Bajaj Finance Limited (BAJFINANCE) Earnings Call Transcript & Summary

Ladies and gentlemen, good day, and welcome to Bajaj Finance Q3 FY '21 Results Conference Call hosted by JM Financial Institutional Securities Limited. [Operator Instructions] Please note that this conference is being recorded. I now hand the conference over to Mr. Karan Singh from JM Financial. Thank you, and over to you, sir.

Thank you. Good evening, everybody, and welcome to Bajaj Finance's earnings call to discuss the third quarter results. To discuss the results, we have on the call Mr. Rajeev Jain, who is the Managing Director; Mr. Sandeep Jain, who's the Chief Financial Officer; Mr. Atul Jain, CEO, Bajaj Housing Finance; Mr. Anup Saha, Deputy CEO; and Mr. Deepak Bagati, who's President, Collections. May I request Mr. Rajeev Jain to take us through the financial highlights, subsequent to which, we can open the floor for Q&A session. Over to you, sir.

Thank you. Thank you, Karan. Thank you for hosting us. Good evening to all of you. I have 2 more colleagues that I missed. Ashish Panchal, who's my Head of Strategy and runs set of businesses; Fakhari, who's our Chief Risk Officer. Okay. I'll be referring to the presentation that's uploaded on the Investor section of the website of the company. I'll be running through 7, 8 pages. I hope to take between 20 and 25 minutes to go through the -- to points that we have outlined for you. And post that, we'll be open for Q&A. Let's quickly jump to panel #4. If I look at the quarter that went by, in our assessment, as a company, it is a quarter marked by granular business recovery. Essentially, all lines of businesses have. So growth has been structural across, other than auto finance business. And given the kind of credit losses that they are taking, it's ought not to grow that business. Otherwise, been reasonably structural. Significant improvement in risk metrics in Q3 and tracking implementation of our business transformation plan, which I'll cover in some degree of detail, and we have provided an update on that. And putting into motion a plan for pre-COVID growth and financial performance from next fiscal onwards. I just want to remind the objective of us taking some of the actions that we're taking, including front-loading our loan losses is with the purpose of ensuring that as we exit this fiscal, we don't have to think about 2021 for us as a company. As a result of the plan that we laid out and which we've been putting into motion over the last 3 quarters, AUM came in, in the current quarter at -- on a year-on-year basis 1% down at INR 143,550 crores. OpEx to NII was better on a year-on-year basis, was 32.3% versus 33.8%. PAT was down 29%, and there are a set of items that I'll cover, which have caused this. ROE on an annualized basis is looking like 12.9% to 13%. And net NPA, this is a pro forma -- on a pro forma basis came in at 0.19%. This is the -- based on the Supreme Court -- Honorable Supreme Court status quo, it's at 19 basis points. But otherwise, it's at 122 basis points. Very quickly on to Panel 5. I talked about AUM. Core AUM growth, essentially, if I had the IPO receivable that was there in last quarter of tad below INR 1,000 crore plus ECL that we take, if you add that, the core growth in the quarter was just a tad below INR 8,000 crores. On an average, if I take last 4 quarters pre-COVID, the number used to be between INR 9,000 crores and INR 9,500 crores every quarter, we think at a fundamental level, as we exit by -- in Q4, we should start to get closer to anywhere between INR 9,000 crores and INR 9,500 crore for us as a company as we get into Q4. AUM growth is back, in general, as I said earlier. It is a secular recovery across all lines, other than AF. We booked 6 million loans against 7.67 million loans. Most businesses are between 85% and 100% of last year volume in Q3. Even in Q3, if I look at, let's say, fundamentally between October and December, if I look at point number four, while on a quarter basis, our B2C businesses, SME, rural B2C and mortgages were 81%. If I look at December, they were at 90%. So clearly, we are quite confident as a company that we'll get back to INR 9,000 crores and INR 9,500 crores or maybe a little higher growth as we exit -- as we enter Q4 for us as a company. In Q3, the urban consumption businesses were at 86%. That's in volume terms. On value terms, they're at 90% to 93%. Rural at 100%, in value terms, 103% and 104%. Credit card origination is at 102% on a year-on-year basis. E-commerce at 107%. AF, I talked about, given the kind of credit pressures that they're in, at 62%. And that's likely to remain so for a little while. Mortgage disbursements were at 90%, but the AUM growth was structurally lower on a year-on-year basis. In Q3, they grew by INR 770 crores against INR 3,700 crores in Q3 last year, mainly caused by significant portfolio attrition caused by the kind of pricing that we're seeing between the top 3 lenders in the mortgages space -- 4 lenders actually in the mortgages space. We've taken the pricing action, and we're starting to see most of the INR 770 crores growth that you're fundamentally seeing essentially came in little higher, came in, in the month of December. So we should go back to a stronger growth momentum in terms of net AUM as we get into Q4. Commercial business grew 15%. LAS business degrew by 22%. We are quite hopeful that commercial business will continue to -- given how well the businesses that we lend to have done through this phase, we are quite confident that even next year should be a good year for them. Company acquired 2.2 -- that is, I think, a good -- good part of the quarter was that we acquired 2.2 million new customers, given the environment. Overall franchise stood at 46 million customers. And the best franchise stood at 25.35 million customers for us as a company. Overall, franchise grew by 15%. But net growth, because a whole lot of customers slipped out as well as a result of delinquency and default, so gross formation was -- or addition was 15%, net addition was 8%. Existing customers contributed 64. In general, they contribute 68 to 70. That's an area that we're working on, and we should make progress as we get into -- in current quarter or into the next fiscal. Overall, the new origination, clearly that we're looking at is looking significantly better than the pre-COVID origination. Whatever is coming through the door is anywhere between 20%, 30% lower than what used to come through the door pre-COVID. But it's only all but natural given the kind of cleansing at a financial system level that's happened. Just as an update, we have received -- we've talked about a second partnership for credit card business, given our ambition to be among the top 3 or 4 card issuers in India. We've received -- we have partnered with DBS Bank, and we have received RB approval to distribute co-branded credit cards in partnership with them as well. RB will continue to be a strategic partner for us as a company. We continue to believe that we will grow that relationship, and this is a structural addition rather than anything else from an ambition and from an orientation standpoint. NIM quickly, so that's on balance sheet. On NIM, overall margin profile remains pretty steady at pre-COVID levels, other than mortgages, which I talked about. Net interest margin was lower, mainly caused by, as you can see there, the interest reversal of INR 450 crores. That on a year-on-year basis used to be at INR 83 crores, so that's INR 360 crore interest reversal. There will be some degree of slip-through in quarter 4 as well. But from Q1 onwards, that should go back to INR 83 crores, INR 85 crores -- between INR 83 crores and INR 100 crores on a quarter-on-quarter basis. Cost of surplus liquidity, we started to wind down. That's the next section I'll cover. It was at INR 213 crores. It was at INR 83 crore. If you actually look prior in Q1 and Q2 of last year, that number used to be INR 50 crores. So even INR 83 crores was an elevated number. This number either will go back to INR 83 crores or may actually even go lower as we get into the next fiscal. That brings me to liquidity management. So clearly, both those actions on top, both being in a way, have stress-tested the model as they go back -- as they revert back to their normalcy, should add to the NIM on a -- which is how it was prior to COVID. Liquidity, we remain sufficiently liquid, represented 11.6%. Given the favorable and stable market conditions, we have now started to dial down our liquidity buffer. If you recall, last quarter, it was 18%. We've dialed it down to 11.6% -- or 22%, sorry. We'll go back to our 7% to 8% liquidity buffer that we used to maintain as a company by March '21. To drive that, we already started to action. In the last 45 days we've paid down close to INR 6,600 crores of prepayment is what we have done between middle of December and early January. As a result, the consolidated cost of funds came in at 7.78%. It will further go down to 7.5% by March '21. It also assumes that CP book which used to run at 7%, 8% -- 8% to 10%, today is virtually running at 4%, 5% -- 4%. It will also go back to normalcy. We have also tested one more model very clearly over the last 9 months or especially in the first 6 months, that the company has a natural hedged liquidity profile built in. Given the kind of rapid churn of the balance sheet, we do generate anywhere between INR 5,500 crores and INR 7,000 crores of natural liquidity in the balance sheet on a month-on-month basis. So I think that's something that's got tested very, very hard through this crisis as well. Deposits continue to grow as part of our intention to granularize, came in at just a tad below INR 24,000 crores, year-on-year growth of 18%, and stood at 19% of consolidated balance sheet. We expect to end the year between INR 25,000 crores or INR 25,500 crore for us as a company. Majority, 77%, 78%, 76% of that is retail, 24% is corporate. That ratio will keep growing to provide greater stability. OpEx came in at INR 1,390 crore, lower by 9% Y-o-Y. So NIM is down 5%. OpEx is down 9%. So clearly, our actions on OpEx have helped. We were quite brutal in the way we managed OpEx. And we think it is the right thing to do, despite a INR 60-odd crores of higher recovery commission costs, given the flow-through. So it's actually down by technically 11%, 12%. It was also, of course, caused by lower business volumes. There's no denying that. But some of the actions that we took have now been institutionalized as we go ahead and should deliver sustained savings. As I've said in Point 18, some of the costs are transient, like employee costs, we're starting to roll back. We rolled about majority of that in -- between November and December, and we will roll back fully from February onwards. But some of the business transformation are changes in OpEx are structural in nature, like call center costs, travel costs, A&P costs, training costs, they are not going to come back or may come back only to the extent of 40%, 50% from where they were pre-COVID. Let's just jump to credit cost, the elephant in the room. Loan losses were at -- bagged at INR 1,352 crores against INR 831 crores. We also did a principal write-off of INR 1,970 crores after looking at the -- applying various lens as a company in terms of clients who should be written off and who -- so various tests were applied by risk and by finance and by independent auditors to write-off INR 1,970-odd crores on account of COVID related debt. As of December, we're still carrying INR 800 crores of overlay provision as a company. Loan losses, just to highlight, we are continuing to roll to front-load basis a lifetime loss estimate rather than a flow-through estimate. That's just for highlighting the point. We do expect that as we complete the current fiscal, that in the next fiscal, we will go back to what has been our average 150, 160, 170 basis points of credit cost of average assets is what we expect to go back to from -- starting next fiscal. And this is not accounting for how recoveries would pan out. If recoveries are better in FY '22, it's difficult to pin a finger on it at this point in time, we may experience in the next fiscal a lower net loan loss to average assets. But at this point in time, the only thing that I'm looking at is how do we go back to -- revert back to average loan loss to average assets rather than anything else without accounting for recoveries. Recoveries will be addition. We did experience, quite honestly, I mean, it looks like we've seen many cycles in 9 months fundamentally. April was in a way one cycle. July was another cycle. And October and November, December was another cycle. In 9 months, it feels like, as a financial service professional, we've seen 3 cycles at a fundamental level. The degree of volatility of up and down is just quite unprecedented fundamentally and not experienced by anybody. I can very confidently say that. At this point in time, clearly, we -- so we continue to observe significant improvement portfolio quality in Q3 -- in Q -- in collection efficiency in bucket 0, I do read some of the analyst reports which talks about general collection efficiencies, I would just guide that there are bucket 0 efficiencies, bucket 1 and 2 efficiencies, and there are bucket 3 and above efficiencies. That's really how fundamentally we internally look at it. When we look at those clients who are current and they went into default, which is called bucket 0 efficiencies, they are actually back to pre-COVID or better than that, okay? So that's level 1 point. The early bucket collection efficiencies, clients who are in bucket 1 and 2, because of the kind of share flow through that's happened from 0 to 1 and 1 to 2 are significantly better than pre-COVID, which is really what is resulting in lower credit card guidance, which is really -- even in bucket 3 and above, we are experiencing the same thing. We have -- if it sustains, it could fundamentally mean lower credit cost. But I would wait for some more time before I could confidently say that it will result in lower credit costs in the process in the next fiscal, but I would wait for a quarter to make that point. Overall, we have provided a guidance of that in Q2, that whatever we have taken we are residually left in our assessment with anywhere between INR 2,600 crores to INR 2,900 crores. At this point in time, we have taken since then in this current quarter, INR 1,352 crores. We think the fourth quarter residual credit cost in our assessment could be INR 1,200 crores and INR 1,250 crores. As I said earlier on the previous point, if they continue to be markedly better, the number could be INR 50 crores plus/minus. Given we've gone so far with it, we, in general, our management view would be, we might as well take it and we can -- if it comes back, it would drop into the P&L as we get ahead. I'd rather add the 95th percentile, not -- or 97th percentile, not get to wait to the 100th. So that's really where we are fundamentally headed. And I've articulated that if they remain better, there could be reduction, but the reduction could be plus/minus INR 50 crores to INR 75 crores. Gross gen -- 24th point is the resolution plan. At this point in time, INR 2,040 crore, of that just INR 950 crores is mortgages. INR 523 crores will get cleaned up essentially -- or sorry, against that, INR 2,040 crores, we are holding provisions of INR 400 crores overall. Mortgages and AFR secured accounts, B2B, it's one single large account, which is part of the resolution plan. So that's the resolution plan framework. Gross NPA and net NPA, 25th is -- while it's a published number, given the interim order of Honorable Supreme Court, if you took a pro forma view, fundamentally, we are at this point of time, ending Q3 at 286 basis points of gross NPA and 122 basis points of net NPA. If you -- if we break that up, the B2B business net NPA is up by 7 basis points post the cleanup, SME is up by 48 basis points, B2C is 89 basis points and order finance businesses 348 basis points. Fundamentally, other than auto finance business, which has an underlying collateral, we expect between Q4, Q1, we should largely be reverting back to the pre-COVID levels of net NPA for us as a company is really what our internal risk models at this point in time are saying. Jumping, if you aggregate this entire conversation between balance sheet AUM growth, NIM, liquidity, OpEx and the credit costs, takes us to the last point of profitability. Overall, contracted 28% from a pretax standpoint to INR 50 crores and INR 55 crores, mainly caused by INR 520 crores of loan loss provision, interest reversal of INR 367 crores and additional liquidity cost of INR 130 crores. I've talked about the actions that we are fundamentally taking in each one of these lines, and they should result into natural addition to profitability as we get into next fiscal for us as a company. I don't think there is any in our -- at a fundamental level, that's really where they're headed to be at a design level. Capital, clearly, we are good for next 3 years from a growth -- as a company. The Tier 1 capital itself is at 25%. So we are very, very well placed. So far, in Q1, we talked to you guys about -- talked to street about zero-based budgeting. We provided a short update on business transformation that we are up to. This is the first time we are fundamentally articulating over the next 5, 6 panels as to what we are headed to become over the next -- or how the company will run over the next 5, 7 years. To us, this is not a -- this is a fundamental change in the way company will be run. So under the hood, there is a lot changing. The point #1 that I would fundamentally want to articulate and once the hood changes, everything else over a period of time will change any which ways. Where we are headed is that clearly, COVID led us to structurally question the way our business is conducted, not run. The way it's conducted and where we are headed is that we want to build for us as a company, the only channel infrastructure in this country for financial services products and services. It's not -- so that's the position that we are taking as a company because we are a regulated business, we are not a pure payments business. Regulated business, not a payments business, which means omnichannel is what would work. You will need to transition between offline to online in a seamless manner, is really on a realistic basis, given the nature of our business, that business will run. And that's really what we are up to. So there is a consumer side of the business, and there is a, within the company, how data will move and how apps will get -- multiple apps will get created to deliver to the consumer an omnichannel experience. It was very clear that if you want to do this, it's not about building an app. Most people mistake it for that. The structural transformation is in changing our operating processes. If you have to change your operating processes, followed by that, you have to change your core technology stack. And that's really what we've been up to. That process will get completed. That's underway. That process will get completed by May '21. What that will lead to is to creating apps. So app will follow. The process, the hardest part in this entire frame has been changing the way processes would run, which means customer first rather than company first. I think that's the fundamental change in methodology or approach that we are headed to build. And that means the stack itself is customer first rather than company first. That is the hardest part in this entire transformation that we have to deliver. Point #31, I talked about omnichannel customer. As I said, we are in a highly regulated business. Customers, in a click of a button, 10% of customers may get money, 90% will need a call in 5 minutes. 90% will need that somebody calls them and takes them, helps them complete the journey in the next 5 minutes, so that the loan happens in 5 minutes. That's really where the omnichannel framework starts to play out. We are in the process of developing or significantly transforming 4 large productivity app in ecosystems within the company, they are inward-facing. But they will seamlessly talk to the customer asset and employee assets on a real-time basis, which will be sales one app, merchant app, collections app and partner app. And so that should also go through by May '21. We do believe, fundamentally, that once these are delivered, this will require much lower headcount addition as a proportion of growth as we get into the next few years. And in our business, if I look at our P&L stack, linear -- linearity of headcount addition to growth is something that always troubled me at a design level. We deliver this, that would dramatically get solved. That does not mean headcount addition will not happen, but the nonlinearity of headcount to growth will go through a significant transformation as we get into the next fiscal. We will launch Bajaj Pay for consumers sometime in Q4. We will launch Bajaj Pay for merchants, mainly meant for our merchants. We are not looking at competing in where the big elephants are fighting. We have 103,000 merchant ecosystem. We move 7%, 8% of their commerce. We think we have an opportunity to double that volume in the medium term, just staying with that organized ecosystem so that we further ring-fence it. Within a single app ecosystem, customer will be able to access, we are creating 5 proprietary marketplaces. EMI Store which existed, but now "it is integral" to our B2B business, the way customers will start to experience it in the next 30 days' time. We're building an insurance marketplace, which will go live by -- along with the overall new ecosystem, we are building an investments marketplace, and Bajaj Finserv Health is already live. These are 5 proprietary marketplace ecosystems using our group companies and the broking app in the last half that we are creating which should allow customers to buy whole lot of financial services products and services through a single app ecosystem across electronics, insurance, investments and health. In addition, we've identified, what we call, adjunct Affinity app ecosystem that we will integrate between now and May as we build this out. I think the first phase of this will get launched by July, second phase by August and third phase by September end. We think by September end, we should deliver what I would call a reasonably solid business. Having said that, let me make one point that it's not -- by next year -- our next year growth as a company is not contingent on it. This is the way company will be run. I think I want to emphasize and reinforce that point. This is not for next year. This is the way company will be run rather than needed for growth at a fundamental level. We will use the second half of next year to optimize, optimize, optimize, sharpen, sharpen, sharpen. And this is really how we will, as we go ahead, run the company just as the last point. This is a new section we've added. Clearly, if we have to build a moment of truth business for us as a company and we believe that 46 million customers someday will go to 75 million or 100 million customers in this country and our strategy is required in cross-sell, we got to deliver much better customer experience, that's one; much reduced friction, number two, so that purchase and post-purchase experience improve. So every quarter, henceforth, we will share as to what we are doing on customer experience for us as a company because at the design level, we got to structurally change this piece of the hood to deliver a moment of truth company for us. So there are a set of actions that are here. Point number 37, I covered. Clearly, the self-service infrastructure will see significant augmentation as we deliver the business transformation. The largest part of the rate actually will be on self-service infrastructure for most, if not all. Some of the items parallelly to -- at a fundamental level that we're doing is like an IO process that we have created to strengthen customer grievance. He also has a mandate to tell us what changes that we make so that customer friction reduces. Call monitoring infrastructure, we fundamentally have created now all calls by March that are made by our collections agents fundamentally will be recorded calls. At -- that's really where we are headed. We are already at 35%, 40% of the calls are recorded fully. They will get to 100% at a design level. We're proactively moving into DRA certification. That's something that we're doing as well. So that's -- we've now moved, we've taken a decision that NPS was used to gauge customer experience. We are headed to gauge it for our collection processes. We are already seeing 3%, 4%, 5% response rates from customers on this. So clearly, we are changing a lot under the hood. It's been a terrible crisis. But I think it has put through some of the structural changes that we needed to make, to create next 10 levels of growth momentum for us as a company. Let's move. Let's jump to Panel 19. The only business that we fundamentally added through this period is medical equipment financing business. We think it's a large opportunity across SME and commercial. In the SME now and in the commercial in the medium term, for us as a company. We've done B2B tires, and we are now disbursing INR 15 crores or INR 18 crores a month. This should grow as we move ahead. Panel 35, very clearly, just on liability mix, from 13 at 0 to overall, on a stand-alone basis of Bajaj Finance, deposits are at 24%. On a consolidated basis are at 19%. ECB is at 6%. Move to -- that's Panel 37. Customer franchise. I talked about it. At the top of the funnel, grew by 15%. At the bottom of funnel, grew by 8%. And we added 2.2 million customers. Panel 45, that's pro forma. This is not the -- as I said earlier, this is a pro forma number rather than the reported number. This is also reported, but this is a pro forma reporting. At a design level, as you can see, the largest impact has been in the auto finance part of the business. Rest of the businesses, we expect between Q4 and Q1 to revert to pre-COVID levels from a net NPA standpoint for us as a company. That's -- I'm left with last one panel. This is on Panel 48. This is stage-wise provisioning. Gross assets -- I mean, let me just take 2 minutes to make -- to provide you with level of clarity on this panel, so that we are all on the same page. Fundamentally, what you see below is Stage 2 and Stage 3 provision, Q3 versus Q3 last year. If you add INR 5,892 crores and INR 4,194 crores, that adds up to INR 10,000 crores of Stage 2 and Stage 3 which used to be INR 6,000 crores of Stage 2 and Stage 3, gives you -- gives us a differential of INR 4,000 crores of Stage 2 and Stage 3 differential. Against that, this Stage 2 and Stage 3 provision itself is INR 1,800-odd crores. That leaves us with INR 2,300 crores of plus/minus INR 50 crores, INR 2,300 crores of residual provision for us as a company. As we guided, we will take INR 1,250 crores in the fourth quarter. That's one part. We expect majority of that, as you can see, is essentially coming from auto finance. You will see recoveries of close to INR 1,000 crores, should take us back closer to -- we have additional Stage 1 provision, as you can see here in this panel, of INR 1,405 crores versus INR 770 crores. That's a INR 700 crore Stage 1 provision that's fundamentally sitting there. So that's the last point. Largely, if not more, should more than cover, at a design level, for us as a company as we take Q4 provision, that we are fully secured and not accounting for recoveries in next year. If any certain impact was to appear in any one line of business, the recoveries from next year should be fully -- should be reasonably sufficient to take care of any -- from a mitigating impact standpoint. So that's really the full recon at a design level between Stage 2, Stage 3 asset last year versus this year, what we will do in Q4 and where we'll get to by Q1 for us as a company. That's what I would call 105% recon, not 100% recon on as a company. Last panel. As you can see, the B2B businesses have started to revert back to -- in Stage 1 to pre-COVID, 2-wheeler and 3-wheeler is what we'll take 2 odd quarters. B2C will take -- B2C and SME have some distance left. Will happen in -- as a result of the provisions that we'll take in Q4 is really what this fundamentally reflects. That's the quarter and some that will give you guidance on Q4, so that we can all hopefully forget about the current fiscal and get back to some degree of normalcy as we get into next year. That's for me, happy to take questions.

[Operator Instructions] question is from the line of [indiscernible] from Elara.

Question is, sir, what was your stock of flexi loans at the end of the quarter? And how much was converted during the quarter?

Sandeep Jain is a flexi expert. Let him answer.

I think flexi is history now. I think we are not converting too many existing terminals into flexi.

Since Q2.

Since Q2. Q2 was the last quarter when we had decent amount of conversion that had happened. Otherwise, from quarter 3 onwards, what we are doing is, we are acquiring flexi in the normal course of business across salaried personal loan, business loan, loan against securities, loan against property, so on and so forth. That's purely new origination that we have been doing in past as well. As far as the portfolio quality of flexi is concerned, we are seeing no difference in the portfolio performance versus what we are observing pre-COVID level. So absolutely nothing to worry about on the flexi side.

Okay. But the stock would be similar to what was in the last quarter?

Would have marginally gone up because of the new acquisition that we have done in the current quarter, which is the new business underwriting that we have done.

Got it. And in terms of write-offs, what was the segment that saw the maximum write-off?

The write-offs were across businesses, across auto finance and particularly 3-wheeler business, where we have chosen to write-off certain set of customers where the receipts were not coming for the last couple of months. We had B2C businesses wherein we had customers who have not paid us for the last 5, 6 months, including moratorium period. Similarly, we had customers in B2B, who had not made the repayments in the last 6, 7 months, including the moratorium period. So the customers who had shown last 6 months of non-repayment track record, these were the customers we assume were badly impacted by COVID-19 and may not be able to come up to the terms in terms of repayments even in quarter 4 and subsequent period, and we chose to write them off in the current quarter itself.

Next question is from the line of Antariksha Banerjee from ICICI Prudential Asset Management.

Can you hear me?

Yes, we can hear you.

So on your Panel 39, where you give the distribution reach. I think there's been a drastic reduction in the retail spend stores, so -- and some of the other stores as well. Can you tell us what's the rationalization strategy followed here?

So fundamentally, look, we have observed closures as well. So in -- say, in digital product stores, the 1,000 number is 1,000 stores have closed. So that's it, right?

Yes.

So in some places, it's closed. But the larger part of the conversation is, if you recall that the where fundamental -- so that is one -- that's 1,000. So just at a macro level, the digital products has enclosures. So you should read 1,000 as having closed. But the biggest chain here is the retail EMI card stores, which has gone down to 15,000 stores. If you recall in Q1, we had paused. In Q1 and Q2, we paused that business. We had articulated the REMI and Wallet are the 2 point of sale businesses that we are -- that we have not restarted. We have restarted retail EMI cards, but we are focusing on an average ticket size of INR 15,000. So wherever the ticket size is less than INR 15,000 on an average, that's where we are not servicing them at this point in time. So that's the 15,000 difference.

And that's likely to remain from here, if I understand?

Once we launch, let me make a point -- once we launch merchant, and that's really what that team is working on, that as we launch Bajaj Pay for merchants, we intend to bring this back. But in a completely new design, that was leading to see -- we did not shut it because we had a credit cost problem. We shut it because the model was becoming completely linear. We cannot then serve 200,000 merchants if we had a design on retail EMI card stores for which we will have to do linear staffing. As Bajaj Pay for merchant emerges, it creates a nonlinear way to grow this pool is really what we'll wait for and we'll build that out from second half of next fiscal onwards.

Any rough idea of how much cost would have been saved due to this alone?

No, this doesn't -- I mean, this is a -- this is not a save item that way. This is a restructuring item. This is restructuring the business model to serve for the future. You want a specific number, the number will be INR 10 crores, INR 12 crores in a quarter. That doesn't -- we don't know -- it's not to reduce cost. It is to structurally ask a question that how will it be run if it was to scale.

Got it. Okay. The second question is on the DBS tie-up. So just to understand the broad contours of the way you do this credit card business remains the same with this partner as well. Is that right?

Yes, very much.

Okay. And just last one, data keeping question. What are the incremental deposit costs that you're incurring on whatever you're raising increments?

That's -- you mean wholesale, retail? Retail is published, but you could raise today, you could place a fixed deposit with us at 6.6%.

Yes, 6.6% is the pricing that we are offering for 3-year tenor. Blended one, blended pricing at a different tenor level would come at 6.45%, 6.5%, yes.

6.5%.

The next question is from the line of Kunal Shah from ICICI Securities. .

Yes. So particularly with respect to this RBI's ownership and the corporate structure guidelines or maybe conversion into this banking, what would be our view? And finally, maybe there are -- maybe RBI has indicated that there would be some kind of a change in the structural framework for NBFCs. So how prepared would we be with respect to some of the aspects, wherein clearly, we see the regulatory arbitrage being there between banks and NBFCs? So how would we structure, definitely nothing is out as yet, but how are you particularly looking at it?

Kunal, your question is the answer. Once it is out, at least I'll be able to respond to the question. Otherwise, it would be hypothesis. And based on that conclusion, it will not be fair.

But generally, in terms of the operational flexibility, do we see a lot being there under the NBFC structure, which might not be there under the bank?

No, I don't think so. I think -- no, I don't -- I believe that the arbitrage will continue to just go away. And if that's -- because I think they've constantly been harmonizing. I think that's the word constantly being used. They've been harmonizing the arbitrage. And I am fully supportive of that. I believe that for the growth -- sustainable growth of the sector, it cannot run our arbitrage. It should run on its own feet and should not purely be built on arbitrages. So I am fully supportive of whatever decisions they make. We will optimize our business model, and it should be with minimum friction is what I would believe to all stakeholders.

Kunal, RBI has always come out with guidelines, which made the sector more resilient. I think that's what RBI is currently looking at as well. And we welcome the move that RBI will make on this count.

Sure. Okay. And secondly, in terms of this distribution, which we highlighted, particularly, say on the digital side or lifestyle stores, but does it any which may impact the servicing or maybe -- because you said it's not based on the customer behavior and it's purely in terms of the closures. But do it -- does it impact in terms of the asset quality to us or no, not really? Maybe even if it's shut down, it doesn't matter in terms of the behavior of the customer.

Yes, yes. So in a way, I have a view that when you see 1,000 digital product store shutdown, 2 things. Consumer durable, you don't see much movement because they are harder to build and shut. Okay? There is a digital product. In general, we do observe that in good times, more much room. And in bad times, they shut down. The entry barriers and exit barriers are very low. So this clearly reflects that right now, it's a bad time. But 12 months down the line, as times turn, you may see it may increase again. You walk in, whether you walk into Vijay Sales or you walk in, in general, to a smaller store, our underwriting models in general don't change to the extent of 95%. So really doesn't change. But there is evident consolidation, at least in consumer durable, in my mind.

Sure. I understand that. But in terms of our customer behavior, okay, which we would have originated from those stores, is the behavior different from the shutdown stores and existing stores or it's?

No, no. That is not different. No, no. Among the top 3 retailers, I can tell you, one of them has -- is below our national average. Now -- so that's the -- and they're one of the largest. So it really doesn't matter.

Sure. And one last question in terms of restructuring, is there a further pipeline wherein we are still yet to, maybe we would have evaluated it, but yet to finally get restructured or it is the -- or this the final number?

Yes, this is the final number. Actually, since 15th December, we are -- we have not seen -- I mean, we are done.

Okay. So there are no additions to it. Okay. This is helpful.

The next question is from the line of Kuntal Shah from Oaklane Capital.

You have said something that it's a moment of truth for the company, which is equivalent to a crisis on whose outcome everything depends. You also said the digital initiatives are not contingent to the growth. So basically, you are relying on the process improvement to ride out through the crisis, is that the correct understanding? And can you explain this in detail?

Yes. Yes.

And secondly...

In fact, they will have a bigger near-term impact. They will have a bigger near-term impact.

Okay. For a couple of quarters?

Yes, by -- and as I said, by May. In fact, I believe, in the next fiscal, the bigger impact of that will come from, I made that point, through the productivity apps that we are creating rather than the consumer app. And that's where the omnichannel frame is, in a time like this, just got back to business, if we can start to generate 2.2 million customers in a quarter, you should safely assume that we can -- or doing 6 million loans, we'll be back very soon to doing what we were doing pre-COVID. That's omnichannel frame, but to serve them is really what we see a 3 in 1 frame will be. To mind their wallet is really what we see 3 in 1 or the business transformation frame will be. So I think that's really where, Kuntal, we are headed.

And one question I had is, when we compare your numbers and your front-loading of credit cost, 4.1% on nonconsolidated basis versus others adjusted for the size comparable. Most of the financials choose to write-off losses over time and provide as their own. I understand you want to get rid of it and start with things late, but the difference is just too start to ignore, like can you just highlight that is it overcautious stance? Or you are seeing that kind of thing and others aren't?

I can't speak for others, Kuntal. I can speak for myself. These decisions are based on millions of customers. We bank 14 million customers at this point in time. It's based on that we're making decisions. So this is not based on a -- this is a very, very large sample strategy across 2,000 cities in India. And we are making database and fact-based decisions. We do believe, philosophically, that P&L must fully reflect the state of the business. That's a philosophical point. The earlier point is a fact-based point. And I can't speak for others. We want our P&L to fully reflect at any point of time the position of -- or the state of the business.

But Rajeev, since you don't have a banking license and you bank with all the other major banks, your customers, their customers overlap. So where's the disconnect?

Kuntal, you're asking me or telling me? Because if you're asking me, I can't -- I look at private sector banks, nationalized banks, we do lend to some customers who have corporative making accounts given our deep distribution. They -- I've not seen a structural difference. We last did bureaus we work with to see whether our customers, are they defaulting with us or are they defaulting structurally or in morat, structurally they went into morat with us or they went into morat with broader banking system. Conclusion was 97%, 98% if they were in morat with us, they were in morat with -- we broke that up by public sector banks, private sector banks. Conclusion was, not reasonably, completely identically on our base. So that's all I would say.

The next question is from the line of Nischint Chawathe from Kotak.

Am I audible?

Yes.

Okay. So Rajeev, when we look at the panels for monthly collections, somewhere you mentioned that December '20 is net of provisions. What does this really mean?

Yes. Sandeep Jain?

We lost in between, Nischint.

Nischint, right?

Yes, yes, Nischint. So when you say December '20 collections are less, which is in those monthly collection efficiency panels, that's the footnote.

Identically leads -- Nischint, identically reads to Panel #45 to these numbers. So wherever it's written off, it's written off against that. Identically reads to that. That reads identically to INR 90 crores, INR 70 crores written off. That shows in Stage 2, Stage 3. Identically, reads to the last 3 panels. That's how -- correct me, Sandeep Jain.

So what you're saying is -- okay, let's go to the page number 49, and what you're saying is December '20 numbers are net of provisions excluding Stage 1?

Yes.

I'm just trying to understand what does this mean.

Yes, yes, yes. Of course. Of course. Because I've provided, I've written off, I have to -- then there's nothing left. That's all I have to do. Stage 1, not accounting for, rightfully, as you said, Stage 1. So Stage 1, INR 1,400 crores, which you see in previous panel -- let's go to previous panel which is INR 1,405 crores is what I would call -- in fact, as that improves, Nischint, and I said earlier, this number has to circle back to on an average asset basis to 60 basis points rather than 110 basis points. So eventually, there is that INR 700 crores fundamentally that's sitting there. In addition, Nischint, as I said earlier, we will take in Q4 anywhere between around INR 50 crores to INR 1250 crores. That's largely, as I said, it's 105% reconciliation.

And if I look at Stage 1, Stage 3 loans, if I look at it on a quarter-on-quarter basis, last quarter was, I think, Stage 2 plus Stage 3 was around INR 13,000 crores. This quarter is around INR 10,000-odd crores, and you have written of approximately INR 3,000-odd crores. So broadly, I think if I have to think about it on a quarter-on-quarter, there is no improvement between Phase II plus Phase 3 loans, if I have to look at both at once?

No, Nischint, we have written off INR 1,970 crores in current quarter, not INR 3,000 crores. So there is INR 1,000 crore to INR 1,200 crores of recovery, which means customers who are in Stage 2 have come back to Stage 1 by making the payment of their installment, not only for the current month, but probably also for the previous months.

And we would expect maybe a higher recovery in the next quarter, and that's where you're kind of confident about the provision guidance?

I won't comment on rollback number as to how many customers or what amount will roll back from Stage 2 and 3 to Stage 1. I won't comment on that. But from the portfolio that we have written off or we may write-off in Q4, do we see certain amount of recoveries to take place in next year? Probably the answer is yes. Can I put a number to it? I can, but probably we'll wait and watch as to how these recoveries work out as we get into quarter 1 of the next year, and then we'll be more confident in terms of giving you a number.

Nischint, let me -- since you have a question, let me reconcile one again for you. Okay? Last year, same time, we were at INR 6,000 crore. Same -- so we are supposed to be at 6, we're not supposed to be 0. That's level 1 let me articulate, right? You're not supposed to go back to 0. We were last year at 6, we are at 10. Differential is INR 4,000 crores. Last year, we had Stage 2, Stage 3 provision of INR 2,000 crores. We have Stage 2, Stage 3 provision of INR 3,800 crores. Okay. That leaves us with INR 1,700 crores. It's all here on this Panel 48. We will take in the current quarter, INR 1,200 crores and so on and so forth. So that's fundamentally on this panel. Full recon. Forget about the recoveries. We are not talking recoveries. We're talking what we will do in Q4, where this number is and where it will go to. We want to get back to 6,000 by Q1, fundamentally. Adjusted was, the only number that we go to Q2 at a design level, maybe auto finance business because there is underlying collateral value that's sitting there. See -- so the INR 4,000 crores differential, even if you take 400 businesses, very little. But take for AF, which structurally over the last 20 years has delivered between 45% and 50% recoveries on repossession, if you take that, as I said, it's 105% recon rather than 100% recon.

Nischint, in noble times, one takes a cut-off at 95% level that if I'm 95% sure that money will not come. Or other way around, if I have 5% out -- 5% only probability of recovery, then I choose to write-off. In the current time, given the extreme stress scenario that we have gone through, probably that threshold has increased from 5% recovery to 10% recovery. So wherever we are seeing that there is less than 10% probability of recovery, rather than keeping it in the GMP and carry it as provision, we'll just write it off.

Sure. Got it. This is helpful. The second question is just trying to understand a little bit about Bajaj Pay in terms of what is the thought process around it and how does it integrate with the overall margin?

PPI business focused on our merchants and our customers. Open loop universal QR infrastructure allows our customers to use the universal QR and the UPI infrastructure for -- to do their day-to-day activities, allows the customer in a single interface to use credit card, UPI, EMI card, depending on the network and QR. That's really what it would do in a single interface infrastructure.

How does it integrate the overall -- the rest of the business in terms of lending?

At a checkout page, you will see these 4 options. Based on the merchant, an EMI may appear. Based on merchant, it may not appear. Based on a reward that we may offer to a merchant, it may appear. Based on a merchant, it may not appear. That's how it will integrate with a broader B2B business. Today, you can walk into a store and use your reward points on a co-branded card and convert into margin. That's how it will integrate into the business. Today, monthly, INR 5 crores to INR 8 crores gets converted from reward points into margin money by customers to buy a product. This is how it will integrate at a fundamental level. And many more ways as we deliver this between July and September.

The next question is from the line of Prashanth Sridhar from SBI Mutual Funds.

Yes. Just a clarification on a data point. What is the amount of flexi loans that will be part of the Stage 3 INR 4,000 crores in the restructuring INR 2000 crores?

I don't have that number handy. I can come back to you separately. But I think once the customers move into Stage 3, it really doesn't matter whether it is flexi or non-flexi. That's the only point that I will put on table. We may have certain set of customers who may be in Stage 2 and Stage 2 with flexi facility from us. But from provisioning point of view, from exposure point of view, they are no different from a normal term loan customer.

Okay. Sure. Fair enough. And we saw the jump in terms of converted to...

Given the numbers which we're at and to the earlier question on flexi, if it was standing out, we would at least have the number, let me make that point. And we tried many numbers, most numbers. If you're not tracking it, you should be reasonably -- you should reasonably believe that we don't think it's the elephant in the room.

That's a fair point. But I think you had spoken about it earlier. Just any feedback on the traction in terms of how much gets converted into flexi this quarter? I think we have the number for the previous 2 quarter.

No way. We are not tracking even that. So that's what Sandeep responded even earlier.

See, see, if you look at the commentary that we have given in quarter 1 and quarter 2, there were significant amount of customers whom we wanted to offer flexi loans who had not availed it earlier. So we had stock sitting out there. We had field forces who did not have enough business in the market to do because of lockdown situation, and that's where flexi conversion took place. I think as we go back into normal business, the flexi conversion is not going to be area of focus. And any which ways, the set of customers that we wanted to convert into flexi product have already got converted.

The next question is from the line of Abhishek Murarka from IIFl.

Yes, can you hear me?

Yes, Abhishek.

So a few questions. One, in terms of the DBS tie-up, the pool of customers that is available to RBL and DBS, is that going to be kept exclusive? Or is it a larger pool and hence the flows that were going to RBL earlier are not going to shrink? If you could give some understanding of the kind of business flows or customer base monetization that you're looking at, it would be great?

Yes. So it would be a little premature, Abhishek. At this point in time, we just received an approval. At first, we have to work with them as a bank to -- they don't have an active credit card business in India. So they have to put together the infrastructure and the technology platform to launch in India. I think that's the first step. So I think that's first thing that has to happen and largely to be done by them. We'll assist them in whichever way we can. I think by the time it gets launched, it will be -- we are in January, maybe by 1st of July. So maybe we will provide an update, Abhishek, sometime in Q1.

Sure, Rajeev. But sorry, just to ask, just conceptually, does this sort of cannibalize the customer flow to RBL? Or does it not cannibalize that flow, the pool that is available to RBL?

No. As I said earlier, Abhishek, the RBL is our strategic partner. We built out with them over the last 4 years a very good business. And they'll continue to -- it's a special relationship and it will continue to be strategic. And it will continue to grow the way it's grown over the last few years. And we see a very large opportunity to be able to serve both the manufacturers as a company for the medium term, definitely.

Sure. Sure. So that's clear. The other question is, Rajeev, when I look at this panel on home loans and LAP and I see the sequential reduction in those buckets that you reported, it's still much, much higher than pre-COVID levels. And I just wanted to know what are the structural reasons for stress in home loans? Because LAP, I understand, could we due to some businesses which are seeing stress. But home loans, considering it is very high-quality underwriting cross-sell to existing customers, why should that portfolio see much higher delinquency?

Because it is the largest outflow for a customer. That's exactly the point that we are -- we all cannot forget that there was a COVID crisis. Wait, Atul wants to make -- So I think it will -- so clearly, you will see. But at a fundamental level, Abhishek, also remember this is super secured. So eventual loss given default as customers go back to normalcy should be -- will be much, much lower. This and AF are the 2 businesses, because they are long tail, in fact, mortgage will take the longest to go back to normalcy to that extent because you have to work through the customer. Even on bounce rate, so far, they're still higher than they were at pre-COVID levels. So we'll have to just work through.

Sure. And last bit, just a few -- a couple of disclosure-related questions. Can you give the movement of NPA? And what is the ECL years outstanding that you have? How much in PCL years 1.0 and 2.0?

ECL years. Okay. Yes

Yes. So Abhishek, I can't give you NPA movement because of the standstill that has been announced by -- or that has been ordered by Honorable Supreme Court. So -- and that's the reason why we chose not to publish that NPA movement walk for the current quarter because fundamentally, I'll not allow any customer to be classified as NPA in the current times. So that's point number one. Point number two, what was your second question?

The ECL years outstanding? The ECL years disbursement that you would have done?

I think that will be about INR 600 crores to INR 700 crores of outflow that we have done.

Outstanding?

Yes.

Okay. And just in the movement of NPA, even though you cannot disclose that, but just on a pro forma basis, the slippages, would that be close to INR 4,200 crores, INR 4,300 crore, if we were just to assume that nearly, in 2019, INR 70 crores was written off?

Yes. If you track the movement in the Stage 3 outstanding and add back to it, the write-off number, that will be the number that will get worked out.

The next question is from the line of Krishnan ASV from HDFC Securities.

My question is a little more structural. You have generally had a lot of wins to your credit from the time that you have been in -- I mean, engineering the turnaround at path. Unfortunately, one constituency that you haven't had great success with is customer friendliness, right? So just wanted to understand, you have mentioned a few things around business transformation and what you want to do in order to become more customer-friendly. But are we reaching a point where you might want to admit that it's not possible to be both customer-friendly and investor-friendly?

No, I don't agree at all. So I mean, just to set the record straight, we run one of the highest NPS in the lending business in India. Structurally, we never published it. Now we start to publish. If you go back to Panel 10. I don't believe that you will make a choice. It's not -- I think -- so that's level 1. We run an NPS -- pre-COVID, we run an NPS of 47%, 48%. 64% of the customers are existing customers. We think it will go back to 68%. So structurally, it isn't -- it's not a choice between either at all, I must say that. You have to be straightforward, you have to be clear. That's -- so I don't think it's a choice. But it's a choice to make. I have said this on Panel 10, if you go to it, that we believe that if we -- see, we actually get as a starting relationship a very small share of his wallet. Let's face it. He's taking 8-month loan at no interest. I'm getting a very small share of his wallet. He's got a personal loan wallet. He's got a home loan wallet. He's got an insurance wallet. He's got a credit card wallet. He's got a used car wallet. I want that part of his wallet. He's got an FD wallet. I want that part of the wallet. If I did not serve him that any which ways, he will not do business with me. So I have to actually perform is not an option, but to serve him right to get that part of the wallet. And that we've done reasonably successfully across businesses for many years. Now should we do much more and that's why the Panel 10 has appeared. Let me make that point that at the size of our franchise that we will get to, we cannot even make -- we have to be far more serious about it as we get it.

Yes. I mean, so the reason I'm asking that -- say, the reason I'm asking that, Rajeev, is, at some point of time, you did have aspirations to be a bank. We have recently had one adverse regulatory development where a large bank was told not to add new customers if you can't be fair to your existing customers. So I'm just wondering -- I mean, is there something that you could do to fix the perception of not being customer-friendly? Maybe you are. But maybe there is a perception that lies out there that you're not very customer-friendly. I mean, can something be done -- is something being done to address that perception?

Yes. I mean, as I said, that's why we do NPS. And NPS tells us that on a year-on-year basis, over the last 3 years, we are making significant progress. So I have to say that -- my experience tells us that our share of our customers wallet have already gone up. So let me -- 3 years ago, we didn't have dedicated customer service branches. Today in top 28 cities in India, I have only dedicated customer service branches. On an average, each branch has 8 to 10 cashiers. There are 10,000 square feet on an average infrastructure. So we can go into detail, but -- I don't know how to fight perception, but when I ask our customers at an NPS level, done independently, that is not what we are seeing. Or that is not what we observed. Because at the fundamental level, let me make a point. What is the core product? The core product is how quickly am I able to give him the core product, which is a loan. If that takes long, I can be whatever, he's not going to like it. If that is quick and reasonable, he's going to like it, and that's why NPS goes up. As friction goes down, we will see performs like it or not NPAs go up. More and more friction goes down, more and more NPS will go up. Less and less questions I ask and more and more I work for him, less and less -- more and more, he's going to be happy. That's really how we all behave, and that's really how customers behave. So now then comes to the next question, which is an investor-related question, what is the price? Prices -- now price between fair, reasonable, low, that's a matter of judgment because therein comes at a fundamental level, what is the risk-adjusted pricing for that customer. So here the conversation starts to get a little more complex, we have to be at it and continue to solve for it.

Great. Just one other question. This was extremely helpful. So just one other question. You have generally been -- I mean you have creamed out almost anything that happens offline in terms of durable financing, mobile financing, electronic appliance financing. I just wanted to understand how is that market share -- how is that share of the wallet or share of the transactions that happen over the online journey? How is that tracking further down?

7%, 8%. That's far largely determined by the new through-the-door customers that are coming offline, level 1. And that's why I said 2.2 million new customer acquisition was an important metric in Q3. Gets converted, in general, 90% of the clients land up taking an EMI card. And of them, those who choose to use, on e-comm, get access to it. That's really how the virtuous cycle fundamentally works. So more and more customers keep coming through the offline ecosystem. More and more customers keep buying EMI card. More and more customers keep using the EMI card based on their choice, whether in offline or in online. So as you can see, in the previous quarter, offline business grew by 86%. E-com grew by 107%. So we'll continue to -- we'll stay focused on the customer rather than on the channel.

Thank you very much. Ladies and gentlemen, due to time constraint, that will be the last question for today. I will now hand the conference over to Mr. Karan Singh for closing comments.

Yes. On behalf of JM Financial, I would like to thank Rajeev Jain and the senior management team of Bajaj Finance for joining us on the call today. Thank you.

Thank you. Thank you all. Thank you. Sorry to be late. Thank you.

Thank you very much. On behalf of JM Financial Institutional Securities Limited, that concludes this conference. Thank you for joining us. You may now disconnect your lines. Thank you.