SBI Cards and Payment Services Limited (SBICARD) Earnings Call Transcript & Summary

Ladies and gentlemen, good day, and welcome to SBI Cards Third Quarter FY '21 Earnings Conference Call. [Operator Instructions] Please note that this conference is being recorded. I now hand the conference over to Mr. Ashwini Kumar Tewari, Managing Director and Chief Executive Officer of SBI Cards. Thank you, and over to you, sir.

Thank you. Good evening, everyone. On behalf of our company, SBI Cards and Payment Services, I extend a very warm welcome to you, and thank you for joining us today for the investor call from third quarter results of FY '21. I'm also joined by all my senior management here, who are there to answer your questions after the call. Hope you're all keeping safe. To start with, I wish you a very happy and prosperous 2021. The initiation of vaccination with the start of this fresh year had brought new hopes along with the learnings of 2020. So 2021 indeed is a special year, a year of new hope and aspirations. Before going any further, we would like to extend profound gratitude towards everyone, especially our employees who helped us sail through this unprecedented situation and the challenges of 2020 through their commitment and relentless efforts. I also wish to thank all of our customers for supporting us and continuing their patronage with us during this difficult time. The business trends in Q3 have been encouraging. The credit card industry has registered a positive growth Y-o-Y for the first time in November '20 after the COVID disruptions in terms of spends. It is significant to note that SBI Card achieved this milestone of higher spends Y-o-Y a month earlier than the industry. The overall retail spend surpassed the pre-COVID-19 levels. Q3 FY '21 daily average spend was higher by 125% sequentially over Q2 and 117% versus Q4 FY '20. Spends have grown by 8% year-on-year to INR 37,797 crores for Q3 FY '21. Our average retail spends have improved across most of the categories, apart from travel and entertainment. Especially the online spends have shown a significant growth. The online spend proportion in the overall spends has increased to 53.4% for the 9 months of this financial year compared to 44.2% for FY '20. Furthermore, what's heartening to note that with unlocked measures well underway during Q3, even the offline spends have gained momentum. And therefore, if you would note, about a quarter back, our online spend proportion was 54%. It has come down, not in absolute terms, but in percentage terms because offline has picked up. We have steadily been growing our market share in terms of cards-in-force and spends and happy to share that our market share for cards-in-force has grown to 18.8% as of November '20 from 18.1% in quarter 3 of last year. This is as per the last RBI data available. And for spends on YTD basis, as of November '20, it has grown to 20.1% from 17.8% in the 9 months of the previous year. The cards-in-force has witnessed a healthy 15% growth. The cards in force currently stands at 11.5 million as of Q3 FY '21 from 10 million in Q3 FY '20. This improvement paves the path towards recovery of the business environment. In Q3, our new account sourcing was at 918,000 accounts, which was 134% sequentially over Q2 and at 107% versus Q4 of FY '20. In fact, in the month of December, we did almost 340,000 accounts, which is the highest run rate since beginning till date. Q3 also witnessed some significant partnerships, which we aimed. For instance, the partnership with Paytm, which is India's leading digital financial services platform, aims to cater the rise of next-generation credit cards. Another significant partnership was with BPCL for the launch of BPCL SBI Card OCTANE, which is a premium version of the existing card, to offer a vastly differentiated and enhanced value to the mass affluent customer segment. Before I get to the financial performance of Q3 FY '21, let me take this opportunity to reiterate some key facts about SBI Cards' business and the environment. Our business fundamentals are robust. We have in place healthy and compliance financial and corporate governance principles, and this forms our core strength. We have continued to grow the business by leveraging on our strengths. The support of the parent is one of them and capitalizing on India's favorable economic and demographic changes, including its strong macroeconomic performance, rising affluence, increasing consumer demand, rapid urbanization and the growth of e-commerce platforms. These are secular factors and would not change in the near term. We are continuously monitoring the external environment. 2021 has started on a positive note with launch of COVID-19 vaccine, which may have favorable socioeconomic impact. While this indicates a positive future outlook, we remain cautious and observant on the opportunities and challenges. Let me now take you through our financial performance for quarter 3 FY '21. On profitability, while the business environment continues to be uncertain and challenging, the company has performed steadily and delivered profit after tax of INR 210 crores for quarter 3 FY '21 and INR 809 crores for 9 months ending FY '21. The receivables have grown by 4% year-on-year to INR 25,749 crores in Q3 FY '21 from INR 24,776 crores in Q3 FY '20. This is the first time that the receivables days has outgrown the pre-COVID levels. Total income for the quarter is at INR 2,540 crores, which is largely flat compared to Q3 last year. And on a YTD basis, total income is at INR 7,245 crores. Increased business uptake in terms of new accounts and spends in quarter 3 required higher investments in terms of acquisition costs and cash back expenses. When the festival period of financial year '21 fell completely in Q3, it is to be noted that festival has -- festival period in the last year overlapped between Q2 and Q3. So therefore, all the cash-back expenses were taken in Q3 this time. And we believe that the benefit of business growth may follow later, consistent with our business model. We're already seeing this happen in Q3. Our enhanced focus on collections also required hard outlays. For quarter 3 FY '21, while our net revenues grew by 3%, the operating expenditure was higher by 7%, driven by the factors mentioned earlier, which is the higher cash back costs and the higher collection costs, which led to contraction of earnings before credit cost by 3%. However, for the 9 months of FY '21, we have positive leverage of 8%, leading to higher earnings before credit costs by 13% Y-o-Y. The credit risk situation continues to be impacted by macroeconomic variables surrounding us. Post COVID, the assessment of credit risk in the financial services has become complex, and is also influenced by the RBI moratorium and the Supreme Court orders on NPA standstill effective August 31. To cover ourselves for future credit risk, the overall management overlay stands at INR 1,113 crores as on December '20. This is over and above the base provisions of INR 940 crores. Our GNPA is at 1.61% compared to 4.3% at Q2. And pro forma GNPA, including the Supreme Court standstill, is at 4.5% compared to 7.5% in Q2 FY '21. Net NPA for the period, therefore, is at 0.56%, and pro forma net NPA is at 1.6%. The performance of RBI resolution book was not available in the previous quarter. In this quarter, we have better information on that book. And as of December '20, 33% of RBI RE book is delinquent between 30 days and 90 days. On this book, which is not yet NPA, from a provisioning perspective, we have provided the same stage 3 ECL levels. Further, NPA standstill matter is still at the Supreme Court and is ongoing. And from a provisioning perspective, we continue to provide it at stage 3 ECL levels. The SC standstill book has caused a portion of our book not getting edged and be included in the pro forma GNPA. If this situation remains the same, the SC standstill book will continue to adversely impact our pro forma GNPA numbers for the next quarter, as in the normal course of business, this book would have been written off and moved out of GNPA. For the quarter ended December '20, the return on average assets is at 3.3% and the ROAE is 13.8%. For the 9 months ended FY '21, the ROAA is at 4.3%, and ROAE is 18.5%. On liquidity and capital adequacy, our liquidity position continues to be strong during quarter 3. Our capital adequacy ratio for the period ended December '20 is 23.7% as compared to 19.2% at Q3 FY '20. In Q3 FY '21, our Tier 1 ratio has moved to 19.8% from 15.4% at Q3 FY '20. Our credit ratings remain excellent with A1-plus and AAA ratings by CRISIL and ICRA for both short-term and long-term borrowings. These strong credit ratings have been recently reaffirmed by the rating agencies and reflect our robust business and financial fundamentals. Our recently implemented sustainability policy continues to strengthen our commitment towards a better society. Our digital intervention to paperless communications enabled a saving of over 1.96 lakh trees as part of ESG initiatives during the year. Our social intervention for community development remains strong. I would like to reiterate that our business operations have continued during COVID lockdown and have steadily started generating growth. We remain committed towards supporting our colleagues and customers for remaining financially robust. Today, SBI Card is the largest pure-play credit card issuer in India. This is RBI report. SBI Card continues to be the second largest credit card issuer in India, both in terms of number of cards outstanding and amount of credit card spend. We're also the largest co-brand credit card issuer in India. Our availability with products and services supported by a strong technological background, advanced risk management and data analytic capabilities, customer centricity, values of trust and transparency and strong lineage will continue to drive our growth. Thank you very much for your time, and I will request the operator to please open the line for any specific questions you may have on the performance of the company. Thank you.

[Operator Instructions] The first question is from the line of Dipan Mehta from Elixir.

My question relates to the surprising drop in the NPA, which is quite impressive. So can you give us some flavor as to why we are -- I mean how we managed to reduce the pro forma NPA so sharply in just 3 months' time? That's the first question, sir.

Okay. So on the NPA piece, actually, there's a mix of both. We have written off almost INR 652 crores during this quarter 3. And also, we have managed to recover and resolve, through the RBI resolution plan and other schemes, a sum of almost more than INR 400 crores. So that is the reason why you see this sharp fall in the pro forma NPAs, including the Supreme Court standstill. I would, however, point out to the fact that RBI book, which is -- which we have much clearer visibility, though it's not NPA, but in the next quarter, some portion of that might turn delinquent. So therefore, we have already provided for that. If you include all of that, then we are still better than what we were at previous quarter. Aparna, would you like to add anything?

You covered everything.

Right, sir. And my second question relates to operating costs, which have increased significantly quarter-on-quarter. So if you can just throw some light on that, sir, from INR 1,109 crores or INR 1,348 crores, sir?

Yes. So there are 2 elements to the operating cost, which -- and which we -- which I alluded to in my earlier comments. The first one is that the -- we had the festival season in this period, and we had 2 or 3 big campaigns. One was the Filpkart campaign, Big Billion Day campaign, which was a 6-day campaign. And we also had other campaigns around the festival season. So a lot of those cash back offers, which last year was spread between Q2 and Q3, this year were all in the Q3 period. So you have seen as part of the operating costs, marketing costs and the cash back costs have gone up significantly, that is one. And this will help us going forward when these spends, which have gone up significantly, will convert to EMIs and will yield us interest income. So that's the first part. The second part is the collection costs, which have gone up significantly in both Q2 and Q3 because the collection machinery is all geared up and working on the account. In the coming quarters, this collection cost is also likely to come down because for those accounts which we have been trying earlier and which we have not succeeded in recovering, we are tampering away, and we are only -- we are not assigning it both to all channels. It is only limited and focused teams, which will work on these assets. So therefore, these costs are both likely to come down to the normal levels in the quarter 4 and going forward. Nalin, anything else you want to add?

No, sir. I think you've covered it.

The next question is from the line of Nishant Shah from Macquarie.

Congrats on the good set of numbers. Sharp improvement in asset quality. I had a few questions on the asset quality front itself. In the RBI RE book, so just so I understand this correctly. New RBI restructurings are INR 600 crores -- or INR 613 crores this quarter. And INR 377 crores is the EMI refunds received from the accounts which were earlier restructured. Is that correct?

Yes. So we have given the walk here. So if you -- so that walk is Slide 17. So if you -- I'll just repeat it for the benefit of everyone, and then I'll answer your question. So if you recall, we were at INR 2,108 crores as at September. We had fresh bookings of INR 613 crores, so that took us through INR 2,721 crores. Then we recovered -- because these are being paid, not everything is closed, some are closed. And -- but others are paying. And that number is INR 377 crores. And that brought us to INR 2,344 crores. So that's the RBI walk through. So this INR 377 crores number, if you see, is people who have paid their balances and some have even closed it, that is the number.

Okay. And this INR 777 crores number, which is 33%, I assume, of the INR 2,300 crores.

Yes, right.

So these are the accounts within the INR 2,300 crores, which are in 30- to 90-day, but they have not slipped there but have exhibited some stress. And on that, you've already provided for 65-odd percent provision coverage, correct?

That is -- as we provide on our NPAs, the same proportion we provided, yes.

Perfect. Could I press you for the 1- to 30-day number within this INR 2,344 crores?

So we have not declared this 1- to 30-day number. What I would only like to point out is that a significant portion of the book, which is close to the half the book, has paid all our installments. Now we have not declared this number. So for me to disclose it in the call wouldn't be in order. But I can assure that 50% of the book has paid almost all the installments, 3 or 4, whatever due.

Perfect. And on a separate -- sorry?

Nishant, technically, everything other than the INR 777 crores is stage 1, okay? So it is all less than 30 days past due?

Yes. Yes. Okay. Yes. No, I was just hoping to see how many of them have paid each and every installment so far. Fair enough. That's helpful. And second one, question on your yield. So there has been kind of a sharp kind of a decline in the yields this quarter. Is the assumption right that this is largely because, although that has been broadly same, it's because of the period end versus daily average kind of differential in the loan [ balance ]?

So the yields have come down actually for 2 or 3 reasons. I mean -- one is that in Q2, we still had a large proportion of the revolver book. Though they were delinquent, but the interest was still accruing. So therefore, we were -- we had a higher yield. Now a large portion of that is converted to the RBI resolution, which is at a much lower rate of interest. So as this book runs off, and you've seen this runoff of close to INR 375 crores in the quarter 3. And if we hope similar number, we'll have going forward as well. So therefore, as this book runs off, and our EMI book, which is typically at 18% to 20%, that goes up, and the revolver behavior returns, because you have to remember, 9% of the book, which was largely revolver suddenly went off and at a much lower yield. So therefore, that is the reason. And we hope that this will come back as this ratio reverts back to our usual ratio of transactors and revolvers and EMI. Currently, the transactors are at a slightly higher percentage of almost 33%, which in the usual run used to be around 27%, 28%.

Got it. Perfect. And just one last question. One of our competitors has been barred from acquiring new cards. Is that a material benefit? Or is that nothing to bother ourselves with?

So while we never comment on what the competitors are doing or what they have been exposed to, I mean, it's a very logical way to look at this, which is that if somebody who's a leading market player is not sourcing anything, those customers who need those cards will go to the alternatives. And which could be one -- which could be us or could be the other players as well. So in that indirect sense, maybe there could be some gain, but we don't look at the competition to decide our strategy. We do it independently.

The next question is from the line of Jaimin Shah from RWC Partners.

Great improvement in asset quality, a good surprise from you. Two questions on my end is, can you talk a bit on how these RBI restructured or the pro forma NPAs, in terms of behavioral change, what you see when you go for collection? Why were they not paying in September and suddenly, they have started to kind of pay? I'm not undermining your effort, but I'm just trying to understand their -- at their level, how was the indebtedness and they were able to now repay? That's question one. I'll go to question 2 in a minute.

Yes. So Jaimin, thank you for your question. Actually, this is a very nuance question. So before September, what was happening is, one, that the RBI -- we did not have a tool basically. We only had the easy payment plan, which are in-house thing. And the other tool was restructuring, which would have led the account to become NPA and, therefore, the -- colored the credit bureau score of a person. So people were really very wary of whether to go in for a restructuring, which would qualify them as NPLs. And once we got the tool, we reached out to these people. We told them that this is a proper way in which your credit bureau profile would stay protected, though the restructured flag will be up and you get to pay at a very lower rate of interest, which is a fixed rate of interest, 14% and 16%. So therefore, that -- and then we pushed it very, very much. That really helped to bring around a lot of people who initially also had the thoughts that maybe they will get a significant relief from the courts. And that thankfully also came during October, November. So whatever relief had to come, did come. So our maximum sourcing of RBI RE, if you ask me, happened in September and October. Therefore -- November, December was very, very tepid. So all the bulk of it happened in September and October. So many of these people who are sitting on the sidelines and maybe thinking that they would get a large waiver from the courts or from the government realized that it is not going to happen. And secondly, some of the people at least got time to resolve their cash flows while they got this time. So mix of those 2, I guess. I mean, we have no clear answers because when we talk to the customer, they always ask for whatever concessions we can give them.

Yes. Okay, fine. This is helpful because as -- given the September quarter, there were kind of question marks on our sourcing, and we had probably gone too far to kind of get the customers and which is kind of showing up in numbers. But this is -- this kind of allayed that concern. So that's very helpful. The other question was more on the growth and the spend. Just wanted to understand how the behavioral changes have happened on spends. Last time we spoke that once you gave the card, it takes about 9 to 12 months to get activated. Is that change? Are people more kind of inclined to start to spend now given that this is trend in terms of digital payments, et cetera? And also wanted to get a sense on this Google Pay linkage we got. Any -- what are the metrics we have seen since the linkage? And any thoughts how we can kind of push the spends in even on the lower ticket items?

Yes. So on the spends, we have told in the last quarter and many other calls that we saw a secular shift happening, which is the online spends increasing. So online spends have gone up to as high as 54% of the overall spends and 114% higher than pre-COVID. We have seen this trend sustaining. The number for Q3, as I mentioned in my comments earlier, was 53.5%. And that is not because the spends have come down. It is only because the POS has started going up. So therefore, the offline spends have also started going up. And overall, the spends are back to pre-COVID levels. What has changed is in between categories. So while I'll give Girish the opportunity to explain, but I can high level tell you that for categories like apparel, which was not a very popular online purchase, has gone up by as high as 539%. So that's a massive Y-o-Y shift. And similarly, there are other categories. And on the activation, et cetera, I'll ask Girish to respond. Girish, please go ahead.

Yes. Thank you, sir. So there are 3 key trends as was -- as has been mentioned. One is that the online spends continued to be above 50% plus. Secondly, as you see from Q2 to Q3, the movement in point-of-sale, point-of-sale was down 67%, and now it's down only 42%. So that's a very, very positive movement in the point-of-sale. And this is despite the travel agent, and that category is still down at 42%. Overall, we see a much -- POS coming up as close to what pre-COVID levels as it was earlier. The other piece, which is very hearty to note is that the EMI conversion remains very, very strong. We see a lot of conversion which we were seeing earlier from spends to EMI, even though the departmental stores as a category and fuel as a category has seen a larger growth, but as a percentage of spends, we see that the EMI conversion remains very, very strong. The last piece is around contactless and consumer behavior for online. So more than now 1/4 of our transactions are contactless. And with INR 2,000 going up to INR 5,000 as per RBI guideline as a minimum, we believe that, that should be very, very helpful because almost 100% of our cards that we -- subscribers are contactless cards. And we also, as you were mentioning, the Google Pay partnership and host card emulation technology that we have for our customers. So we believe that's going to give a flip. So online, very strong. POS coming up, has already picked up quite well. Travel, still not there. We need to get. Once travel comes up, we will see further more growth in that category. And on the Google Pay piece, as for the question you asked, we are seeing a very good traction. We see more than now 50,000-plus transactions on a monthly basis on the Google Pay platform. Good small ticket size, which is the category that we wanted to enter. Even -- as of now, not all customers -- all merchants are accepting cards there because some merchants still accept only UPI as a platform benefit. Hopefully, as it grows, we will see the traction there, and it's growing month-on-month.

The next question is from the line of Mahesh M.B. from Kotak Securities.

Sir, I've kind of a slightly long question. So just kind of bear with me as I kind of highlight this. If I go to Slide #17, the -- I'm just trying to understand how have you broken up this management overlay provisions. I'm just trying to understand if you are right on this or not. If I minus pro forma gross NPAs minus pro forma net NPAs, I get a number of 2.9%. And if I minus the gross NPAs and net NPAs on a reported basis, I get roughly about 1%, which is sitting there. So the difference between the 2 essentially is the provisions, which you may have created for the management overlay, plus the INR 777 crores, which is there in the 30 to 90 days, of which you've made 65% provision, that also is in the management overlay. Is that understanding correct?

Yes. So anything -- see, other than the NPA, which is a declared NPA, everything else is a pro forma. So whether it's the Supreme court standstill or the RBI RE which we have provided, which is not yet NPA also. So that is how it is. Aparna?

That's correct. His understanding is correct.

Your understanding is correct.

Okay. So if I go with that, with that line of thought, then essentially, you have made roughly about, let's say, INR 480-odd crores pertaining to the revised gross and net NPA and about INR 500 crores for the RE related provisions, INR 777 crores. Is that right?

Yes, 65% of INR 777 crores, so that comes to around that same number.

Yes. So Mahesh, the only thing is, 10%, in any case, one has to provide. So 55% will be the overlay. So around about INR 430-odd crores.

Additional is 55%, correct.

Perfect. So then, is it what extra you have in the balance sheet is nothing. So you just have a risk level of INR 100 crores. Is that understanding correct?

So there is -- after covering the SC standstill and RBI RE, there is still some provision available.

But it's not much, right? It would be INR 100 crores.

It will be somewhere around plus INR 200 crores.

Plus INR 200 crores. Okay. Okay. This is useful, sir. Just one last question. I just wanted to check, is there any possibility that you could start disclosing what is your profile of utilization based on various customer segments at a later date? This is kind of a feedback because we still are struggling as to which part of the portfolio is seeing the maximum stress, this kind of angles from a credit score perspective or from a utilization of card perspective? Some incremental data would help.

Yes, Mahesh, we don't disclose that. But on a generic level, we have been disclosing in the calls that the self-employed and the category C, which is basically MSME and smaller companies, that is the segment which has seen the maximum stress. In addition to this, there are certain sectors like airlines and hotels and entertainment, those sectors continue to be stressed. So whether they were in salaried category or any other category, if they have a jobs problem, that category is showing some stress. But large stress pools are self-employed and category C. I mean we call it category B now, but earlier we used to call it category C. So those are the sectors.

And one clarification from Aparna. This INR 777 crores which is there, you started off with the restructuring exercise and the customers didn't even pay even one installment? Or did they attempt to make some installments and they have kind of completely given up? How should we read into it?

No, no, no. It's not that entire book, I mean, they have not paid anything. Some of them have paid 1 installment, some 2. They are delinquent for more than 30, 30 to 90, that's all. It's not that this book has not paid at all. That's not the case.

Okay. So there is some evidence of an interest to pay?

Correct, correct.

The next question is from the line of Anuj Singla from Bank of America.

Sir, first question, again, on Slide #17. When I talk about the RBI RE book, can you confirm how many payment cycles we have seen in this book till date? And will it be fair to assume the -- ex of the INR 777 crores of book, which is kind of delinquent, the rest of the book is kind of seasoned and we can expect that payment trend to continue. Is that a fair assessment?

So that is largely a fair assessment because see, while we have not disclosed these numbers, cycle numbers, but the bulk of the enrollment happened towards the end of August, September and sometimes -- something in November. I mean November -- sorry, October. So about, as you see, the INR 613 crores, which happened after 30th of August, that number was largely done in October. So that has seen only about maybe November and December, 2 cycles. Prior to that up till September 30, whatever was the book, has already seen 3 cycles. And of course, September and all has seen 4 cycles. So therefore, those who have paid 3 to 4 installments together, and as I mentioned earlier, without giving a number, that's close to 50%. So therefore, we can safely say that if somebody has paid 3 to 4 installments, and many of these have only a term of 6 months, so they would -- in the next couple of months, they would actually have paid the full thing. So from that perspective, it does seem that the book will continue to behave in a similar manner. However, since the economy outside is still a little uncertain, we cannot give a firm guidance as to this will continue. Our initial observation is good. And hopefully, this will continue like this.

Okay. Understood. Sir, second question is, again, this 9% of book. What is the tenure on a -- maybe on a weighted average basis? We offer 2 plans, 24 months and 12 months. So what is the weighted average tenure for this book.

Do we have that number?

I don't think we have the weighted average, but we offered only 12 and 24 months. So...

It'll be somewhere in between.

Somewhere in between. I mean, we don't have it readily available.

Within it, what I'm trying to get at the drag on the NIMs or the yields. When can we expect that to normalize, that was the key intention. So when we...

What is happening -- actually, see, a couple of things are happening. One is that a lot of people are enrolled for, let's say, 12 or rather 24 months, they are paying upfront. And we have no penalty for paying more or earlier. Some of them are coming back and saying that you reopen a line once we pay upfront, et cetera. So we are seeing all shades. So by -- I think we'll have a much clearer picture by quarter 4, by which time, most of these people who have the money would have either paid up or settled, and we would have a clarity on that. At this point, a weighted average number would still be a little early, in my view.

Okay. Understood. And sir, you also talked about the receivable mix, the revolver has gone to a multi-quarter low of 29%. How should we look at this number? And on the corresponding yield trajectory, when we look at maybe in FY '22, what kind of maybe uptick in this number should we be expecting?

Yes. So 2 things. One is that we -- as I mentioned earlier, that we had lot of spend from the festival side in the quarter 3, especially around Diwali and that festival, between Durga Puja and Diwali. That has converted to assets in the usual way that we have, and that is converting to EMIs. So that will see -- the effect of that, we will see in quarter 4. Again, during this time, the Republic Day sale, we are again partnering with Amazon, and we are seeing fantastic results, actually, much more than what we had budgeted for. So all of this is giving us hope that this spend will convert to the EMI. So EMI should trend back to the normal levels, which was around 30% or thereabouts. What was that number? 32%. So we are hoping that from 29%, it should trend back to maybe 30% plus in Q4. And going forward, it should trend back to that original number. On the revolver side, actually, again, this -- right now, it's very, very low. But as we -- what our hope is that as this RBI RE people continue to pay up, and many of them actually have paid, let's say, 6 months installment, we are starting to reopen, we will start to reopen their lines. And these are the people who used to revolve periodically, not always. And hopefully, this -- all of these actions together and the new customers which we are getting, which will start to get seasoned, they -- all of that will start to trend back the revolver number. Because if you add the RBI RE book, large portion of which was revolver earlier, plus the revolver today, we are back to 38%. So our hope is that as the RBI RE book trends down, we reopen cards, we get more customers, we will have that number, but this will -- the revolver book will probably get back to where it was sometime in the next year. Not going to happen very, very quickly.

Yes, sorry, I was on mute. Sorry, can you hear me?

Now we can.

Sorry, last question. Recoveries, INR 125 crores is, again, very significantly higher Y-o-Y, Q-o-Q. Is it a function of the front-loading of credit cost we have done and write-offs we have taken? And so should we expect this number to remain elevated given our focus on collection or -- and our write-offs we have taken? Or maybe in FY '22, we can see an elevated number on this. Is that a fair assumption, given that we've already taken the write-offs early in this cycle?

Yes. So it is a mix of both, actually. So one is that the write-offs are higher. So therefore, the pool itself is higher. But you have to look at that in the business as usual, we never had a 1% recovery rate. It used to be much lower than that. So therefore, the 1%, which is an improvement over the previous quarter, actually tells us that the focus on recovery and the collection efforts have also played a significant role. So both these things together. And our hope is to continue with this number till we have a significant pool of written-off accounts to continue to accept and continue to focus on these customers. Anu, would you like to say something else?

What I will say is that the recovering efficiency has gone up [indiscernible].

Sorry, ma'am, your voice is not audible.

Yes, just hold on. I mean we have the Collection Head here, so I would ask her to actually pitch in. Yes, Anu. Go ahead, please.

So the recovering efficiency has gone up on account of collection efforts as well as the fact that there was a write-off also that has happened in quarter 1. So quarter 1, since the lockdown was there, and that write-off actually happened because [indiscernible] So when write-off has come, we have begun to actually strike from that and can strike better, in fact. So both have added to the recovery efficiency to go up, and we hope that since the write-off happened, we have taken quarter 3 a little higher, the fact that we'll try to maintain this consistency in quarter 4 as well.

The next question is from the line of Anand Laddha from HDFC Mutual Fund.

Sir, if you can give some color on your RE book? Or what the learnings we had in terms of what was the sort of category of customer who didn't pay or who took the restructuring, where self-employed customer or salaried? What was the utilization level on their cards? Was there any early signs that you could have noticed that these customers are borderline customer or risky customers? And all this should -- the learning which we have, we can have, if you can share that? And what's the outlook on credit costs for next year? Given the fact this year we'll be closer to 10% to 12% of credit cost, should it go back to a normalized level of 6%, 6.5% credit cost next year?

So Anand, see, subject to the economy, because that is something we cannot predict. Hopefully, it will continue to improve and continue to be better than what it has been. Subject to that, the credit costs should be lower. Because we are cleaning up this year whatever was there. And if the Supreme Court hopefully lifts the stay, we will be able to clear up the book as well, that's one. Second is, in terms of whether these customers were border customers or whether they had any inclinations, I think, I'd mentioned in my earlier calls, that almost 54% to 55% of these customers in the previous 24 months, had not missed anything, maybe missed a couple, one installment at the most. So they were largely good customers. So the surprising thing is that while on a whole, the self-employed segment and the cat B segment did see more delinquencies, there is no pattern to this, that these kind of customers were likely to default. We don't see that pattern at all. In fact, in the last week -- I mean, last 15 days of December, our -- most of the senior management team, full-time employees, they were all out on the street trying to meet all the customers who are not paying up, whether they are in RE or any other segment. And we heard different stories. People had lost jobs, and they were struggling with education, struggling with paying fees for their children. I think good people who are middle class people, and people have gone to the villages. I mean all kinds of shades. So we really don't have a visibility onto what the signals could have been on particular kind of customers. The learning is that we need to continuously be in touch with customers, continuously monitor them. And the portfolio basis, look at all the alternate data. And that is the models we are building now to incorporate all the alternate data, which could be leading indicators rather than the default occurs and then we get to know. So I think a lot of learning in terms of what to do to monitor things. But in terms of, if we could have guessed, these were the customers who are likely to default, no. Because the credit scores were good. They were all paying well for the last 2 years. And we always considered them to be good customers. They still are. I mean they're not shying away, they're not running away. They just have a situation.

Our LGD assumption on 65%, 66%...

Sorry to interrupt you. Sir, may I request you to speak a little louder?

Sir, our LGD assumption on 65%, 66%. Despite these customer being good customer, good credit score customer, we still believe we will have an LGD of this percentage? Or do you think this LGD will come down for us?

So as I mentioned, so this INR 777 crores, it is not that all this entire book has not paid anything, no. Some have paid 1 installment, some have paid half installment, some have paid 2 installments also. Just that for -- as on December 31, they were overdue by 30 days or more and below 90 days. So therefore, we don't expect this book to all go -- all become delinquent and be written off. No. We do believe that we will be able to recover a part of this book. However, as a matter of caution, we have provided this 65%. It doesn't mean that we believe that it will be 65% delinquent going forward, no. Just a caution and just to comfort that we have provided extra. If we are able to recover, well, we'll use that provision elsewhere. We have done that in Q2 and Q3. Whatever we have written off, we have not -- whatever provision was released from recovery, we've not written it back. We have used it and kept it as an overlay as well.

Lastly, sir, this time, the other income was higher. Could be because of recovery from return of asset. If you can give any color of what sort of recovery can we expect going forward? Or this quarter, what we saw was a one-off?

So you saw the RBI book. So INR 377 crores was recovered and resolved from that book. So therefore, the recovery is something we will continue to see going forward, especially from the RE book. On the SC standstill, it is a little tricky because that is already a seasoned book, and people have not been paying. So there, possibly, we'll have to have more of settlements and more of extraction rather than actually -- or legal rather than actually a recovery. Nalin?

So I'll just add to what Mr. Tewari said. INR 377 crores is the normal collection that we've received from RBI RE. The other income that you see is the recovery from the written-off account that we have collected. And Anu had spoken about it, earlier about this.

The next question is from the line of Ajit Kumar from AMBIT.

Just one question. In one of the earlier calls, you had highlighted that you have tightened the credit criteria for self-employed segment and sourcing from external channels. But the share of self-employed segment in new sourcing has been consistently increasing in the last 3 quarters. It was 21% in 1Q, 22% in 2Q, 24% in 3Q. So sir, even within salaried, the share of government and PSU segment in new sourcing has been coming down consistently in the last 3 quarters. Similarly, sourcing from external channels have also increased from last quarter. So why is this happening, actually?

Yes. So you're right. The self-employed proportion in the new sourcing has gone up from 21%, 22% to 24%. Now out of this 24%, the open market is just 4%. 20% is from Banca. And Banca, as we have been explaining time and again, that Banca is all prequalified through the Shikhar program, which we run, where the data, if there's certain qualifying criteria, like, let's say, the liability balances or housing loan account, et cetera, from the bank, goes to the bureau and after prequalification, then it comes to our executive and the bank executive, who together market it. And that is -- and the performance of this segment in Banca is borne out by the delinquency levels. The Banca delinquency levels overall and also within this segment are far lower than what they are for open market. So while the proportion is going up, it is only in Banca. Not in open market, where we have cut down the segment significantly through whatever tightenings we have done and those tightenings remain in place as of now.

Okay. And the share of government or PSU segment in the new sourcing within the salaried part, that has also been coming down in the last few quarters?

I think that's marginal. That's not very significant. In fact, what is also happening is that we are actually engaged with a lot of government agencies or government ministries, et cetera. This is an ongoing program. So we hope to have this back on, not as of a conscious policy, just because the self-employed has gone up a little bit, is it's fallen. Absolute number is going up.

Okay. Okay. So in terms of share, in 1Q, it was 47%; in 2Q, it was 42%; in 3Q, it is 38%, actually?

Yes. I think this number -- because see, the first -- quarter 1 and quarter 2, the sourcing itself was so low. So therefore -- and since the approaches to various marketplaces, et cetera, were all closed, so therefore, all we were sourcing was from government, PSU and all these. Therefore, percentage -- incremental percentage was higher. But as the market opened up, in Q3, we had a full range. I mean, most of our places are open. So therefore, the incremental sourcing you see is higher in the self-employed category. But as we go forward, I think it will revert back to the overall theme, and which is also reflected in the cards-in-force, that is not changing too much.

[Operator Instructions] The next question is from the line of Shweta Daptardar from Prabhudas Lilladher.

You partially answered my question previously, but I would just like to have some color on. So you mentioned that incremental sourcing is now back to pre-COVID levels from the open market channel. So what is the customer acquisition rate differential between Banca and open market channel, if you can quantify that?

Yes. So that percentage terms for this year, it has been in the range of 42%-58%. And if you look at Q3, it's 48% -- sorry, it is almost -- how much is that?

48%-52%.

52%-48%. Then that's --- sorry, it is -- that's the incremental, that's the sourcing. Fresh sourcing. 48%-52%. So therefore -- but in certain quarters, it was pretty skewed. So it is starting to revert back to the mean, which used to be 50%-50%, but still Banca is more.

No, sir. So I think I'll split the question, again. So 2 points here. You just mentioned in the previous question that it's getting back to the pre-COVID levels. But then the kind of -- you also mentioned that it transpires into better asset quality when it comes from Banca channel. So -- but if I look at your pre-COVID numbers as well, your SBI sourcing was definitely slightly lower. So can you throw color on this? And secondly, what is the customer acquisition cost and the differential from open market channel and from the Banca channel?

So in terms of acquisition costs, clearly, Banca is much, much lower. I don't think we have given this cost. Have you given this cost? We have not given this cost, but much lower. So therefore, from an acquisition perspective and from a delinquency perspective, the Banca channel-sourced customers are clearly much, much better. Where we have an advantage in open market, I mean, we have been a significant open market player, and we still are, is the fact that the sourcing -- I mean what we do from open market are customers who are carded. So therefore, they spend more, they revolve more. And therefore, they are more profitable customers in a way. So we have a good mix of open market, which gives us profitability, though the delinquency could be a little higher; and Banca, which is much lower delinquency and also a much lower cost. So we have this mix. The challenge is to reduce the open market delinquency and increase the Banca spends and profitability. So that's what we are working on.

Okay. Okay. Sir secondly, again, a slide reference question. So you mentioned that credit cost will be coming down in the following year. But if I remember your Q2 commentary, you had mentioned that credit cost would slightly remain at elevated levels in the second half of the year. So now that they are down almost to 12%, what could be the ballpark number as we exit this current fiscal? Some guidance there?

No, no, we never give guidance. So therefore, I can't give you a number. What I can tell you is that now you have, one is that the Supreme Court standstill book is there, which is a book, we say, that it's not a very performing book. So therefore, there, we'll have to have whatever losses we need to take. With the usual NPAs which accrue in the normal course of business, their delinquency levels are back to pre-COVID levels, so -- in fact, lower. So therefore, the NPA accretion will be lower than what used to happen earlier. And finally, to the RBI RE book, as we have shown this time, 33% are overdue by 30 days, doesn't mean that all of them will go back. So therefore, as a mix of all these 3 factors, I think, our credit costs are going to trend lower. I will not predict because the economy is not known, how this is going to behave or if something might happen in the economy, that we don't know about. But in terms of what we have, the stock of what we have and the flow which is happening through the fresh acquisition, I think, we are reverting back to the mean, which is pre-COVID levels. It might take a quarter or 2 depending on the Supreme Court opening it up for us. But it's not going to go up, that's all I can say.

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

Sorry, I missed out in terms of the flow of this pro forma GNPLs in terms of how much has actually slipped? And what was the recovery and what is getting into the RBI RE pool?

So we have given all this on Slide 17, Kunal. But just to recapitulate, on the RBI RE, we had INR 2,108 crores in September, we are now INR 2,344 crores. And that had INR 613 crores of new bookings, but a recovery of INR 377 crores. So RBI RE, at the moment, 33% is overdue by more than 30 days, which is INR 777 crores. And we have provided at 65% for that book, though we believe that not all of this is going to go bad. The Supreme Court standstill is around that same number, which was, I think, INR 649 crores last time? What was it?

INR 762 crores to INR 749 crores.

Yes, INR 762 crores to INR 749 crores. So it's not much change. We have recovered a little bit because we've not been able to write-off anything. But that is a book which we believe -- I mean, the recovery or the extraction will be low. It is -- and therefore, this book, if we are allowed to, will qualify as NPA and maybe we'll ultimately write it off or settle at. And then there's the usual NPA number, which we have already seen revert back to the pre-COVID levels or even lower than that. So that accrual of fresh NPA is lower. So that's the overall thing.

Okay. So when we look at from, say, 7.46% to 4.51%, the decline which is there, maybe that would have either been the new booking into RBI RE or recoveries, and there are no fresh accruals is what you are projecting? So I wanted the walk for this 7.46% to 4.51%.

Okay. So see, the 7.46% had 2 components. One was the NPA component. And the other was the RBI -- the Supreme Court standstill at that point of time. And now we have made a write-off of INR 682 crores -- INR 652 crores or INR 682 crores?

INR 648 crores.

Sorry, INR 648 crores in this period. So from the book, which was NPLs as on 31st of August, a significant portion, which is INR 648 crores has been written off. And that is a reason for reduction. The other reason for reduction is the recovery which we have made, which is just short of INR 400-odd crores, which is a mix of both recovery, closures, all of that. So therefore, INR 648 crores of write-off and about INR 400 crores of recovery. That's the walk which you have in this.

The next question is from the line of Dhaval Gada from DSP.

Sir, congrats on a good performance on the asset quality front. Just 2 questions. One is related to the EPP number that you shared last time. What is the update on that front? And if we carry any provisions against that? And the second -- I'll ask the second question afterwards.

Okay. So the EPP book, see, we don't provide for that. That's about 4.5%, is the usual -- no. Sorry. The book is INR 500-odd crores. Provision wise, I think, we just provide...

Provide it like a standard asset.

Standard asset, which is about 1%.

And anyway, EPP is, to book an EPP, you have to pay everything overdue. So anyway, the account becomes current to be able to be seen as EPP. So that is actually our standard book.

And what we have seen so far from that book is that the payments are regular. So we don't have any concerns on that book.

Okay. And the second question...

Yes. Go ahead, please.

Yes. And the second question is just a recalculation. So last quarter, on a pro forma basis, we had about INR 1,790 crores of gross NPA. And this quarter, I think we have about INR 1,160 crores of gross NPA. And you mentioned that INR 648 crores is the write-off, INR 400 crores is the recoveries. And therefore, the implied number of slippage would be like INR 400-odd crores, INR 400 crores to INR 450 crores?

Fresh NPA, you mean?

Fresh, fresh, yes.

Yes, it's about that number, yes, you're right. Because that you can evaluate from that. Correct.

The next question is from the line of Harshvardhan Agrawal from Infina Finance.

Sir, just one data-keeping question. The ECL provision that we have on our book, which is of 8%, what will be the absolute number of that?

8% ECL.

I didn't get the question.

He says, what is the absolute number for that 8% ECL, the last line.

8% of the overall outstanding.

Overall outstanding, right.

So that would be close to around INR 2,000 crores?

Yes.

Yes.

And all this provision that we have discussed for RBI RE, NPAs and EPPs, et cetera, so all those provisions will be part of this INR 2,000 crores, right?

Yes, correct.

Yes. Just to clarify, like we said, EPP is not any additional provision. EPP is provided as per the standard book.

Right. But whatever provisions we have on that EPP would be part of this?

Will be part.

Yes. Yes.

The next question is from the line of P. Kothari from Pictet.

This is Prashant Kothari from Pictet. My question was the market share gains that you're having, especially on the spend side, which has been quite phenomenal this year. If you can you just help us understand what is leading to these gains? Are there any specific campaigns which are helping us? And also, on the flip side, who are the ones who are losing it? Is it the private sector banks? Or is it the foreign banks? Who is losing on that share of spend?

Okay. Girish, you can answer that.

So we, early on in the COVID period itself, recognized that the spends are moving online. And given the technology and the analytics capabilities that we have built over a period of time in the organization, we were able to figure out and see that the consumer behavior is changing. So we started tieing up with a lot of those players where the consumer -- those consumers are going. Our offers online increased drastically in this period. In fact, we have done more offers in the COVID period than in pre-COVID scenario. Also, if you see the data on Slide 10, we also recognized that there was a consumer movement towards not only paying, let's say, new categories, for example, insurance as a category which was already there, but it just exploded during that period of time. So we capitalized on that and got that benefit. The other is like apparel as a category because clothing, though it was subdued in quarter 2, but quarter 3, it suddenly picked up because -- and people were unable to go to point-of-sale scenarios. And quarter 3, the online sale of apparel and that category, just fashion as a category just picked up. So we tied up with a lot of those players and partners in this category. So that has helped. One, the market share being lost is basically either some players who are multinational and were -- their portfolios were very, very dependent on travel as a category and were catering to that. So there are players who have lost market share in that. And mostly, that are multinational. If you look at, a couple of Indian players have also lost market share. But they would have lost because they would have blocked a large majority of their customers and the balance customers were not active enough. So that would be the reason. Otherwise, it is restricted to people who are very, very biased towards travel as a category. Apart from market share, I think what we have also done is, we have targeted -- we utilized our capability to target customer one on one. And have sent a lot of segmented offers during this period, which we have seen has also helped us not only in activating the customer, but also increasing the spend.

So I think you can look at the RBI data for November, which is the last available data. So for 9 months -- or rather 8 months, everybody has lost in terms of spends, everybody has shown a negative growth, but some have shown a less negative growth. And we are amongst those players who have shown a less negative growth. There are other players, both domestic and foreign, who have lost -- who have shown more negative growth. And as Girish explained, largely because they were inclining more towards travel or maybe some of them have blocked a lot of their cards from spending. That's the reason.

So just as a -- would you think that this is -- all these gains are not sustainable as those travel kind of open up and as well as those blocks which have been put in place are also released, do you think you'll lose some market share going ahead.

So something would definitely revert. We cannot say that this is going to continue. The only thing is that our engagement with the online players and the customized offers which we have been rolling out, and they have worked. So I think we have a little bit of a head start. So in that sense, while maybe this thing will not continue in the shape, we'll surely be more than what we were there. And as we continue to work towards more hyperpersonalization, I think, we'll be able to sustain a lot of our gain, if not the full one.

The next question is from the line of Karthikraj Lakshmanan from BNP Paribas Mutual Fund.

Congrats on a good set of numbers. Just on the non-RBI RE book, how is the over 30-day overdue book compared to, say, last year? Has it come down? Or is it in line?

So the 30-day DPD book, which we haven't declared here. But as I mentioned earlier in my comments, actually, it is behaving better than last year. So it is -- the numbers are lower than last year. And this is largely a result of the fact that one book got cleaned off a little due to the moratorium blocks and those who ultimately did not pay. And secondly, our credit filters, which we have tightened significantly, twice over, once in May and then later on in October again, so those are also having an effect in terms of bringing down that delinquency level. So they are much, much lower than last year in all segments.

Right, sir. And just in terms of the gross new additions in the last 4 quarters is close to 27-odd lakhs and the net addition is almost 15 lakhs, 14.5 lakhs. So the difference is largely -- is it the NPA accounts or the inactive ones? How should we read that? And is the number in line with normal years? Or is it higher this time around?

So this is a mix of both. These are attrition, which is -- attrition is, again, both. One is people who actually request for a closure, which is a voluntary attrition and those whose account become NPLs and delinquent and therefore, those customers go up the book, so that is involuntary addition. So that has gone up a little. But I think the traditional numbers, whatever there, will revert back in the next year. At this point, the attrition levels are slightly higher, especially in the quarter 3.

Ladies and gentlemen, due to time constraint, we'll take the last question from the line of Manas Agrawal from Bernstein. Due to no response, we move on to the next participant. The next question is from the line of Pranav Gupta from Aditya Birla Sun Life Insurance.

Could you hypothesize the average tenure of the RBI RE book?

Pranav Gupta, sorry to interrupt you. Your voice is not coming clear. May I request you to come in a better reception area.

Is it better now?

Yes, yes, much better.

Yes. Just wanted to get a sense on the average tenure for the RBI RE book. Why I'm asking this question is because once this book starts running off, only then will you see your reversion to the usual revolver percentage of the overall spend. So could you give us a sense on that? And the second question is on the corporate side. So spends have bounced back pretty well, even though travel and the usual spend categories might not have normalized. So if you could give us a sense on what are the new categories of spends that are growing here for this to surpass pre-COVID levels as well? Those are the 2 questions.

Yes. So the RBI RE, the average tenure, when we began this, was around 18 months because 24 months and 12 months were the 2 categories. However, as we -- as I explained earlier, we are seeing a lot of different behavior. People are prepaying, people are paying in advance and closing the accounts or paying off some installments to get the card opened. So we can see all of that. And therefore, it may not stay till 18 months. That's our sense. We've already seen this book run off from INR 2,700 crores to INR 2,344 crores, so almost close to INR 375 crores is run off in one quarter itself. Going by this logic, hopefully, in the next 6 quarters, we should run this book entirely. But however, we will wait. This is too early to actually, I mean, decide on anything. On the second question on corporate spend, I'll ask Manish, our Sales Officer -- Chief Sales Officer to respond.

Yes. Yes, the corporate spends have risen, but the composition of the spend has changed. See, we are working with our corporates to look at new use cases, which are primarily around utility payments, online expenses, training, marketing expenses, online tax payments done through the payment aggregators, that has taken off in a big way. We're working with some of the good corporates, big corporates who are adopting to these corporate payment methods. And that's what has increased the quantum. So yes, the T&E is low, the travel agencies who buy airline tickets, that is low. But what we term as B2B spend, obviously, we remain pretty tight on the credit parameters for this. But that has built up nicely. So when we look at this quarter and the recent months, it's actually higher than the pre-COVID quarter. So that is what is coming up. Corporate travel is still some months -- more months away because a lot of companies will continue to be, in our opinion, working from home for some more period of time. And international travel was happening on air bubble only, though there is a recent media article which says government is asking airlines to do domestic flights to full capacity. But that has not got resolved, it's just a media item that was seen. So that will take some time to come back. And one more important thing is that our expectation is that these new use cases have more permanence to it. So it's not that this is, for the time being, the corporates are doing this because people who have -- the companies who've adopted this have been continuing to do this for the past several months, and we expect that to continue and the travel piece to come and add back.

Sure. Sure. Just lastly on the EPP. So I just wanted to understand that in an earlier question, you kind of alluded to the fact that a customer has to pay all his overdues to convert his balance into EPP. So essentially, what is he converting here? Is he converting a large spend which he has incurred in this month? Or is -- I mean I just wanted to understand what is the conversion which is happening here on the EPP?

Aparna?

So effectively, it is balance. So everything that is overdue. So if you have missed the last 3 payments, so technically, you will have 3 minimum dues overdue. So you have to pay all the overdue amount. And whatever is the balance on the card, that is converted to installments at the lower interest rate.

So would he -- so see, in your example, the past 3 months bills are due. Would he have to pay minimum dues on those past 3 months? Or would he have to clear the entire balances on those past 3 months to convert into EPP?

He has to pay the minimum due because technically in a credit card, what's billed and the aging and everything works on minimum due. So he has to pay 3 minimum dues to become current. And then whatever is the balance is converted to installments.

Just paying the minimum due for each one, that's how it is.

I will now hand the conference over to Mr. Ashwini Kumar Tewari for closing comments.

So thank you, everyone, for your comments and questions. They give us a lot of insight, and we will prepare ourselves accordingly for calls to come in, in the coming days. I hope all of you have a very good evening. Thank you.

Thank you very much. On behalf of SBI Cards and Payment Services Private Limited (sic) [ SBI Cards and Payment Services Limited ] that concludes this conference. Thank you for joining us. You may now disconnect your lines. Thank you.