← Throughline · holding view Deep analysis ↗ Q4 FY26
BAJFINANCE Bajaj Finance Ltd · NBFC Q4 FY26 · concall
Pattern: ai capex roi p

Karnataka 'political risk' retired after Q1FY26.

3 weak · 6 clean — pushback across 3 of 9 Q&A turns

Focused evidence 3 of 9

Kuntal Shah · Oaklane Capitalweak

What kind of capex or investment we are going into AI infra? Can you give some color and what kind of ROI and payback you are envisaging? And when it will start impacting the P&L via cost saving or faster turnaround?

Consumers and employees both should experience in the current year what AI does for them. Walking into 3,000 stores or 2,700 across branch — gold loan branches, customer service branches and in-stores, we will deploy close to 2,700 cameras, customer identification, reduce friction, faster turnaround, lower cost. AI call center agent is one-third of the cost, straight. We at a point had 5,000 outbound voice agents — that number we have brought down, and 30% of them are now AI voice agents. On overall capex — we don't publish capex. Otherwise it will become a new added metric and we anyway publish hundreds of metrics. But we are not pulling back any stops in investing whatever it takes. We will hand to Anurag. Anurag: We are parallelly investing deeply on the security and the compliance infrastructure as well. RBI has released a FREE-AI framework. We have consumed and are in process of implementing the explainability, auditability and the transparency what the FREE-AI report talks about. Rajeev added: You will experience on the app very soon, AI summaries will start to appear because there is 120,000 Figmas. We have 71,000 employees — they'll all transitioning by end of May to Copilots. Anurag: Last year we processed on peak of Diwali on a single day 600,000 loans, which without AI we had capacity to process only 100,000 loans on a single day. So you might see next Diwali, we'll process close to 1 million loan accounts on a single day.

Shubhranshu Mishra · PhillipCapitalweak

How do we split the opex into cost of acquisition and cost of collections? In risk management, how many people do we have on rolls for collection and how many feet on street of collection agencies? On back calculation basis our growth and ROA metric, profit growth would be upwards of 35% versus 22% AUM growth — looking at 35-40% profit growth in FY27?

On opex, the results sheet carries additional breakups. The sourcing cost to the extent it is variable in nature gets accounted as EIR in the interest income line, so that's not separately identifiable. But the recovery commission and the cost we incur for collecting money from customers and doing DMS debt management services is shown as fees and commission expenditure separately. On profit growth for next year — 22% to 24% growth in AUM should lead to a little better growth in profitability if we deliver operating efficiency growth of 25 to 40 basis points that Rajeev has talked about, and if we see improvement in loan loss to average AR, even with marginal moderation from NIM perspective. There is a possibility that the P&L growth may be faster for the next year than the balance sheet growth. I don't think I've done a math internally which suggests a 35% number, but I'll be more than happy to be positively surprised. Rajeev added: we have tailwinds subject to geopolitical stability, we should be in good place.

Bharat Shah · BCS Capital Ideasweak

If we see evolution of Bajaj Finance pre-COVID vs post-COVID — out of the 4 levers (growth, margin, cost intensity, credit cost), pre-COVID every element played like a good note in symphony, return on asset and return on equity tended to be buoyant. Post-COVID that equation has been middling, the crescendo has not been rising but undulating. Now with AI in play, are we really poised for the rising crescendo rather than the falling crescendo?

Your observation is correct. Pre-COVID to now, the company is 4x, INR140,000 crores balance sheet — we are INR510,000 crores. Through this period if you take return on assets and return on equity, FY19 22.5%, FY20 20.2%, and we have remained between 19% and 20%. COVID was a big shock. Post-COVID there was a massive boom which we leveraged by expanding lines of businesses, expanding locations very rapidly, investing very deep in digital transformation. ROE in FY23 went to a peak of 23.5%. Good times lead to what I would call in boom cycles, you take a set of decisions, which necessarily should be more calibrated. In the last 18 months we've calibrated those responses. You should see us from here on in a reasonable cruise mode. At INR510,000 crores we are by far the largest, an important player in overall credit. It's our fiduciary responsibility to ensure we are highly resilient. We will be a bulletproof business while delivering a 20%, 22% ROE and a 22% to 24% balance sheet growth. From here on, lot more cruise mode. I want to go back to FY20 credit costs — now the 3 MOB, 6 MOB, 9 MOB lower than FY20 and balance sheet is 3.5x. The profit growth principally should be higher than balance sheet growth — every incremental dollar of differential goes to strengthening the balance sheet because we're just living in continuous crisis. Last year we had Indo-Pakistan war for 4 days, tariff for 4-5 months, and February-March Iran-U.S. war. We have to remain open for business. The best way is to just keep bullet proofing the business — it needs tremendous agility.

▸ Other Q&A (6)
Abhishek Murarka · HSBC

On credit cost guidance — net loan provision of INR 1,866 crores works out to 165 bps in 4Q. You're guiding for 145 to 160 bps next year, full year. And for the annual number, INR 7,934 works out to 175. So on an annual basis, 175 goes down to 145 to 160? Also, in this 145 to 160, you have not accounted for any kind of adverse impact of geopolitical issues — if anything comes up second half, then this may go up?

Yes, that's perfect on both points. I would just say the more important point or a meta point in this is we are entering the year on credit costs with tailwinds. Headwind over headwind virtually would consume twice the fuel. Thankfully, we are entering the year with tailwinds. We have momentum. We can navigate even if the environment weaken. I'm also making a point that we have tailwinds on credit cost. We don't have headwinds. So that itself is a big help. Sandeep added that some level of conservatism in terms of doing the number crunching and giving a guidance of 145 to 160 has been done.

Shreya Shivani · Nomura

On the long-term guidance of ROA inching towards 4.3 - 4.7 percentage — can you help us get into the details over the longer term? Also for FY27, as per my math, you may need a little bit more NIM expansion (or at least stable/flat NIM) to achieve the ROA guidance for FY27?

See, the answer to the second part is, no. If you go to Panel 35, which is a 19-year long snapshot, since FY20, 4.2%, 4.1%, FY22, 4.2%, FY23, 5.3%, 5.1%, 4.6%, 4.3%. I'll add a dimension of ROE. ROE guidance of 20 to 22 last 6 years, adjusted for COVID, and during this period the balance sheet moved from INR115,000 crores to now INR510,000 crores. So it should give you confidence that we can continue to sustain the ROA, at times by expanding margins, at times by reducing cost, at times by reducing credit cost. These are the 4 levers — growth, margin expansion, operating expense and credit costs. We orchestrate these 4 to deliver an outcome of 4.3% to 4.7% ROA and 20% to 22% ROE to investors on a long-term basis. Sandeep added: on 4.6%, the core financial performance ROA for last year was 4.6% and guidance for next year is 4.4% to 4.6%. There will not be a need for NIM expansion in the next year to deliver that — we have to grow nicely (22% to 24%) and remain anchored on opex to NTI and loan loss to average AR.

Kunal Shah · Citigroup

On the growth guidance, 22% to 24% — apart from SME which is almost bottomed out, ex of that, and the newer businesses, where do we see we would be able to grow at a pace faster than the overall AUM growth? And which segments do you still believe could continue to lag, particularly on our core segments?

So Kunal, the idea would be to maintain the portfolio mix. That's going to be the agenda for next year as well. The current year number of 22.4% AUM growth is impacted by two things: one, the captive 2-wheeler 3-wheeler financing business has seen an attrition of 60%. That's almost INR6,500 crores of attrition we recorded in the current year. Second, MSME, the growth used to be 25%, 28%, 30%, has been brought down to 6%. These 2 things should provide tailwind on the overall growth number next year — captive dial-down impact will be much, much lesser, and MSME should come back into double-digit momentum from second half onwards. Having said so, there are a few businesses like gold loan financing, tractors, CV — they may grow faster because of low base. Gold is INR18,000 crores in the balance sheet (3.5% of book today), and given doubling down on branch presence, it should look towards 5% kind of number in the next year. Rajeev added: if you plot Panel 63 of composition of our balance sheet over last 5 years, you will not see material movement — base is large; reasonably diversified portfolio.

Piran Engineer · CLSA

Can you share some portfolio cuts or what you're seeing in the environment that gives you the confidence to pick up on MSME growth? And in new consumer products like car loans, gold loans, what percentage of business comes from existing customers and how many from new customers?

In Q3 I made a point that I'm not necessarily a very big believer in macro. We are a micro business. At a design level, MSME is a very, very large sector. You just have to pick what you can digest from a risk standpoint. Even at this point of time, based on our assessment, we have pruned business INR100-odd crores. We used to do INR1,800 crores monthly. We brought it down to INR1,400 crores. We have further taken cuts in the current month. The guidance next year is based on stable environment and macro stability. Our business is micro, not macro. On new businesses contribution from existing customers: in 2-wheeler (3 years old open-architecture), ETV performs significantly better than NTB — 50% of 2-wheeler customers we bring is an ETV customer. Between 43% to 45% of new car financing customers comes from ETV (INR450 crores to INR500 crores volumes a month). At India level, 35% of all new car finance and used car finance per bureau data comes from this 120 million customers. Personal loan our market share is 8.5% to 9%; 45% of personal loans takers in India come from our 120 million franchise. Risk, wallet and operating leverage all 3 require us to focus on this metric across all our 32 lines of businesses.

Viral Shah · IIFL Capital

On PCR coverage across Stage 1 and 2 — versus last quarter, you have further shored it up, almost highest levels barring COVID. Is there cushion built up that may unwind through next year, or does PCR remain or reduce? And on the credit cost guidance of 1.45% to 1.6% — the change versus 165 to 175 basis points you guided last quarter for FY27, is this purely because of the change in presentation of recoveries?

Your assessment is correct on the second point versus 165 to 175 that Rajeev had said — that becomes 145 to 170 because of the change in the presentation. As regard the ECL provisioning across Stage 1, Stage 2, Stage 3, this is a bottoms-up exercise that we do every year. We are moving from yearly exercise to quarterly exercise. The idea is to focus on resiliency of the balance sheet and the provision coverage ratio. I have reason to believe the numbers should remain in this corridor. We get opportunity, we will further up the NPA on provision coverage ratio to ensure that we are bullet proofing the balance sheet from any kind of potential impact that one could witness maybe 3, 5, 7, 10 years down the line. So resiliency and bullet proofing of the balance sheet are the 2 most important parameters that we also look at when we do a complete remodeling of ECL.

Omkar · Shree Investments

In the Investor meet you guided that it's not about question of growth but how quickly you can compound. Given current balance sheet size, you have guided 22% to 24% for current FY. How does it compare with overall long-term guidance of 17% to 19%, 20%?

The long-term guidance at this point in time is between 22% and 24%, 25% kind of corridor. Based on current performance and momentum, we are reasonably confident that 20% to 24% should be achievable for next year. It's year-by-year. So long-term commitment continues to remain anchored at 20% plus number for FY22, 23, 24. For next year, 22% and 24%, and then build on top. Rajeev added: for INR100,000 crore addition in balance sheet, the growth in total credit if denominator is total credit in India (gross bank credit plus NBFC credit minus bank lending to NBFC), our market share would have moved from 225 basis points to 250 basis points. If India total credit grew by 12%, we have grown by 22%. Given our relative market share, we should be aspiring to grow 2x of what the system growth is to meet our ambition to be among top 5, 6 financial services lenders in this country over next 5 to 7 years. Take customer centricity and AI transformation — the compounding should become easier. INR100 invested in us principally generates a 20%, 22%, fixed deposits give you 7%, we are giving you INR20 for the last 10, 11 years.

▸ Prepared remarks (5 blocks)
Good evening or good morning, depending on the geography that you're in, and welcome to BFL Q4 earnings call. The investor deck for Q4 has been uploaded on our website. I'll focus on key updates for Q4, year gone by, and management assessment before open the floor for Q&A. I expect to take between 20 to 25-odd minutes. Overall, I would say, from a company standpoint, it was a pretty strong quarter across all key metrics of volume, new customer addition, AUM growth, credit cost, profit growth, return on assets and return on equity. During the quarter, the company crossed the milestone of INR 5 lakh crores AUM. It's a proud moment for us as a company to end the quarter at approximately INR <strong>510,000 crore</strong>s. In Q4, AUM grew by 22.4% and PAT grew by 26.7%. In Q4, AUM grew by a record INR 25,500-odd crores. We booked 12.9 million loans and added 3.93 million new customers to the franchise. Overall, the customer franchise as of FY'26, March stood at 119.3 million customers. In Q4 this year, the company has taken an additional ECL provision of INR 142 crores towards management and macroeconomic overlay. The company has also revised disclosure of recoveries against written-off loans. This was earlier presented under other operating income. It will henceforth effective Q4 FY'26 will be reported under loan losses and provisions as a net number. This principally leads to redefining 2 key metrics: opex to NTI and loan loss to average AR. As a result of this change, the opex to NTI is increasing from 33.2%, which is what it has come in to 33.8% as a result of this reporting change. And loan loss to average AR is reducing from 1.75% to 1.65%, this is really how we will report going forward as a company. The gold loan portfolio continued to witness strong momentum. It grew by 115%. The business now contributes to 3.5% overall AUM. We foresee gold loan portfolio to probably cross 5-odd percent of total AUM by FY'27. MSME continue to see muted growth. It grew by 6% only on account of a set of proactive risk actions that we have been taking since Q2 FY'26. We expect that it should come back to double-digit growth or the company growth momentum between Q2 and Q3 of FY'27. Cost of funds improved by 4 basis points in Q4 and came in at 7.41%. NIM remained steady. AI implementation continues to accelerate across the length and the breadth of the company. Full-time employees stood at 71,613 employees. We added just a tad below 1,800 employees in Q4, 50% of them were actually in gold loan and MFI. Loan loss to average AUM as what investors have been used to came in at 175 basis points or 1.75% as against 1.97% in same period last year or Q4 FY25. Stage 2 and Stage 3 assets have sequentially reduced by INR 430-odd crores. GNPA, NNPA came in at 1.01% and NNPA came in at 41 basis points. Provision coverage ratio was at 60% as against 54% same time last year. The captive 2-wheeler and 3-wheeler business, which now contributes to less than 1% of AUM, but even in Q4 accounted for 13% of GNPA and 5% of overall credit cost in Q4.
This book will principally just wind down to less than INR 1,500-odd crores by September '26, it will lead to further improvement in GNPA and should result in lower credit cost for FY27. We expect FY27 to be probably the busiest year from a FinAI transformation in terms of outcomes, inputs and outputs as a firm. We have 203 people in the firm who are dedicated as the AI unit. We are expanding that to 363-odd resources by June'27. <strong>52 million</strong> voice-to-data conversion happened, around 2.3 million text-to-data conversion happened. Almost 100% of our videos are now AI-generated. 27 AI, voice and text BOT's are now live and all customer engagement is going into BOT interface by June'26. We send 260-odd million pieces of communication on a monthly basis. In FY27, we expect to deploy 600-plus autonomous agents and across operations, DMS, HR, technology and risk. At a core-to-core level, it was a strong year for the company across all key metrics. We crossed 2 milestones, INR 5 lakh crores AUM and 50 million new loans. The year so far has started off pretty well. It's likely for the first time, we've never done 5 million loans in a month. Our last peak was 4.6-odd million loans, and that was in festival period. Overall, last year, we booked 52.5 million loans and added 17.5 million new customers on a full year basis. AUM grew 22%. PAT grew 24%, ROA, and opex to NTI showed an improvement of 36 basis points despite rapid expansion of gold loan branches. The Board of Directors have recommended a dividend to shareholders of INR 6 per equity share. So that's a 600% dividend. The FY27 assessment is based on expectations of easing geopolitical tensions and macro stability. In terms of customer franchise, we continue to be confident of adding 15-17 million new customers in FY27. In terms of AUM, we're confident of a 20% to 24% AUM growth, aided by new businesses that we've launched in the last few years as they begin to scale. NIM, we expect marginal moderation. On opex to NTI, we expect to see continued improvement of 25 to 40 basis points from current levels. On credit cost, this is a new credit cost metric, which, as I said we just reported a 165 basis points or 1.65% credit cost to AUM, the revised metric. We expect this to continue to trend down. In terms of profitability, we are pretty optimistic about the profit growth. ROA came in as of fourth quarter at 4.6%. We continue to guide for 4.4% to 4.6%. Return on equity came in at 20% in the current quarter, we continue to guide for 19% to 20%. And GNPA and NNPA, we expect them to remain range bound and well within our long-term guidance. Just quickly on BHFL, company has already reported its performance. They had a good quarter on disbursements, AUM, asset quality, opex and profit, their disbursements and AUM growth remained strong at 23%. Their loan approvals are up a record 37%, opex to NTI continued to improve for them, came in at 19.2% as against 21.8%. PBT for them in Q4 grew by 20% and PAT excluding onetime reversals, grew by 20% and asset quality continue to remain very, very healthy with GNPA and NNPA at 27 and 11 basis points respectively. BFSL though small, continues to grow, delivered a 77% growth in AUM and their profits were up 50%. We added 124,000 new customers in Q4 and delivered an ROE of 10.5%.
Yes. And just to help you with the math on <strong>4.6%</strong>, the guidance metric for the next year on ROA. And that's the reason why Rajeev when he presented the financial statement and numbers, he said, look at the core financial performance. If you look at core financial performance for last year, the ROA came in at 4.6%. And the guidance that we are giving for next year is 4.4% to 4.6%. I don't think that there will be a need for NIM expansion in the next year to be able to deliver that. We have to ensure that we grow nicely, which is where the 22% to 24% guidance has been given. And we have to ensure that we remain reasonably anchored on the operating performance of the company, which is opex to NTI as a metric and loan loss to credit -- loan loss to average AR. As we deliver that, I think there is enough and more headroom available in the ROA profile of the company.
Cost of funds improved by <strong>4 basis points</strong> in Q4 and came in at 7.41%. Deposit book came in at INR 68,533 crores. It contributed to 16% of consolidated borrowings. Opex to NTI, going back to the earlier point that I made, came in at 33.8% in Q4 as against 33.6%. These are new numbers. Otherwise, it came in at 33.2% an improvement of 10 basis points versus Q4 last year. There was a marginal increase sequentially, primarily due to cost of new Labour Code and accelerated gold loan branch expansion.
As regard the ECL provisioning across Stage 1, Stage 2, Stage 3, this is a bottoms-up exercise that we do every year. We are moving from yearly exercise to quarterly exercise. These are bottoms-up calculation done by product, where we are ensuring that we remain true to the model. The idea is to focus on resiliency of the balance sheet and the provision coverage ratio. I have reason to believe the numbers should remain in this corridor. We get opportunity, we will further up the NPA on provision coverage ratio to ensure that we are bullet proofing the balance sheet from any kind of potential impact that one could witness maybe 3, 5, 7, 10 years down the line.
Watch next