Mexico and Colombia. We acknowledge investors increasing focus on the importance of profitability. And while we continue to make our decisions based on long-term value creation, we are not disconnected from our investors in the need to sanity check the profitability engine of the entire model as we invest for the future. The current context of macro volatility reinforces our focus on cost diligence on different fronts. This should contribute not only for us to achieve our goal of becoming the lowest cost manufacturer in our markets but also shorten the path to very high profitability. And with that I turn the call to our CFO, Guilherme Lago who will discuss the operating trends for the second quarter. Go ahead Guilherme.
Guilherme Lago, CFO
Thank you, David. Now before moving on to the quarterly results, let's recap our simple yet powerful value-generating formula. First growing our customer base across Brazil, Mexico and Colombia and turning customers into active customers. Second, expanding the average revenue per active customer or ARPAC through both product upsell and product cross-sell. And third, delivering all this growth while maintaining one of the lowest cost operating platforms in the industry through best-in-class cost to acquire, to serve, of risk and of funding. Now let's take a deep dive into the quarterly results of our business. We added 5.7 million customers during the second quarter, in line with the net adds of the first quarter and mainly through organic channels. This brought total customers to 65.3 million by quarter end, a 57% year-on-year increase making us likely the fifth largest financial institution in Brazil in terms of number of customers. We are also pleased with Mexico and Colombia continue to grow at faster rates compared to Brazil and have contributed together with over 700,000 unique customer additions this quarter. Importantly, we continue to drive customer acquisition while increasing the monthly activity rates to 80%, up from 72% a year ago and 78% in the prior quarter, marking the 90th consecutive quarter of higher monthly activity, another testament to our ability to continue to grow our ecosystem while driving customer engagement. Now moving to an analysis of our customer cohorts. These three charts show how increasing engagement of our customer base and a higher number of products per active customers continue to drive up ARPAC. When looking at ARPAC in the chart on the far right one can see that, while we reach the monthly ARPAC of $7.8 in the quarter, mature cohorts are already at $21. One can also see that all cohorts maintain healthy growth trends. This combination of higher engagement and more products per customer has proven powerful in terms of customer monetization. ARPAC expansion has helped fuel our triple-digit revenue growth. We also saw another quarter of exceptional revenue growth, 230% year-on-year on an FX-neutral basis, reaching a record high quarterly revenue of nearly $1.2 billion. This growth is the result of the compounding effects in two areas: first, a growing number of monthly active customers, not only customers, but monthly active customers. Second, higher levels of product upsell and product cross-sell which continue to expand our monthly ARPAC. Although our monthly ARPAC has expanded strongly over the past quarters, we believe we are very far from our ARPAC potential. The ARPACs of incumbent banks are still far above ours at $40. Notwithstanding our ARPAC potential, it's important to note that with the advantages in our cost pillars, we can afford to serve customers with lower ARPACs and still post very healthy unit economics. Moving on to the progress of our cards business. Purchase volume increased to $20 billion, up 94% year-on-year on an FX-neutral basis. Again, this is the result of more product upsell, cross-sell and the continued engagement of our customers. In the second quarter, as David mentioned, we estimate that we became the number four cards player in Brazil, continuing to surpass some of the most traditional incumbent financial institutions in the country. In our consumer finance portfolio composed of credit cards and personal loans we continue to grow at a healthy pace on an FX-neutral basis albeit slower than in prior quarters. The lower level of growth stems mostly from personal loans. This quarter, we decided to reprice and moderate the growth of our personal loan portfolio aimed at strengthening its credit resilience in the context of a more uncertain short-term outlook for the Brazilian economy. I also make two observations to contextualize this trend. First, FX negatively impacted our balance figures. If we adjust the portfolio figures on an FX-neutral basis, our portfolio would have grown by around $300 million in personal loans and over $1 billion in credit cards. Second, originations in personal loans remain largely unchanged in FX constant terms quarter-over-quarter. In the context of a short duration portfolio, similar origination levels yield lower growth rates simply because part of the originated volume replenishes the amortizations in the period. Let's now look into two things in more detail. One, the evolution of our credit card interest-earning portfolio; and two, the evolution of our personal loan origination. With regards to the evolution of our credit card interest-earning assets in Brazil or IEP, it's important to contextualize, that we have been consistently expanding our credit card IEP, mainly as a result of launching new features that allow our customers to use their credit cards as financing means. These features include the ability to finance bank slips, individual purchases into installments and large transfers in all cases using credit card limits. As you can appreciate, IEP arising from installments, including those arising from these new financing features have outpaced the growth of IEP from revolving. And they are still way below the industry average. This has been a change in behavior we have incentivized, rather than a change in behavior suggesting credit degradation. Finally, it's worth highlighting that these IEP figures capture the entirety of our credit card asset class. We do not commingle IEPs of credit cards and personal loans. If you were late in credit card, you are ineligible for a personal loan. Now, moving to the originations of our personal loans. The graph in this slide shows the evolution of our personal loan origination, along with the respective monthly interest we charge on them. One can note two things in the second quarter of 2022. First, we continue to reprice our personal loan origination, aimed at increasing our credit resilience and offsetting the impact of higher interest rates. Second, originations for our personal loan book remained largely stable on an FX-neutral basis. Going forward, the pace of origination in personal loans will hinge on the short-term outlook of the Brazilian economy and the credit performance of our cohorts. As of today, we are assuming for the third and fourth quarter of this year levels of origination similar to that of the second quarter on an FX-neutral basis. We continue to advance on our strategy of building a robust local currency deposit franchise, which we use to fund most of our operations. We ended the quarter with a loan-to-deposit ratio of only 24%. Deposits continued to grow at a fast pace during the second quarter, up 87% year-on-year on an FX-neutral basis, with an average funding cost below CDI, Brazil's risk-free rate. We added nearly $700 million in deposits over the last three months, closing the quarter with a total deposit balance of $13.3 billion. On a quarterly FX neutral basis, deposits would have grown $1.9 billion quarter-over-quarter. In addition, we recently introduced changes on the yield remuneration for short-term deposits within the new account. At the same time, we launched other features in our ecosystem, such as the money box, which we expect will help optimize our deposits franchise, faster cross-sell and upsell of our investment platform and mitigate potential attrition. These changes combined can contribute to a reduction in our cost of funding over time. Since we touched again on the topic of costs, in this slide, we remind you that one of the key competitive advantages of our platform rests on its very low cost to serve. This quarter, our average monthly cost to serve was $0.80 remaining below $1, while our monthly ARPAC reached $7.8. I remind you that our cost to serve continues to be 85% lower than those of traditional banks. Going forward, we expect our cost to serve to remain at the dollar level. Moving down the P&L, we delivered another quarter of record high gross profit, up 109% year-on-year on an FX-neutral basis to $364 million. Note that, our gross profit margin has decreased in the quarter as a result of the following factors: first, growth driving credit loss provisioning. Under IFRS, we have to front load the recognition of our credit loss provisioning, whenever a loan is booked. So, the faster we grow our credit book, the more short-term pressure it brings to our gross profit margin. Second, interest on cash balance. As interest rates go up, we earn more money on our large pool of cash balance, even if it's partially or fully offset by higher funding expenses. In other words, higher rates drive our revenues up, but leave our gross profit largely unchanged. This pushes down our gross profit margin, as it increases the denominator of the gross profit formula. As our credit portfolio matures and interest rates stabilize, the gross profit margin is expected to converge to those of the more mature cohorts. Operating leverage is a key component in our platform and can be observed in two fronts. First, as we expand our credit portfolio, we optimize the use of our large and low-cost deposit base and expand our net interest margin or NIM, as can be seen on the chart on the left-hand side. Second, as our overall revenue levels increase, we further dilute our low-cost operating platform and improve our efficiency levels as can be seen on the chart on the right-hand side. We expect both trends to continue and compound over time, allowing the company to achieve high NIMs and best-in-class efficiency ratios. Moving on to the bottom line, our recurring profitability confirms, again, that we are on the right path with our earnings formula. We reported adjusted net income of $17 million. To sum up this section, it's important to emphasize that despite our undeniable growth orientation we will never leave behind our cost discipline, as we believe that the long-term winners in the Latin American financial services industry will be the lowest cost manufacturers. So we aim at being the lowest cost manufacturer across the four cost pillars of retail banking: cost to acquire, cost to serve, cost of risk and cost of funding. I would like to highlight two important operating trends that exemplify our cost discipline. First, we expect a slower personnel growth in the second half of the year after advancing on the staffing of our new geographies during the first half of 2022. Second, we expect a reduction in our cost of funding over the coming quarters, as a result of the change to the new account remuneration. Now, I would like to turn the call over to Youssef, our President and Chief Operating Officer, who will walk you through our asset quality performance and credit underwriting approach.
Youssef Lahrech, President & COO
Thank you, Lago. Let me now walk you through a few key indicators that track the asset quality and overall health of our credit portfolio in the second quarter of 2022. As a reminder, we make underwriting decisions to optimize the return, resilience and payback of our credit originations. Let me start by reinforcing two key features of our credit business. First, we underwrite mostly unsecured credit through credit cards and personal loans. These products are expected to have higher loss rates and shorter durations than secured credit products and they are priced accordingly. We seek to optimize return for the amount of risk we take, not minimize risk. Second, as Lago mentioned, we have taken management actions to reprice our products and bolster the resilience of our credit underwriting. As a result, we expect risk-adjusted margins of 50% to 60%. These actions have resulted in even more resilient portfolios. These charts illustrate the unit economics of each of our core credit products, credit cards and personal loans, both expressed as an annualized percentage of receivables. On the left-hand side, you can see the unit economics of our credit card product. Revenues consist of non-interest and interest revenues, which together amount to 27%. Deducting expected losses of 11%, we end up with a net lending margin of 16%. This results in a post-tax return on assets of around 7% and an ROE in excess of 80%. On the right-hand side, you have the unit economics of our personal loan product. The interest revenue yield is about 58% annualized, with a cost of funds of approximately 11%. The expected annualized losses for the portfolio are about 22%, yielding a net lending margin of 25%. This results in a return on assets of about 14% and a return on equity in excess of 120%. Not only are the returns of these products attractive, but their resilience is also strong. They can both withstand more than a doubling of losses and still be profitable. We are, of course, in the business of taking credit risk and always with an eye on earning commensurate returns and always ensuring the resilience of those returns in the face of uncertainty. With this, let's now turn our attention to asset quality trends. Overall, leading indicators namely NPL 15 to 90 have remained stable and asset quality has followed its post-pandemic normalization course as discussed in prior calls. Now, before I discuss the specific trends, let me make a few observations on the metrics themselves. First, it is important to note that NPL 90+ is both a lagged metric and also more of a stock than a flow metric, given that assets continue to accumulate in that bucket until write-off. Hence, to get a sense for the latest trends in credit performance, we find early delinquency metrics like NPL 15 to 90, the chart on the left to be more informative. In the appendix of the presentation, we provide an illustration of how this lag and the stock versus flow dynamic result in different behaviors of 90+ delinquency ratios compared to 15 to 90 ratios. Second, in Q2 of 2022, we made a change in the write-off methodology for personal loans to better align it with recovery expectations as per IFRS guidelines. This resulted in anticipating write-offs for nonperforming personal loans from 360 days to 120 days of delinquency. This change has two distinct impacts on NPL metrics. First, it reduces 90+ NPL ratios by virtue of eliminating 121 to 360-day delinquent loans from both the numerator and the denominator of that calculation. Second, it increases NPL 15 to 90, as it reduces the denominator of that ratio. The charts on the slide incorporate these changes, as if they had been implemented since the fourth quarter of 2017, for ease of comparability across the whole time series. For credit cards, write-off timing has remained unchanged at 360 days. For both products, we apply a partial write-off methodology, which means that only the expected recovery part of written-off loans is kept on balance sheet under Stage 3. Lastly, I want to emphasize that this change does not affect the P&L income statement, as these write-offs had already been fully provisioned for under our expected credit loss methodology. Let us now turn to the actual NPL trends. NPL 15 to 90 as a leading indicator is showing a stable picture, suggesting that the post-COVID normalization cycle may have run its course. 90+ NPL ratio increased from 3.5% to 4.1% in line with our expectations. The normalization cycle is still working its way through the 90+ ratio as the delinquency inventories continue to flow through later stage buckets. Our baseline expectation for the rest of 2022 is that our early delinquency ratio of 15 to 90 will remain largely stable, absent any substantial changes in either of our underwriting strategy or the environment. As 90+ is a lagging indicator and more of a stock than a flow metric, it will mechanically take more time for it to stabilize, so we expect it to continue to rise over the coming quarters. However, under IFRS 9, this expected mechanical increase in 90+ rates does not represent a P&L overhang, as we have already provisioned for the expected delinquency flow-throughs under our expected credit loss methodology. We provide more information about this in the following slide. Let me recap the impact that expected credit loss or ECL as a loan loss provisioning methodology has on the consumer finance business with high growth rates, as is the case for Nu. For IFRS 9, loan loss provisions must be recognized when a loan is granted, even before any revenue associated with that loan is accrued. This results in an intentional timing mismatch between revenues and costs. For this reason, the higher our consumer credit growth rate, the more provisions we have to book upfront. And as Lago mentioned earlier, this negatively impacts gross profit and gross profit margins during periods of high growth. And as growth rates normalize, vertical gross profit margins are expected to converge over time to those of mature cohorts. With that context, let us turn our attention to NPL provision formation in the quarter and its drivers. Our provision balance grew by $129 million on an FX-neutral basis or 15% quarter after quarter, after taking into consideration the change in write-off methodology. Excluding the impact of the change in write-off methodology, around 85% of the provision built in the quarter or $207 million was driven by the growth of the portfolio. The remainder, 15% or about $39 million was driven by the changes in macroeconomic assumptions. In the context of IFRS 9 and given the short duration of our portfolio, risk add-on provisions tend to be far more sensitive to changes in 15 to 90 delinquency, rather than in 90+ delinquency rates. And as discussed earlier, early delinquency, our leading indicator has remained largely stable. In summary, the growth of our portfolio has remained the dominant driver of provision changes. To wrap up, I want to reinforce that we are confident in our credit strategy, both in terms of our ability to underwrite and our ability to price adequately for risk. Our models are continuously enhanced and our credit framework is designed to be resilient to the ups and downs of macro cycles. Having shared these data and perspectives on credit and asset quality, let me now turn the call back to our Founder and CEO, David Velez for his concluding remarks.
David Vélez, Founder & CEO
Thanks, Youssef. To summarize, our performance in the second quarter illustrated the distinctive strengths of our platform, which we believe provides a unique combination of exceptional growth with compelling unit economics and operating in one of the largest and most profitable banking markets globally. First, exceptional growth. Our customer base grew to over 65 million customers and nearly 60% growth, driven by organic channels. Our monthly active customer base grew even more reaching 52 million customers, a 75% year-on-year growth. Mexico and Colombia are already moving the needle. They have contributed over 700,000 customers this quarter and we believe we have already become the number one issuer of new credit cards in both markets. Our ARPAC grew to $7.8 at 105% year-on-year FX neutral growth, driven by continuous product cross-sell and upsell. Our purchase volume grew to $20 billion, a 94% year-on-year FX-neutral growth making us already the fourth largest player in Brazil. Our revenue grew to $1.2 billion in the quarter, a 230% year-on-year FX-neutral growth, driven by both customer adds and ARPAC expansion. It is hard to find a company that is compounding this level of growth at our scale. Second, compelling unit economics. We believe we have become the lowest cost manufacturer of our industry combining best-in-class cost to acquire, cost-to-serve, cost of risk, and cost of funding. While our ARPAC more than doubled over the past year, our cost-to-serve remained flat in the same period exhibiting the very high operating leverage potential of our platform. Our core consumer finance products, credit cards and personal loans, presented above industry average profitability with ROEs in excess of 80% and 120% respectively, even in this more uncertain environment proving the robust levels of credit resilience. Our core market Brazil generated positive earnings, which are expected to compound over time and allow us to continue to invest into new products and geographies. It is hard to find a fintech of our scale with such a compelling profitability structure. We're very proud of what we achieved this past quarter and even more excited about what lies ahead of us in the coming quarters. We would like to take your questions now. Thank you very much.
Operator, Operator
We will now begin the question and answer segment for investors and analysts.
Jorg Friedemann, Investor Relations Officer
Thank you operator. We are collecting questions and our first question comes from the line of Jorge Kuri from Morgan Stanley. Could you please open his line? Thank you.
Jorge Kuri, Analyst
Hi everyone. Congrats on the great results. Could I ask you about the growth in the portfolio? You've made some comments about the slowdown particularly on personal loans being driven by the environment that you're seeing in Brazil now. Maybe we can just double-click on that. Brazil is actually growing more than what consensus expected. The economy will probably grow almost 2.5%. Unemployment is now at a seven-year low. The consumer is getting more handouts from the government. And your early delinquency kind of points to a risk on scenario if you will. I mean there's evidently nothing that's going wrong with the portfolio probably maybe because you're slowing down. But given that your personal loans particularly are in such early stages of growth where the cross-sell vis-à-vis your total clients, your credit card clients is still small, it is a bit surprising in the context of everything that I said that the slowdown is so sharp versus the previous quarter. And I am looking at FX-neutral numbers and according to my calculations on a quarter-on-quarter basis, the growth really slowed down basically was cut in half. And so just wondering if you can provide a little bit more color on all of these. Thank you and again congrats on the numbers.
Guilherme Lago, CFO
Jorge, thank you so much for the question. Look I would start by referring you to slide 35 of the earnings deck because you will note that a material portion of the decrease in the velocity of our growth in our credit book actually stems from the FX devaluation. If you look at our portfolio growth on an FX-neutral basis, you will see that the growth remains quite healthy. You see that our credit portfolio went on an FX-neutral basis from about $7.9 billion to $9.2 billion. Now, we are certainly cautious about the environment. Now, being cautious does not mean stopping to grow. In the second quarter, we grew personal loans originations by about 7% quarter-on-quarter on an FX-neutral basis and almost 2.5 times year-on-year also on an FX-neutral basis. So, we believe we are remaining growing at paces that are materially higher than the market. The second thing is that the growth in personal loans and the growth in credit card is by no means constrained by demand. Like our customers account for about one-third of the consumer finance profile in the country. The growth is also not constrained by product. We believe that our credit card and personal loan products have the best NPS in the markets and some of the best conversion ratios in the market. It is also not constrained by capital and liquidity Jorge. Now, we have plenty of capital and liquidity available in our balance sheet. So, it is only constrained by our credit underwriting appetite, which is 100% under our control. And we have chosen to be slightly more selective in the second quarter, but we've chosen to actually be also repricing our products more aggressively, as you may note in the presentation that we have shared with you. Going forward, we do expect that we will remain to grow at levels that are much higher than the average growth pace of the market, but we will continue to be quite select and conservative as we have been over the past seven to eight years. I would say, that irrespective of a potential decrease in the velocity of growth in the second quarter, we continue to grow our customer base very aggressively. Our customer base already accounts for about 35%, 36% of the entire adult population of Brazil. It accounts for about one-third of the entire consumer finance pool in Brazil, so we have plenty of room to cross-sell in the future when the market proves itself.
Jorge Kuri, Analyst
Thanks for that Lago, that was very clear. And if I may follow up on the point you made about increasing rates for the personal loans business, which is on Slide 16 of the presentation. With an average rate of 4.6% in this quarter up from 4%, how much of this is a mark-to-market of the increase in funding costs? And how much is actually going to increase the net NII of the product? And has more been done in the third quarter, so we should expect sort of like the NII of the product to go up over time?
Guilherme Lago, CFO
Yes. So, just a small correction. It's 5.6% in the second quarter. But I think, the growth in pricing that we did in the second quarter of the year more than offset the increase in funding cost and should therefore, help us expand our risk-adjusted margin. I think for the first time we are providing disclosure, if you take a look on Slide 20, of the expansion of our net interest margin. So notwithstanding a scenario in which interest rates are going up, we have proven that we can put to work our very large low-cost deposit base and actually expand by taking our net interest margin from about 5.2% to almost 10%. We expect this trend to continue and compound over time, not only as we lower our cost of funding, but also as we increase our interest-earning portfolio IEP and therefore, increase the carry of our positions.
Jorg Friedemann, Investor Relations Officer
And our next question comes from the line of Thiago Batista from UBS. Operator, could you open his line. Thank you.
Thiago Batista, Analyst
Hi, guys. Thanks for the opportunity. Very good top line, not so good asset quality. I have a question on the asset quality. Trying to understand, why Nu has changed the write-off methodology for personal loans this quarter, if there's any strong motivation to do this right now.
Youssef Lahrech, President & COO
Hi, Thiago. This is Youssef. Thank you for the question. So a couple of things on the write-off methodology change. First of all, it has nothing to do with asset quality in the quarter. It is basically us following IFRS 9 guidelines. So IFRS 9 doesn't specify exactly the timing of write-off for a particular asset class or type of loan. However, it provides that you must write off a loan when there's no more reasonable expectation of partial or full recovery of that asset. So when we started the personal loans business a couple of years ago, we didn't have any actual experience with recovery rates. So as a default, we started with write-off timing that was similar to credit cards i.e. 360 days. Now two years later, we actually have actual recovery rate data to base our write-off methodology on, and we find that 120 days is a more appropriate reflection of the IFRS guideline. The other thing to note is writing personal loans at 120 days is generally, best practice amongst fintechs and financial institutions globally and we're following that practice. Now as I mentioned earlier, it's also important to know that these write-off methodology changes have no bearing whatsoever on our P&L or income statement because the assets written off had been already fully provisioned for in earlier periods.
Thiago Batista, Analyst
No, very clear. And just one very small follow up. If I'm not wrong you mentioned during the call that you guys are expecting a more stable asset quality going forward. Is this the right message on what you guys are expecting for the legacy ratio going forward?
Youssef Lahrech, President & COO
Yes, Thiago. It’s important to differentiate between two aspects. One aspect is lead indicators of asset quality, such as the 15 to 90 delinquency rates we disclosed, which have remained stable for a few quarters. This indicates that the normalization to pre-pandemic risk levels may have been reached concerning early delinquency indicators. In contrast, the 90+ NPLs represent a lagging metric and are still adjusting to the trends shown by early delinquencies. I anticipate that this will continue to rise over the next couple of quarters as it aligns with the early delinquency data.
Jorg Friedemann, Investor Relations Officer
And our next question comes from the line of Mario Pierry from Bank of America. Could you open his line, please operator.
Mario Pierry, Analyst
Hi, guys. Good afternoon. Thanks for taking my question. Let me ask you a question on deposits. You showed that you continue to grow your deposit base at a healthy pace. But we saw a meaningful slowdown. I think it rose above $3 billion in the previous quarter and this quarter it was less than $1 billion. At the same time you changed the remuneration of your deposits. So I was wondering has that impacted your ability to grow your deposit base? Clearly, we do see the benefits right of lower funding costs. But do you have any data that you can share with us what has been your – the impact on your ability to grow your deposit base since you changed the remuneration of the accounts?
Guilherme Lago, CFO
Sure, Mario. Thanks for your question. So I think if you – you're probably referring to Slide 17. And if you take a look at the slide, the apparent decrease in the velocity of our deposits from $12.6 billion to $13.3 billion stems mostly from the FX devaluation in the quarter. On a quarterly FX-neutral basis so same FX, deposits would have grown $1.9 billion quarter-over-quarter, right? So it would have grown largely at the same pace as the prior quarter. So just wanted to flag that FX devaluation plays a role in the way that we present those numbers. So that's one thing. The second thing is that no I think the changes to the NuConta remuneration to which you alluded has not had any impact so far in the velocity with which we are – we continue to accumulate deposits. We have now done the rollout of the NuConta remuneration for a relatively small portion of our customer base and we are doing no massive AB testing. We have not observed as of today any difference in any of the relevant KPIs including the amount of deposits, including engagement, including NPS. We will continue to monitor this very closely going forward. And the speed at which we roll it out to the entire base will largely depend on those observations. But so far we are quite encouraged by what we have seen.
David Vélez, Founder & CEO
Just wanted to add very quickly on the deposit product. One thing that's important to consider is this was a project that we worked on over a year that had us a goal to ultimately create a better deposit product at lower cost. And part of the challenge was to redesign the product so that customers could have access to investment products that have higher than 100% CDI, as we got to integrate a new investment product. And so ultimately what we're looking for is we should see an increase in deposit flow at a lower cost. That's at least our goal. The deposit product today is much better than two months ago as customers have access to a number of different functionalities. And so ultimately we're – it's very early days. We're still trying to monitor basically daily all those different metrics. But if things play out that's where we're ultimately going.
Mario Pierry, Analyst
Okay. No that's clear. And let me ask a follow-up question, right? It has been asked already about the NPLs and your risk appetite, you're diminishing the growth of personal loans. But I was wondering if you're making any adjustments on your credit cards. Are you reducing credit card limits? Is there anything that you're doing? Clearly you're seeing – even though you showed the NPLs are fairly stable do you think it's more prudent to be a little bit more cautious on personal loans? Are you also going to be a little more cautious on credit cards? Are you seeing anything that worries you on the credit card book?
Youssef Lahrech, President & COO
Yes. So just to address the question on our current stance in both credit cards and personal loans, I would say, we're consistent across both. We're not reducing credit card limits at the moment. What we're doing is we're erring a little bit more on the side of building resilience. If you look at the unit economics chart for both credit cards and personal loans in the earnings presentation, that's basically what we're targeting at the moment in terms of returns ROA, ROE margins and resilience. So it's resulting in growth that has not increased as much if you will compared to prior time periods but is absolute levels of growth that are fairly stable about $5 billion in personal loans we've added in the quarter which has been consistent with the prior couple of quarters and about five million new customers, which is also consistent with prior time periods. So we're basically taking a slightly more cautious approach in terms of resilience but not really taking any drastic approach. We're not seeing any worrisome signs. And if you look at the delinquency metrics we went over if anything they show signs of stability in the early delinquency metrics. So it's more of an abundance of caution given the uncertainty in the environment looking forward, not necessarily based on anything worrisome that we've seen in the data so far.
Mario Pierry, Analyst
Okay. Thank you very much.
Jorg Friedemann, Investor Relations Officer
And the next question comes from the line of Ashwin Shirvaikar from Citi.
Ashwin Shirvaikar, Analyst
Hello and congratulations from me as well. My first question is obviously your front-loading loan loss provisions is required. But your early delinquency indicators show stability as you pointed out in the 15-90 range. And while the mathematical impact, as it relates to NPL 90+ is there to see. I was trying to think ahead if you were to see continued stability in delinquency indicators, could you provide a framework for how to think of gross margins heading into next year?
Guilherme Lago, CFO
It's a good question. We believe that there are two main factors affecting our gross profit margin: the growth rate of our credit portfolio and interest rate changes. When interest rates increase, they have an impact on both the numerator and denominator of the formula, leading to a decrease in gross profit margin. Looking at our cohort data, we believe that once these factors stabilize, we could achieve gross profit margins around 60%. However, I'm not indicating that this will happen in 2023 or 2024. This reflects the steady state gross profit margin of a mature cohort.
Ashwin Shirvaikar, Analyst
Understood. Thank you for that. And just staying on the topic of scaling, as I'm looking at the various operating expense lines, should we look at G&A as a primary source of operating leverage, or are there opportunities in customer support and ops using automation and other techniques, or I would imagine you probably want to continue increase marketing at a robust pace.
Guilherme Lago, CFO
Yes, that's a great question. For the first time, we are sharing a key metric that we closely monitor at Nubank, which is the evolution of our efficiency ratio with and without share-based compensation. Please refer to slide number 20. Since the beginning of last year, we have improved our efficiency ratios from around 90% to 58% when share-based compensation is included, and from about 70% to 50% when it is excluded. We believe we are just beginning our journey towards greater efficiency. The improvements will primarily come from two areas. First, we expect our cost to serve to remain stable in dollar terms, while our average revenue per active customer is projected to continue growing, as we have observed in previous quarters. Second, our general and administrative expenses will increase at a significantly slower rate compared to the growth in our gross profit and revenues. In fact, if you analyze our G&A, you’ll find that 50% to 60% is driven by our headcount, and the rate at which we will grow our headcount in the next two years should be much lower than the rate in the past two years. This will also provide a substantial boost to our operating leverage moving forward.
Ashwin Shirvaikar, Analyst
Thanks, very clear. Thank you.
Jorg Friedemann, Investor Relations Officer
And our next question comes from Tito Labarta from Goldman Sachs. Could you please open his line? Thank you.
Tito Labarta, Analyst
Hi, good evening, and thank you for taking my question. I want to ask about the revenue side, specifically regarding the healthy ARPAC trends. Referring to the chart on page 11, your more mature cohorts are around the $20 mark, although it seems like they might be starting to peak. Do you believe there's still potential for growth in ARPAC for these mature cohorts? Additionally, looking at the chart, it appears that in about 30 months, many of your cohorts could be nearing the $20 level. Is it reasonable to think that in about three years, a significant number of your cohorts should be around $20 or possibly higher?
Guilherme Lago, CFO
Tito, we believe our mature cohorts have not reached their peak yet and still have significant potential for growth. The expansion of ARPAC is driven primarily by product cross-sell and up-sell. We are effectively increasing both up-sell and cross-sell each quarter, even within our more mature cohorts. The mature cohort shown on slide 11 highlights the penetration of Credit Card and NuConta, but the adoption of personal loans, investments, insurance, and a variety of other products we plan to introduce is far from where we anticipate it could be. There is considerable opportunity for cross-sell and up-sell ahead. For perspective, the ARPAC for traditional banks' retail operations is around $40. While we do not expect to reach that exact figure due to the absence of certain fees we won't charge, there remains a substantial gap for us to close from our current average of $7.8 to the $40 level of these incumbent banks.
Tito Labarta, Analyst
Great. That's very clear Lago. Thank you. And on the terms of the total cohorts. Is like another three years a reasonable assumption for when those begin to mature and get to the same level as your current mature cohorts?
Guilherme Lago, CFO
I think in general, we expect to see the maturation of our cohorts to happen slightly faster than that of our older cohorts, simply because if you are a customer of Nu today, you join an ecosystem with plenty of products that you didn't have if you had joined Nu five years ago. So the ability for you to consume more products is much higher and greater today than it was three four five years ago. And therefore, the speed with which the monetization happens over the next two years should be slightly faster than the speed with which we saw over the past three years.
Tito Labarta, Analyst
That's great. That's very clear, Lago. Thank you. Can you also discuss the primary banking accounts, which are around 55%, and whether the newer cohorts are achieving this milestone faster? Would there be additional growth as you introduce more products?
Guilherme Lago, CFO
I think the upside, the primary banking relationship we do expect that it will continue to go up. I personally believe, however, that the upside from product cross-sell and up-sell is even more pronounced than our getting marginally more primary banking relationship even though both trends should continue to grow and compound.
Jorg Friedemann, Investor Relations Officer
And our next question comes from the line of Neha Agarwala at HSBC. Thank you.
Neha Agarwala, Analyst
Hi. Congratulations on the results. Thank you for taking my question and for the additional disclosures; I find them very helpful. I noticed your chart showing the NPL metrics under the previous methodology, which indicates that the NPL ratio increased by 120 basis points quarter-on-quarter, with about 60 basis points attributed to product mix. This suggests some deterioration in asset quality. Is this one of the reasons for your more conservative approach regarding asset quality? If we consider the previous methodology, was the deterioration consistent with your expectations? Additionally, you mentioned that the change in methodology aligns more closely with what other global fintechs follow. How do other players in Brazil approach their personal loans methodology? What would be the average for the system if we were to compare your numbers to the overall system? What would be the best way to do that? Thank you.
Youssef Lahrech, President & COO
Thank you for your question, Neha. I'll address your inquiries one by one. Regarding the trend and quarter-on-quarter movement in the non-performing loans (NPLs), you're right that when looking at the 90-plus category under the previous write-off methodology mentioned in the appendix, there was a nominal increase of about 120 basis points, which translates to roughly 60 basis points in a consistent product mix. This change is largely expected. If we refer to the left part of that slide, we see that 15 to 90 went up by a similar amount last quarter, specifically 110 basis points nominally and about 80 basis points when adjusted for mix. Essentially, the current movement in the 90-plus category reflects the mechanical impact of what occurred a quarter ago in the 15 to 90 range, which is not surprising. You also inquired about the write-off methodology and how ours compares to other players in Brazil. I cannot specifically comment on the methodologies used by other financial institutions or fintechs in Brazil. However, I am aware that some players adhere to Brazilian GAAP, which is different from the IFRS 9 regime that we follow. Some of these institutions have 360-day write-off policies, but it’s important to consider that their portfolios typically consist of longer maturities and more secured loans, making it not a straightforward comparison.
Neha Agarwala, Analyst
Perfect. And if I can follow-up on the credit card receivables, I think the interest-bearing portion of the credit card is now higher as you showed in one of your slides. Should we expect that to continue rising further, or should we expect some stability there? And I think you mentioned previously that, this is mostly being incentivized it's driven by you and you are not seeing any worsening in terms of people not paying or more people being late in the credit card payments. Could you confirm the trend? Thank you.
Guilherme Lago, CFO
Thanks, Neha. Yes, I think you were referring to slide number 15 in which we showed the interest-earning portfolio of our credit card against the market. And yes, I think both of your assertions are correct. We have increased our interest-earning portfolio over the past four, five quarters mostly as a result of actions that we have taken and things that we have incentivized more specifically the launching of new financing products and new financing features. With our credit cards, you can basically finance specific purchases. You can finance boletos, which are called bank slips. You can now also finance fixed transactions. So we have been adding those features that allow our customers to finance a greater number of their transactions and hence foster the increase in IEP. Now, even by doing so, we are still quite behind the average of the Brazilian market. We started I believe a year ago being about 50% of the market. Now, we are around 60% of the market. We expect to continue going up. We don't guide and it's very hard for us to be more precise on if and when we're going to catch up with the market. But we do expect this to go up, over the course of the coming years; not necessarily coming quarters, but coming years. To your second question, which is, is the increase in IEP a sign of credit degradation? Our observation given the AB testing that we have done and we've done a few control groups is no, is that most of the increase in IEP comes from our launching new financing products not necessarily as a signal of overall credit degradation in our books.
Neha Agarwala, Analyst
Perfect. Thank you so much.
Jorg Friedemann, Investor Relations Officer
And our next question comes from the line of Eduardo Rosman from BTG Pactual.
Eduardo Rosman, Analyst
Hi, everyone. Can you hear me?
Guilherme Lago, CFO
Yes.
Eduardo Rosman, Analyst
Hi, guys. So congrats on the numbers. My question here is just trying to think about your potential future profitability, right? You showed two very interesting graphs in your presentation that show the potential or the expected ROE for your credit card and personal credit at around 80% and 100%. When we look to Itaú's results, right naturally Itaú is very different from Nu right. But Itaú's perceived as the premium retail bank in Brazil. They break it down the retail and the wholesale units. And the retail units used to deliver a 30% ROE in the past then that went down to 20% and now it's at 16% because it has been suffering from higher provisions. So just trying to think about here is, do you think trying to look to that metric it's a good one? Not as such the 16%, but trying to look how much let's say, a leading kind of a retail bank is delivering in retail is a good metric for us to follow. I know that, you mentioned that your ARPAC is going to be smaller than the big banks, but you're going to be more efficient as well. So just trying to think here about how you think about the business and that comparison. Thanks.
Guilherme Lago, CFO
Hey, Eduardo, thank you for the question. Look, it's very hard for us to provide you with a specific number or reference. However, I would discourage you from eventually anchoring the profitability of our business in the profitability of an incumbent bank. However, well run this incumbent bank may be, we just believe that we have been with digital banking redefining a new category in which we will have a completely different cost base of an incumbent bank, and also a completely different revenue stream of an incumbent bank to a large extent. Of course, we still believe that we will have consumer finance as our core. But if you take a look at our four cost pillars, cost to acquire, cost to serve, cost of risk, and cost of funding and we see the examples of other digital banks around the globe that have also started with consumer credit, they have been able to obtain levels of profitability and returns materially higher than those of incumbent banks in the markets in which they operate, even in markets that have substantially lower NIMs than Brazil. So I wouldn't necessarily compare our target profitability with those of a local incumbent bank. And of course, we will try to gain the game or win this game on the cost base. We believe that the winners in the long term in Latin America will be the lowest cost producers. And we want to be as David mentioned the lowest cost manufacturer in the banking industry in Latin America.
David Vélez, Founder & CEO
Hi, Eduardo, this is David. I want to add a point to Lago's response. We view our balance sheet and capital management as one focused on maintaining a small balance sheet that allocates its capital to high return on equity operations. When it comes to financial products that fall below a specific narrow threshold, we prefer to act as distributors rather than manufacturers. For instance, we have a partnership with Creditas to provide secured lending for auto and home purchases. In these cases, we choose to be a distributor to uphold high ROE on capital, which differs from many banks in Brazil that have a weighted average of diverse retail businesses with varying ROEs. It's essential to highlight that we aim to keep a capital-efficient strategy, being disciplined in capital allocation and allowing for both distribution and manufacturing roles.
Jorg Friedemann, Investor Relations Officer
And our last question comes from Jamie Friedman from SIG.
Jamie Friedman, Analyst
Hi. Thank you for the opportunity. Lago, in your response to Thiago's earlier question on slide 11, you suggested that the number of products per active customer is, in your view, more significant than the primary banking account anchor. I'm curious about why you see it that way. If you did not express that, I apologize. However, if you do believe that, could you share your reasoning?
Guilherme Lago, CFO
Jamie, what I meant is that if you look at the product usage among our customers, we see that over 60% have a credit card, more than 80% have a bank account, but only around 5% have personal loans. An even smaller percentage of our customers engage with crypto, marketplace, or insurance. Therefore, the opportunity for us to capture our fair market share across these product categories is likely to result in a significantly larger value creation than just improving our primary banking relationship from 55 to 65. That's the main point I wanted to share. However, there is indeed a strong relationship between being a primary banking customer and our ability to offer more products to them. I'm not suggesting that primary banking customers aren't an important metric. I was simply highlighting the substantial product opportunities we still have to pursue in the categories we currently operate in, as well as new categories that we plan to introduce in the upcoming quarters and years.
Jamie Friedman, Analyst
Okay, that makes sense. I understand now. Lago, in your prepared remarks, I think you mentioned that similar levels of origination lead to slower growth rates due to the need to replenish the amortized loans. Can you clarify if that's accurate and explain what you meant? Are you referring to growth in percentage terms?
Guilherme Lago, CFO
Yes, growth in percentage terms. And I would draw your attention to slide 35, which basically shows the evolution of our credit portfolio on an FX-neutral basis. And the underlying thesis there is that, our credit portfolio has a very low duration. The weighted average life of our personal loan business is about six to seven months. So, there is a volume that we need to originate every month just to replenish the personal loan that is being repaid or prepaid every month. That is what I meant when I made that remark. Thank you.
Operator, Operator
The Q&A section is now closed. I would like to turn the call over to Mr. Jorg Friedemann, at Nu Holdings.
Jorg Friedemann, Investor Relations Officer
I appreciate the participation of all of you. And if you had any questions that were not responded, the IR team will be following up with you off-line. Thank you once again, and hope to hear from all of you soon. Thank you.
Operator, Operator
The Nu Holdings conference call has now concluded. Thank you for attending today's presentation. You may now disconnect.