PicS N.V. (PICS) Earnings Call Transcript
August 24, 2026
Earnings Call Speaker Segments
Thanks, operator, and hello everyone. This marks the third earnings call we've held since going public, and I'm pleased to report yet another quarter of solid performance throughout our platform. Ahead of discussing the results, I'd like to offer a few remarks regarding the change in our CFO role. In early August, we shared that Andre Cazotto has taken over from Rodrigo Couto as Chief Financial Officer. A planned succession process led to this change, and I have full confidence in how strong and continuous our leadership team remains. During a pivotal stage of PicPay's development, Rodrigo was central — bolstering our finance organization, heading our Sarbanes-Oxley readiness efforts, and playing a key part in our successful IPO in January. He's been an outstanding partner, and I'm pleased that he'll keep collaborating with us as special adviser until year-end. Cazotto carries more than 20 years of experience across payments and financial services, and he joined PicPay in 2021, where he led the capital markets work stream for our NASDAQ listing, Investor Relations and M&A. His institutional knowledge runs deep, and he holds strong relationships with those who have a financial stake in our company. Cazotto, I'm certain you're the right fit for this position. Welcome aboard, and best of luck as we begin this new chapter side by side.
Thanks, Eduardo. It's an honor to take on this position at a moment this exciting for the company. Over the last few weeks, I've been working side by side with Rodrigo and our teams to make sure the handover goes smoothly. What really stands out to me is how solid our financial base is and how much discipline this business brings to how it runs. I'm eager to guide the next stage of PicPay's financial journey.
Thank you, Cazotto. Now let's turn to how we performed in the second quarter. The second quarter is something I feel great about. One image on this slide captures everything, showing substantial beats on essentially all metrics. Our credit portfolio hit BRL 31.9 billion, which is 3% beyond the top of our guidance. The cost of risk landed at 3.9%, sitting inside the range we had guided. Revenue came to BRL 3.7 billion, exceeding guidance by 3.6%, while net interest income stood at BRL 2 billion, 5.4% over our guided range; however, profitability is where the genuine narrative lies. Gross profit was BRL 1.25 billion, surpassing guidance by 8.4%, powered by operating leverage. Adjusted net income hit BRL 283 million, beating guidance by 15.5%, which mirrors robust top line momentum alongside ongoing cost discipline. That captures it. Across every dimension we met our commitments, with especially notable outperformance on the profitability measures that count the most. Let me begin with operating metrics, which continue scaling steadily. Total accounts climbed to BRL 70.4 million, a 10% rise year-over-year and 3% quarter-over-quarter. Quarterly active clients expanded to BRL 45.4 million, demonstrating persistent engagement throughout our base. Consolidated TPV totaled BRL 167.6 billion, 27% greater than the prior year and 7% up sequentially. Wallet and banking TPV amounted to BRL 142.6 billion, increasing 19% year-over-year and 6% quarter-over-quarter. Total cash in was BRL 136.4 billion, expanding 17% against a year ago and 9% sequentially. That works out on average to over BRL 45 billion each month. Deposits rose to BRL 35.8 billion, 45% higher year-over-year and 10% above last quarter; this strongly indicates growing trust and principality in our franchise; and active insurance policies totaled BRL 11.1 million, 63% beyond last year and 9% over Q1, as our insurance vertical keeps scaling quickly across all metrics, delivering consistent sequential growth on top of already robust comparables. Now to financials. This is where the monetization engine truly demonstrates its strength. Total revenues were BRL 4.1 billion, up 67% year-over-year and 17% above last quarter. That's a top line expanding rapidly. But allow me to point out what sits beneath it. If we exclude derivatives and hedge accounting, managerial revenues were BRL 3.7 billion, up 59% year-over-year and 17% sequentially; that acceleration comes from secured and partially secured credit origination, deeper card engagement and a richer fee-based product mix. ARPAC rose to BRL 92 per active client, 52% above where we stood a year ago and 14% over Q1. Excluding hedge accounting, ARPAC was BRL 83.3, which shows that even on a like-for-like basis, we monetize every client considerably more. Gross profit was BRL 1.25 billion, up 48% year-over-year and 14% higher sequentially. The distance between revenue growth of 67% and cost growth at a fraction of that is the operating leverage this model was designed for, and that leverage appears plainly in our unit economics. Cost to serve was 21.3% per active client, up 13% year-over-year, though just 5% sequentially. One thing to keep in mind is that this number contains BRL 0.70 for each client tied to opportunistic spending on marketing campaigns for seasonal events, which we pulled forward out of the third quarter. If you strip out that pull-forward, cost to serve comes to BRL 20.6, which works out to just a 1% increase quarter over quarter. Let me frame what that means. Revenue per client rose 67% year-over-year, whereas cost to serve was up only 13%; for each real we put into serving our clients, we bring in more than BRL 4 of revenue. That is the leverage built into this model. Adjusted earnings before taxes came to BRL 291 million, 174% higher year-over-year and 17% higher sequentially. That shows a business scaling efficiently and converting top line growth into bottom line results. Adjusted net income stood at BRL 283 million, up 135% year-over-year and 67% above the prior quarter. The sequential climb from BRL 169 million to BRL 283 million shows strong top line momentum, ongoing cost discipline and the favorable tax benefit from Brazil's Lei do Bem incentive program for technology companies. Let me pause on this slide, because it captures a deliberate structural change in PicPay's revenue mix. Total revenue of BRL 4.1 billion breaks down this way: 29% from unsecured credit, 24% from secured and partially secured products, 24% from fees and commissions and 23% from float and hedge accounting. The number that matters: 71% of our revenues now come from streams with no or lower credit risk — float, hedge accounting, fees, commissions and secured and partially secured credit. That compares with 63% only 12 months ago. I'll repeat that, we are expanding total revenue 67% year-over-year while at the same time building a business that is fundamentally more resilient. A more diversified revenue mix, together with a larger share of collateralized credit revenues, lets us balance growth across more mature collateralized portfolios while treating intentional risk as a lever, growing through small and progressive limits on cards, Buy Now Pay Later on loans and selectively moving into slightly higher risk clusters within private payroll loans, all while keeping the same risk appetite and targeted risk-adjusted returns. If you look at the 3 revenue engines separately across the last 5 quarters, secured credit revenues hit BRL 1 billion, up 158% year-over-year and 23% sequentially. The path from BRL 391 million to BRL 1 billion in 12 months tells the story of our payroll loan franchise arriving at meaningful scale. Unsecured credit revenues came in at BRL 1.2 billion, up 40% year-over-year and 11% above the prior quarter, growing at a strong, deliberate but measured pace. Noncredit revenues reached BRL 1.9 billion, up 57% year-over-year and 19% higher sequentially. This is fees, commissions, float, hedge accounting, insurance and acquiring, all capital-light, all compounding quarter after quarter. Three engines, three growth vectors and every one of them getting stronger. On returns, let me take you through the 2 charts on this slide. First, adjusted net income, BRL 283 million, up 135% year-over-year and 67% sequentially. This marks a significant acceleration in profitability as we scale the business. Next, adjusted ROE came in at 20.2%, compared with 15.5% a quarter earlier. Our R&D tax incentive program, Lei do Bem, gave both metrics a lift and helped drive the quarter's solid results. Turning to credit: PicPay card TPV hit BRL 19.5 billion, rising 40% versus a year earlier and 12% quarter over quarter. As our vintages mature, card usage keeps getting stronger, pushing spend per user higher. Consumer loan origination came to BRL 4.8 billion, 78% above the year-ago figure and 7% higher than the prior quarter. Our total credit portfolio stood at BRL 31.9 billion, 99% greater than a year ago and 14% up sequentially. The consumer book accounts for 93% of that total, while SMBs and others make up the other 7%. Within our audiences and ecosystem business unit, the portfolio we've assembled allows users to handle the bulk of their everyday needs inside PicPay. Having additional reasons to engage daily lifts engagement, opening doors to cross-selling financial products and raising customer lifetime value. Whether it's shopping and food delivery or travel, entertainment, telecom and urban mobility, we address the main journeys of daily life. iGaming stands out as one example. Within only 1 year, we assembled a high-margin operation serving more than 2.7 million clients through lucky numbers, national lotteries and themed World Cup games, all folded into our ecosystem. This wider everyday ecosystem makes us more relevant, deepens engagement and reinforces the financial relationship we hold with our customers; in our small and medium businesses segment, momentum is genuinely building everywhere. New small and medium business accounts hit 85,000 per month during the first half of 2026, versus 27,000 in the first half of 2025 — a threefold acceleration. Supply chain finance is expanding rapidly. Origination reached BRL 1.05 billion this quarter, compared with BRL 40 million in last year's final quarter and BRL 693 million only one quarter back. The direction is unmistakable and the unit economics look attractive; we're also introducing tap on phone to individual consumers, which turns 70 million PicPay users into prospective merchants. That's a distribution play placing us in a distinctive spot within the payments value chain; and we've just rolled out our marketing AI agents. SMBs can now build self-serve ads, and our platform pinpoints the most relevant customers inside the merchant's geographic footprint before delivering those ads. First-week results: 10,000 opt-ins, 1,500 campaigns and 1.7 million individuals reached — AI empowering SMBs to lift sales through our base of more than 70 million customers. Danilo, please share more about our consumer finance product highlights.
Thanks, Eduardo. I'm glad to provide an update on where we stand and what we're prioritizing. Our guiding principle hasn't changed: take great care of customers, create offerings that people find valuable, and expand in a disciplined manner. Our everyday banking operations keep advancing, which shows that customers trust us more and engage more deeply when it comes to payments, credit and routine perks — all backed by consistent execution, careful risk oversight and a strong customer experience. On our investment platform, customers can now access over 280 products, investment funds and fixed income among them; in addition, we introduced a brokerage service that lets customers trade stocks right inside our app; and we're phasing in the Epic segment for our existing customer base. This quarter, the offer achieved 23% penetration among the eligible base. Cards under the Epic brand represent 14% of total card TPV, and 80% of users are actively taking advantage of benefits like Amazon Prime, Einstein telemedicine and Sem Parar toll tags. In Brazil, convenience is what counts. Whether someone is settling a bill, consulting via telemedicine or driving through a toll, it needs to be fast and dependable. AI agents are increasingly at the heart of what we're doing as well. No other Brazilian bank has an official plug-in across both the Claude and OpenAI ecosystems — we're the first. Beyond that, we're introducing second-generation WhatsApp and in-app agents equipped with additional tools, memory, Internet access and the ability to execute sequential multistep tasks. This strengthens our appless approach, resolving fragmented journeys wherever users need us, with products and services that fit the context and matter to them. Now for credit; we keep capturing market share by growing our share of wallet among the products our customers use. In private payroll loans we hit 6.4%, in personal loans 2.8%, in card TPV 1.6% and in the credit card portfolio 1.2%. We remain convinced there's plenty of room for expansion. Onto portfolio growth. During the second quarter, our credit portfolio expanded by BRL 3.9 billion. Lower-risk loans and mature credit cards generated 86% of that expansion. The contribution new cards made to portfolio growth nearly doubled versus last quarter, which mirrors our progressive limits approach and the maturing of more recent card cohorts. Now for underwriting strategy and cohort performance. Across two complementary goals — optimizing performance and optimizing growth — we keep carrying out our underwriting strategy. As cohorts mature, progressive limits make up a bigger portion of the portfolio. Across both strategies, NPL creation within the credit card portfolio is performing better than during the equivalent period a year ago and stays fairly steady relative to recent quarters, even once seasonality is factored in. Cohort performance under both strategies has been fairly steady over recent quarters, which speaks to the durability of our models and our hands-on risk management. When it comes to private payroll loans, we started raising originations again in growth clusters, having regained confidence in how operationally mature the product is and having added new features beginning in the fourth quarter of 2025. That's pushing the growth strategy mix higher. More recent cohorts show the intentional, incremental risk we took on to speed up growth, all while staying inside our approved risk appetite and targeted risk-adjusted returns. Even though we aren't observing meaningful warning signs of credit quality weakening inside those same risk buckets, over the next several quarters we anticipate that portfolio metrics will show the effects of deliberate additional risk appetite in private payroll loans, along with aging and maturation of cohorts. Among those metrics are 90-plus NPL Stage 3 and cost of risk measured against the entire portfolio. Once newly originated loans represent a reduced portion of the total outstanding book, the dilution impact they have on such measures will naturally shrink. Cazotto is going to give more granularity on how these dynamics work during the following session. Let's now take a closer look at what's happening in our private payroll loans business. This quarter the portfolio hit BRL 7.2 billion, comprising over 3.6 million contracts, with employer risk that is well diversified. Marginal ROEs we expect remain appealing, underpinned by pricing adjusted for risk and revenues tied to credit. On top of that, we're observing improved ARPAC and cross-selling metrics among these customers, which bolsters other revenue streams; our confidence stays intact regarding our capacity to grow this operation while generating sound ROEs and risk-adjusted returns; at this point I'll hand things over to Andre Cazotto, our CFO, so he can walk through our financial results.
Thanks, Danilo. Now I'll take you through the way our delinquency numbers have moved and explain what's behind those curves. The chart on the left shows early NPL, our term for loans running 15 to 90 days past due. It was 8.4% in the first quarter, then fell to 7.5% in the second quarter. Two things explain that quarter-over-quarter drop: a helpful seasonal trend during those months, plus stronger results from newer vintages. Over on the right, NPL over 90 days increased to 9.8% for the quarter, and Stage 3 reached 12.9%. You should look at those 2 metrics together. Everything in NPL over 90 days also falls within Stage 3. That means the loans behind that metric have already been classified as credit impaired and provisioned for accordingly. Stage 3 represents the wider classification, since it also includes other credit impaired exposures that might not yet be more than 90 days past due, but have already been flagged as deteriorated. Put differently, no further credit risk exists outside Stage 3. Everything has already been recognized and provisioned within that bucket. What is pushing these later-stage metrics higher is mainly portfolio aging. As our products and vintages season, a bigger share of the book naturally moves into later stages of delinquency. That is a mechanical and expected pattern in a fast-growing portfolio, rather than evidence of deterioration. Also worth mentioning, the way these numbers evolve is going to reflect a choice we've made on purpose: we're taking intentional risk in private payroll loans. That's a deliberate portfolio decision, and we're fine with seeing higher delinquencies because we believe we'll get substantially stronger returns once you adjust for risk across the product's lifetime. As the portfolio continues to season and the strategy reaches maturity, our expectation is that NPL over 90 days and Stage 3 will both move gradually toward a level that's more stable. One more thing that deserves emphasis: over 75% of the Stage 3 portfolio has already been provisioned, which shows solid coverage relative to expected losses and supports the conclusion that our provisioning framework is adequate. Let's turn to the next page. On this slide, we're looking at what drove the quarter-over-quarter changes in both NPL over 90 days and Stage 3 as we went from Q1 '26 to Q2 '26. Let's take NPL over 90 days first: it went from 80.9% (sic) 8.9% to 9.8%, which works out to a net rise of 93 basis points. Portfolio aging was the biggest factor, adding 318 basis points as older vintages naturally seasoned and rolled into later delinquency stages. The Desenrola program partly counteracted that, delivering a 117 basis point improvement. Seasonality chipped in 50 basis points, in line with what we typically see during this period. Another point to keep in mind: because we originated fewer new loans than in earlier years, the dilution impact on this metric was less pronounced. In other words, the denominator expanded at a slower rate, which helped push the ratio higher. Product mix, along with other items, provided modest offsets — 34 basis points and 7 basis points, respectively. Turning to Stage 3, the figure moved from 12.7% to 12.9%, which works out to a net increase of only 27 basis points. The same basic forces are at work here, but one notable distinction deserves attention. Aging contributed 184 basis points — substantially less than the 318 basis points recorded for NPL over 90 — and that is entirely by design. Stage 3 is a pre-NPL measure, so it picks up credit deterioration at an earlier point in the cycle. Because of that, the aging effect still pressuring NPL over 90 days was already partly absorbed within Stage 3 during earlier quarters, which leaves a smaller incremental aging impact. Seasonality contributed 41 basis points, and the Desenrola program subtracted 46 basis points. Origination subtracted 117 basis points, while product mix and others delivered further offsets of 26 and 9 basis points, respectively. All together, these waterfall charts confirm what we told you earlier. The trends we're observing with NPLs are largely a function of how our vintages are maturing and normal seasonal patterns, which reflect a choice we made on purpose to accept risk within private payroll lending — it isn't a sign that the portfolio's fundamental quality has deteriorated. Moving to the following slide; looking at the left side, combined Stage 2 and Stage 3 formation has continued its positive trajectory, easing to 4.9% during Q2 compared with 5.1% in each of the prior two quarters. Over on the right side, Stage 3 formation came in at 3.65% for the second quarter, a decline from 3.9% in the preceding quarter. The Desenrola renegotiation program's effect accounts for the bulk of this improvement. Most of the loans that were reworked under this program remained on our books and were less than 360 days overdue. On a gross basis, the renegotiated exposure totaled about BRL 520 million. Thanks to an average discount of roughly 50%, the outstanding balance dropped by around BRL 260 million; that reduction immediately lowered the balance that feeds into Stage 3 formation, which is why the ratio improved to 3.65%. Had Desenrola's impact been excluded, Stage 3 formation would have come in at approximately 4%, largely in line with earlier quarters. That underlying figure also reflects how the portfolio naturally ages as our products and vintages continue to mature. Let me now walk you through how the portfolio breaks down by stage, along with the coverage levels we're seeing. In the second quarter, Stage 3 remained at about 13% of the total credit portfolio. As for coverage, the levels still look comfortable to us, with Stage 2 plus Stage 3 coverage coming in at 62.7%, while Stage 3 coverage stood at 74.1%. Stage 3 coverage dropped to 74.1% in the second quarter from 77% in the first quarter. That decrease was mainly driven by the Desenrola renegotiation program. Loans that were renegotiated through Desenrola come with an FGO guarantee, and the operations guarantee fund covers 50% of the outstanding exposure. Because of this guarantee, expected recovery on those loans goes up, which in turn brings down the LGD applied to those exposures. Because a lower LGD translates into lower provisional requirements, including these loans mechanically pulled down the overall Stage 3 coverage ratio. So the drop from 77% down to 74% isn't a sign that portfolio quality got worse, nor does it come from us changing provisioning standards or shifting our risk appetite. What's really happening is mostly a mix effect: the Desenrola portfolio carries a lower LGD, and that's backed by the FGO guarantee. Once that effect settles down, we'd anticipate Stage 3 coverage returning to the high 70% range over the next few quarters. On the credit risk management side, three metrics matter most to us: loss absorption, cost of risk, and portfolio coverage. Taken together, they paint a picture of a book that is tightly managed and whose reserves are steadily building. In the second quarter, our loss absorption ratio came in at 56.5%, a level that rests well within the 40% to 60% band we set for ourselves. As for cost of risk for the quarter, it came in at 3.9% in Q2. That matches the guidance range of 3.7% to 3.9% that we provided when the quarter began. Compared with the first quarter's 3.7%, the sequential rise was driven primarily by our private payroll loan book getting older on its own, since loans originated in earlier periods continue to season and progress through the provisional cycle — something that happens mechanically and was anticipated, given the pace at which this product has expanded across the last few quarters.
That gain got partly counterbalanced by a BRL 59 million favorable effect from Desenrola, amounting to roughly 5% of what we spent on credit overall during the quarter. Now, turning to credit loss allowance expenses and total coverage: CLA expenses came in at BRL 1.2 billion for the second quarter, versus BRL 974 million in the first quarter, which lines up with how quickly the portfolio is expanding. Even more significant, total portfolio coverage stayed put at 13.9%, the same as the previous quarter, which underscores that our provisioning levels remain adequate as the book keeps growing. Stable coverage alongside rising absolute provision balances shows a disciplined, consistent way of managing credit risk. Let's go to the next slide, on operating leverage. The trend really says it all. Net revenues hit BRL 4.1 billion in the second quarter, rising 17% from the prior quarter and 67% from a year earlier, when we posted BRL 2.5 billion in the second quarter of last year. During that same stretch, adjusted operating expenses — which leave out stock-based compensation expenses — climbed to BRL 955 million, a small fraction of how fast revenue grew. What that produces is ongoing, steady gains in our adjusted efficiency ratio, which fell to 44.8% in the second quarter from 46.9% in Q1. That's a 210 basis point improvement sequentially. AI is a major force behind this. We can already see and measure what it's doing to our operations. Headcount hasn't moved since October 2025, and the 10% growth we'd initially expected for 2026 is not going to happen. Productivity gains are flowing straight into margin expansion instead of extra hiring. Going forward, we see AI as a big and accelerating source of operational leverage, which makes the efficiency path on this slide a floor rather than a ceiling. Now on to financial margin expansion. Net interest income came to BRL 2 billion, 18% higher than the previous quarter and 65% higher than a year ago. Our net interest margin was 19.4%, up from the 18.7% we reported in the first quarter. Margin from credit products totaled BRL 2.1 billion, expanding 18% sequentially and 81% year-over-year; this measure reflects the complete economic contribution of our credit operations, counting revenues from products tied directly to credit origination, like credit card interchange and credit insurance, while leaving out cash remuneration and derivative revenues, so it gives a clearer picture of the actual margin our lending activity generates. Net interest margin from credit products landed at 27.8%, up from the 27.2% reported in the first quarter. Just as important is the margin from credit products after losses, which hit BRL 980 million in the second quarter, 14% above the prior quarter and 68% above a year earlier. The net interest margin after losses remained steady at 12.1%. Let's turn to funding on the next slide. Our funding base expanded 10% quarter-over-quarter, reaching BRL 35.8 billion in the second quarter, 45% higher year-over-year than the BRL 24.8 billion recorded in the second quarter of '25. The slight quarter-over-quarter rise in funding cost, moving from 94% to 96.2% of CDI, is mostly attributable to our new FIDC issued in May 2026 — a securitized vehicle collateralized by our FGTS portfolio, through which we brought in BRL 1.2 billion. Then, in July and August, we completed further capital markets deals, bringing in resources via promissory notes and debt security issuances, in line with our approach of steadily broadening where our funding comes from. Those deals gave our balance sheet an extra boost and sharpen our ability to keep up the fast expansion of our credit portfolio in a disciplined, cost-efficient way. We'll keep tapping a range of funding channels — digital platform deposits, third-party platforms, creditory funds and capital market instruments — actively pursuing the most efficient funding options out there to back our growth goals; on the capital front, our position stays solid, with a total capital ratio of 17.6% and a common equity Tier 1 ratio of 15.6% for the second quarter; worth noting is that roughly BRL 450 million, equal to about 1.7 percentage points of both our total capital and common equity Tier 1 ratio, is still parked at our holding company in the Netherlands and hasn't been pushed down into the operating entity. Now that the Kovr acquisition has closed, we foresee capital consumption of about 150 basis points in Q3. Even once that hit is absorbed, we're still comfortably clear of our internal capital appetite thresholds, and we anticipate finishing the year with a total capital ratio near 14% and a common equity Tier 1 ratio between 12% and 12.5%, levels giving us meaningful headroom over regulatory requirements and fully backing our growth ambitions. Last, on the following slide, we're introducing guidance for the third quarter of 2026. Just like our earlier guidance, the numbers here are for PicPay stand-alone operations and leave out any contribution from Kovr. We foresee our total credit portfolio hitting roughly BRL 34.7 billion. Quarterly cost of risk should stay in the 3.9% to 4.1% band. Turning to revenue, managerial revenues should come in at around BRL 4 billion, while net interest income ought to hit approximately BRL 2.1 billion. Gross profit is guided at about BRL 1.3 billion. On profitability, we anticipate robust pretax earnings growth. IFRS earnings before taxes is guided at roughly BRL 360 million, 34% higher quarter over quarter, and adjusted EBT at about BRL 378 million, up 30% versus the second quarter 2026. At the net income line, though, some context on the sequential dynamics matters. IFRS net income should land at approximately BRL 255 million, 5% lower quarter over quarter, and adjusted net income at roughly BRL 265 million, 6% under the second quarter. That drop isn't the result of any operational weakening — quite the reverse. During the second quarter, we enjoyed a sizable positive effect from the Lei do Bem, a recurring tax incentive that this year was heavily concentrated in the second quarter, sharply lowering our effective tax rate for that period. During the third quarter, our effective tax rate returned to a normal range, matching what we saw in Q1 of this year. So with that said, I'll turn things over now to Eduardo Chedid so he can share some final thoughts.
Appreciate that, Cazotto. Ahead of my final thoughts, there's a milestone I'd like to call out that stands on its own merits. Once we'd secured the green light from the insurance regulator, the antitrust authority and the Central Bank, the deal to acquire Kovr closed on August 3. This goes beyond a simple M&A deal. It represents a strategic step-up in our insurance ambitions. Kovr comes with a full-service insurtech platform offering more than 100 products, a senior executive team boasting over 20 years of proven insurance experience, and established distribution channels that dovetail with ours. The economics look compelling, and from August to December 2026 we anticipate a meaningful incremental boost to PicPay's bottom line; yet what truly fires me up even more is the strategic fit and the opportunities lying ahead in the years to come. For those insurance products Kovr sells via our channels, we'll now capture the complete economics and gain the ability to build more tailored products for our client base. On top of that, roughly 70% of Kovr's business comes through high-quality partner distributors, and we anticipate that channel will continue to perform. Thanks to Kovr, we now possess the product development speed, the underwriting expertise and the distribution reach needed to transform insurance into an even more substantial recurring earnings stream. We're keeping Kovr independent and reinforcing its partnerships; this marks merely the start of a new phase, and it's already signaled by a brand change. Kovr is now KEV. Allow me to close with 6 points that capture where we stand. First, the macro outlook. Although delinquency is still elevated, recent trends indicate stabilization. The economic backdrop keeps providing meaningful support for credit quality. The labor market stays highly resilient, with unemployment close to historic lows and over 103 million people holding jobs. Real wage income hit roughly BRL 380 billion as of June, rising 3.6% year-over-year. Notably, net formal job creation clustered in income brackets earning up to 2x the minimum wage, with more than 160,000 new positions added in that segment. Though economic activity is gradually decelerating as anticipated under contractionary monetary policy, growth remains in positive territory. Our scenario doesn't envision a sudden employment deterioration, but rather a progressive normalization with the labor market at historically strong levels; this mix of elevated employment, resilient income and moderate expansion lowers the risk of a systemic deterioration in households' repayment capacity and puts PicPay in a strong position for the quarters ahead. Second, asset quality. Our portfolio stays resilient by design, underpinned by greater exposure to secured and partially secured products, disciplined underwriting and robust risk management in line with our credit fundamentals of a balanced portfolio, loss absorption ratios between 40% and 60% and ROEs above 30%. The rise in NPL over 90 days mirrors portfolio aging and intentional risk taken in payrolls, not deterioration. Early delinquency improved to 7.5%. Kovr's ratios are robust and the underlying quality of our origination remains strong. All of this while keeping the same risk appetite and targeted risk-adjusted returns. Third, private payroll loans. This product is scaling with attractive economics. Since launching, we've surpassed 3.6 million contracts, with marginal ROEs that remain quite healthy, and over-30-days NPL metrics that are steady across both the standard portfolio and the growth portfolio. That's what's underpinning profitable expansion in partially secured lending. Number four, noncredit revenue. It grew 57% versus a year ago, which highlights how strongly our broader platform monetization is performing outside of credit-related revenue streams. Number five, small and medium businesses. Scale, relevance and customer traction are all building in this segment, and its potential to add meaningfully to future growth keeps rising. And number six, Kovr. Our insurance ambitions get a boost from this acquisition, opening up chances to build products and deepen penetration, capture extra economics inside our customer base and also via distribution partners. A meaningful, recurring contribution to earnings growth should come out of it. Last, we exceeded guidance across every major metric in this quarter. Our confidence is high that we can produce sustainable, profitable growth and build long-term value for shareholders. Urgency will guide our actions, yet quality and trust will never be sacrificed for it. Thanks — and at this point we'll open the line for questions. Operator?
Cazotto, nice work on the new position. There are 2 things I'd like to ask. The first concerns Desenrola. You were pretty clear, correct, about the 117 basis point gain for NPL and roughly a 5% drop in credit cost. That works out to approximately BRL 59 million. Did Desenrola bring any other advantages? And from what I gather, the program got extended, right? So in the third quarter, should we anticipate more gains coming from that renegotiation initiative? I'll save my second question for later.
Okay. Thank you, Mario. Thanks for asking. I'll attempt to restate what we already conveyed during our conference call. So yes, regarding cost of risk, Desenrola produced a favorable effect of roughly BRL 59 million, which corresponds to about 5% of our overall cost of credit for the quarter. On the NPL above 90 days, the program brought the ratio down by roughly 117 basis points, partly counterbalancing the effect of portfolio aging and seasonality. Desenrola additionally assisted with stage 3 formation. As we stated during the conference call, approximately BRL 520 million in loans underwent renegotiation on a gross basis, with an average discount of 50% applied, essentially shrinking the outstanding balance by about BRL 260 million, which directly lowered the balance feeding into Stage 3 formation. Consequently, the ratio fell from 3.9% to 3.65%. If you set aside the "Desenrola" effect on stage 3 formation, let's say the ratio would land near 4%, more or less consistent with earlier quarters. For Q3, yes, we do anticipate some extra favorable effect from Desenrola, though it will be smaller. I believe the second quarter gave us a considerably larger impact.
All right, I believe that's clear. So, one thing on the results that caught us off guard was the funding cost. It landed somewhat above what our model assumed; and if we look at your deposits, right, those are expanding more slowly than your loans; and you mentioned, right, that you're tapping other funding sources. Could you just elaborate a bit on the capacity to keep growing deposits from the clients you already have?
Mario, I appreciate the question. Yes, I'd say we feel quite at ease with the amount of deposits we're bringing in on our platform. We're broadening what we can do. As we mentioned during the conference call, the modest rise in our cost of funding, from 94% to 96.2%. That's mainly a result of issuing our FIDC, which is essentially backed by our FGTS portfolio; we're also reaching additional funding avenues in capital markets; we're boosting, shall we say, the principality of our customer base; so our platform is becoming increasingly transactional. We keep growing the cash at roughly 20% year-over-year. So increasingly, we are turning more cash into deposits. So we feel quite at ease and will keep expanding our deposit franchise down the road. We expect to hold, shall we say, the funding cost at about 95% of CDI over the next quarters; so we do believe we can keep growing, access new funding lines, and still produce a very healthy cost of funding across our operations.
Fantastic outcomes. Congratulations, Cazotto, on taking on the new position—I'm really eager to see what you accomplish there. One thing I'd like to ask about is AI. You presented quite a few really compelling offerings on the agentic AI front. Could you perhaps, Cazotto or Eduardo, share some thoughts regarding those offerings? We thought they were quite fascinating.
Dan, Eduardo here. I'm able to weigh in, though I believe I'll hand that over to Danilo, since he's the one actually leading our AI efforts, and he can provide you with a more, shall we say, thorough grasp of every aspect we're working through.
Okay, so to start with, our approach to AI actually rests on 2 separate pillars. Customer-facing products make up the first pillar. The second one, which I'll touch on briefly, relates to our operational AI capabilities. The agents we brought up during the presentation were tied mostly to that first pillar, you know, the customer-facing products. Those are the ones where our aim is really to give our customers power wherever they happen to need us. So we talked about the consumer agents, which are now in our second generation, and they can do more than just respond to questions — they can actually carry things out, such as paying bills, Pix transactions, managing savings, renegotiations, and so forth. Naturally, all of that happens with the user confirming every transaction. But we're now putting out more than 70 tools to our customers, and the agent can actually carry out several sequential tasks on the customer's behalf, right? On top of that, it now works across multiple channels too. Not just inside our app, but on WhatsApp as well. And like we said, we became the first Brazilian bank to have our own plugin available on both Anthropic and OpenAI's official plugin store — so that covers the consumer side. We also brought up the small and medium business agent, the marketing agent, which can actually build and send out ad campaigns for our small and medium business clients, aimed at our customer base, using the geolocation of the business. That one just launched, and we shared a few of the early big results. But we also have agentic agents supporting our internal operational capabilities, right? So starting at the beginning of the year, what we did was build our own proprietary platform, our AI harness, around what we see as the key to pulling value out of AI agents. So we created a platform with model routing, caching, and plenty of governance layers. Thanks to that, we managed to cut our token costs by 70% between the beginning of the year and now. And that lets us keep access to the top frontier models without total token costs growing, because of that; and we're applying it in different areas; so we're using it in credit; we're just now rolling out our proprietary foundation model for personal loans underwriting. That model should deliver something around 15% to 20% improvement over the previous one. We also have agents and people leveraging our internal platform for product development. So AI helps with coding, designing, quality assurance. These days, roughly something around 90% of our employees are actually on our AI platform, with most of them on it daily. Since the beginning of the year, we've had 30% of employees contributing with deployments, real deployments for PicPay products. And most of them, myself included, couldn't contribute without AI — we aren't actually coders. Yet that has now become achievable. As a result, an increasing number of individuals are pitching in with actual offerings. Plus, deployments at the start of the year were twice as many, thanks to those productivity gains. We're also accomplishing some genuinely strong work in-house, and we expect to gain from it, along with making the most of our operational efficiency.
Fantastic, that's really impressive, and once more, congratulations on such strong results.
Nice work on yet another quarter of steady results, and congrats to you, Cazotto, on the new position. My first question is about the private payroll loan. This product has kept expanding very quickly. Although it's viewed as a secured offering, we've noticed delinquency patterns getting worse, and among certain cohorts, the NPL ratio is beginning to approach the level seen in unsecured personal loans; that's based on Central Bank data, correct? So I'm asking, how do you weigh the risk-reward tradeoff for this private loan right now? And might you become more selective or tweak your risk appetite for this product down the road? Also, why do you believe the consolidated Central Bank data shows quicker deterioration for this product? That's my first question, and I'll ask my second one afterward.
Gustavo, this is Eduardo Chedid. To begin with, I believe we haven't observed any worsening at the same risk profiles. What we've experienced, and you can see it in the presentation, is that we are deliberately moving into higher-risk profiles, which means on average you'll notice NPLs rising, yet that isn't deterioration within the same risk profiles. Looking ahead, and this is, I'd say, the philosophy we've applied across all products. Any benefit we're gaining from our new models, we're not, shall we say, putting that toward additional growth, but essentially keeping origination flat while capturing the model improvements, so that asset quality stays, let's say, under control. If you examine the Central Bank data, I think it also mirrors the fact that, looking at the earlier product, it only, let's say, served very large companies. Now this product, and that's largely because of the new approach, is being extended by people to smaller companies too, and that's an advantage of the centralized system. So as you bring in more companies and more of their employees, let's say, from smaller firms, comparing the old product with what's happening now is quite difficult.
Sorry, Gustavo — let me just add something here. When we talk about this intentional approach of embracing incremental risk within very particular, carefully chosen customer segments, it's crucial to point out that our framework for risk hasn't changed at all. We're still aiming for a loss absorption ratio in the range of 40% to 60%, along with ROEs that exceed 30%. That point really deserves emphasis.
All right. I just have one more question, actually 2 more, about this private payroll loan. Have the operational problems we witnessed a few months back gotten any better? And if you could, I mean, could you tell us what cost of risk level this product carries?
Gustavo, regarding the operational problems, I would say we passed through roughly 3 phases, correct? So the initial phase, when the operational problems were enormous. At the start, FPDs were coming in at about 17%; after that we went through a cycle, a second momentum, essentially in which we cut originations back drastically so that we could watch those operational problems get resolved. A portion of them got resolved through the centralized system. A portion of them got resolved through workarounds that we ourselves put in place. Simultaneously, we also — I believe we are now on the third or fourth distinct concept — fourth evolution of the concession model, which likewise assisted us in essentially reaching that third phase, namely growing the product. Sure, you're bringing up any, shall we say, significant gains between last quarter and this one, I would not claim that. And when you examine guarantees along with the automatic, shall we say, payroll linkage, we are still not underwriting as though they carried real weight; which therefore means we continue to view those as insufficiently meaningful for us to factor them in.
When it comes to the cost of risk, it's essentially consistent with other peers, shall we say, that have recently reported this figure. Therefore, we can state that it's approximately mid- to high teens on an annual basis.
Still on the topic of asset quality, let me wrap this up. What trajectory do you foresee for NPLs stage — sorry, for 90-day NPLs and stage 3 — across the coming quarters? Should we anticipate additional normalization as the portfolio seasons? Or do you think the present levels already broadly reflect the underlying credit performance?
Let's put it this way: we continue to expect the aging effect to keep weighing on NPLs, correct? So by year-end, we anticipate NPLs over 90 days will sit closer to, say, the low teens, essentially converging toward something comparable to what our stage 3 over total credit portfolio shows. Keep in mind that stage 3 is a pre-NPL metric and it's pretty much absorbing, shall we say, every credit impaired item we have in the model. So essentially, we think this could convert to a level similar to what we currently have on stage 3 over the total credit portfolio by the end of this year. That said, we are not observing deterioration. It's essentially portfolio aging, the growth we have on the private payroll loan that is still maturing; we were, let's say, early adopters of this product, likely the second company prepared to operate private payroll loans in Brazil. So naturally, that portfolio keeps aging and impacting this metric going forward.
I've got just one more question on private payroll. Given the growing worries about the macro backdrop, rising unemployment has turned into a significant risk we've been talking over with certain investors regarding private payroll lending, especially since it carries a longer duration. So might you weigh in on what unemployment level would start to make you worry more about this product's profitability? Where would that level sit at which profitability would come nearer to breakeven, in your assessment, depending on an unemployment increase?
All right. I appreciate the question. My answer will be, I'd say, not the one you anticipate, though I'll aim for the same conclusion, correct? So when you examine how we approach credit, it rests mainly on the loss absorption indicator, correct? Thus for private payroll loans, for those vintages to reach breakeven, we could endure a rise of as much as 70% in the delinquency rate on these products; that means my expected losses might genuinely increase 70%, and I'd still be sitting at a breakeven position; now regarding unemployment, and looking forward, what market consensus and Focus projections show at present is that they foresee unemployment staying fairly steady and rather healthy, essentially climbing from 5.4% to roughly 6% across 2027, which somewhat strengthens our belief in a structural floor beneath household income instead of any abrupt deterioration in the labor market. Even supposing we consider the gloomiest scenarios in the Focus Survey, unemployment would top out at levels similar to what we witnessed during 2024, somewhere between 6.5% and 7.2%, which means that even in stress scenarios, we're discussing historical levels that did not signify a severe hit to credit or to households or to household repayment. Naturally, we continuously and dynamically review those projections. And as I mentioned to you, we are right now directing gains from our concession model more toward maintaining origination levels than toward expanding originations. So that's my take on it.
Got it, that's very clear. And could I ask one more thing — what do you anticipate for the bottom line in 2026, given that Kovr is going to be consolidated now? It would really help us if you could give some idea of how much Kovr might ultimately add during the second half of the year.
All right. Let's turn to the Kovr acquisition. Kovr is now known as KEV. Consolidation of KEV begins August 3. We expect a net income contribution somewhere in the range of BRL 80 million to BRL 100 million for the August through December period — that's essentially what we can tell you about Kovr for those 5 months left in the year. And we did also provide third quarter guidance, so that's about all we're able to share at this point.
I just wanted to — I had a quick follow-up regarding ARPAC. It's still extremely robust once again, and your monetization rate keeps getting better. Could you maybe walk back through the approach, meaning the strategy going forward and perhaps how some of that aligns with the product road map and the mix shifts you're observing in the business. Obviously, things are heading the right way, but I just want to be certain I grasp the cadence for how that advances from here.
So I'd say it's more of — roughly the same narrative going forward. It's still largely propelled by greater penetration of our products and chiefly that rather than new clients, but very heavily focused on cross-selling those products and primarily, say, credit and insurance products into our user base; that's what's mostly fueling ARPAC growth. Meanwhile, you can observe that our cost to serve is expanding at a far slower pace. ARPAC is growing at 52%, whereas cost to serve is growing at a 13% rate year-over-year. And the bulk of that cost to serve growth comes mainly from adoption of new products. So I'm just attempting to give you additional proof points that this is what's truly driving all of that ARPAC growth. If you examine more mature cohorts, you'll also notice ARPAC more than doubling when compared to the average ARPAC, which only strengthens the thesis, namely that cross-selling more of those products, particularly credit products, will be the key driver for further raising ARPAC ahead.
Great; and then just very quickly regarding Kovr. I heard what you said about what it added from August through December when it comes to net income. Is there also a revenue figure that you're tying to that, so we can be sure we're setting the right baseline in the model?
Perhaps we can provide that to you and to everyone else. We're not — that number isn't available at this moment, but we can pass it along afterward.
I've got 3 questions, and they're quick ones. The first concerns operating expenses. In 2Q there was something of a jump. I think a few extraordinaries were in there, plus some additional marketing spend you took on during 2Q, and your guidance suggests a sequential decrease in 3Q. Could you give us a little more color on the OpEx growth trend we ought to expect moving forward? And what exactly were those one-offs in 2Q? Question two is about risk-adjusted margins. On the reported figures, they fell 20 basis points. Yet once you adjust for the Desenrola benefit, the risk-adjusted margin is probably near 11.3%. What path should we, what path should we assume over the next several quarters? And at what level should that risk-adjusted margin settle for you? Question three concerns the write-off policy. Could you remind us what your write-off policies are? And has anything changed on that front recently?
So Neha, thank you for your question. I'll begin with the last one. Our write-off policies are unchanged, 360 days for credit cards as well as personal loans; regarding the risk-adjusted NIM, we anticipate stabilization in Q3 relative to what we delivered in the second quarter, near 12.1%. As for efficiency, we had the anticipation of BRL 30 million in marketing expenses during this quarter. We chose to anticipate owing to the World Cup, we made a decision to speed up certain initiatives in marketing for very specific products such as the iGaming platform that we have, along with other initiatives where we spotted an opportunity to accelerate; so we ought to expect some, let's say, improved benefit from reduced marketing expenses versus the second quarter in Q3; regarding overall efficiency, we're on a very healthy trend. As you can see in the quarter, our efficiency ratio came in at roughly 44%, declining more than 200 basis points sequentially; we expect that trend to keep going forward. We're seeing AI speed up our operating leverage opportunities. Headcount has been pretty much flat since October 2025. If you recall, we expected to grow headcount by roughly 10% this year. That isn't happening thanks to AI and all the initiatives we have. So we think we can deliver our efficiency ratio near low 40s, high 30s by the end of this year, taking into account many, let's say, initiatives we have, including AI opportunities on the, let's say, personnel expenses, but on tech expenses too.
Great. Then one more thing I wanted to check. Regarding the risk-adjusted margins, you said we should anticipate roughly 12.1%; so when I take out the Desenrola gain, my math gives about 11.3%. So your expectation is a bounce back in 3Q and that it holds near the 12% level. Is that accurate?
That's right. For Q3, we anticipate risk-adjusted NIMs to come in essentially flat on a sequential basis, more or less. Desenrola does affect us somewhat in Q3 too. However, as we mentioned, that effect is somewhat smaller than what we saw in the second quarter. Still, yes, we look for this ratio to land around 12%, 12.1%, basically consistent with the prior quarter.
So what's going to be the catalyst behind that? Given the guidance you've provided, the cost of risk is going to keep creeping up each quarter. So what would it be — and deposit costs are likely to stay at roughly the same level, with not a lot of improvement going by what you said earlier — so what's going to be the thing that drives the better or steady NIMs, correct, if we leave out the Desenrola effect?
Right. It's essentially a mix effect. As we mentioned, we're expanding a bit less in more, shall we say, segments where we're comfortable when it comes to taking on incremental risk; so we do have this risk-adjusted strategy in the business. Funding costs ought to be a bit lower than in this quarter. So we came in at 96.2% of CDI during the quarter. We expect Q3 to land more around 94%, 95%; so that's one more, shall we say, gain that we can observe on this risk-adjusted NIM.
Once more, well done, Andre. I'd like to follow up regarding the expansion you're observing in payroll loans — what attach rates are you witnessing for other offerings after those loans are made, such as credit cards and other products?
Eduardo speaking. To begin with, consider our, shall we say, typical ARPAC for customers who hold that product: it stands at 8.9x what we see for the typical PicPay customer. Two things drive that — the product itself, and also, as you mentioned, the attachments, which he essentially receives. Then, when it comes to cross-selling, the figure runs 30% above what we record for, on average, our customers. That spans a wide range of products, yet on average it's 30% above our typical customer. Hence it likewise, shall we say, lifts not just ARPAC, but the take-up of additional products too.
Alright everyone, I appreciate you joining us once more for our third call. Our confidence in the full-year numbers hasn't wavered. We closed the Kovr deal, and that marked a significant step forward — for this year, yes, but equally for the years ahead. And hey, maybe we'll catch you off guard again next quarter. Thanks.
That wraps up the call for PicPay. Thanks for joining us, and have a good evening.
Read the full transcript via the API
You're viewing the first half of this call. Get the complete PicS N.V. transcript - plus 132,000+ transcripts from 7,000+ companies, speaker segments and full-text search - through the EarningsAPI REST API or hosted MCP server.
Get an API key View API docs →For developers and AI pipelines
Programmatic access to PicS N.V. earnings transcripts and 132,000+ others is available through the
EarningsAPI REST API and the hosted MCP server.
Quarterly plans from $162 - full transcripts, speaker segments, full-text search,
and the /api/v1/transcripts/recent polling endpoint for ETL pipelines.