Poonawalla Fincorp Limited ($POONAWALLA)

Ladies and gentlemen, good day, and welcome to the Poonawalla Fincorp Limited Q4 FY '25-'26 Earnings Conference Call. We have with us today on the call Mr. Arvind Kapil, Managing Director and Chief Executive Officer; Mr. Sunil Samdani, Executive Director; Mr. Shriram Iyer, Chief Credit and Analytics Officer; Mr. Harsh Kumar, Head, Artificial Intelligence and CHRO; and other senior management officials. [Operator Instructions] Please note that this conference is being recorded. I now hand the conference over to Shabnum Zaman, Company Secretary of Poonawalla Fincorp Limited. Thank you, and over to you.

Thank you. In line with good corporate governance practices, please note this presentation may contain forward-looking statements regarding the company's future business prospects, strategies, estimates and profitability. But it is important to note that these statements are based on certain expectations, assumptions, anticipated developments and are subject to various risks and uncertainties. The actual results may differ significantly from what is stated in these forward-looking statements. Risks and uncertainties related to these statements include fluctuations in earnings, our ability to manage growth, competition, economic conditions in India and abroad, changes in laws, rules and regulations relating to any aspect of the company's business operations, general economic, market and business conditions, attracting and retaining skilled professionals as well as government policies and actions. Now I would like to hand over to Mr. Arvind Kapil, Managing Director and CEO of the company.

Thank you, Shabnum. A very good evening to everyone, and thank you for joining us. I'm pleased to report that quarter 4 financial year '26 marks a significant inflection point for our organization. Our strategic focus on scaling our 6 new business lines while maintaining rigorous cost discipline is now yielding tangible results. We have successfully expanded our return on assets, a clear indicator of our improved earning power and quality of our portfolio. This performance is underpinned by a meaningful shift in our efficiency profile. We have driven operating leverage effectively, bringing down our OpEx to AUM ratio down to just a year ago. Our results demonstrate that our deliberate investments in technology and our specialized branch network are not just growing the top line, but are structurally enhancing our profitability. I will begin by outlining our financial performance for the quarter and year and subsequently talk about the key drivers underpinning them. We closed the year in AUM of around INR 60,348 crores. This year, this reflects a year-on-year robust growth of 69%. The contribution to AUM from the customer businesses has substantially increased. The 6 new businesses are responsible for it. We've launched and they're contributing close to 24% disbursement in this quarter versus 20% last quarter. We have successfully operationalized 400 gold branches. The commitment we had made has been met. At this juncture, let me cover and reflect on the 6 vectors of our performance, which are an outcome of the nature and granularity of our growth. Our NIM has expanded sequentially by 43 basis points from an 8.62% in quarter 3 to 9.05% in quarter 4. In our quarter 1 financial year '26 call, we had guided that we will reduce -- we will restore, sorry, 9% NIM levels in 3 to 4 quarters. Happy to report back that we have achieved the same in 3 quarters. A significant contributor to the NIM going forward is our disbursement yield, in my opinion, which has already gone up by 40 basis points. It is a milestone that underscores the successful integration of our new product verticals and digital businesses. We plan to continue to optimize this by prioritizing segments with best risk reward dynamics. There has been a quarter-on-quarter decline in credit cost from 2.62% to 2.51%. This improvement is further validated by our 6-month outcome, 6 MoB 30-plus, which continues to further trend positively. Our GNPA for quarter 4 stands at 1.44% versus 1.51%. Credit cost declining trend is in line with the structural strength of our calibration and collections. It has started to play out as an advantage. And let's remember that GNPA is lead indicator of where the credit cost is heading for. OpEx to AUM has declined from 4.76% in quarter 4 last year to 4.13% this quarter on the back of productivity gains in our new businesses. We believe that this is a structural shift in the OpEx levels that will sustain and further reduce over time. I will explain this aspect in detail shortly. The culmination of these moving parts is a 70% sequential growth in PAT, reaching INR 255 crores for the quarter. So you can see the profits are moving in a certain trajectory. The credit cost is moving in a certain trajectory, and our OpEx to AUM is structurally moving to a lower floor. Our ROA has moved to 1.81% this quarter versus 1.2% last quarter. And if I compare it with the quarter ending March '25, it's a difference -- it was around 0.78%. We believe it is the strongest evidence of our structural operating leverage. We've crossed in my limited view, a threshold of our investments in tech, collections and new business lines with this slowly, steadily, gradually compounding. The 1.81% is a new baseline, I believe, and we should, in my limited view, grow strength to strength in a couple of quarters. Among all our 6 vectors in my limited view, ROA, I would treat as our North Star metrics. Let me now spend some time talking about the structural drivers underpinning this growth. I quickly give you a sense today. Firstly, I want to talk about our product mix and distribution. We like to think of our portfolio as a multispeed growth engine. Our consumer products like personal loan, consumer durables. I would like to call them our high-velocity engines that introduce consumers and households to our franchise. If you look at education loan and gold provide the right balance of yield and resilience, while our MSME and pre-owned cars are calibrated business for us that enable us to participate in the growth story of the country. Let me briefly talk about our new products portfolio. Prime personal loans. PL Prime, we introduced in August '24 continues to demonstrate strong momentum. We ended March '26 with a monthly disbursement a run rate of approximately INR 468 crores. We continue to see that 70% plus customers have bureau scores of 750 plus employed with Category A corporates earning INR 75,000-plus net take-home. In short, credit resilient salary profiles. In quarter 4, 33% of disbursements were processed through a fully straight-through digital processing. This has increased from 28% in quarter 3. We are very, very excited about the digital calibration and the momentum it's bringing and the cost efficiencies and operating leverage. We will move towards 35% to 40% over the next few quarters, adding strength to our operating leverage. PL Prime has transitioned from a linear cost model to a scaling digitally, where we build AUM without a corresponding increase on the overhead, just to give you a ballpark sense on the digital contribution increasing. A quick sense on the gold loans. Our gold loan footprint as well as portfolio size has significantly scaled up. We've stayed on course with our earlier commitment on opening 400 branches as of March. They've been operationalized. Total disbursement in quarter 4 is close to INR 819 crores, a fairly decent step up. More than 90% of branches are in Tier 2, Tier 3 locations. These storefronts and customer front is expected to deliver strong lifetime value to our franchise. We will aim to open branches along similar lines as last year. While opening dedicated gold branches creates a front-end investment phase, the long-term unit economics are very favorable for sustained profits. These branches act as localized profit engines and once at scale, require minimal incremental cost to maintain. A quick sense on the consumer durable business. Consumer durable business for us continues to scale efficiently, having onboarded 12,500-plus retail outlets across 240 locations. We have set up a plan to onboard around 12,000 outlets at the beginning of the year, which we surpassed. Seamless digital journey has enabled the business to disburse over 54,000 cases in a single month. Conceptually, we view durables as the anchor product of the emerging middle-class homes. So it creates a virtuous cycle, wider penetration leads to richer behavior data for us, which fuels smarter underwriting. Based on these beliefs, we aim to at least double the consumer durable customer count over financial year '27, basically strengthening our overall customer acquisition for the company. Commercial vehicle, I think a strong belief in last-mile logistics and infra push in the country, we're consistently scaling up the business metrics. Average monthly disbursals are approximately now INR 125 crores in quarter 4, grown by around 18% quarter-on-quarter. Our disbursement yield has also grown by 40 bps in portfolio compared to 3. We have concluded financial year '26 with INR 1,000 crores of disbursements. UCV continues to be the primary focus, comprising 70% of the quarterly and annual disbursements. Education loans of 1 year since our launch. We've built a strong momentum processing over 22,000 files as of March 26, our education has crossed approximately INR 900 crores in disbursals, driven by a distribution network now of around 500-plus consultants and strategic partners, firmly aligned with our original vision. We are also seeing a healthy traction in the instant sanction process, which now contributes to nearly 20% of total sanctions. Secondly, I want to touch upon our risk calibration and collection performance in my perspective which will be covered by Shriram at length. A quick snapshot in this area is, as I said, our credit cost declined further to approximately 2.51%, improving sequentially from 2.62%. The 6 MoB 30-plus book delinquency declined by 30 basis points to a new low, reinforcing the underlying strength of our portfolio. Incrementally, monthly cohorts across products are showing healthier delinquency roll rates than the previous months. To give you a flavor, the 12 MoB 90-plus of cohorts originated post September '27 -- '24, sorry, has seen improvements of over 50% compared to the cohorts originated 12 months prior to September '24. This performance is primarily on the back of our fine-tuned credit models. We are able to continuously monitor and refine the performance of our models in sync with signals received from payments data or collections visit and fueling back to the credit guys. Similarly, our collection capability continue to be sharpened via near real-time digital feedback loops and AI-led models and treatment strategies. We are in a position to dynamically update our models in a matter of days versus weeks, reflecting agility in our treatment strategies. Third fundamental point, let me explain our point around structural efficiencies and operating leverage. Our operating expense to AUM ratio has declined from 4.76% to 4.13% year-on-year, a few drivers of this to give you a sense. First, bulk of our planned investments in new products, physical infrastructure, technology, human capital that we've made over the past few quarters are now a source of operating leverage due to productivity gains. Our digital capabilities across the company and website customer journeys have matured and are leading to improvements significantly in productivity, cost of acquisition and organic business boost and optimized digital performance funnels. Even on our website for that matter, we have seen a 47% improvement in keyword ranking leading to a 51% improvement in share of voice over the last financial year. We continue to hold the performance levels across nonpaid search with approx 7 million traffic in March alone. Our AI investments are also beginning to improve our productivities. We are now live with 42 off 76 AI projects planned, which Harsh will cover in detail. This quarter, we have launched multiple initiatives that our Head AI will talk about in detail shortly. Important to note this, when we look at OpEx to AUM, it's important to contextualize it within the current year strategic investments. And along with robust profits, we're making investments as well to keep a sustained profits as a model which we're trying to achieve. Our investments are building a more profitable mode for the future, and we are deliberately continuing to invest behind the dual engine strategy. The first engine is focused on operational fortification, strengthening collections, upgrading technology and improving core execution capability. The second engine is geared towards market expansion, franchise deepening, including investments in new gold branches and consumer durable segments. On the first front, technology, our investments are moving beyond digital towards building an AI-led intelligence layer. We're strengthening our back-end engines to enable a more predictive underwriting and sharper decisioning. Importantly, we're leveraging AI to identify customers with highest propensity for subsequent borrowing, thereby accelerating cross-sell and materially shortening conversion cycles. On the collections infrastructure, we're consciously decoupling collections into distinct strategic vectors. We are investing in robust tech-enabled collection system to ensure a rapid acquisition supported by strong and scalable recovery framework. On the second front, we view consumer durable as a critical first touch point in the customer life cycle, enabling us to build early relationships and drive long-term cross-sell opportunities. By financing essential household needs, we effectively earn a seat at the table for thousands of new households. Becoming their first point of engagement with formal credit. This creates a high velocity acquisition funnel for us. We view this as a fundamental -- foundational investment, one that strengthens the long-term quality and depth of our franchise. Similarly, on the gold loan branches in parallel act as a high-yielding kind of a safety world for lack of another word, if one may use that term, while also providing strong risk buffer. While we've set ourselves an internal benchmark to close the next financial year at a lower OpEx to AUM ratio than our current levels, you may see fluctuations quarter-on-quarter for 10 to 25 basis points based on our investment trajectory and clustering of our branch opening. And this is a similar kind of guidance we gave last year as well. But I think the plan is that every March end, we should structurally move to another level and build the strength for the company on OpEx to AUM getting lower. In closing, let me summarize our trajectory remains focused on sustainable high-quality growth. The steady improvement in our credit profile with credit cost moderating to 2.51% and even more encouraging trends of 6 MoB 30-plus data reflects the health of our lending ecosystem. We're fairly looking confident on credit cost from here on. With asset quality improving across Stage 1, 2 and 3, we are operating from a position of strength, if I can say that. We've moved past the lifting of initial business setup and now firmly in the phase of harvesting operating leverage. We remain committed to our investment philosophy in the tech and collection to further fortify this momentum and strengthen new businesses, which are adding considerable value to our measurable metrics across business and risk. Thank you for your continued confidence in our vision. We stand committed to a focused strategy of creating long-term predictable, sustainable profits. With that, I will now hand the call over to Shriram.

Thank you. Thank you, Arvind. Good evening, everyone. India's growth momentum during the financial year 2026 remain anchored in domestic demand supported by consumption resilience and reinforced by policy measures. For Poonawalla Fincorp, the step-by-step execution of risk calibrated framework across origination, risk containment and collections is translating into a structurally stronger portfolio. This approach is delivering better incoming cohorts, lower embedded volatility and sustained improvements in collection efficiency over the cycle. Focusing on the asset quality, let me give you a glimpse of our key trends. The GNPA has shown a sequential improvement to 1.44% in Q4 FY '26 versus 1.84% in Q4 FY '25 and our NNPA, which was 0.85% in Q4 FY '25 is down to 0.74% in Q4 FY '26. Over the last 4 quarters, there has been a steady quarter-on-quarter improvement in Stage 1, Stage 2 and Stage 3 composition of assets, emphasizing our calibrated approach to portfolio expansion and strengthened debt management practices. Our Stage 1 composition in quarter 4 FY '26 is at 97.5% versus 96.3% in Q4 FY '25. Stage 2 composition in Q4 FY '26 is at 1.1% versus 1.85% in Q4 FY '25. On Stage 3 composition in Q4 FY '26 is at 1.44% versus 1.84% in Q4 FY '25. There has been significant improvement across all the stages, Stage 1, 2 and 3. The quarterly credit cost has improved to 2.51% for Q4 FY '26 versus 2.62% for Q3 FY '26 versus 3.14% for quarter 4 FY '25 last year. I would further like to highlight a few critical areas that reinforce our commitment to delivering best-in-class credit costs in the industry. First and foremost is a credit by design framework. We are meticulously building our portfolio skewed towards secured and unsecured products that inherently have lower risk and low risk cohorts like salaried profiles. Across products, 6 MoB 30-plus shows a downward trajectory compared to previous quarters. Key strategic changes for SME products like business loans and preowned cards have supported improvement in 6 MoB 30 plus. The early monitoring indicator of 3 MoB 30 plus for consumer loans continued to improve quarter-on-quarter, strengthening our confidence. Our deliberate choices in product mix and disciplined risk calibration focuses on customer segments that have inherently lower risk and more stable behavior patterns. PFL's risk management framework continues to be in alignment with banking standards as reflected in our 30-plus delinquency, which benchmarks us favorably against peers. Sequential improvement in the 6 MoB 30-plus for the last 4 quarters is a testimony of risk first framework implemented. 6 MoB 30 plus as of Q4 FY '26 is 1.05% versus Q3 FY '26 at 1.3%. The 12 MOB 90 plus for cohorts originated post September '24 has seen an improvement of over 50% compared to the cohorts originated 12 months prior to September 2024. The company is undergoing a fundamental strategic pivot towards a future-ready portfolio by gradually shifting its asset mix towards higher velocity, low probability of default segments such as loan against property, personal loans to salaried profiles of top corporates, gold loans and education loans to insulate the book from cyclical volatility. The front-end asset selection is bolstered by a transition from traditional recovery to a predictive AI-driven collection engine, which is creating significant operational leverage and a continuous feedback loop with underwriting for real-time risk calibration. By prioritizing quality at risk approach at the ground level, the firm is successfully stabilizing credit cost at a new lower baseline to deliver sustainable portfolio. It gives me confidence to share with you all that the seasoning of our portfolio is to our satisfaction, and we expect with fair confidence that from here on, we'll move strength to strength on our portfolio growth and quality. The second key strategic focus is on enhanced collection efficiency. Over the last financial year, our investments on enhancing the tech stack used by collections and manpower investments to be ready for scale across products and buckets have shown positive returns. I would like to share a few performance stats that will give you a glimpse of the impact of these investments. Our current bucket flow has shown quarter-on-quarter improvement, and we have achieved a 2x improvement compared to the previous financial year. The above has led to sequential improvement in Stage 1 portfolio to 97.5% in Q4 FY '26 versus 97.4% in Q3 FY '26 versus 96.3% in Q4 FY '25, reflecting strong early-stage collections and disciplined credit sourcing. There has been a quarter-on-quarter improvement in slippage ratios since September '24. Sequentially, Stage 1 slippage ratio has improved by over 15% in Q4 FY '26 versus Q3 FY '26 and Stage 3 slippage ratio has improved by over 11% in Q4 FY '26 versus Q3 FY '26. Slippage ratios have improved due to a portfolio calibration and improved collection efficiency across product categories. As we further prepare for scale, the team is focusing on GenAI voice bot-led calling with intelligent human handoff, behavioral AI nudges, speech to text AI integration for allocation, optimization and agentic workflows to auto trigger events. A few of the other AI, GenAI and automation use cases like multilingual voice bots for digital communications, copilot-based insights for collection manager, humanless allocation management, campaign strategy builder and call quality monitoring are supporting smarter day-to-day operations for continuous improvement cycle. Third key focus area is on continuous enrichment of our in-house proprietary models used across the digital and nondigital journeys. As we speak, 18% of the AUM is via digital journey driven by AI/ML models used across the different credit swim lanes. The multilayered AI/ML risk models leverage the volume, velocity, variety and veracity of data to enable early risk detection, industry-level risk representation and vintage level structuring. Usage of these risk models also augment the physical underwriter decision process. Across diverse models, consumption use cases, we have started our journey with implementation of Gen 2 models for credit risk and Gen 3 model versions for debt management to dynamically optimize calibration in response to shifts in portfolio mix. The credit AI for underwriter focuses on productivity improvement, enhancing accuracy and standardization. The Q4 FY '26 marks implementation at scale of credit AI projects across all major products, covering personal loans, business loans, professional loans, preowned car loans and equipment loans. For example, in personal loans, the current installed capacity of our headcount as of March '26 will now be able to process 1.2x the piles. Methodically, strengthening each stage of the credit life cycle from origination discipline to on-book risk management and collection, the organization is creating a sustainable improvement cycle, building AUM with better cohorts, lower portfolio volatility and sustainable gains in collection efficiency. Thank you so much, and I'm handing over to Harsh Kumar.

Thank you, Shriram. Good evening, everybody. AI continues to be our core driver for the operating model for transformation that we have taken up, a key source of long-term differentiation for Poonawalla Fincorp. Before I take you through the specifics of this quarter, I want to set up 3 things that in our assessment define an inflection point in our AI journey. First, the scale of AI usage across organization has crossed an important threshold. It is now meaningful to be measured across multiple metrics. Second, several of our AI platform we have spoken about in earlier calls have moved to full production. They are now generating measurable outcomes, not just productivity increase, and I'll walk you through a few of them. And third, our AI architecture has begun a clear strategic shift. We are moving from point solution to agentic system, AI that does not merely assist user, but reasons, executes, monitor, improves outcomes within govern boundaries. Let me start with scale. Enterprise-wide AI token consumption has increased more than 100x year-on-year and now stands approximately at 30-plus million tokens per month. Based on our current deployment pipeline, usage trends, we expect the number to grow multi-fold over the next year and AI deepens into workflow decision-making across the organization. Additionally, we now have 76 AI projects identified, 42 deployed with 34 under development across business and functions. Each of these projects may have multiple agents acting in concert. We are introducing token consumption as a metric because in our assessment, it is the cleanest measure of how AI is actually being used inside the organization. The infrastructure cost of running these tokens, model inference, compute, cloud is in our assessment, materially small related to the scale of measured business savings or productivity gains that these platforms are now generating. The ratio of measured business benefit to AI infrastructure cost is decisively positive, and we expect that ratio to expand as scale increases because inference costs are largely fixed at the platform level, while application across more workflows is incremental. Sathi is one of the credit AI platform that Shriram has already elaborated on, and thank you, Shriram for that. So I'll now take you through that, but we had also announced customer service part that I wanted to cover, having gone live in FY '26. I'm pleased to confirm that the platform is now live in production. The platform's design capacity is to autonomously resolve 80% to 85% of customer interactions. That remains the steady-state target, and we expect to reach that level over the course of next 2 quarters. Customer wait times have already reduced by approximately 35% to 40%. This single platform when at design capacity will materially reduce our cost to serve while improving customer experience. Beyond just credit and customer service, their internal productivity platforms have moved to full production at scale in this quarter. Build Buddy and AI engineering copilot enabled our IT teams to deliver 6 new product builds within a single quarter with measure productivity uplift of 70% to 80% across development. The more important benefit, however, is the compression of release cycles, which directly accelerates time to market for the new businesses we are scaling. We continue to maintain high levels of governance and oversight to mitigate the risk of vulnerabilities for credit risk. DART Genie, our natural language insight engine is now active use across operations, HR and finance. It is saving approximately 400 man hours annually in analysis time that was previously spent on manual dashboard adhoc reporting request. We expect adoption to expand significantly in coming quarter as we extend the platform into customer service and risk. PAI@HR, our AI-powered employee assistant is constantly growing and is now autonomously resolving close to 90% of HR queries. Average resolution time has compressed from 24 hours to under 10 seconds. To give context to the 90% employee interaction per month, which vast majority would have previously acquired HR team intervention. High-volume operational tasks like cab booking, key employment letters are now executed with 0 humans. Our AI-led hiring platform has compressed offer release to now an average of under 1 minute and scaled monthly hiring capacity now dependent on compute alone. We have already pushed the system and without increasing further compute improved from 68% increase in capacity with 81% reduction in hiring related operational costs. And importantly, no compromise on candidate quality. As we continue to expand our branch network, the new product teams, this platform is what allows us to hire at scale without proportionate operational track. In March 2026, we launched an AI content factory, a designed creative studio where prompt engineers, creative specialists work closely to create hyper-personalized communication, creative and media. Within a few weeks, we have witnessed a 12x increase in our communication output with significant cost saves. This will enable us to run multi-segment campaigns across media types across cohort at scale. This brings me to what I described as an important strategic shift of the quarter, our move from point AI solution to engine system. Just to explain a point solution executes a single defined task for a person and agentic RPA, which is what we have launched, works across multiple people, context, take sequence of action, monitor those outcomes and actions, ensure that system and agent build to work are governed within the boundaries itself. This is a shift that allows AI to scale beyond individual use case and become operating infrastructure, not a tool but a substrate on which workflows run. Similarly, we have launched autonomous API testing agents. This embeds agentic intelligence into our engineering life cycle. It independently generate test kits, validate API behavior, proactively identifies anomalies. This will improve release quality, accelerate development cycles and reduce manual testing effort, directly reinforcing the engineering productivity gains we are seeing from Build Buddy. DIY bot creation tool. This we launched primarily as a no-code platform that allows business team to independently build, deploy domain-specific conversational agents for their teams. It has already begun to drive bottom-up innovation across enterprises, improving query turnaround time and democratizing access to institutional knowledge. We believe this will be one of the highest leverage platform over the coming year because it shifts AI from centrally built capability to one that organization can extend organically. On the agentic road map for FY '27, these 3 platforms are foundational. They are not totality of our agent strategy. Every one of these platforms with multiple agents will operate within seven-sutra governance framework within our existing model risk discipline. Alongside these foundational platforms, several focused AI solutions have been deployed over this quarter to strengthen frontline execution, sales and digital support. Over the next few quarters, we will deploy agentic AI across our funnels to improve relevance and guide customers appropriately through their loan journeys. Three agents are in pipeline to be delivered in Q2 FY '27 are expected to improve our digital loan conversion rate by over 15%. Our data foundation across structured unstructured signals will be closely coupled with our journey on the cloud. And this will significantly improve AI experience delivery to our sales channel and customer while keeping a tight control on our cloud costs. For our sales and credit legal technical teams, we have launched multiple bots, ask for education loan called Build Buddy for business loan and LAP assistant for loan against property. We standardize the interpretation of policies and process across frontline, reducing operational risk in precisely the product that are scaling fastest. For services and tax workflow, we have deployed sake assist and taxation. The reduce friction in the back-office workflows that are critical to running regulated lender at scale. Speed, consistency and compliance has improved. And for leadership development scale, we launched Lean Forward, an AI-powered coaching assistant that provides our managers personalized on-demand guidance for feedback, people decision and development planning. Before I close because AI skills, governance is what underwrites the credibility of entire program. All our AI projects operate inside the seven-sutra governance framework. And of course, we are further strengthening our AI governance processes, which we have adopted earlier in alignment with the principle laid out by the regulator. Every agentic deployment, including 3 platforms that I described in previous section, operate within explicitly defined boundaries, scope of action, escalation trigger, human loop checkpoints or any decision that affects credit collection, customer or stakeholder outcomes. On model risk specifically, our 50-plus AI model, which Shriram has already spoken about, support trade decisioning setting inside our model risk framework. On data privacy and DPDP Act, our architecture is designed with explicit data classification, consent management and data localization principle. Sensitive customer data is handled within governed environment. We have control in place to manage exposure to external model providers and those controls are reviewed by information security team. On vendor concentration, our architecture is deliberately multi-model. We are not building dependency on any single large language model provider. The agentic platform and DIY bot creation tool are designed inherently to be model agnostic at the orchestration layer. This protects us against vendor pricing changes, capability gaps and continued risks. And on auditability, every agent action is locked and is reversible. To summarize the point, we see governance not as an overlay on AI, but as a precondition for scale. Thank you. I'll now hand over to Mr. Sunil Samdani, our Executive Director.

Thank you, Harsh, and good evening, everyone. Let me quickly take you all through the financial highlights for the quarter. The assets under management stood at INR 60,348 crores. On the liability side, as part of our debt strategy and in line with our projected AUM growth, we continue to diversify our liability book, focusing on long-term borrowings. Hence, the share of borrowings from long-term sources has gone up by approximately 3% from 83.42% to 86.50% quarter-on-quarter. This number was 61% in Q4 of FY '25. Our net interest income, including the fees and other income continue to grow healthy, standing at INR 1,276 crores for Q4 of FY '26, which is up 18.2% quarter-on-quarter and 78.5% year-on-year. On a full year basis for FY '26, our NII, including the fees and other income stood at INR 4,029 crores, which is a growth of 49% year-on-year from INR 2,708 crores last year. The cost of borrowing for the quarter stood at 7.63% versus 7.65% in Q3 of FY '26. The OpEx to AUM stood at 4.13%, a reduction of 28 bps quarter-on-quarter. The pre-provisioning operating profit during the quarter was at INR 695 crores, a 31.6% increase quarter-on-quarter. For the full year of FY '26, the PPOP is INR 1,934 crores, which is a growth of 36% from INR 1,417 crores a year earlier. The asset quality improved quarter-on-quarter with GNPA at 1.44%, a reduction of 7 bps quarter-on-quarter and 40 bps year-on-year and net NPA stood at 0.74%, a reduction of 6 bps quarter-on-quarter. Our provisioning coverage ratio stood at 49%. Our profit after tax stood at INR 255 crores during the quarter, 69.6% growth quarter-on-quarter. This is despite company making significant investments in new businesses, branches, AI and technology. For the full year of FY '26, the profit after tax is INR 542 crores. Our debt equity ratio stood at 4.67x. This is before the capital raise of INR 2,500 crores, which we did in April of 2026 through the QIP route. Post capital raise and basis March '26 balance sheet, the pro forma debt equity would stand at 3.7x. On the capital adequacy ratio, we continue to remain healthy and comfortably above the regulatory requirements at 16.83%, of which the Tier 1 capital is at 15.90%. With successful INR 2,500 crores of capital raise, the stimulated capital adequacy ratio stands at 20.74%. This is basis the March '26 balance sheet and the capital raise that we talked about, and this gives us enough headroom for growth. Our liquidity coverage ratio at 181% is comfortable. On the liquidity front, our surplus liquidity of INR 7,590 crores stood at the end of March 31, 2026. Thank you. And now I would like to open the floor for question-answer session.

[Operator Instructions] Our first question comes from the line of Chintan Shah from ICICI Securities.

Congratulations on another strong quarter. So sir, firstly, on this yield, so yes, we have started giving the disbursement yield range. So also, sir, could you help me with what would be our disbursement is for the quarter and what would be our book yield for the quarter? Just trying to understand the difference between the disbursement yield and the book yield. Yes. That's the first one. And secondly, on this asset quality, the asset quality has been continuously improving. But given the geopolitical situation, do we envisage any risk to that? And particularly, are we looking to make or did we think of making any overlay or provisions just to strengthen our provisions in terms of any risk if it emerges, given that we are also almost 50% portfolio is unsecured. So in that context, yes.

Yes. Thanks, Chintan. I think let me address your second point first. I think on the credit side, if you look at -- you open Shriram, but our exposure in my limited view, remains well within the defined risk tolerance. We do a lot of -- internally, we do a lot of worst-case scenarios, stress test modeling across not just the portfolio, but across liabilities and a whole lot of disbursal yields, which is why if you notice, normally, liability yields in the industry changes fast and asset repricing par is very tough to get. But in our case, if you notice the way we've built the model structurally to make it stronger, of course, we didn't know the world is going to come in, but we wanted to fundamentally make it stronger. And with that intention, you start seeing that our disbursal yield in quarter 4 has gone up by 40 basis points and your portfolio yield will gradually do the catch-up. But the more your disbursal yield goes, what does it show? It shows that if you were approximately, I'd say, 15.56% was your disbursal yield a quarter before approximately. And if it goes to a 15.96% and next quarter, it goes inching higher, that means you have the pricing power in a business, which is not a very easy visible side across the industry, across any company normally. It's the strength of the construct, which it indicates in my limited view. And I think that's the way I would read it. We see a lot of promising increase in our disbursement yield leading to subsequently, obviously, the portfolio will catch up. But we have a healthy pace of growth in this financial year plan. And as far as the calibration is concerned, see, the very fact we diversified across consumer durable, gold, these are investments. And if you look at PL Prime, the digital side is giving us the first right to refuse despite our size. Look at loans against property, we've structurally do over INR 50 lakhs. We don't do micro LAPs. We've kept ourselves disciplined on the credit calibration. If you look at education loans, commercial loans, even if you look at business loans and preowned cars, you'll be surprised, unlike our overall growth rate looks very high, there are products like business loan where we've asked the team to operate right now on low teens to mid-level teens on that SME, MSME. But if you look at LAP, we are doing a healthy robust growth because it's over INR 50 lakhs. And if you look at preowned cars, which is also an industry relatively more riskier, we've again said low teens. So we've calibrated ourselves in investments and balancing of the profit so that you can sustainably grow faster. But Shriram, would you like to add something on the war side, which he's asking and impact that you see?

Yes. As Arvind said, if you look at our assets under management, the focus is on the low-risk assets like loan against property, personal -- loan against property, we don't do less than INR 50 lakhs. And we've seen that the portfolio is below INR 50 lakhs, tend to be a higher portfolio at risk, and that is a portion which we don't do. And we also focus on gold loans. Our exposure to -- is only to the top corporates, employees where we give personal loans, education loans. So all of these assets, in my view, is that these are very less vulnerable to the headwinds in the external environment. I don't see any risk. And if you look at your GNPA quarter-on-quarter has improved, the slippage ratios have improved. My 6 MoB has significantly improved, even my 90-plus 12 MoB, if you see, there has been a significant improvement of 50% over. So if you look at overall from a portfolio point of view, I don't see any risk here. Hence, these assets are less vulnerable to headwinds in the external environment. And hence, the management maintains a cautious optimism and continues to focus on monitoring and recalibrating our portfolio if required.

Correct. So we are closely watching, Chintan. I mean it's not to say that war should be ignored. Without a doubt, we're closely watching the environment. But consciously, somehow we wanted to make a more moderate risk model. So kind of vulnerability probably index is not the right word. Our portfolio could be a little more crafted in a manner which looks more solidity to it, but we are closely watching the environment.

That is very detailed. And so sir, just one last question, if I could squeeze in. So yes, in terms of this fixed MoB 30 plus, it has been continuously on a declining mode, and that is around 1.05%. So what could be a steady state number year which we would be looking at probably at this level, it could settle down.

Yes. So if you look at -- it's coming down quarter-on-quarter, as you've seen the numbers, it would be range from. But if you actually look at the products such as gold loans, personal loans prime, all of these assets when they start having a larger share into the overall AUM, these numbers will trend downwards. So that is something which I can tell you. But I think you must remember one thing, like I said in the conversation, which I was speaking, GNPA normally, in my limited view and experience, normally shows you the lead indicator of what's coming ahead. So as your mix of gold will increase, you had, for example, barely, I think I don't remember the exact number, but let's say, INR 900,000 crores book of gold, but you have 400 branches. So obviously, your March number would have been close to a substantial number. And you can well imagine how the contribution of gold will go up. Now you know the industry cost of gold. So with -- these are strategies we shared with you. Every word of what we shared over the last 23 months in my limited view has been executed precisely before or on the time. And that credibility, you can trust us that we have normally keep adequate margin of safety when we talk. And of course, we are moving steady and steady. The credit cost and the portfolio, the way we stand for multiple reasons gives us fair confidence that the portfolio strength should get stronger and stronger from here.

[Operator Instructions] Our next question comes from the line of Kaitav from Anand Rathi.

Congratulations on a good set of numbers, sir. Number one question is on the fee income trend that has been trending very robustly. So if you have some guidance around that, that would be the first question.

Yes I think on the fee income side, we all, as a management team comes from -- we've handled fairly large businesses. So we understand the various vectors of fee income, whether it's your processing charges, whether it's your insurance businesses, whether it's various cross-sell businesses or we plan to launch some new stuff. So I think you will see a fair amount of strength coming in this year. As we've already done a lot of effort in launching all our businesses, things are stabilizing, distribution are stabilizing. It's becoming more a regular calibrated growth from here on is sort of a massive amount of launches that our effort went in. So a lot of focus will go on to ROAs. And I think all of us are well aware that our fee income is a very important component of ROAs. And ROA fundamentally for any company, valuation professionally has to be given the due respect of being a North Star. I mean I just used it to say that in our health, a lot of our decisions will be based on ROA because as a company also we'll do robust profits plus investments. So that's the way we believe sustained profits get created. You cannot lower investments just to further boost profits because that's not the way sustained profits happen. But fee income between various vectors is, I'll be honest with you, it's very strongly under our focus and the entire team is working on it, but there's no guidance that we give on these things. Let me have something in the back of my pocket. Broadly, we've given you everything on it.

Sir, the second question was we have always focused on technology and AI.

Right. Yes, sir. Sorry, we can't hear you.

How AI and tech has been progressing so far to the best of your knowledge, where you are in that journey, if you can spend 2, 3 minutes on that, that would be grateful on what we can look forward to in the next couple of years.

See for us, AI and digital, let me put it as 2 vectors, which could give you value. On the digital side, if I share the figures with you, we are well over 30% of our entire business like a 480 kind of 500 kind of number that we are looking at now. Imagine a 30% or 40% of that gradually becomes fully digital of this scale in a personal loan prime corporate India taking from us can give you a sense of the stuff we are planning to -- this is something which we've executed. This is not something which we are trying to execute. But now how does this scale over the next few years? Business loan is calibrating SME ticket sizes. And right now, we are very, very strict on its calibration. But within 6, 9 months, once we are very solid on that stuff. We will come up with some interesting turnaround times on the digital side of the business loan as well. We are looking at digitizing a lot of business, even if it's halfway through on the front side with the customer. We all come with distribution background. The turnaround time for us is going to be key. So that's where the digital path rest. On AI, you have to appreciate that the idea of giving you these 75, 76 projects is to give you a sense that we haven't launched 2 businesses or 4 businesses which only did credit underwriting. Today, for example, to give you a sense of the output, a single product like a personal loan near prime, we are not hiring new underwriters despite our growth rate being substantially robust this year. So we've kind of frozen a manpower of last year. And because of our ability with AI and the way our credit is calibrated, we have successfully managed to grow the operating leverage there. Similarly, one by one, all products will start seeing that value. You'll have to take a product at a time and start doing it. And if you look at across the organization, whether it's a small initiative or large, you'll find the 76 bringing in a culture where whether it's a finance department doing automation, operations trying to do automation plus AI or a department like business trying to figure out which parts of it can he use AI. There's a cultural focus on the fact that we've got to create an operating leverage with both AI and digital. So it has to be net impactful. I'm not into technologies which don't change my life yet. So for me, every step that we take, either culturally builds our efficiency of innovation, and there will be some products which could far exceed the others on the impact. we are very clear that we'll keep moving forward with this. All these projects, I think, should give you confidence that culturally, we are extremely rich and agile as a company in these areas.

Our next question is from the line of Jay Betai with NBIE.

Congratulations on a good set of numbers. Sir, my question pertains to AI and the returns we are generating. So if you can share some color that how are we focusing on increasing our ROA going ahead first? And second question is on disbursement. I would like to know the disbursement number for the full year as well?

Let me start with, I think the most interesting one is the ROA. ROA, I think we've taken a 1.81%. I see -- if I look at the NIMs, in my limited assessment, looking at the environment, I think our NIMs in my assessment overall are looking positive and accretive. So I think that's a strength, which gives me confidence that there's -- we are very confident as I see through the future of the next few quarters, right up to the 4 quarters, ROA should gradually start moving strength to strength. We don't give any intermediary guidances. But like you can see, we never gave a guidance of 1.81% either. But step by step, we are building on the businesses. I'll keep balancing investments and profits. But I see the ROAs from here on moving strength to strength for probably a couple of quarters or probably a couple of years. And I think we've reached that point that our strength is emanating out of our pricing power. If you see our disbursement pricing, 40 basis points, you're well aware, is not easy to increase on the NIM side, on the disbursement yield side. And with the growth rate that we have, probably this year, disbursement will be 1/3 of our book. So I think that will also give substantial strength to the NIMs, adding to the ROAs. We have also fairly positive despite our investments with minor fluctuations in quarter-to-quarter as you cluster the branches. But I think structurally on the OpEx to AUM, we should leave -- we are hoping to reach a lower level by March end despite the fact that our OpEx to AUM is actually substantially improved than even we had anticipated, to be honest. And it's -- the productivities are kicking in and stabilizing. giving us the confidence. Credit cost also one of the vectors, which adds to the ROA is looking from here on, looking fairly robust, both by design and by calibration and I must mention by collections. We are making substantial investments in collections, both in terms of focus and technology for us to see the results. Jan to March has seen a substantial strength on the collections, and we are hoping to keep it robust. So that's, I think, one part of it you asked.

Disbursement number for the full year.

So I think we've given a guidance of AUM of INR 35,000 crores to INR 40,000 crores. We could be probably a notch better, but it depends. We'll watch closely how this environment plans out. But directionally, I think we would like to look at a INR 35,000 crores to INR 40 crores and commensurate disbursements along with it. And that's the balance that we'd like to keep. There's something else.

Just one more thing. If we see on Slide 22, we -- for the longer-term horizon, we have some negative ALM. So if we factor in the amount raised of INR 2,500 crores, so do we see -- how do we see that gap bridging out?

So it gets bridged. With the capital raise, it has got bridged.

Thank you. That was our last question, ladies and gentlemen. On behalf of Poonawalla Fincorp Limited, that concludes our conference. Thank you all for joining us. You may now disconnect your lines.